Endless, overshadowed by the more structured Blonde, raised a strange question: why was he building a staircase?

A clue appeared in the magazine released alongside the albums, featuring an introduction by artist Tom Sachs.

Sachs is known for leaving the process exposed in the final work. His studio operates with near-militant adherence to defined procedures, challenging the romantic notion that art emerges from chaos. His work often embodies bricolage, which is constructing something from a diverse range of available materials.

That ethos echoes throughout Endless. It opens with a monotone poem by Wolfgang Tillmans (who shot the now-iconic Blonde cover) about livestreaming one’s life. The project moves from Aaliyah covers to southern hip-hop to Johnny Greenwood strings, Trent Reznor synth textures, and Jasmine Sullivan vocals. It feels assembled from multiple worlds, yet unmistakably its own.

Taste is having a moment in tech. Greg Brockman recently called it a core skill in the AI era. But taste is often misunderstood.

Popular usage treats taste as having “good” or “correct” aesthetic sensibility; my view is that taste is simply an opinion. The more experiences and references you have, the stronger opinions you can form.

When someone says “you have good taste,” they often mean “you like the same things I do” or “you like the things you’re supposed to like.”

Some may prefer Monet, Sargent, Rockwell, or Picasso. As the expression goes, “it’s a matter of taste.”

Craft is something else.

Craft is acting on your taste. The goal of an artist is to inhabit a world that is completely their own. To create something only they could create and return to the regular world for a moment to share the creation. The question of “why create?” is then a pertinent one.

Plato’s allegory of the three beds offers a useful metaphor. God creates the ideal bed. The carpenter builds a bed in its image. The artist paints the carpenter’s bed, a representation of a representation.

But perhaps great artists and inventors are not painters in this analogy. Perhaps they are carpenters: devoted to the discipline of making, striving toward an ideal they may never fully reach.

Viewed this way, Ocean’s staircase begins to make sense. We as fans receive a polished final product, but it is the ritualistic commitment to the act of creation that makes it possible. The goal may be to produce a finished staircase or album, but excellence is forged in a monk-like devotion to craft itself.

The most obvious parallel in technology is Steve Jobs. His father taught him that the unseen back of a cabinet should be finished as carefully as the front. Calligraphy influenced the Macintosh’s typography. The Apple I motherboard was designed to be visually elegant despite being hidden.

Jobs framed taste not as objective beauty but as exposure:

Less discussed is his admiration for Sony. He viewed the company, and co-founder Akio Morita, as a North Star for design and consumer-friendly innovation. The Walkman influenced the iPod. He even considered running Mac OS X on Sony VAIO hardware. He admired that Sony was committed not only to researching consumer technology, but to building it beautifully.

Apple products often felt like attempts at the platonic ideal of a computing device. That devotion to craft produced enduring dominance for the company and led to it becoming the most valuable in the world for a period of time.

This ethos persists in modern technology companies. Stripe, for instance, feels crafted at every level, as if striving not only toward the ideal payment processor but toward the ideal technology company.

ASML builds machines few will ever see, engineered to tolerances that border on the absurd. Hermès limits production because the craft cannot be rushed. Nintendo obsesses over how a button feels under the thumb. Ferrari optimizes not for efficiency, but for emotional precision. SpaceX’s Starship reveals visible weld seams, which, like Sachs’ work, make the process of construction legible.

In software, Mercury and Ramp bring the same devotion to clarity and workflow design. Their products feel considered and thoughtful rather than purely functional.

Craft reveals itself across mediums. Akira redefined animation through obsessive hand-drawn precision and environmental detail that still feels modern decades later. More recently, works like Expedition 33 demonstrate that even small teams can produce cohesive, authored experiences and worlds that resist formula. You could feel the care in the soundtrack. Every detail clearly mattered. Nothing was accidental.

In AI research, latent space describes how concepts cluster by proximity. Craft operates similarly. Certain objects feel adjacent despite belonging to different domains: True Detective Season 1, a Birkin, a Patek Philippe Nautilus, a McLaren, the iPhone, and Blonde. They share a devotion to excellence that transcends category. It’s hard to describe, but they have a certain aura. They radiate brilliance and intention. They were clearly made on purpose and with great care.

What does exceptional craft teach technology?

• Craft as the end, not the means.

• Maximize exposure to great work across a variety of domains.

• Taste is having opinions, craft is the chance to create something uniquely your own, something that did not exist before. A chance to be the carpenter.

Today, there is much discussion of “AI slop,” content clearly generated by prompting a machine. But slop did not begin with AI. Humans have produced plenty of it.

Formulaic four-chord pop dominated the 2000s. Frank Ocean’s music felt like a rejection of that template, and Endless felt like a statement embracing the path of the carpenter.

It is the pursuit of an ineffable ideal. The willingness to push beyond comfort. The discipline to care about one’s work in painstaking detail. The creation of a standard that endures beyond the moment whether in art, companies and products, or culture.

In an age obsessed with scale, the most enduring companies behave like artisans, building toward an ideal they may never fully reach.

Taste is aesthetic preference, but Craft is disciplined devotion to excellence.

Everything is Computer by Ryan McEntush

https://a16z.com/everything-is-computer/

Consumer electronics underpin all of the most important technologies of our time. And everything, absolutely everything, is based on the indispensable paradigm of the smartphone. An electric vehicle is a smartphone with wheels. A drone is a smartphone with propellers. A robot is a smartphone that moves. The differences matter, of course, but the family resemblance is impossible to ignore.

This explains why so many electronics companies — especially in China — now appear to be building many different types of products at once. From the outside, this looks like reckless sprawl. From the inside, it is straightforward reuse. The products change; the components do not. A company that builds smartphones at global scale already understands batteries, sensors, compute, thermal management, wireless stacks, and high-volume manufacturing. To build electric vehicles, it doesn’t need much more than that.

Everything that can be electrified will be electrified, because electric systems are the native substrate for code. Power electronics become the transmission, motors become the engine, and software becomes the differentiator. Across land, sea, and air, mobility is shifting to battery-electric and hybrid architectures. Rockets are the lone exception — chemical propulsion still dominates on thrust-to-weight ratio — but even these systems are increasingly going electric. Starship carries hundreds of kilowatts of power electronics and the same Tesla batteries you’d find in a Model 3 or Powerwall. The rocket equation stays chemical; everything around it becomes electrical

Today, the United States lacks serious modular middle firms that bind together the electronics ecosystem. Exceptional figures like Elon Musk have succeeded despite this by aggressively plugging into global supply chains and then pulling critical modules in-house over time. But that is not a reproducible strategy for the world ahead. We cannot make national success contingent on finding a hundred more Elons able to vertically integrate down to a screw. If we want the default choice to be American — fast, competitive, and reliable — we need to rebuild the missing layer. Reclaiming our electro-industrial future starts with building an American modular middle.

Standing Out in 2026 by Lulu Cheng Meservey

In 2025, the din of direct-goers turned into cacophony as even the stodgiest of companies started factory-farming for “CEO content” and “storytelling.” The edge now came from winning attention, primarily through video. Cue the avalanche of corporate podcasts and cinematic launch videos. Some built quality dedicated audiences while others, in blind Icarian pursuit of engagement, went the way of Talk Tuah.

In 2026, narrative alpha will come from doing real things.

That means:

• Putting in real effort to create beautiful durable things that exhibit taste and craftsmanship, with zero tolerance for slop. As Thomas Paine wrote during the American Revolution: “What we obtain too cheaply we esteem too lightly; it is dearness only that gives everything its value.”

• Showing real evidence for real outcomes, not flashy claims or vagueposts

• Designing real world events and artifacts, leaving people with memories that far outlast the cheap “impressions” generated by brainrot content troughs

• Showing up as real humans, with real flaws and foibles, instead of ultra-polished personas following AI scripts

• Forming real relationships that will weather time and tide

Many Small Steps for Robots, One Giant Leap for Mankind by Packy McCormick and Evan Beard

Not Boring by Packy McCormick

Many Small Steps for Robots, One Giant Leap for Mankind

Welcome to the 1,179 newly Not Boring people who have joined us our last essay! Join 256,826 smart, curious folks by subscribing here…

Read more

3 months ago · 99 likes · Packy McCormick

If you believe there’s a massive set of economically valuable tasks waiting on the far side of some threshold, then the optimal strategy is to straight-shot it. Lock your team in the lab. Scale models. Scale compute. Don’t get distracted by deployments that might slow you down. Leap.

If you believe, like we do, that there is a continuous spectrum of economically valuable jobs, many of which robots can do today, then the best thing to do is to get your robots in the field early and get to work.

Each deployment teaches you where you are on the gradient. Success shows you what’s stable, failure shows you where the model breaks, and both tell you exactly what to work on fixing next. You iterate. You take small steps.

It’s widely accepted in leading LLM labs that data is king. The optimal data strategy is to climb this spectrum one use case at a time. You don’t need “more” data. What you really want is diversity³, on-policyness⁴, and curriculum⁵. Climbing the spectrum iteratively is the strategy that best optimizes for these three dimensions of good data for any given capital budget. Real deployments on your bots get you on-policyness (nothing else can), the market intelligently curates a curriculum, and both provide rich and economically relevant diversity.

We’ve learned this lesson over years of deployments.

Whenever robotics evolves to incorporate another aspect of the job spectrum between automation and autonomy, it also unlocks another set of jobs, another set of customers, another chunk of the market. One small step at a time.

Take screwdriving. It is much easier to use end-to-end AI to find a screw or bolt than to try to put everything just so in a preplanned and fixed position. Search and feedback is cheap for learning systems. Our robot can move the screwdriver around until it feels that it’s in the right place. It wiggles the screwdriver a little. It feels when it drops into the slot. If it slips, it adjusts. And when our robots figure out how to drive a screw, it unlocks a host of jobs that involve screwdriving. Then we start doing those and learn the specifics of each of them, too.

We learn on the job and get better with time. Many of these robots are imperfect, but they’re still useful. There’s no magic threshold you have to cross before robots become useful.

Vanity Fair Piece About TBPN

https://www.vanityfair.com/news/story/the-technology-brothers-have-silicon-valley-in-their-thrall

A great piece about TBPN. I’m a huge fan of them and also got a little shoutout in this piece.

What if they could bottle the energy of a Silicon Valley insider group chat? They liked being able to bounce ideas off each other without things getting too crowded—plus they could juice their networks for a call-in show dynamic, even booking big-name guests over DM while live on camera.

In an early video for YouTube, they printed out a post by a popular anonymous X account, @NetCapGirl, and analyzed it on camera. She couldn’t resist retweeting their video, getting them free exposure to 100,000 new viewers. Onto the next viral account. Rinse, repeat. “Our first growth hack,” Coogan tells me.

2026: This is AGI by Pat Grady

The AI applications of 2023 and 2024 were talkers. Some were very sophisticated conversationalists! But their impact was limited.

The AI applications of 2026 and 2027 will be doers. They will feel like colleagues. Usage will go from a few times a day to all-day, every day, with multiple instances running in parallel. Users won’t save a few hours here and there – they’ll go from working as an IC to managing a team of agents.

Today, your agents can probably work reliably for ~30 minutes. But they’ll be able to perform a day’s worth of work very soon – and a century’s worth of work eventually.

What can you achieve when your plans are measured in centuries? A century is 200,000 clinical trials no one’s cross-referenced. A century is every customer support ticket ever filed, finally mined for signal. A century is the entire U.S. tax code, refactored for coherence.

The ambitious version of your roadmap just became the realistic one.

Gavin Baker Invest Like The Best Interview

A wide ranging interview covering the economics of AI, discussions of potential bear cases, the impact to adjacent tech industries and lots more. I’m always interested in hearing what Gavin has to say as he brings a really unique perspective to investing in tech across public and private markets.

On Google’s TPU Economics

A couple of things. So, one, for whatever reason, Google made more conservative design decisions. And part of that is, so, Google, let’s say, the TPU ... So, there’s front end and back end of semiconductor design, and then there’s dealing with Taiwan Semi.

And you can make an ASIC in a lot of ways. What Google does is they do mostly the front end for the TPU, and then Broadcom does the backend, and manages Taiwan Semi and everything. It’s a crude analogy, but the front end is, like, the architect of a house. They design a house. The back end is the person who builds the house. And then managing Taiwan Semi is stamping out that house like Lennar or D.R. Horton.

And for doing those two latter parts, Broadcom runs a 50% to 55% gross margin. We don’t know what on TPUs. Let’s say in 2027, TPU I think consensus estimates may be somewhere around $30 billion. Again, who knows? $30 billion I think is a reasonable estimate. 50% to 55% gross margins. So, Google is paying Broadcom $15 billion. That’s a lot of money.

At a certain point, it makes sense to bring a semiconductor program entirely in-house. So, in other words, Apple does not have an ASIC partner for their chips. They do the front end themselves, the back end, and they manage Taiwan Semi. And the reason is they don’t want to pay that 50% margin.

So, at a certain point, it becomes rational to re-negotiate this, and just as perspective, the entire opex of Broadcom’s semiconductor division is, round numbers, $5 billion. So, it would be economically rational now that Google is paying ... If it’s $30 billion, we’re paying them $15 billion. Google can go to every person who works in Broadcom semi, double their comp, and make an extra $5 billion. In 2028, let’s just say it does $50 billion. Now it’s $25 billion. You get triple their comp. And, by the way, you don’t need them all.

On Potential Bear Cases for AI

There’s one really obvious bear case and it is just edge AI, and it’s connected to the economic returns to ASI. In three years, on a bigger and bulkier phone to fit the amount of DRAM necessary, and the battery won’t probably last as long, you will be able to probably run a pruned-down version of something like Gemini 5 or Grok 4, Grok 4.1 or ChatGPT at 30, 60 tokens per second, and that’s free. This is clearly Apple’s strategy. It’s just, “We’re going to be a distributor of AI, and we’re going to make it privacy-safe and run on the phone. Then you can call one of the big models, the God models in the cloud, whenever you have a question.” If that happens, if 30, 60 tokens a second at a 115 IQ is good enough, I think that’s a bear case.
Other than just these scaling laws break. In terms of if we assume scaling laws continue, and we now know they’re going to continue for pre-training for at least one more generation ... and we’re very early in the two new scaling laws for post-training, mid-training, RLVR, whatever people want to call it, and then test-time computed inference ... we’re so early in those, and we’re getting so much better at helping the models hold more and more context in their minds as they do this test-time compute.

That’s really powerful, because everybody’s like, “Well, how’s the model going to know this?” Well, eventually, if you can hold enough context, you can just hold every Slack message and Outlook message and company manual in a company, in your context, and then you can compute the new task and compare it with your knowledge of the world, what you think, what the model thinks, all this context. It may be that just really, really long context windows are the solution to a lot of the current limitations. That’s enabled by all these cool tricks like KV cache offload and stuff. I do think, other than scaling laws slowing down, other than there being low economic returns to ASI, edge AI is to me by far the most plausible and scariest bear case.

On Power Being A Bottleneck

Having watts as a constraint is really good for the most advanced compute players, because if watts are the constraint, the price you pay for compute is irrelevant. The TCO of your compute is absolutely irrelevant because if you could get 3X or 4X or 5X more tokens per watt, that is literally three or 4X or 5X more revenue.

If you’re going to build an advanced data center costs 50 billion, a data center with the ASIC maybe costs 35 billion. If that $50 billion data center pumps out 25 billion of revenue and your ASIC data center at 35 billion is only pumping out eight billion, well, you’re pretty bumped. So I do think it’s good for all of the most advanced technologies in the data center, which is exciting to me as an investor.

So as long as power is a governor, the best products are going to win irrespective of price and have crazy pricing power. That’s the first implication that’s really important to me. Second, it is in the only solutions to this, we just can’t build nuclear fast enough in America. As much as we would love to build nuclear quickly, we just can’t. It’s just too hard. NEPA, all these rules, it’s too hard. Like a rare ant that we can move and it could be in a better environment, can totally delay the construction of a nuclear power plant, one ant. That is America today.

Humans need to come first. We need to have a human-centric view of the world. But the solutions are natural gas and solar. And the great thing about these AI data centers is apart from the ones that you’re going to do inference on, you can locate them anywhere.

So I think you were going to see, and this is why you’re seeing all this activity in Abilene because it’s in the middle of a big natural gas basin and we have a lot of natural gas in America because of fracking. We’re going to have a lot of natural gas for a long time. We can ramp production really fast. So I think this is going to be solved. You’re going to have power plants fed by gas or solar. And I think that’s the solution. And you’re already, all these turbine manufacturers were reluctant to expand capacity, but Caterpillar just said, “We’re going to increase capacity by 75% over the next few years.” So the system on the power side is beginning to respond.

Gavin also shared his thoughts on the Nvidia/Groq deal.

The Prison of Financial Mediocrity

A thought provoking piece on the current financial dynamics the young generations are facing. This piece succinctly captured what seems to be a super pervasive sentiment amongst millennials and gen z: anxiety about the level of uncertainty for how AI will shape the future combined with a widening wealth gap that leaves many feeling they can’t attain their “dream life” (which may be skewed because of social media) without taking on greater financial risks.

On the breaking of the old “deal” of work

The implicit deal used to be simple: show up, work hard, stay loyal, and you’ll be rewarded. Companies offered pensions. Tenure meant something. Your house appreciated while you slept. The system worked if you trusted it.

That deal is dead.

Staying at one company for 20 years is now a career liability, not an asset. Wages grew 8% while housing costs doubled and debt payments for young people increased ~33%. The math doesn’t support patience anymore.

Looking at the bigger picture, I used to think it was bad, but with the advent of AI and the economic impact they are going to have (even with current technologies), I think it is only going to get significantly worst.

On AI and Social Media Creating Anxiety

Previous generations had limited visibility into how others lived. You compared yourself to your neighbors, your coworkers, maybe some celebrities in magazines. The reference class was narrow. Now the reference class is infinite. A 25-year-old making $70k is constantly fed content from people their age making $2mn, living in Bali, “working” four hours a day. The baseline for “enough” keeps moving.

You never catch up. No matter what you achieve, social media will show you what you’re missing. The spread between your life and the life you “should” have is maintained algorithmically, forever uncollapsible.

So you have AI shrinking your timeline AND social media ensuring you never feel like you’ve arrived. The pressure to escape, NOW, FAST, before it’s too late, compounds daily.

The anxiety is pervasive. Every white-collar worker has done the mental exercise: “Could AI do my job? When?” And most of them don’t love their answers. Even if they think they’re safe for now, “for now” keeps getting shorter.

The Numbers and the Trades

Prediction Markets: Polymarket and Kalshi did $10B+ in volume in November 2025 alone. Combined annual volume is approaching $40B. In 2020 this was essentially zero. The growth rate is vertical.

Sports Betting: Legal sports betting revenue went from $248M in 2017 to $13.7B in 2024. Gen Z and Millennials account for 76% of betting activity. Activity on online sportsbooks rose 7% year-over-year for both cohorts.

TransUnion’s report identified these gamblers as “speculators”: urban renters, heavy users of crypto apps, concentrated in mobile trading platforms. Young people locked out of traditional wealth-building, seeking asymmetric payoffs in the only markets that offer them.

If this diagnosis is right, that a generation of economically locked-out young people will continue seeking agency through high-variance financial products, then you want to be long anything that serves that demand.

The platforms win regardless of whether the users win. You are looking for platforms that doesn’t care if you win your bet or if your prediction is right. You are looking for businesses that extract fees from activity, and activity is growing.

Dan Wang 2025 Letter

https://danwang.co/2025-letter/

A very thoughtful and wide ranging piece reflecting on the year, touching on topics such as culture in Silicon Valley, the future of AI, the arms race between The US and China and much more.

On AI Future and SV Culture

While critics of AI cite the spread of slop and rising power bills, AI’s architects are more focused on its potential to produce surging job losses. Anthropic chief Dario Amodei takes pains to point out that AI could push the unemployment rate to 20 percent by eviscerating white-collar work.

I wonder whether this message is helping to endear his product to the public.

The most-read essay from Silicon Valley this year was AI 2027. The five authors, who come from the AI safety world, outline a scenario in which superintelligence wakes up in 2027; a decade later, it decides to annihilate humanity with biological weapons. My favorite detail in the report is that humanity would persist in a genetically modified form, after the AI reconstructs creatures that are “to humans what corgis are to wolves.” It’s hard to know what to make of this document, because the authors keep tucking important context into footnotes, repeatedly saying they do not endorse a prediction. Six months after publication, they stated that their timelines were lengthening, but even at the start their median forecast for the arrival of superintelligence was later than 2027. Why they put that year in their title remains beyond me.

It’s easy for conversations in San Francisco to collapse into AI. At a party, someone told me that we no longer have to worry about the future of manufacturing. Why not? “Because AI will solve it for us.” At another, I heard someone say the same thing about climate change. One of the questions I receive most frequently anywhere is when Beijing intends to seize Taiwan. But only in San Francisco do people insist that Beijing wants Taiwan for its production of AI chips. In vain do I protest that there are historical and geopolitical reasons motivating the desire, that chip fabs cannot be violently seized, and anyway that Beijing has coveted Taiwan for approximately seven decades before people were talking about AI.

Silicon Valley’s views on AI made more sense to me after I learned the term “decisive strategic advantage.” It was first used by Nick Bostrom’s 2014 book Superintelligence, which defined it as a technology sufficient to achieve “complete world domination.” How might anyone gain a DSA? A superintelligence might develop cyber advantages that cripple the adversary’s command-and-control capabilities. Or the superintelligence could self-recursively improve such that the lab or state that controls it gains an insurmountable scientific advantage. Once an AI reaches a certain capability threshold, it might need only weeks or hours to evolve into a superintelligence.

And if an American lab builds it, it might help to lock in the dominance of another American century.

On China’s Market Competition

Probably the most underrated part of the Chinese system is the ferocity of market competition. It’s excusable not to see that, given that the party espouses so much Marxism. I would argue that China embodies both greater capitalist competition and greater capitalist excess than America does today. Part of the reason that China’s stock market trends sideways is that everyone’s profits are competed away. Big Tech might enjoy the monopolistic success smiled upon by Peter Thiel, coming almost to genteel agreements not to tread too hard upon each other’s business lines. Chinese firms have to fight it out in a rough-and-tumble environment, expanding all the time into each other’s core businesses, taking Jeff “your margin is my opportunity” Bezos with seriousness.

Third, western elites keep holding on to a distinction between “innovation,” which is mostly the remit of the west, and “scaling,” which they accept that China can do. I want to dissolve that distinction. Chinese workers innovate every day on the factory floor. By being the site of production, they have a keen sense of how to make technical improvements all the time. American scientists may be world leaders in dreaming up new ideas. But American manufacturers have been poor at building industries around these ideas. The history books point out that Bell Labs invented the first solar cell in 1957; today, the lab no longer exists while the solar industry moved to Germany and then to China. While Chinese universities have grown more capable at producing new ideas, it’s not clear that the American manufacturing base has grown stronger at commercializing new inventions.

On Geopolitics and Potential for Improvements

The United States isn’t so good at celebrating its history either. 2026 is the 250th anniversary of the country’s founding. Where are the monuments to exalt that history? Most of the planned celebrations look small bore. Why hasn’t the federal government built a technological specimen as sublime as the Golden Gate Bridge, the Hoover Dam, or the Apollo missions? Probably because planning for any project should have commenced 10, 20, or 30 years ago. No president would have gotten around to starting a project that has no chance of being completed in his term. Lack of action due to the expectation of long timelines is one of the sins of the lawyerly society.

But American problems seem more fixable to me than Chinese problems. That’s why I live here in the US. I made clear in my book that I am drawn to pluralism as well as a broader conception of human flourishing than one that could be delivered by the Communist Party. The United States still draws many of the most ambitious people in the world, few of whom want to move to China. Even now a significant number of Chinese would jump to emigrate to the US if they felt they could be welcomed. But this enduring American advantage should not excuse the US from patching up its deficiencies.

I learned recently that the Bay Area has 26 separate transit agencies; is it really a triumph of democracy to have so many unconsolidated efforts? I wonder whether we can accuse the California government of subverting the will of the people by making so little progress on its high-speed rail, which was approved by referendum in 2008; California rail authorities take more pride in creating jobs than doing the job. I am tempted to use the language from American foreign policy at home. Why talk about American credibility only in terms of combat? Why shouldn’t the failure to deliver on big projects, after spending so much money, constitute a more severe blow to the credibility of the American project? Is the state of the US defense industrial base really deterring adversaries?

I won’t belabor issues with American public works or manufacturing. I’ll suggest only that the US ought to be acting with greater curiosity on how to do better. It doesn’t have to become China; but it should better study China’s successes. There is a 21st century playbook for becoming an industrial power and China has written it. This playbook consists of infrastructure development, solicitation of foreign investment, industrial subsidies, and the creation of industrial ecosystems. I hope that the US will stop attributing all of China’s successes to stealing. If such a program would be sufficient for building a world-class industry, then American spooks should dedicate their formidable capabilities to extracting Chinese industrial secrets. The reality is that there is little to be learned from blueprints. By failing to recognize China’s real strengths — the industrial ecosystems pulsating with process knowledge — the US is only cheating itself.

Roundup

AI’s trillion-dollar opportunity: Context graphs

The Bitter Lessons: Thoughts on US-China Competition

A Guide to Claude Code 2.0

Modeling for Fundamental L/S Equity

Thanks for reading! Be back same time next week and please never hesitate to send me interesting things to read! My dms are open.

Michael Burry made waves this week with news that his hedge fund has terminated its SEC registration, possibly signaling that he is shutting it down, and for a tweet raising concerns that big tech companies may be understating depreciation expenses on GPUs by assuming longer useful lives for the machines. Burry argued that these changes will lead these companies to overstate earnings by $176B from 2026 through 2028.

SemiAnalysis highlighted the counterargument that the useful live of GPUs has actually increased.

Dr. Burry’s claim is predicated on an assumption that the NVIDIA product cycle is now 2-3 years, which is far lower than the useful life of the assets. We believe this is a fatal flaw in the argument. The new accounting, while beneficial to the companies in the short-term, is also predicated on real operational experience in datacenters.

Back in 2020, when Microsoft, Meta, and Google increased the useful life from 3 to 4 years, we were still in the year 2 BC (Before ChatGPT). Now, in present-day 3 AD (After Da Launch of ChatGPT), the increases in useful life have proved beneficial to CAPEX-hungry hyperscalers. What began changing in IT equipment in 2020 that has continued through to 2025? The answer is reliability, and incentives.

Server OEMs such as Dell, SuperMicro, HPE, Lenovo, and Cisco have long sold servers with a standard warranty between 3 to 5 years in length. The 5 year warranty is more expensive, of course, but many extended warranty options exist for 6 and 7 years. The price goes up of course, but all it takes is the vendor stocking enough spare parts to make service calls on worn out nodes. Meanwhile, networking equipment vendors such as Cisco, Arista, Aruba and Juniper have experimented with lifetime warranty on their switches. Storage vendors have offered the same - just pay a yearly support contract and they’ll keep swapping worn out drives. Think of it like a car: the high end of the market might lease and upgrade their Benz every 2 years, while others drive their 20 year older beaters for the price of gas and insurance.

QF Research also had thoughtful tweets about economic reality of H100s, rebutting Burry’s thesis.

As many people highlighted, another common rebuttal is the fact that A100s are still being rented, 5 years after they were purchased, though the going profitability remains a question. I also enjoyed this tweet about the different incentives that various parties have to depreciate GPUs over different timelines.

What does this all mean?

As we highlighted in a previous piece, a few important points to note:

• Big tech is in the middle of becoming an asset heavy sector. The major platforms will likely spend close to $400B of capex in 2025, and that figure is expected to keep rising through 2026 and 2027. Roughly half of that spend goes directly into servers, chips, and accelerators. These companies are no longer software businesses with light capital footprints. They are infrastructure companies building global compute utilities.

• This shift matters because of the concentration of the index. The Mag7, with Broadcom substituted for Tesla, now account for roughly 36% of the S&P 500. At about 30x 2026 earnings, Wall St. expects the group to compound earnings at about a 20% CAGR over the next two years.

• Burry’s point follows from that setup. If hyperscalers are assuming longer useful lives for GPUs than the economic reality justifies, then depreciation expenses are understated and earnings are overstated. Since compute hardware accounts for 50-60% of data center capex, even small changes in useful life assumptions can move earnings by tens of billions. If the group that represents over a third of the index reports lower earnings than projected, or even if the market begins to believe the earnings are overstated to a meaningful degree, it could pull down the entire market.

• The counter argument is equally straightforward. The useful life of modern accelerators actually appears to be increasing. GPUs do not drop to zero after two or three years. Demand for older models like the A100 remains strong. Hyperscalers can reuse them for fine-tuning, inference, embeddings, internal workloads, and third-party cloud capacity. If the industry continues to see reasonable payback periods and positive ROIC in years 3–5, the accounting treatment becomes far less important. The real economics are determined by cash returns, not depreciation schedules, so consistent positive ROIC would matter more than whether the useful life is listed as four years or six.

From a financial perspective the most important question is the incremental ROIC that hyperscalers earn on each new dollar of AI capex. The group is on track to deploy more than $400B of AI-related capital expenditures in 2025. ROIC will appear in operating profit over time, but operating profit will be affected by the opex required to run these systems, including energy, networking, staffing, and data center operations. Depreciation can influence reported earnings, but the cash economics flow through free cash flow. This is where replacement cadence becomes relevant. If hyperscalers eventually move toward an eighteen month purchase cycle for new accelerators, the higher frequency of capex will show up immediately in free cash flow regardless of the assumed useful life. The hardware does not need to pay back before the next refresh, but it must reach economic breakeven over its productive life. The hurdle is set by the cash flows the hardware produces, not by the accounting schedule that allocates its cost.

The second factor is the relationship between replacement cadence and capacity expansion. These two variables reinforce each other. If total compute capacity can continue to scale through both hyperscalers and neocloud providers, older GPUs can remain productive even as newer hardware comes online. Many of these older accelerators will have already achieved economic payback. They can be redeployed for fine tuning, inference, embeddings, smaller models, and internal workloads. In that scenario a faster replacement cadence does not imply a shorter useful life. The newest chips expand the compute base and older chips continue to generate positive cash flow at lower margins. The constraint is not only the hardware but also the availability of power, cooling, land, and networking. If those constraints bind, capacity expansion slows and new chips begin to crowd out older ones. Economic life then compresses even if the accounting life remains long.

The third question is where value accrues within the stack. At present most value creation sits with the semiconductor supply chain, particularly NVIDIA, TSMC, and the upstream and downstream partners that support them. The hyperscalers capture value next, but only after deploying significant capital and operating large physical footprints. Consumers enjoy a large surplus because products like ChatGPT, Claude, Cursor, and Gemini offer significant capability at relatively low prices. The application layer sits in a more uncertain position. The industry has not yet demonstrated how much of the value will remain with software companies once the cost of infrastructure is considered. At this stage the hyperscalers remain price takers in accelerator supply, and the application layer has not established durable high-margin economics relative to the scale of investment being made beneath it. JP Morgan estimates that to return 10% on AI investments through 2030, there will need to be $650B in revenue in perpetuity, the equivalent of $34.72 per iPhone user.

SemiAnalysis recently published a detailed cost breakdown for H200 clusters that helps frame these questions. The total capital cost per eight-GPU server is roughly $246,000, which works out to about $1.06 per GPU per hour before operating expenses. Once electricity, colocation, networking, and on-site operations are included, the total cost of ownership rises to about $1.41 per GPU per hour. These figures highlight why incremental ROIC and replacement cadence matter far more than depreciation policy. If the hardware produces enough cash flow over its productive life to cover this all-in cost, the investment case holds. If not, the underlying economics will appear in free cash flow regardless of the useful life assumptions in financial statements. A faster refresh cycle would increase the capital cost per GPU hour unless total capacity grows quickly enough to keep older hardware productive at lower margins.

This context explains why Burry’s comment stirred quite a bit of discussion. His tweet brought about a useful debate about the assumptions that sit underneath the largest capital investment cycle in modern technology. Useful life affects reported earnings, but the long term economics will be shaped by incremental ROIC, replacement cadence, capacity constraints, and the distribution of value across the stack. These variables will determine the actual return profile of AI infrastructure investment over the next decade.

Satya Nadella Interview - How Microsoft Thinks about AGI

Dwarkesh Patel and Dylan Patel sat down for a very illuminating interview with Satya Nadella where they discussed Microsoft’s new data centers, the business of AI, their partnership with OpenAI and more. I always enjoy everything that Dwarkesh and Dylan put out and this interview was both timely and informative from one of the most central figures in the AI arms race.

On the new Fairwater 2 data center:

Scott Guthrie

We’ve tried to 10x the training capacity every 18 to 24 months. So this would effectively be a 10x increase from what GPT-5 was trained with. So to put it in perspective in the number of optics, the network optics in this building is almost as much as all of Azure across all our data centers two and a half years ago.

Satya Nadella

It’s got like five million network connections.

Dwarkesh Patel

You’ve got all this bandwidth between different sites in a region and between the two regions. So is this like a big bet on scaling in the future, that you anticipate in the future that there’s going to be some huge model that will require two whole different regions to train?

Satya Nadella

The goal is to be able to aggregate these flops for a large training job and then put these things together across sites. The reality is you’ll use it for training and then you’ll use it for data gen, you’ll use it for inference in all sorts of ways. It’s not like it’s going to be used only for one workload forever.

On AGI:

Dylan Patel

When you look at all the past technological transitions—whether it be railroads or the Internet or replaceable parts, industrialization, the cloud, all of these things—each revolution has gotten much faster in the time it goes from technology discovered to ramp and pervasiveness through the economy. Many folks who have been on Dwarkesh’s podcast believe this is the final technological revolution or transition, and that this time is very, very different.

At least so far in the markets, in three years we’ve already skyrocketed to hyperscalers doing $500 billion of capex next year, which is a scale that’s unmatched to prior revolutions in terms of speed. The end state seems to be quite different. Your framing of this seems quite different from what I would call the “AI bro” who’s like, “AGI is coming.” I’d like to understand that more.

Satya Nadella

I start with the excitement that I also feel for the idea that maybe after the Industrial Revolution this is the biggest thing. I start with that premise. But at the same time, I’m a little grounded in the fact that this is still early innings. We’ve built some very useful things, we’re seeing some great properties, these scaling laws seem to be working. I’m optimistic that they’ll continue to work. Some of it does require real science breakthroughs, but it’s also a lot of engineering and what have you.

That said, I also sort of take the view that even what has been happening in the last 70 years of computing has also been a march that has helped us move. I like one of the things that Raj Reddy has as a metaphor for what AI is. He’s a Turing Award winner at CMU. He had this, even pre-AGI. He had this metaphor for AI, it should either be a guardian angel or a cognitive amplifier. I love that. It’s a simple way to think about what this is. Ultimately, what is its human utility? It is going to be a cognitive amplifier and a guardian angel. If I view it that way, I view it as a tool.

But then you can also go very mystical about it and say this is more than a tool. It does all these things, which only humans did before so far. But that has been the case with many technologies in the past. Only humans did a lot of things, and then we had tools that did them.

On where value will accrue:

Dwarkesh Patel 00:20:02

I guess the reason to focus on this question is that it’s not just about GitHub, but fundamentally about Office and all the other software that Microsoft offers. One vision you could have about how AI proceeds is that the models are going to keep being hobbled and you’ll need this direct visible observability all the time.

Another vision is that over time these models which are now doing tasks that take two minutes, in the future, they’ll be doing tasks that take 10, 30 minutes. In the future, maybe they’re doing days worth of work autonomously. Then the model companies are charging thousands of dollars maybe for access to, really, a coworker which could use any UI to communicate with their human and migrate between platforms.

If we’re getting closer to that, why aren’t the model companies that are just getting more and more profitable, the ones that are taking all the margin? Why is the place where the scaffolding happens, which becomes less and less relevant as the AI becomes more capable, going to be that important? That goes to Office as it exists now versus coworkers that are just doing knowledge work.

Satya Nadella 00:21:07

That’s a great point. Does all the value migrate just to the model? Or does it get split between the scaffolding and the model? I think that time will tell. But my fundamental point also is that the incentive structure gets clear. Let’s take information work, or take even coding. Already in fact, one of my favorite settings in GitHub Copilot is called auto, which will just optimize. In fact I buy a subscription and the auto one will start picking and optimizing for what I am asking it to do. It could even be fully autonomous. It could arbitrage the tokens available across multiple models to go get a task done.

If you take that argument, the commodity there will be models. Especially with open source models, you can pick a checkpoint and you can take a bunch of your data and you’re seeing it. I think all of us will start, whether it’s from Cursor or from Microsoft, seeing some in-house models even. And then you’ll offload most of your tasks to it.

So one argument is if you win the scaffolding—which today is dealing with all the hobbling problems or the jaggedness of these intelligence problems, which you kind of have to—if you win that, then you will vertically integrate yourself into the model just because you will have the liquidity of the data and what have you. There are enough and more checkpoints that are going to be available. That’s the other thing.

Structurally, I think there will always be an open source model that will be fairly capable in the world that you could then use, as long as you have something that you can use that with, which is data and a scaffolding. I can make the argument that if you’re a model company, you may have a winner’s curse. You may have done all the hard work, done unbelievable innovation, except it’s one copy away from that being commoditized. Then the person who has the data for grounding and context engineering, and the liquidity of data can then go take that checkpoint and train it. So I think the argument can be made both ways.

On Azure:

By the way, the other thing is that I didn’t want to get stuck with massive scale of one generation. We just saw the GB200s, the GB300s are coming. By the time I get to Vera Rubin, Vera Rubin Ultra, the data center is going to look very different because the power per rack, power per row, is going to be so different. The cooling requirements are going to be so different. That means I don’t want to just go build out a whole number of gigawatts that are only for a one-generation, one family. So I think the pacing matters, the fungibility and the location matters, the workload diversity matters, customer diversity matters and that’s what we’re building towards.

The other thing that we’ve learned a lot is that every AI workload does require not only the AI accelerator, but it requires a whole lot of other things. In fact, a lot of the margin structure for us will be in those other things. Therefore, we want to build out Azure as being fantastic for the long tail of the workloads, because that’s the hyperscale business, while knowing that we’ve got to be super competitive starting with the bare-metal for the highest end training.

On Chips and OpenAI Partnership:

Dwarkesh Patel

You mentioned how this depreciating asset, in five or six years, is 75% of the TCO of a data center. And Jensen is taking a 75% margin on that. So what all the hyperscalers are trying to do is develop their own accelerator so that they can reduce this overwhelming cost for equipment, to increase their margins.

Dylan Patel

And when you look at where they are, Google’s way ahead of everyone else. They’ve been doing it for the longest.

They’re going to make something like five to seven million chips of their own TPUs. You look at Amazon and they’re trying to make three to five million[Lifetime shipment units]. But when we look at what Microsoft is ordering of their own chips, it’s way below that number. You’ve had a program for just as long. What’s going on with your internal chips?

Satya Nadella

It’s a good question. A couple of things. One is that the thing that is the biggest competitor for any new accelerator is kind of even the previous generation of Nvidia. In a fleet, what I’m going to look at is the overall TCO. The bar I have, even for our own… By the way, I was just looking at the data for Maia 200 which looks great, except that one of the things that we learned even on the compute side… We had a lot of Intel, then we introduced AMD, and then we introduced Cobalt. That’s how we scaled it. We have good existence proof of, at least in core compute, how to build your own silicon and then manage a fleet where all three are at play in some balance.

Because by the way, even Google’s buying Nvidia, and so is Amazon. It makes sense because Nvidia is innovating and it’s the general-purpose thing. All models run on it and customer demand is there. Because if you build your own vertical thing, you better have your own model, which is either going to use it for training or inference, and you have to generate your own demand for it or subsidize the demand for it. So therefore you want to make sure you scale it appropriately.

The way we are going to do it is to have a close loop between our own MAI models and our silicon, because I feel like that’s what gives you the birthright to do your own silicon, where you literally have designed the microarchitecture with what you’re doing, and then you keep pace with your own models. In our case, the good news here is that OpenAI has a program which we have access to. So therefore to think that Microsoft is not going to have something that’s—

Dylan Patel

What level of access do you have to that?

Satya Nadella

All of it.

Dylan Patel

You just get the IP for all of that? So the only IP you don’t have is consumer hardware?

Satya Nadella

That’s it.

The entire interview is definitely worth watching. Probably the most illuminating picture into the mind of someone whose job it is to try to position his company as the winner of the AI arms race.

CoreWeave Earnings

CoreWeave remains one of the most interesting companies in the AI ecosystem because it represents a new category of infrastructure provider. It is not a hyperscaler and not a traditional cloud platform. It is a neocloud. The company was built for the AI age with a model centered on GPU utilization rather than general-purpose compute, and its scale-up has occurred entirely during the peak of the AI arms race. In March 2025 CoreWeave raised $1.5B in its IPO at a valuation of roughly $23B. The stock is up about 90% since then, valuing the company at about $37B. The company has also relied heavily on debt financing and has accumulated more than $19B of total debt plus operating leases as of Q3. In July CoreWeave closed an additional $2.6B secured facility, part of more than $25B in combined equity and debt raised in less than two years.

The Q3 results highlight both the growth and the uncertainty of the neocloud model. Revenue for the quarter reached $1.36B, up from $584M a year ago. For the first nine months of 2025, revenue totaled $3.56B compared with $1.17B in the same period of 2024. The company is scaling quickly, but the cost structure remains complex. Cost of revenue was $369M in Q3 and $944M for the year to date. Technology and infrastructure expense was $747M in Q3 and $1.98B for the first nine months. These lines include power, networking, bandwidth, operational labor, and hardware depreciation that is embedded across several categories instead of appearing as a separate line item. As a result, it is difficult to determine the true economic margin of the business from the income statement alone. CoreWeave reported an operating loss of $237M in Q3 and a net loss of $110M, which contributed to a year-to-date net loss of $715M.

The balance sheet gives a clearer view of the scale of investment. Total assets increased from $17.83B at year-end 2024 to $32.91B at the end of Q3 2025. Property and equipment grew from $11.91B to $20.66B in the same period. The company is adding nearly $9B of hardware and data center assets in only nine months. Total liabilities also expanded to $29.03B, reflecting a growing mix of current and long-term obligations. CoreWeave’s capital structure is therefore highly leveraged, and the long-term return on these invested dollars has not yet been demonstrated.

This combination of rapid growth, heavy capital spending, and an untested margin structure makes CoreWeave an important company to monitor. It is the closest real-time example of what a pure AI infrastructure provider looks like in practice. The sustainability of the model will depend on several factors. Utilization must remain high. Replacement cycles must align with the payback period of the hardware. Capacity must grow fast enough to keep older GPUs productive at lower margins. Operating scale must eventually lead to efficiency improvements. If these conditions hold, the neocloud model could mature into a viable alternative to traditional hyperscalers. If they do not hold, the economics may ultimately favor only the largest platforms with the power, land, and balance sheet required to support multi-billion-dollar annual capex cycles.

Either way, CoreWeave offers a clear window into how AI infrastructure economics may evolve. Their earnings will be an extremely useful proxy into GPU pricing.

On their earnings call, they provided useful insights into the state of AI:

In Q3, we beat expectations, delivering revenue of $1.4 billion, up 134% year-over-year. We added over $25 billion in revenue backlog in the third quarter alone, bringing us to over $55 billion in revenue backlog to end Q3, almost double Q2 and approaching 4 times year to date. Further, CoreWeave has reached $50 billion in RPO faster than any cloud in history. These results demonstrate the deep confidence customers have in CoreWeave, the company they trust as their essential cloud for artificial intelligence.

We continue to scale aggressively even as the industry remains capacity constrained. We expanded our active power footprint by 120 megawatts sequentially to approximately 590 megawatts, while growing our contracted power capacity over 600 megawatts to 2.9 gigawatts. This leaves us well-positioned for future growth, with more than 1 gigawatt of contracted capacity available to be sold to customers that we expect to largely come online within the next 12 to 24 months.

For context, I found this tweet helpful in understanding the costs involved in building a 1 GW AI data center

CoreWeave also noted their deployment of GB200 and GB300s as well as customers renewing H100 rentals within 5% of the original price:

In the third quarter, we continued to deliver many of the initial scaled deployments of the GB200s, while once again being first to market, this time with the GB300s, further highlighting our incredible track record of operational excellence.

CoreWeave’s industry leadership is unmatched. We are the only cloud provider to submit MLPerf inference result for GB300s, setting the benchmark for real-world AI performance. And just last week, SemiAnalysis once again recognized our dominance, awarding CoreWeave its highest possible distinction, its Platinum ClusterMAX ranking, for the second time ahead of more than 200 providers, including the hyperscalers and emerging neoclouds. No other cloud has achieved this once. CoreWeave has done it twice, underscoring yet again that CoreWeave stands alone at the forefront of the AI cloud.

Demand for AI cloud technology remains robust across generations of GPUs. For example, in Q3, we saw our first 10,000-plus H100 contract approaching expiration. Two quarters in advance, the customer proactively re-contracted for the infrastructure at a price within 5% of the original agreement. This is a powerful indicator of customer satisfaction, as well as the long-term utility and differentiated value of the GPUs run on CoreWeave’s platform.

CoreWeave uses a 6 year useful life for their GPU depreciation. Customers re-upping older models at close to the original price lends credibility to the idea that GPUs have a longer useful life than 2-3 years but this is also during a supply shortage, so the economics when demand and supply are more balanced and the industry is more mature remain to be seen. This is area to monitor closely and one we will be actively covering.

Also worth mentioning is that in their AI Capex Report, JP Morgan predicted $1.5T of funding for AI data centers from the high grade bond markets over the next 5 years.

High Grade capital markets will look very different. We think High Grade markets could absorb $300 billion of AI/data center related paper over the next year, and are assuming $1.5 trillion of funding from High Grade bond markets over the next five years. AI/Data center capex-related sectors already represent 14.5% of the JULI Index, which is larger than the US Banks component of the index. Mathematically, if anything like our forecast plays out, AI/Data Center related sectors could represent north of 20% of the market by 2030. Historically, lumpy issuance from select sectors is not “new” to High Grade. Recent examples like Healthcare (‘21-’24) and Telecom (‘16-’19) have been manageable and in both cases led to 15-20bp of underperformance from the impacted sector.

So there will be more debt involved in the financing of AI data centers and it remains to be seen how valuable the equity in a company with a model like CoreWeave’s will be.

Warren Buffett’s Final Shareholder Letter

Warren Buffett wrote his final letter to Berkshire’s shareholders on November 10, 2025. As always, it is filled with timeless wisdom and insights from an incredible lifetime. The collection of his shareholder letters is an amazing book and one that I would highly recommend for anyone interested in investing.

Looking Ahead and Roundup

This week all eyes will be on NVIDIA’s Q3 earnings and Alphabet’s rumored release of Gemini 3. It seems like the news cycle will continue to be dominated by all things AI for the foreseeable future.

This week saw Cursor raise $2.3B at a $29.3B valuation.

Anthropic announced a $50B data center investment and that it disrupted the first reported AI-orchestrated cyber espionage campaign.

Sundar Pichai posted a cryptic tweet that seemed to imply Gemini 3 will be released this week.

Bill Ackman offered some dating advice and quickly sparked a new meme.

Wall St. expects $54B in revenue from NVIDIA this week, with $48B coming from data centers as production ramps up for their Blackwell systems.

Citrini had a great tweet about reflexivity as it pertains to CDS spreads widening for CoreWeave and Oracle.

As always, thanks for reading and we’ll see you next week.

All eyes were on OpenAI CFO Sarah Friar this week as she spoke about the US Government potentially becoming more involved in funding AI development. At the Wall Street Journal’s Tech Live Conference she said the following that then made waves around the internet.

She would later follow up with a LinkedIn post saying:

I want to clarify my comments earlier today. OpenAI is not seeking a government backstop for our infrastructure commitments. I used the word “backstop” and it muddied the point. As the full clip of my answer shows, I was making the point that American strength in technology will come from building real industrial capacity which requires the private sector and government playing their part. As I said, the US government has been incredibly forward-leaning and has really understood that AI is a national strategic asset.

Sam Altman later followed up in a tweet saying:

I would like to clarify a few things.

First, the obvious one: we do not have or want government guarantees for OpenAI datacenters. We believe that governments should not pick winners or losers, and that taxpayers should not bail out companies that make bad business decisions or otherwise lose in the market. If one company fails, other companies will do good work.

What we do think might make sense is governments building (and owning) their own AI infrastructure, but then the upside of that should flow to the government as well. We can imagine a world where governments decide to offtake a lot of computing power and get to decide how to use it, and it may make sense to provide lower cost of capital to do so. Building a strategic national reserve of computing power makes a lot of sense. But this should be for the government’s benefit, not the benefit of private companies.

The one area where we have discussed loan guarantees is as part of supporting the buildout of semiconductor fabs in the US, where we and other companies have responded to the government’s call and where we would be happy to help (though we did not formally apply). The basic idea there has been ensuring that the sourcing of the chip supply chain is as American as possible in order to bring jobs and industrialization back to the US, and to enhance the strategic position of the US with an independent supply chain, for the benefit of all American companies. This is of course different from governments guaranteeing private-benefit datacenter buildouts.

Which was followed by the post receiving a community note with a quote from OpenAI’s letter to the government:

“To provide manufacturers with the certainty and capital they need to scale production quickly, the federal government should also deploy grants, cost-sharing agreements, loans, or loan guarantees to expand industrial base capacity and resilience.”

A thoughtful tweet from Dean Ball shed some helpful light on the situation:

4. In an, again, entirely separate public interest comment submitted to the White House (downstream of a request for information that, incidentally, I drafted while I was in government) late last month, OpenAI discussed broadly the notion of reducing the cost of capital for manufacturers in the AI data center supply chain. We already do this for semiconductor manufacturing through the CHIPS Act.

5. Lowering the cost of capital for manufacturers of strategic goods is not at all a “loan guarantee.” Consider natural gas turbines. That industry has gone through brutal boom and bust cycles in recent decades. If you run a natural gas turbine manufacturer, or are a long-term investor in one, or loan money to such firms, you are going to be weary of too much expansion for fear that the AI bubble will pop. This slows down supply expansion for a good that we really do need to power AI in the near term. So what do you do?

6. Well, one thing you could do is have the federal government serve as buyer of last resort of future turbines. You write a contract that says “if the manufacturer makes X turbines over the next five years, the federal will pay Y price for Z number of turbines if no other private-sector buyer emerges at or above price Y.” That way, the manufacturer can go to its investors and lenders and say, “don’t worry, we’ve got a buyer for turbines if we expand.” And perhaps the lender is willing to offer the manufacturer a lower rate of interest--a lower cost of capital. I myself advocated for precisely this policy when I worked for the Trump Administration (though it didn’t make it into the AI Action Plan, sadly). There are many, similar schemes one could imagine.

7. This idea involves the government taking limited, pre-defined risk. The political economy problems with this are non-zero, but they are far smaller than the regulatory capture that would ensue from the US government guaranteeing untold billions of OpenAI debt.

8. As I read OpenAI’s public interest comment, I interpret them to have been talking much more about the kind of thing I describe in item (6) rather than the loan guarantee for OpenAI debt. They are referring them to manufacturer cost of capital in that comment; I don’t think OpenAI refers to itself as a “manufacturer.”

What to make of all of this? For one, at a meta level, OpenAI’s dominance of the finance and tech press surrounding AI has seemed to serve to make the company synonymous with AI in the mind of the public. One word from their CFO created global headlines.

But at a deeper level I think the story reflects something more important and that is that AI is now unquestionably a national strategic industry and inevitably, the US government will play an increasingly important role in the AI supply chain.

At a cursory glance, I would tend to agree with the sentiment that their CFO misspoke and the term “backstop” was the wrong word for the suggestion that the government was going to become increasingly involved in the AI arms race. As Dean points out, industrial strategy at a national level to lower manufacturer’s cost of capital would make sense for one since AI is becoming increasingly a national security issue, and also because the downstream effects, such as more manufacturing, more electricity producing assets, a better power grid etc., will hopefully help the American people in different ways for the foreseeable future. The cynical read is that of course regulatory capture is a potential strategy for the various companies involved at the frontiers of AI and the situation is certainly one to monitor.

On another note, Friar said that there are no near term plans for OpenAI to IPO. One could argue that more valuable startups going public would be a net positive for Americans who have a portion of their wealth invested in domestic stock indices. But with their aggressive spending plans, they seem to be waiting until they reach a more mature stage financially, with revenue growth and margins that would satisfy investors in a public offering (though there is little doubt that OpenAI would be one of the biggest IPOs of all time.)

All of this speaks to the scale of the capital required to build the future. But another part of the picture is how consumers are behaving right now, and earnings this quarter paint a more complicated story

Consumer Earnings

Every earnings season is a lens through which investors attempt to understand the broader macro environment. While tech earnings have focused attention on future capex, model training cycles, and the AI infrastructure race, the consumer side of the economy tells a different story. Investors are looking closely at restaurant chains, delivery platforms and consumer brands to diagnose the underlying health of household spending.

For instance, many are calling “peak slop bowl” as Chipotle, Cava and Sweetgreen all disappointed investors and saw their stocks drop.

Chipotle reported third quarter revenue of about $3.0B, up 7.5% year over year, but same store sales increased only 0.3% versus expectations of roughly 0.7%. Restaurant level operating margin fell to 15.9% from 16.9% a year ago. The stock dropped as much as 19% following the release.

Cava reported revenue of $289.8M, up 20% year over year, and same restaurant sales growth of 1.9%. Even with earnings of $0.12 per share versus estimates of roughly $0.11, the stock declined about 2% after the call. Investors focused on slowing traffic and cautious guidance.

Sweetgreen reported revenue of about $172M, down 0.6% year over year. Same store sales declined 9.5% and traffic fell 11.7%. Full year guidance was cut to a revenue range of $682M to $688M with a same store sales decline of 7.7% to 8.5%. The stock declined around 8% in response.

Door Dash fell 17% following its third quarter report. Revenue grew 27% year over year to $3.4B and total orders grew 21%, but EPS came in at about $0.55 versus estimates of roughly $0.68. Rising costs, higher planned spending in 2026 and questions about unit economics drove the negative reaction.

Celsius produced one of the most dramatic moves. Revenue of $725M represented growth of more than 170% year over year, and non GAAP EPS came in at $0.42. Despite the strong headline numbers, the stock fell more than 20% due to concerns about distribution channel changes and near term margin risk.

McDonald’s reported third quarter revenue of $7.08B, slightly below estimates, and EPS of about $3.18 compared to expectations of $3.33 to $3.35. Comparable sales grew 3.6% globally and 2.4% in the United States. The company highlighted persistent weakness among lower income consumers and a widening gap between high income and low income spending patterns.

e.l.f. Beauty reported quarterly revenue of $343.9M, up 14.2% year over year, but short of analyst expectations of roughly $367.9M. EPS came in at $0.68 versus estimates near $0.57. Despite the EPS beat, forward guidance disappointed and the stock fell more than 20% in after hours trading.

Toast reported revenue of $1.55B, up 24.8% year over year, and net income of $0.13 per share. The company continues to grow at a strong rate, but results relative to expectations have been mixed across recent quarters. While revenue growth was solid, the stock move was limited as investors looked for clearer signs of sustained operating leverage.

Uber reported solid operating results with revenue of $13.5B, up 20% year over year, and adjusted EBITDA growth of more than 30%. Monthly active consumers grew 17%. Even with strong reported numbers, the stock reaction was muted because gross bookings guidance for the next quarter came in toward the low end of expectations.

Robinhood was one of the few bright spots. The company reported revenue of $1.27B versus expectations of about $1.19B and EPS of $0.61 versus estimates of $0.51. The stock rose about 4% after hours.

You can also see the market wrestling with two very different interpretations of these results. One argument is that fast food, especially fast casual, expanded too quickly and pushed prices beyond what consumers were willing to pay. As BuccoCapital (a great follow) pointed out, brands have been trying to raise prices while simultaneously becoming more commoditized. Toast reported higher gross payment volume per location, which suggests consumers are still eating out but are becoming far more selective.

The other argument focuses on valuations. As Wasteland Capital (another great follow) pointed out many of these restaurant stocks were trading at forward price to earnings multiples of thirty five to one hundred times. They were priced at a premium even to the fastest growth names in tech and artificial intelligence. The recent declines may have less to do with a sudden deterioration in the consumer and more to do with a bubble deflating.

You can also see this pressure reflected in consumer sentiment data. Recent University of Michigan numbers show that sentiment among Americans ages eighteen to thirty four is now significantly more negative than the general population. Historically, younger people are more optimistic, but over the past few years the relationship has flipped. As Joe Weisenthal (yet another amazing follow) noted in September, in thirty years of data we have never seen such a sustained period of relative gloom among the young.

Broader sentiment is weakening as well. New University of Michigan survey data shows that about seventy percent of respondents have a poor view of current United States government economic policy. Overall consumer sentiment has now fallen to its lowest level in more than three years. This is consistent with what companies are reporting in their earnings calls. Consumers are still spending, but they are becoming more selective, more price sensitive and more skeptical of economic conditions

As Bloomberg reported in September, the top 10% of earners account for half of all spending.

This lends credibility to the idea that we’re in a “K-shaped” economy. Per Lisa Shalett at Morgan Stanley:

“Decoding this conundrum may hinge on the so-called K-shaped economy,” she wrote, “a concept that captures the widening chasm between the ‘haves’ and ‘have-nots.’”

Then Shalett said the situation is actually even worse than what Zandi produced: “For the U.S. consumer, wealth concentration has produced a situation where the top 40% of households by income account for approximately 60% of all spending; those households, in turn, control nearly 85% of America’s wealth, two-thirds of which is directly tied to the stock market, which has climbed more than 90% in three years.” She calculated that spending by the wealthiest households was growing 6x-7x faster than for the lowest cohort.

Even Federal Reserve chair Jerome Powell talked about seeing the pattern at October’s Federal Open Market Committee meeting. “If you listen to the earnings calls or the reports of big, public, consumer-facing companies, many of them are saying that there’s a bifurcated economy there and that consumers at the lower end are struggling and buying less and shifting to lower cost products,” Powell said last week. “But at the top, people are spending, at the higher income and wealth.”

The idea of a K shaped economy has many interesting implications for markets moving forward but beyond just markets, one wonders what this may mean for culture and politics. These points feel particularly relevant when viewed alongside the current macro data.

Peter Thiel Interview

Peter Thiel gave a thoughtful interview with The Free Press about how he sees the current political environment. He highlighted student debt and housing as two major drivers of why so many young people are becoming disillusioned with capitalism. Points that definitely feel relevant to understand against the current backdrop of macro data.

If you graduated in 1970 with no student debt, compare that to the millennial experience: too many people go to college, they don’t learn anything, and they end up with incredibly burdensome debt. Student debt is a version of this generational conflict that I’ve talked about for a long time.

The rupture of the generational compact isn’t limited to student debt, either. I think you can reduce 80 percent of culture wars to questions of economics—like a libertarian or a Marxist would—and then you can reduce maybe 80 percent of economic questions to questions of real estate.

He had written the below email in 2020, urging the leadership team at Facebook to take a serious look at what was driving a growing interest in socialism.

There are many themes that could be developed more here; let me make a few quick points for now:

Nick -- I certainly would not suggest that our policy should be to embrace Millennial attitudes unreflectively. I would be the last person to advocate for socialism. But when 70% of Millennials say they are pro-socialist, we need to do better than simply dismiss them by saying that they are stupid or entitled or brainwashed; we should try and understand why. And, from the perspective of a broken generational compact, there seems to be a pretty straightforward answer to me, namely, that when one has too much student debt or if housing is too unaffordable, then one will have negative capital for a long time and/or find it very hard to start accumulating capital in the form of real estate; and if one has no stake in the capitalist system, then one may well turn against it.

When asked where to look for answers, in typical Thiel fashion he suggests the solutions may lay outside the Overton Window

So the idea is: Maybe we should look for solutions outside the Overton Window. And that includes some very left-wing economics, socialist-type stuff. I don’t think those ideas will ultimately work, but they’re more than whatever was on offer. Cuomo did not have a plan for housing. He didn’t even think it was an issue. And of course, he’s been in politics and government for many, many years, so it’s hard not to ask: Why is he going to do something now, if he hasn’t done anything before? So, I’m not optimistic about Mamdani, but if you score it relatively, this is the kind of thing that’s going to happen if you look outside the Overton Window for solutions.

It is always interesting to hear what Peter Thiel has to say, and his comments feel especially relevant to the sociopolitical climate we find ourselves in today. With pessimism toward capitalism rising, gambling surging in popularity, and this weekend’s conversations about stimulus checks and fifty year mortgages, it feels useful to understand the cultural and economic forces shaping this moment. There are extrapolations to be made about future economic and market trends, but even beyond that, it helps to have a clearer sense of what people are responding to in a period that can feel increasingly tense and polarized.

Roundup

There was a lot happening this week. Here are a few more pieces that I found interesting

A fascinating look at what it’s like inside Cursor by Brie Wolfson. Another brilliant piece from Colossus.

Compelling Mission + Hardcore Technical Problems + Winning + Excellent Recruiting = Off-The-Charts Talent Density

Normally, the best talent doesn’t readily accrue to a company so early in its life. But because Cursor has all the magic ingredients, it could hire exceptional people from the outset.

On the product and engineering side, Cursor is building at the intersection of the most interesting challenges in UX and machine learning. (The work on Cursor 2.0, including its own custom model and a new UI dedicated to agent work flows, is a recent testament). On the go-to-market side, Cursor is one of the fastest growing companies of all time from a revenue perspective—it went from $0 to $100mn ARR without a sales team, and the one that’s since been installed is determined to add another zero before the end of 2025. The #closed-won channel, in which a Slack bot alerts the company to newly closed sales victories, is a near-constant stream of notifications.

Tesla shareholders approved Elon Musk’s almost $1T pay package

The first tranche of stock gets paid out if Tesla hits a market capitalization of $2 trillion. Tesla’s current market cap is $1.54 trillion. Awards tied to market cap gains are paired with operational achievements.

The next nine tranches would be awarded if Tesla’s value increases by increments of $500 billion, up to $6.5 trillion. Musk would earn the last two tranches if the market cap rises by increments of $1 trillion, meaning it would need to hit $8.5 trillion for Musk to get the full package.

Tesla also laid out a series of earnings milestones for Musk, beginning with $50 billion in annual adjusted profit and moving up to $400 billion. In the third quarter, Tesla reported adjusted EBITDA of $4.2 billion.

Other goals tied to the newly approved pay plan include reaching 20 million vehicle deliveries, 10 million active FSD (Full Self-Driving) subscriptions​, 1 million bots (Optimus humanoid robots) delivered and​ 1 million robotaxis in commercial operation. To date, Tesla has delivered more than 8 million vehicles, according to its September proxy statement.

Elon also mentioned that Tesla may need to build their own chip fab

“But even when we extrapolate the best-case scenario for chip production from our suppliers, it’s still not enough,” he said.

Tesla would probably need to build a “gigantic” chip fab, which Musk described as a “Tesla terra fab.” “I can’t see any other way to get to the volume of chips that we’re looking for.”

According to Musk, Tesla’s potential fab’s initial capacity would reach 100,000 wafer starts per month and eventually scale up to 1 million. In the semiconductor industry, wafer starts per month is a measure of how many new chips a fab produces each month.

For comparison, TSMC says its annual wafer production capacity reached 17 million in 2024, or around 1.42 million wafer starts per month.

While Tesla doesn’t yet manufacture its own microchips, the company has been designing custom chips for autonomous driving for several years.

It is currently outsourcing production of its latest-generation “AI5” chip, which Musk said will be cheaper, power-efficient, and optimized for Tesla’s AI software.

“With AI and robotics, you can actually increase the global economy by a factor of 10, or maybe 100. There’s not, like, an obvious limit,” Musk said at the shareholder meeting.

An interesting Wall Street Journal piece about the “end of the microchip era” by George Gilder

You can see the effects of the reticle limit in the ever-mounting complexity of Nvidia-defined vast hyperscale data centers. The result—smaller, denser chips and “chiplets,” each with its own elaborate packaging—is a greater need for ultimate reintegration of the processes for coherent outcomes. The calculation first has to be dispersed among many chips, then recompiled. The effect is more communications overhead between chips requiring ever more complex packages, ever more wires and fiber-optic links.

The result of the inexorable reticle limit is the end of chips. What’s next? A wafer-scale integration model, which bypasses chips altogether. Mr. Musk pioneered this concept at Tesla with his now-disbanded Dojo computer project; the effort has been recreated as DensityAI.

Cerebras of Palo Alto, Calif., used the concept in its WSE-3 wafer-scale engine. The WSE-3 boasts some four trillion transistors—14 times as many as Nvidia’s Blackwell chip—with 7,000 times the memory bandwidth. Cerebras inscribed the memory directly on to the wafer rather than relegating it to distant chips and chiplets in high-bandwidth memory mazes.

Also working on a full wafer-scale future is David Lam, founder of Lam ResearchCorp., the world’s third-largest wafer-fabrication equipment company. In 2010, Mr. Lam founded Multibeam Corp., which created a machine that performs multi-column e-beam lithography. The technology allows manufacturers to bypass the reticle limit.

A new paper on capital allocation by Michael Mauboussin which is “a comprehensive study of how public companies in the U.S. spend money. We extend most of the analysis back to 1970, update the data through 2024, and discuss results for the first half of 2025 where practicable.”

A new Howard Marks memo

As always, thanks for reading. There will be more to come next week as we continue to focus on bottlenecks, thresholds and all the forces driving us toward inevitable futures.

AI Bubble Talk

It’s tough to read anything finance or tech related these days without some discussion of if AI is a bubble or not. With stock markets at an all time high and capital expenditures from big tech companies continuing to increase at an unprecedented rate, bubble talks dominate every corner of finance and tech media. As with any debate, it’s helpful to understand both sides, to learn from people with skin in the game and get perspectives from people who are on the ground with a front row seat, before eventually coming to our own conclusions from which we must make investment decisions.

AI Capex Boom

This piece from Sparkline capital does a great job substantiating the big tech companies transition from being asset light to businesses that spend a much higher % of revenue on capex. They cite a Mckinsey estimate that AI spending will reach a cumulative $5.2T over the next 5 years and a Bain estimate that data centers will need to generate $2T by 2030 for companies to see an adequate return on investment.

The Magnificent 7’s exceptional success has been fueled by their asset-light business models. By leveraging intangible assets, such as intellectual property, brand equity, human capital, and network effects, these firms have generated outsized returns without requiring a lot of capital.

As the top panel of the next exhibit shows, the Magnificent 7 have enjoyed enviable returns on invested capital of 22.5%, compared to just 6.2% for the S&P 493 over the past decade. Their higher capital efficiency is explained by their heavier use of intangible assets. As the bottom panel shows, their intangible intensity has been 3-7 times that of the S&P 493.

The piece addresses other concerns such as how concentrated the stock market has become and how AI infrastructure spending may be propping up the economy, as well as drawing comparisons to past booms such as railroads and the internet.

The future and the near term return on investment from all of this AI spending is at the top of mind for many investors, but the piece also highlights how management of big tech seems to be viewing the AI arms race. Less about immediate ROIC and more about who will emerge as leaders of a future where AI is the most important technology in the world.

If we end up misspending a couple of hundred billion dollars, I think that that is going to be very unfortunate, obviously. But what I’d say is I actually think the risk is higher on the other side.”

- Mark Zuckerberg, Facebook CEO

“I’m willing to go bankrupt rather than lose this race.”

- Larry Page, Google co-founder

The Magnificent 7 have a well-deserved reputation as high-quality, ensconced in an oligopoly across key tech markets.

However, AI appears poised to disrupt this cozy oligopoly, at least in the minds of Big Tech CEOs. As Bill Gates himself has argued, AI collapses their respective markets – search, social media, shopping – into one, and whoever wins the AI race wins all markets. Viewing AI as an existential risk, they have been forced into a bitterly competitive and costly arms race.

The AI arms race resembles the classic “Prisoner’s Dilemma” game theory problem. While the optimal move is for firms to mutually agree to moderate their AI investments, preserving their oligopoly, this equilibrium is unstable, as each firm is incentivized to unilaterally ramp up investment to capture the market; if OpenAI goes all-in, the rest cannot afford to sit idly. This results in a suboptimal equilibrium, in which all firms invest aggressively, even if it leads to over investment and the destruction of the collective profit pool.

Return on capital is on everyone’s mind but I think the above is probably the right way to understand the level and rate of spending that we’re seeing. This is the defining technological arms race of our lifetime and as long as the promise of AGI exists, big tech companies will spare no expense in their pursuit of emerging victorious.

This was a sober and thoughtful read and definitely helpful for investors to understand the reality of big tech companies spending more on capex and what that might mean for future returns.

Gavin Baker Interview

For an argument on why AI is not a bubble, Gavin Baker, founder of Atreides Management, gave an illuminating interview with David George at a16z’s Runtime conference. I always enjoy learning from listening to Gavin and admire how he is able to balance his optimism for technology with a sober grounding in financial reality.

He addresses the comparisons to the internet bubble by looking at both valuations and utilization of GPUs compared to the amount of unused fiber that was built during the telecom bubble.

I do not believe we’re in an AI bubble today. I had, depending on how you look at it, the privilege and the misfortune of being a tech investor during the year 2000 bubble, which was really a telecom bubble. And I think it’s helpful to compare and contrast today to the year 2000. Cisco peaked at 150 or 180 times trailing earnings, NVIDIA is at more like 40 times. So valuations are very different. More importantly, however, is that the internet bubble or telecom bubble in 2000 was defined by something called dark fiber.

If you were around in 2000, you’ll know what that was. Dark fiber was literally fiber that was laid down in the ground, but not lit up. Fiber is useless unless you have the optics and switches and routers that you need on either side. So I vividly remember companies like Level Three or Global Crossing or WorldCom would come in and they say, “We laid 200,000 miles of dark fiber this quarter. This is so amazing. The internet’s gonna be so big. We can’t wait to light these up.”

At the peak of the bubble, 97% of the fiber that had been laid in America was dark. Contrast that with today, there are no dark GPUs. All you have to do is read any technical paper, and understand that the biggest problem in a training run is that GPUs are melting from overuse.

He also touches on the increase in ROIC the big tech companies are seeing as evidence on why we’re not currently in a bubble.

And there’s a simple way to kind of cut to the heart of all of this.

It is to look at the Return on Invested Capital of the biggest spenders on GPUs who are all public. And those companies, since they ramped up CapEx, have seen around a 10 point increase in their ROICs. So, thus far, the ROI on all the spending has been really positive.

There’s an interesting and open debate about whether or not it will continue to be positive, with how much spend we’re going to have on Blackwell. (I personally think it will.) But there’s no debate that thus far, the ROI on AI spend has been really positive. And valuation wise, we’re just not in a bubble.

The interview is substantial and touches on topics from market structure, to business models, the infrastructure vs application layer to humanoid robots and more. One part that is particularly interesting is the statement that Google, not other chip or networking companies, is perhaps the biggest potential competitor to Nvidia.

So Nvidia’s biggest competitor isn’t AMD, it’s not Broadcom. It’s certainly not Marvell. It’s not Intel. It’s Google.

And more specifically, it’s Google because Google owns the TPU chip. And - today at least - it’s the only alternative to Nvidia for training. And maybe the best inference alternative.

And Google’s a problematic competitor. Because they also own a company called DeepMind and they have a product called Gemini. And I think you could argue that they’re the leading AI company today. I think they’ve taken 15 or 20 points of traffic share in the last two or three months. That’s just traffic to Gemini; it does not include search overviews. I suspect on an actual traffic basis, Google is bigger than OpenAI, Anthropic, anyone today. And that business is gonna run on TPUs.

And then we have three other labs that are relevant today. There’s Anthropic; that’s an Amazon and Google captive. Anthropic is really going to run on TPUs and Trainiums. And so you’re left with xAI and OpenAI at the forefront. And if Google is going to a lab like Anthropic and saying, “I’m going to help you fundraise and give you chips, for competitive reasons, it’s very hard for Nvidia not to respond. And as Jensen said, he thinks it’s going to be a good investment.

He continues his discussion of Google’s positioning later on in the interview

I think it is really a fight between Nvidia and the Google TPU. And then something that I don’t think is broadly appreciated is the extent to which Broadcom and AMD are effectively going to market together.

Nvidia is no longer just a semiconductor company, as I’m sure you’ll hear from Jensen tomorrow. It was a semiconductor company, then a software company with Cuda, now a systems company with these rack level solutions. And now arguably, a data center-level company with the level of architecting they’re doing with scale up, scale across and scale out networking.

So the networking, the fabric, the software, it’s all important. And what Broadcom is saying to companies like Meta is, “Hey, we will build you a fabric that can theoretically compete with Nvidia’s fabric, which is a mixture of NVLink and either Infiniband or Ethernet. We’ll build it on Ethernet, it’s gonna be an open standard. And we’ll make you your version of TPU, which took Google three generations to get working. And you know what, if your ASIC isn’t good, you can just plug AMD right in.”

I personally believe most of those ASICs are gonna fail, particularly in the fullness of time. I think you’ll see a bunch of high profile ASIC programs canceled, especially if Google starts selling TPUs externally.

What can we make of all this? It does seem inevitable that the big tech companies continue to ramp capex in the near term as the AI arms race shows no sign of slowing down. It also seems inevitable that the leaders of these companies view the risk of losing the arms race as far worse than overspending. We think this is the proper lens through which to view what’s going on. As far as the bubble talk, there are pockets of the market that are worse than the Mag 7. This chart from BAML shows the Mag7 (with Broadcom instead of Tesla) valuations and projected growth for the next two years. Yes, they are collectively more expensive than the other 493 companies on an absolute basis, but they are also poised to grow earnings 20% vs an average of 5% over the next two years.

This chart from UBS projects capex for the hyperscalers through 2028, and sees Amazon, Google, Microsoft and Meta all spending over $100B each in 2028 in capex. The projections also predict over $400B of 2028 revenue for Nvidia, the majority of which will come from compute with the rest coming from networking.

Satya Nadella said on the BG2 podcast that the biggest bottleneck for AI isn’t compute, but power and the ability to build data centers quickly, near power. This is an active area of research for us as it seems that AI progress will be inevitably constrained by access to power. It fits well within our framework of identifying bottlenecks and thresholds and this week there were several major deals that highlight how central energy is becoming to AI.

Energy Deals

The pursuit of energy solutions to power the AI arms race is something that is at the top of mind for hyperscalers.

This week saw Google and Next Era teaming up to restart a nuclear power plant

NextEra Energy and Google have reached an agreement to restart an Iowa nuclear power plant shut five years ago, the companies said on Monday, in another sign that data-center power demand is renewing interest in U.S. nuclear energy.

The technology industry’s quest for massive amounts of electricity for artificial-intelligence processing has renewed interest in the country’s nuclear reactors, which generate large amounts of around-the-clock power that is virtually carbon free.

It also saw a landmark deal announced between Brookfield, Cameco, Westinghouse and the United States Government to build $80B worth of nuclear power plants across the country.

At the center of the new strategic partnership, at least $80 billion of new reactors will be constructed across the United States using Westinghouse nuclear reactor technology. These new reactors will reinvigorate the nuclear power industrial base.

As a result of this historic agreement, nuclear energy deployment will be a central pillar of America’s program to maintain global leadership in nuclear power development and Artificial Intelligence.

The partnership will facilitate the growth and future of the American nuclear power industry and the supporting supply chain. Each two-unit Westinghouse AP1000 project creates or sustains 45,000 manufacturing and engineering jobs in 43 states, and a national deployment will create more than 100,000 construction jobs. The program will cement the United States as one of the world’s nuclear energy powerhouses and increase exports of Westinghouse’s nuclear power generation technology globally.

Once constructed, the reactors will generate reliable and secure power, including for significant data center and compute capacity that will drive America’s growth in AI.

The partnership contains profit sharing mechanisms that provide for all parties, including the American people, once certain thresholds are met, to participate in the long-term financial and strategic value that will be created within Westinghouse by the growth of nuclear energy and advancement of investment into AI capabilities in the United States.

The construction of the Westinghouse nuclear power plants was also included in the announcement that Japan will invest $550B in the United States.

Energy-related projects appeared to loom large in the list in terms of total scale of business. Westinghouse’s construction of AP1000 nuclear reactors and small modular reactors was expected to be worth up to $100 billion, involving Japanese suppliers and operators including Mitsubishi Heavy Industries. Another small modular reactors project that could involve GE Vernova / Hitachi was framed as also being worth up to $100 billion.

“These are great companies, many of them household names — that you would expect — that are going to build infrastructure and improve the national economic security of the United States,” Lutnick said, adding that other projects can be considered.

“We look forward to engaging with other companies that wish to provide that same sort of foundational position. We’re going to be examining shipbuilding, pipelines, there’ll be an enormous number of other critical minerals. There are a whole variety.”

If the energy bottleneck for AI ends up reenergizing investment in nuclear power, that’s a win-win for the United States and the companies involved. Nuclear has long promised an abundance of clean energy but has been held back for reasons beyond the scope of this piece. This remains an active area of research for us, as energy is one of the defining bottlenecks in the ongoing AI arms race. It’s exciting to see so much focus on scaling the world of atoms to power the world of bits.

Corporate Earnings

Google

With talks of a bubble so prominent in the media, there is an increased focus on earnings of the big tech companies. In our 2023 piece about the future of networking, we noted how Google’s custom TPU’s and purpose built networking and overall data center architecture put them in a unique position as both a cloud service provider and AI model and application developer. Google has been under lots of pressure as of late because they were seen as being behind in the AI race, despite the transformer having been invented by Google researchers. We are believers that Alphabet is one of the best companies of all time and this quarter was great evidence that they are well positioned to capture value in the ever changing landscape of high technology.

Some highlights from their first ever $100B quarter (wild to say)

• Consolidated Alphabet revenues in Q3 2025 increased 16%, or 15% in constant currency, year over year to $102.3 billion. Google Search & other, YouTube ads, Google subscriptions, platforms, and devices, and Google Cloud each delivered double-digit growth in Q3.

• Google Services revenues increased 14% to $87.1 billion, reflecting robust performance across Google Search & other, Google subscriptions, platforms, and devices, and YouTube ads.

• Google Cloud revenues increased 34% to $15.2 billion, led by growth in Google Cloud Platform (GCP) across core products, AI Infrastructure and Generative AI Solutions.

• Total operating income increased 9% and operating margin was 30.5%. Excluding the $3.5 billion charge related to the European Commission (EC) fine, operating income increased 22% and operating margin was 33.9%, benefitting from strong revenue growth and continued efficiencies in the expense base.

• Other income reflected a net gain of $12.8 billion, primarily the result of net unrealized gains on our non- marketable equity securities.

• Net income increased 33% and EPS increased 35% to $2.87.

• With the growth across our business and demand from Cloud customers, we now expect 2025 capital expenditures to be in a range of $91 billion to $93 billion.

So yes the capex numbers are eye popping. But one could argue that $100B quarters with 30% operating margins afford a company such bets. The blog post is full of exciting developments in all aspects of their business including this note about TPU demand

Overall this quarter is about everything one would want to see from the company as an investor. Double digit growth across every part of the business. Strong AI usage and shipping at a high pace. Cloud growth with solid margins. We believe that Alphabet remains uniquely positioned to capture value in the AI arms race as both an infrastructure provider and model and application developer; on top of being an exceptional and reasonably diversified technology company.

Amazon

Amazon also delivered a strong quarter. Showing strong cloud growth indicating healthy continuing demand for AI infrastructure. They reported $180B of Q3 sales (13% increase) and $21.2B of net income compared to $15.3B in Q3 2024. This sent shares up 11% after the report.

“We continue to see strong momentum and growth across Amazon as AI drives meaningful improvements in every corner of our business,” said Andy Jassy, President and CEO, Amazon. “AWS is growing at a pace we haven’t seen since 2022, re-accelerating to 20.2% YoY. We continue to see strong demand in AI and core infrastructure, and we’ve been focused on accelerating capacity – adding more than 3.8 gigawatts in the past 12 months. In Stores, we continue to realize the benefits of innovating in our fulfillment network, and we’re on track to deliver to Prime members at the fastest speeds ever again this year, expand same-day delivery of perishable groceries to over 2,300 communities by end of year, and double the number of rural communities with access to Amazon’s Same-Day and Next-Day Delivery.”

Some other highlights:

• Saw continued strong adoption of Trainium2, its custom AI chip, which is fully subscribed and a multi-billion-dollar business that grew 150% quarter over quarter.

• Launched Project Rainier, a massive AI compute cluster containing nearly 500,000 Trainium2 chips, to build and deploy Anthropic’s leading Claude AI models.

• Announced new Amazon EC2 P6e-GB200 UltraServers using NVIDIA Grace Blackwell Superchips, designed for training and deploying the largest, most sophisticated AI models.

• AWS 20% growth, revenue of $33B and operating income of $11.4 which was about 2/3 of the company’s total operating profits.

Amazon’s ability to execute at their scale is incredible, and while they have been under considerable pressure from investors about their ability to win AI deals, their Trainium chips and plans for $125B in capex puts them in a solid position following these earnings.

The focus on big tech earnings is understandable considering how much of the index they represent (Microsoft, Meta, Amazon, Alphabet and Apple represent about 25% of the index and all reported last week.)

Meta & Microsoft

Meta’s earnings were skewed by a one time tax hit, but the stock fell over 9% as a result of a higher 2025 capex guidance ($70-$72B) and a 300bps decrease in operating margin (43% to 40%) despite $51.2B in Q3 revenue, a 26% increase from last year, which underscores strength in the core business. A cursory read of the move is that the market questions whether Meta can earn a satisfactory return on the capex as the cloud service providers.

Microsoft posted strong results, with revenue of $77.7 billion (up 18% year-over-year) and EPS of $3.72, both ahead of expectations. Growth was led by the Microsoft Cloud segment, which reached $44.4 billion in revenue, up 29%, with Azure and other cloud services growing about 40%. Despite the beat, the stock slipped roughly 3% as investors focused on the company’s steep $34.9 billion in quarterly capex and management’s signal that spending will accelerate further in FY2026. The results reinforced Microsoft’s leadership in AI infrastructure, but also highlighted the market’s growing demand for evidence of durable margins and returns on this wave of investment.

Roundup

Some other interesting pieces from this week:

Bain highlighted a case of a private equity firm that was going to buy a software company. During diligence, they used AI to build a prototype similar to the company’s product and thought it was better so they didn’t buy the company.

Anthropic launched Claude for Excel

Substrate launched with the ambitious plan of not just creating a new lithography tool, but building their own fab in the US. Semianalysis, one of my favorite publications, wrote an excellent piece covering the company.

The Fed cut interest rates by 25 bps bringing the fed funds rate to 3.75-4% but Powell stated that future rate cuts are “far from a foregone conclusion.” And notably, the committee had one member dissent in favor of a 50bps cut and another in favor of no cut.

1x launched NEO, a home robot you can buy for $20k.

And Elon said SpaceX will be building data centers in space.

This week was busy with discussions of the state of the market and whether AI is a bubble, landmark energy deals, corporate earnings, a Fed decision and some exciting startup launches amongst many others. This format is new for me and I imagine it will continue to evolve. I’ll do my best to continue surfacing interesting things I come across in the process of my research and just as a result of being terminally online. As always, my dms are always open for interesting things along these lines!

The idea I’ve come up with is to publish a weekly piece with links to things I found interesting in the prior week. The thesis being that in today’s age of media overload, it’s harder than ever to sift through the noise, so as someone who’s terminally online maybe my curation can help surface interesting things for readers.

It’ll be work in progress, so hopefully it’ll evolve as the weeks go by, but going back to the goal of making publishing more natural on here, I spend a lot of time reading (or listening/watching) things that are relevant to investment ideas and also just things that generally pique my curiosity. Hopefully there will be something for everyone and I’m also hoping it will encourage people to send me more interesting content that I can include in future pieces.

Colossus Profile of Josh Kushner

The New World - Colossus

This brilliant piece by Jeremy Stern is deities worth a read. It’s an incredibly illuminating profile of one of the most unique and inspiring investors of our time and spans from Josh’ grandparents story as holocaust survivors, to his founding of Thrive and Oscar, to in depth stories from the firm’s investments in companies like Github, Stripe and OpenAI. I always love learning more about investors I admire and can’t recommend this piece enough.

There’s a lot of takeaways from the piece, so difficult to highlight any one in particular but what stands out to me about Josh and Thrive is the willingness to rethink what it means to be an investor and a fund from first principles. It is easy to think you have to fit into a box that the world already understands and far more difficult to stay committed to building the exact kind of firm that you think should exist. This piece is filled with the evidence of what makes Thrive unique and it’s definitely inspiring to learn about the success of a fund with such a unique view of what it could be (an investor, an incubator, a company itself, a true “capital partner” to their founders in every sense of the word.)

In 2010, moreover, the unknown Kushner brother’s unknown little firm was making a bunch of large but weird-sounding claims for itself, like that it was a stage-, geography-, and sector-agnostic venture firm that would concentrate all its investments in a very small number of companies; that it was not only an investment firm but also itself a company; that it incubated its own companies as well as invested in others; and that it didn’t just invest and incubate but functioned as a service provider, product creator, and embedded operational commando unit for founders.

Later on in the piece:

Cutler also introduced Kushner to Andy Golden, the legendary head of Princeton University’s endowment. Golden later recalled a happy hour for VCs in Cambridge in 2010, where he saw a six-foot-three, emo-looking kid in a black cardigan standing apart from the group, staring at the floor.

“I think that Andy saw more in me than I saw myself,” Kushner said. “He spent time talking to me about who I wanted to be and what I wanted to build.” One insight Golden vouchsafed to Kushner was that investment firms, as they scale, start to lose a sense of their identity and wind up focusing not on what they’re good at but on the size of their assets under management, in turn leading to a lower cost of capital, less ambitious people, more mediocrity, and lower returns.

In response, Kushner pitched him on his seemingly unwieldy idea for Thrive: an opportunistic vehicle agnostic to stage, sector, or geography, which viewed itself as an enabling technology for the world it wished to see, and which had the capacity to not just invest in companies but to build them. “It was seen as very controversial at the time,” Kushner told me. “I was not trying to be provocative. The idea for Thrive came purely from where we believed the world was moving. It was what felt right to us even though it had never been done before.”

To a writer and not a VC, hearing investors describe their firms as “controversial” and “provocative” can get tiresome, as if they doth protest an awareness of conventional respectability with too much poser-iconoclasm. When told this, Kushner replied, with his customary decorum: “At that moment in venture you were either an early stage firm or a later stage firm. You were either a software firm or a consumer firm. You were either a European firm or a US firm. The idea of having a fund that could build companies, invest in companies, invest in them early or late, and inside or outside the US, it was just deeply unconventional. I feel really grateful that Andy saw it and understood it.

I found this profile to be amazing - in depth, substantial, super well written and definitely inspiring for investors of all kind. It definitely makes a case that being a great investor and a great operator can be two sides of the same coin (something we discussed in our piece, Time ~= Money.) I think what Colossus is doing is super cool and definitely looking forward to reading future pieces from them.

Dan Sundhiem Cheeky Pint Interview

This interview was wide ranging and covered topics from large cap valuations and if we’re in an AI bubble, to vibe trading and why it’s tricky to short bad companies but the part I found most interesting is his discussion about public markets vs private markets. He highlights an interesting point about the effect that stock prices can have on employees (below excerpt from the transcript)

“So, you’ve got Ramp, Stripe, SpaceX, and, you know, maybe one day, of course, all these companies go public, but it would seem as if there’s like a lot of later stage private companies now. Well, like, why do you think that’s happening, and where do you think that’s going to go in a couple of years? Like, are public markets just going to kind of be the laggards, and all the new hot stuff will be private, or will it rebalance one day?

I mean, if I ran a private company like Stripe, I wouldn’t go public. I think the public markets, you know-

It’s kind of ironic because you’re a public markets investor.

Yeah. I think the public markets are kind of problematic at this point. Let’s just take Stripe, for example, and I won’t speak for John, but basically, Stripe grows earnings, cash flow at some amount, value compounds, and they do tender offers, and the tender offers are relatively in line with the value creation.

“Therefore, the people who are working at the company, and they’re creating that value, you get paid for that value because the stock price goes up in line with value creation. Now, what we see in public markets is you take your company public, and depending on what the retail crowd is doing at that day, the stock may trade at some insane value. Most people are high-fiving.

This is amazing. Our stock is trading to 2X where it should be. This is great.

We’re all rich. The problem with that is that you’ve now pulled forward a ton of value. So all the people working at the company now are being overpaid because they didn’t actually create this value.

The stock gave this value. Then the people who you’re hiring, and those people are probably more likely to just cash out because they just made too much money.

You’re robbing future employees to pay the current employees.

“Exactly. Then future employees now have to give them stock options that are as used at a stock price you don’t really believe in. So the stock is so volatile that you’re actually not being paid as an employee based on value creation.

You’re being paid arbitrarily based upon multiples which have nothing to do with the true intrinsic value of the company. So, I think it’s like, obviously, it’s bad to be undervalued as a company, because then you’re issuing stock to employees at too low of a value, and then they don’t appreciate it usually. But it’s pretty bad to be overvalued too.

Because if the stock doesn’t go up, they will definitely come back to you and ask for more options. If the stock goes up way more than it should, they’re not going to come back to you and be like, Oh, you know what? Hey, I made too much money.

So you end up having this asymmetric, I think it’s really not a healthy dynamic to be a public company.”

This is interesting commentary for a number of reasons. D1 is mainly a public market fund that is gearing up to launch a private equity arm. We have seen firms like Thrive, Coatue, General Catalyst and Lightspeed with similar plans that align with our thesis that in the future, inevitably, the best investors will find value across stages, whether the company is public or private.

The public is private debate is one that could have its own post altogether. Dan raises a valid point about the variance of the stock price potentially making it difficult to retain employees, but the other side of that argument is valid as well. There’s no reason that private markets are a “better” valuation. Another point to the contrary would be look at how aggressive Meta has been with their hiring of AI talent. If they weren’t a trusted public company with the ability to offer liquid stock to potential hires, would their hiring spree be as successful? So you join OpenAI and the valuation is more “stable” but you can only sell into tender offers or you join Meta, the stock has higher variance but is far more liquid. I’m not sure, but this topic is super important for understanding the future of capital markets.

I do think there’s a case to be made that healthy capital markets involve companies eventually going public, but at the same time, if you can access capital like SpaceX or Stripe, there’s not a whole lot of incentive to do so. There’s probably also an argument that the more good companies that go public, the better the US investor does on average, as those companies now have a chance to be included in major indices. Anyway, it’s definitely a thought provoking part of the interview and something I’m interested in following and potentially writing more about in the future.

Dwarkesh Karpathy Interview

When looking for information about frontier technology, I’m often reminded of the Picasso quote:

This interview is an awesome lens into how Andrej Karpathy thinks about AI and the current landscape tech, with discussions about agents, reinforcement learning, the path to AGI, how humans learn, the potential impacts of AI on GDP, self driving cars and lots more. I find conversations like these super valuable, especially when so much of the conversation amongst investors is whether or not AI is a bubble at the current moment in time.

In line with the bottleneck framework shared last week, the podcast kicks off with this question

Dwarkesh Patel 00:00:58

What do you think will take a decade to accomplish? What are the bottlenecks?

Andrej Karpathy 00:01:02

Actually making it work. When you’re talking about an agent, or what the labs have in mind and maybe what I have in mind as well, you should think of it almost like an employee or an intern that you would hire to work with you. For example, you work with some employees here. When would you prefer to have an agent like Claude or Codex do that work?

Currently, of course they can’t. What would it take for them to be able to do that? Why don’t you do it today? The reason you don’t do it today is because they just don’t work. They don’t have enough intelligence, they’re not multimodal enough, they can’t do computer useand all this stuff.

They don’t do a lot of the things you’ve alluded to earlier. They don’t have continual learning. You can’t just tell them something and they’ll remember it. They’re cognitively lacking and it’s just not working. It will take about a decade to work through all of those issues.

The next interesting excerpt:

Dwarkesh Patel 01:06:45

Or they’ll propose horizon length. Maybe they can do tasks that take a minute, they can do those autonomously. Then they can autonomously do tasks that take an hour, a human an hour, a human a week. How do you think about the relevant y-axis here? How should we think about how AI is making progress?

Andrej Karpathy 01:07:05

I have two answers to that. Number one, I’m almost tempted to reject the question entirely because I see this as an extension of computing. Have we talked about how to chart progress in computing, or how do you chart progress in computing since the 1970s or whatever? What is the y-axis? The whole question is funny from that perspective a little bit.

When people talk about AI and the original AGI and how we spoke about it when OpenAI started, AGI was a system you could go to that can do any economically valuable task at human performance or better. That was the definition. I was pretty happy with that at the time. I’ve stuck to that definition forever, and then people have made up all kinds of other definitions. But I like that definition.

The first concession that people make all the time is they just take out all the physical stuff because we’re just talking about digital knowledge work. That’s a pretty major concession compared to the original definition, which was any task a human can do. I can lift things, etc. AI can’t do that, obviously, but we’ll take it. What fraction of the economy are we taking away by saying, “Oh, only knowledge work?” I don’t know the numbers. I feel about 10% to 20%, if I had to guess, is only knowledge work, someone could work from home and perform tasks, something like that. It’s still a really large market. What is the size of the economy, and what is 10% or 20%? We’re still talking about a few trillion dollars, even in the US, of market share or work. So it’s still a very massive bucket.

Going back to the definition, what I would be looking for is to what extent is that definition true? Are there jobs or lots of tasks? If we think of tasks as not jobs but tasks. It’s difficult because the problem is society will refactor based on the tasks that make up jobs, based on what’s automatable or not. Today, what jobs are replaceable by AI? A good example recently was Geoff Hinton’s prediction that radiologists would not be a job anymore, and this turned out to be very wrong in a bunch of ways. Radiologists are alive and well and growing, even though computer vision is really, really good at recognizing all the different things that they have to recognize in images. It’s just a messy, complicated job with a lot of surfaces and dealing with patients and all this stuff in the context of it.

I don’t know that by that definition AI has made a huge dent yet. Some of the jobs that I would be looking for have some features that make it very amenable to automation earlier than later. As an example, call center employees often come up, and I think rightly so. Call center employees have a number of simplifying properties with respect to what’s automatable today. Their jobs are pretty simple. It’s a sequence of tasks, and every task looks similar. You take a phone call with a person, it’s 10 minutes of interaction or whatever it is, probably a bit longer. In my experience, a lot longer. You complete some task in some scheme, and you change some database entries around or something like that. So you keep repeating something over and over again, and that’s your job.

You do want to bring in the task horizon—how long it takes to perform a task—and then you want to also remove context. You’re not dealing with different parts of services of companies or other customers. It’s just the database, you, and a person you’re serving. It’s more closed, it’s more understandable, it’s purely digital. So I would be looking for those things.

But even there, I’m not looking at full automation yet. I’m looking for an autonomy slider. I expect that we are not going to instantly replace people. We’re going to be swapping in AIs that do 80% of the volume. They delegate 20% of the volume to humans, and humans are supervising teams of five AIs doing the call center work that’s more rote. I would be looking for new interfaces or new companies that provide some layer that allows you to manage some of these AIs that are not yet perfect. Then I would expect that across the economy. A lot of jobs are a lot harder than a call center employee.

I found his perspective to be insightful and refreshing. It was helpful to zoom out a bit amongst the tiresome bubble conversations and get a ton of substance in one sit down interview.

I’m also a huge fan of Dwarkesh in general and think he’s doing amazing work. I really enjoyed his piece about the AI buildout Thoughts on the AI buildout. It touches on a lot of key aspects of he race to build the best AI systems and current the bottlenecks for the arms race. Of particular interest is the reality facing the energy bottleneck

“For the last two decades, datacenter construction basically co-opted the power infrastructure left over from US deindustrialization. For AI CapEx to continue growing on its current trajectory, everyone upstream in the supply chain (from people making copper wire to turbines to transformers and switchgear) will need to expand production capacity.

The key issue is that these companies have 10-30 year depreciation cycles for their factories (compare that to 3 years for chips). Given their usual low margins, they need steady profits for decades, and they’ve been burned by bubbles before.

If there’s a financial overhang not just for fabs, but also for other datacenter components, could hyperscalers simply pay higher margins to accelerate capacity expansion? Especially given that chips are an overwhelming 60+% of the cost of a data center.

We did some back-of-the-envelope math on gas turbine manufacturers which seems to indicate that hyperscalers could pay to have their capacity expanded for a relatively small share of total datacenter cost. As @tylercowen says, do not underrate the elasticity of supply.”

As this tweet covering an interview with an employee working on Meta’s infrastructure points out, energy is the biggest constraint to the planned data center buildout

https://x.com/rihardjarc/status/1981721050086883834?s=46

An astute reply pointed out 4 key bottlenecks to watch:

1. Energy capacity (operational, 18-24 month lead time)

2. TSM 2nm supply (physical, multi-year constraint)

3. HBM premium pricing (financial, 3-5 year window per expert)

4. Storage infrastructure (SSD/HDD for inferencing)

That’s four separate bottlenecks. Even if you solve GPU supply, three others can choke deployment.

This is an area of interest that is definitely aligned with our overall framework of looking for bottlenecks and thresholds, and one I am actively researching.

Anthropic announced agreement to use Google’s TPUs

Expanding our use of Google Cloud TPUs and Services \ Anthropic

Today, we are announcing that we plan to expand our use of Google Cloud technologies, including up to one million TPUs, dramatically increasing our compute resources as we continue to push the boundaries of AI research and product development. The expansion is worth tens of billions of dollars and is expected to bring well over a gigawatt of capacity online in 2026.

“Anthropic’s choice to significantly expand its usage of TPUs reflects the strong price-performance and efficiency its teams have seen with TPUs for several years,” said Thomas Kurian, CEO at Google Cloud. “We are continuing to innovate and drive further efficiencies and increased capacity of our TPUs, building on our already mature AI accelerator portfolio, including our seventh generation TPU, Ironwood.”

Seems relevant because previously Anthropic was reported to be spending lots of money on AWS’ trainium, so an interesting story to follow both as it pertains to cloud revenue and alternative AI chip solutions to Nvidia’s dominance.

Oracle financing data centers

https://www.bloomberg.com/news/articles/2025-10-23/record-38-billion-debt-sale-nears-for-oracle-tied-data-centers

Banks are preparing to launch a $38 billion debt offering as soon as Monday that will help fund data centers tied to Oracle Corp. in what would be the largest such deal for artificial intelligence infrastructure to come to market, according to people with knowledge of the matter.

JPMorgan Chase & Co. and Mitsubishi UFJ Financial Group are among banks leading the deal, which is split across two separate senior secured credit facilities, said the people, who asked not to be identified when discussing private matters. One $23.25 billion package will go toward financing a data center in Texas and another $14.75 billion facility will help fund a project in Wisconsin, the people said.

Vantage Data Centers is developing both data centers, which are set to be used by Oracle to power OpenAI, Bloomberg has reported. The projects are part of Oracle’s broader effort to invest $500 billion in AI infrastructure alongside OpenAI, known as Stargate.

I found this interesting because it’s tangible development toward’s OpenAI’s super ambitious plan for Stargate and also just an eye popping number for the debt markets.

Origins of Efficiency

I started reading Origins of Efficiency by Brian Potter this week. I’m only a couple chapters in but it’s been amazing so far. The book opens with the story of how the development of Penicillin was as much about manufacturing and being able to produce the medicine at scale as it about the scientific breakthrough that occurred when Penicillin was first discovered.

About The Origins of Efficiency

Efficiency is the engine that powers human civilization. It’s the reason rates of famine have fallen precipitously, literacy has risen, and humans are living longer, healthier lives compared to preindustrial times. But where do improvements in production efficiency come from? In The Origins of Efficiency, Brian Potter argues that improving production efficiency-finding ways to produce goods and services in less time, with less labor, using fewer resources-is the force behind some of the biggest and most consequential changes in human history. The book is punctuated with examples of production efficiency in practice, including how high-yield manufacturing methods made penicillin the miracle drug that reduced battlefield infection deaths by 80 percent during World War II; the 100-year history of process improvements in incandescent light bulb produc-tion; and how automakers like Ford, Toyota, and Tesla developed innovative production methods that transformed not just the automotive industry but manufacturing as a whole. The Origins of Efficiency is a comprehensive companion for anyone seeking to understand how we arrived at this age of relative abundance-and how we can push efficiency improvements further into domains like housing, medicine, and education, where much work is left to be done.

Again this is the first iteration of a weekly “things I found interesting” type of piece so I imagine it will evolve as I do more of these. Feel free to give feedback or ideas about what you’d like to see more of. Ideally I’ll try to surface some niche stuff I find from being terminally online. I’m not sure about the total number of links to include. On one hand I want to provide some of my thoughts about each selected piece but on the other I do come across lots of good stuff throughout the week so maybe it will make sense to highlight a few in particular and then include a collection of other great stuff at the end of the piece. I do hope this will inspire readers to send me interesting things you come across, my dms are always open!

In our experience, the best investments are found when we are in search of the inevitable. It is of course far easier to look back in hindsight and say “it was inevitable that so and so happened” but the exercise of understanding patterns of history and how the forces of scientific discoveries, technological breakthroughs and human behavior collide to drive both innovation and economic growth is essential in building a coherent framework through which to view the world and its many possible futures. History shows us that major economic and technological waves emerge when a constraint gives way or when a major threshold is crossed. Entire new classes of companies and markets can become possible when a physical, computational, or informational bottleneck is broken.

Studying the past allows us to gather many examples and compare similarities in an attempt to build a general framework from the common patterns. There is, of course, no way to know how exactly the future will unfold. So any declaration of “it is inevitable that…” comes with the understanding that there’s a high degree of uncertainty in any prediction about the future. Markets (and reality) are chaotic, random and far from deterministic. Rather than pretending we have a crystal ball in our search for the inevitable, we treat our theses like a scientific hypothesis. Something to be tested against real world evidence and updated as we gather new data.

The origin of this thesis and the name Inevitability Research come from my personal experience with investing. It was 2015, I was in college and had just signed up for a Robinhood account to invest the money I had earned over the previous summer. A physics major, I had been interested in how frontiers of technology became commercialized as businesses. Frustrated by Boston traffic and impressed with Uber’s meteoric success, my curiosity led me down a rabbit hole about self driving cars. I, in my young naivety, became convinced that self driving cars were not just possible; but inevitable. We were inevitably going to build these things. Inevitably, a tremendous amount of capital was going to be invested in making this a reality. I learned about the different technologies involved and the various bottlenecks that would have to be solved in order to bring this dream to fruition. This led me to computer vision and GPUs and the thesis that the company that stood to benefit from this inevitable future was Nvidia. This framework also led to investments in AMD, ASML, Taiwan Semiconductor and Tesla.

The thesis that AI training was inevitably bottlenecked by networking technology led to investments in Arista, Broadcom and Google, as it led to a deep dive on their TPUs and custom data center design that we explored in this piece about the future of networking. AI training was also bottlenecked by the amount of data centers, which led us to Celestica. The thesis that the best luxury companies would continue to be able to compound the value of their storied heritage led to LVMH, Hermes and Ferrari as we explored in Compounding Heritage. An exploration of inevitabilities in a post ChatGPT world led to our piece of Hyperindividualization.

I’m also always looking for historical examples that follow similar patterns. A useful exercise is to write a simple statement and then work backward to understand the dynamics that made it true. For instance: “In hindsight, it was inevitable that once we understood the unification of electricity and magnetism, we would seek to use electricity to power devices, beginning with light.” From there, we can analyze what bottlenecks had to be overcome and what thresholds were crossed to make that progress possible. Engines That Move Markets is a fantastic book that explores various technological breakthroughs, their commercialization and the financial bubble that almost always followed.

As I tried to understand history in terms of inevitabilities following scientific discovery and technological breakthroughs, what I noticed was that while it was a useful exercise to understand history through the lens of inevitabilities, a more repeatable process (and one more applicable to investing) would be identifying bottlenecks that were constraining progress and thresholds that once crossed, could unlock new companies and markets.

For example, in the case of electricity, the economy in the late nineteenth century was bottlenecked by the lack of consistent, cheap illumination. Once the threshold of industrial-scale electricity generation and distribution was crossed, entire industries reorganized around artificial light.

The semiconductor industry is perhaps the purest example of inevitability expressed through physics. For decades, Moore’s Law served as both an observation and a challenge: that transistor density, and therefore computational power, would double roughly every two years. Each generation faced its own bottlenecks in materials, heat, and fabrication, but engineers repeatedly crossed those thresholds through innovation in lithography, chip design, and process technology. Each breakthrough continued to shrink transistors, expanded what was computationally possible and opened entirely new markets, from personal computing to internet and software companies to the modern AI companies today.

Once organizations depended on software and accessing the internet became common in most people’s daily lives, computing itself became bottlenecked by physical infrastructure. Scaling required capital-intensive servers, space, and maintenance. Google, for instance, quickly required more computing power than was available at Stanford. The threshold was crossed when virtualization and elastic compute made it possible to rent computation the way one might draw power from a grid. Cloud computing basically inverted the main constraint facing companies needing compute and storage resources, turning infrastructure from a bottleneck into a catalyst. The next generation of startups was built atop the newfound abundance created by cloud computing. Netflix is a perfect example of a company that was possible now that the bottleneck of needing on prem resources was solved. Snowflake is another example. They uniquely understood that the next generation of data applications was bottlenecked by compute and storage being aggregated.

The same logic applies to modern technology. In the case of Instagram, the founders recognized that mobile applications were bottlenecked by slow cellular networks. Early investors doubted that phones could handle a feed of images on 3G connections, but the founders understood that the threshold would be crossed as networks evolved to 4G and beyond. The same framework could have led one to anticipate that video-first platforms like TikTok would become inevitable once 5G removed bandwidth as a bottleneck.

Studying history makes it apparent that it is essential for the investor in search of the inevitable to spend some time at the frontiers of science and technology. It may have been inevitable to the AI research community that a new era was under way upon reading “Attention Is All You Need” in 2017. It certainly seemed inevitable after the launch of ChatGPT. It seemed inevitable to many early believers that Bitcoin would play an increasingly important role in the global financial markets, and appreciate significantly in value. Growth, new companies and business models and even entirely new markets themselves are downstream of innovation at the frontiers.

Something I learned from studying Nvidia, and later Apple, is that this lens can also be applied to understanding how a founder or CEO sees the world. For Jensen Huang, it was inevitable that customers would want better graphics for video games. The bottleneck was chip design, not demand, and nothing in the laws of physics prevented him from building a processor optimized for graphics workloads. Another example of him betting Nvidia’s future on his view of the inevitable was investing heavily in developing CUDA for scientific computing, which later became the foundation for training deep learning models on GPUs.

Steve Jobs is perhaps the best example of someone who took a definite view of where technology was heading in the future and made it inevitable by building it. Some may have seen Doug Engelbart’s presentation (now titled “The Mother of All Demos”) as just a fun idea of how to use a computer. Many (including Xerox’s management!) dismissed personal computing as too niche to ever catch on, especially compared to the enterprise opportunity. But Jobs saw this future as inevitable and understood that building an operating system that encourages easy and standardized application development, as well as access to networking were the main bottlenecks in creating a device that was simple for the average person to use (which he did, perhaps ironically, during his time at NeXT.)

This is to say, ideally, a founder is building towards what they believe to be an inevitable future. Jensen couldn’t imagine a world without better graphics chips, Jobs without personal computers. Elon believes that for the survival of the human race, it’s inevitable that we become a multi planetary species and that we’d need to be able to reuse rockets that we send into space for various missions. Jeff Bezos saw the inevitability of e-commerce but also offered a crucial viewpoint in terms of how to view possible futures. Instead of trying to predict what would change in the future, he focused on what would not change. For instance, customers would always want to receive their packages sooner, have a wider selection of products at competitive prices, have access to free shipping etc. The common thread being the founders making a bet based on a future they see to be inevitable, and then attacking a bottleneck or threshold that brings them closer to that future.

This piece has mainly discussed our framework in terms of technology investments, which is where I spend a lot of my time. Of course, the stage at which an investment is made informs how tightly the numbers and the story are intertwined. The earlier the stage, the less tangible the numbers are and the more the thesis depends on narrative potential: the story of the bottleneck, how it will be solved, and what will be unlocked once it is.

We are, however, firm believers that this framework is applicable across sectors and stages, and across both public and private markets. Mature companies can unlock growth by identifying and navigating bottlenecks, and when important thresholds are crossed. In fact, the future of an established company can be more inevitable than that of one poised for hypergrowth. A useful way for gauging the inevitability of a more mature company is to try to understand what it would take to compete with them. Warren Buffet once said you could give him all the money in the world and it wouldn’t be enough to compete with Coca-Cola when explaining his thesis for the company. This is a helpful way for understanding the prospects of a more mature company. In the case of an acquisition, the bottlenecks may exist at the organizational level. Or recent thresholds may have enabled the company to enter new markets. Point being, we seek to make this framework generalizable to any and all investments.

A company’s ability to grow revenue, generate cash, reinvest capital at a high rate of return, and do so for a long period of time remains the foundation of any investment thesis. The importance of weighting towards story or numbers of course depends on the stage at which the investment is being made. It is always a balance and never an exact science.

“Once this bottleneck is solved, we can grow by X, improve our margins by Y, or open up a market worth Z that doesn’t currently exist.”

A key component of the work we are doing is the ability to adapt and evolve. If we were to rethink the job of an investor from first principles, it would be obvious that one should have the flexibility to invest across sector,stage and market; wherever bottlenecks are breaking or thresholds are about to be crossed and inevitabilities begin to reveal themselves.

How then, can one navigate the frontiers and later stages fluently, efficiently, and most important, profitably? Curiosity and discipline. Being a student of history and being able to imagine possible futures. Being able to triangulate viewpoints and information amongst people who deeply understand and are well versed in the topic of interest. This part is especially key and one of the many reasons that I find X dot com formerly Twitter to be an invaluable resource. You are one tweet or dm away from the foremost experts on nearly any subject in the world. Sure, there is a lot of noise, but one insightful conversation full of substance is worth all the noise in the world. In that sense, it’s the ultimate “expert network.” This is key for the investor in search of the inevitable because there is no way to stay on top of everything yourself. Seeking meaningful relationships and valuable conversations is integral to this pursuit.

As investors, we’re trying to understand the possible futures and do our best to come up with falsifiable hypotheses about them. Understanding history plays an important role in developing our framework. When discussing investment opportunities with founders and management team, this kind of framing can be useful to understand their thinking.

“In the future, it is inevitable that…” Their answer will offer lots of color about how they view their world and their company’s place in it.

“What needs to happen for your success to be inevitable?” This helps understand the possible bottlenecks and/or thresholds that need to be solved or overcome in order for the company to succeed and grow in a way that offers significant return on investment.

Inevitability, to us, is not about prophecy but about curiosity, discipline and process. The goal is not to predict the future. It is in the search for the inevitable that we identify bottlenecks and thresholds that present interesting investment opportunities.

This essay serves as the foundation of how we think at Inevitability Research. The ideas here will evolve, as all good hypotheses do, but the goal will remain the same: to understand where constraints are breaking, where thresholds are near, and where the future is quietly becoming inevitable.

If you’re building or investing where a bottleneck is breaking or a threshold is near, my DMs are always open.

One key takeaway for any founders or future founders reading this piece: you’re not building a tech business, you’re building a Business. There’s no real divide between finance and technology. Founders are capital allocators, whether they realize it or not. From day one, you’re allocating capital. You’re making decisions about how to deploy time, talent, and money. And once you’re allocating capital, you’re accountable to outcomes. The examples below - Google, Amazon and Mellanox, are tried and true stories of what it looks like to win by building around that mindset. I’ll be publishing follow-ups that unpack how today’s best startups are following similar playbooks and continuing to explore the theme of founders as capital allocators in my never ending quest to bridge the worlds of finance and tech. Stay tuned.

At a high level, it’s simple: you either

1. Go public

2. Get acquired (by another company or a PE fund)

3. Fail

There is a fourth outcome: operate indefinitely as a private company, generating enough cash to cover your lifestyle. But this is rare (and probably unrealistic) if you take institutional capital.

I should note the recent trend of companies staying private much longer than in the past, with liquidity available to employees, founders and investors through secondary transactions. The debate about the sustainability of liquidity via secondary and the overall trend of larger private rounds (dollar numbers that would have been massive IPOs not too long ago) is the topic of another article.

There is far more private capital than at any point in history, the network is perhaps more mature (and just following laws of supply and demand e.g. there is enough demand for Anthropic, Databricks, OpenAI, Stripe shares that these large rounds and secondary transactions can get done) the realities of going public force a different kind of scrutiny on a company and the current FTC regime has made it exceedingly difficult to get M&A deals across the line.

All that said, what’s the point of writing this? Well, I often get asked “what should a technical founder learn about finance to arm themselves when starting a company?” So this is just my two cents on how to keep the Pokemon League in the back of your mind as you build your team and collect gym badges. If you’ll allow me to be grandiose for a moment - I do also believe that healthy capital markets are a key component of a strong economy and business environment in the US (and across the world.)

So you started a company. You’re aware that you need to consider the potential exit scenarios to best prepare yourself for the journey. What might they look like?

The home run. The 5 star outcome.

You’re profitable AND growing like crazy. The north star for this situation would be Google’s IPO.

From $220k to $960m in 5 years and being profitable for 3 of them. You’ve proven your technology. You’ve built a world class team that’s aligned with your mission. You’ve found a proven method for getting distribution and capturing value and you’re looking for additional capital to do this at a greater scale. Google’s IPO should be studied for what an ideal company looks like at time of going public. Google is of course, one of the best businesses of all time, so it’s unlikely that many other companies will be able to chart a similar course but just because you won’t be able to play basketball like Michael Jordan doesn’t mean you don’t study his jump shot.

There are lots of other good examples of companies that were growing rapidly and either profitable or with a clear path to profitability. There (maybe) a few select types of business that just has no other way but to burn billions and billions in order to reach the scale necessary for their ultimate success. Uber pulled this off. But due in part to mis-aligned incentives, the growth at all costs mentality has become the norm and it seems that few companies reach the public markets already generating an operating profit.

There is no inherently right or wrong way. It’s up to you as the founder to understand the ROI of every decision you make. You can tell from reading his early shareholder letters that Jeff Bezos was fluent in this idea. He understood that the key driver of sales growth was going to be reinvestment in the business. And so it was marginally less important to show operating or net margins in the early years, when those dollars could be reinvested towards growing Amazon’s total sales.

The Almighty Merger / The Holy Acquisition

I was trying to relate each outcome to a civ win condition. IPO would probably be a domination victory. In a sense, an acquisition could be any of the others - cultural, science, religion etc. An absolutely unstoppable force of a company probably has no amount for which they would be acquired, but that doesn’t mean an acquisition is somehow a worse outcome. In fact, from a founder’s perspective, when you factor in time invested, it may even be preferable.

So why does a company get acquired? For various reasons and often a mix of a few key strategic factors -

1. Being a Strategic Fit

2. Acquiring the IP and Talent

3. Acquiring an existing customer base

4. Acquiring revenue synergies

5. Being a cultural fit

Case Study:

Nvidia’s 2019 Acquisition of Mellanox (2019, $6.9B)

Often an ideal acquisition will check multiple, if not all of these boxes to some degree. Mellanox was the perfect acquisition for Nvidia because it delivered strategic alignment, key IP + talent, customer expansion, revenue synergy, and cultural fit all in one deal.

People might not remember a time when Nvidia wasn’t the dominant leader in Data Center Products, but the segment represented much less of their business in 2017. At the time, there was a race to figure out what the post Moore’s Law world would look like. There was some consensus that as Moore’s Law slowed, performance gains would come from linking multiple chips together in high-performance computing clusters rather than relying on individual chip advancements.

This shift to distributed systems (like the ones that were being built by hyperscalers to offer their cloud computing services) significantly increased the need for high bandwidth / low latency networking solutions (with the challenge of number of instructions per chip solved by doing the computations in parallel, the focus shifted towards how fast / how much data you could send between nodes in the network and how reliable those connections were.)

Mellanox was the industry leader in high-speed interconnects, with InfiniBand and Ethernet solutions widely used in cloud and high performance computing systems. As enterprise computing environments continued to evolve towards the distributed systems model, Mellanox’s product suite was a clear strategic fit for Nvidia’s expansion further into the data center market.

Rather than developing an in-house solution, acquiring Mellanox allowed Nvidia to offer data center customers a fully integrated compute and networking package. This vision proved to be incredibly prophetic (and profitable) as high speed networking represents one of the major bottlenecks for computing in the AI era, and the combination of Nvidia GPUs + Mellanox Interconnects has become the industry standard for state of the art AI workloads.

It’s probably to the point where it’s taken for granted, but the massive performance improvements as well as a much lower total cost of ownership that came from the Nvidia + Mellanox synergies have definitely accelerated (and probably enabled to some degree) the major advancements in AI we’ve seen in the past few years.

The IP and Talent angle naturally follows from understanding how strong the synergies were with Mellanox as a strategic fit for Nvidia. Mellanox had deep expertise in networking silicon, high-speed data movement, and low-latency interconnects—a highly specialized field that few companies mastered. They had been building at this vision since 1999 and had acquired some key technologies themselves along the way. The engineering team at Mellanox was world-class, with a commanding lead both R&D and production for data center interconnects, InfiniBand, and Ethernet technologies. Nvidia didn’t need to build this expertise from scratch—instead, it acquired both the technology and the people driving innovation.

The story around revenue and synergies naturally follows as well. Nvidia saw a future where “The Data Center is the Computer” and being able to sell an integrated compute + networking solution to customers investing heavily in that vision of the future would position it well in a “picks and shovels for the gold rush” type of scenario (overused I know but is there really a better modern example of this in action?)

To put it into perspective, in 2019, when the acquisition was being finalized, Nvidia’s revenue was almost $12B for the year. Last quarter, their data center revenue alone was $39B. They are on track for more than $160B in sales for fiscal year 2026. That’s over a 13x on a $12B revenue base in 7 years post acquisition. That’s insane. It should be in the business school dictionary next to the definition of revenue synergies. An extraordinary example of revenue synergies in action.

Also to zoom out, Nvidia’s path to dominance in data centers was by no means a given. Nvidia was worth $144B at the end of 2019, a little more than half of Intel’s $260B market cap at the time. Looking at the above financials for Nvidia and Mellanox at the time of the acquisition respectively, it is clear that great CEOs see the potential far beyond a sum of the current revenue bases. As the world increasingly depends on the most advanced supercomputers to power AI, a combined revenue base of $13B at the time of the purchase has grown into one of the biggest and most important companies in the world.

The customer base was an important but less crucial factor of the acquisition. Nvidia already had relationships with its key customers (AWS, Azure, Google Cloud etc.) and while Mellanox did have some key relationships that Nvidia would inherit (Dell, HP, Microsoft, Oracle was once a 10% shareholder, to name a few) the main synergy in terms of customers was being able to sell a much improved, integrated product to existing customers. Importantly, Nvidia was now a one stop shop for crucial enterprise (and now AI) computing infrastructure.

In terms of a cultural fit, both Nvidia and Mellanox were led by highly technical CEOs (Jensen Huang at NVIDIA and Eyal Waldman at Mellanox) who prioritized deep R&D over short-term market pressures. Their cultures valued technical excellence—rather than focusing on rapid iterations, they built fundamental technology for long-term impact.

Neither company optimized for short-term product cycles. Instead, both invested in high-risk, high-reward innovation that would define computing infrastructure for the next decade.This shared long-term philosophy made the integration smooth—Mellanox didn’t have to adjust to a radically different corporate mindset.

Unlike many acquirers that aggressively merge teams, NVIDIA allowed Mellanox to maintain independence while aligning on strategic goals. This helped retain key engineering talent, ensuring that Mellanox continued innovating within NVIDIA rather than experiencing brain drain. Cultural alignment is as important as financial synergy—misaligned cultures can derail even the best deals. If an acquirer respects your company’s autonomy and vision, it leads to better integration, talent retention, and long-term success.

There are other great examples that come to mind for each category. Facebook buying Instagram and Google buying Youtube are great examples of companies buying a startup for strategic fit. Apple’s acquisition of NeXT and Google buying DeepMind are two examples of acquisitions for IP and Talent (of course, in almost every case, there are multiple reasons why an acquisition is made.) Microsoft buying LinkedIn, Amazon buying whole foods and Facebook buying WhatsApp come to mind when thinking of acquisitions that bought a customer base along with the company. Disney buying Pixar is a great example of both revenue synergies and a cultural fit. These aren’t rules that are set in stone or boxes that an acquirer checks when diligence a company, but hopefully a helpful framework for understanding the various reasons behind an acquisition.

Alignment and Incentives

From capital formation to business development and growth to exits and beyond, there are a million moving parts but one of the most important things is understanding alignment. If you raise institutional capital (venture money) it’s helpful to understand their incentives. They raise a fund to invest in companies and eventually have to pay back their own investors. They need companies to have liquidity events in order to generate cash and return it to their investors (and to renovate the kitchen in the Marin house.) The better the companies do, the better the investors do. The downside of this model is most (99%?) startups fail. So they need really big outcomes to generate an above market return on their fund. Not in every case, but often this will skew their deployment of capital towards companies that have a chance at these kind of outcomes. Again not in every case but this also informs their advice to founders - go big or go home. Raise more. Grow faster. Burn $1B for the chance to become a $10B+ company.

This isn’t necessarily a bad thing. If you become a unicorn or decacorn or a hundred billion dollar corn, both you and your investors will (more than likely) get rich. The downside is that it doesn’t really leave room to exist somewhere in the middle (what some VCs call “lifestyle businesses”(derogatory).)

One way to generate that liquidity is going public. We can talk for hours about the “why” this isn’t happening as often as it used to or what it says about the “health” of capital markets but the simple explanation seems to be a lot of deals get done at sky high multiples in private transactions that wouldn’t hold up vs public market comps. Going public is probably what your goal should be as a founder. If you want to build a world changing business, a low cost of capital is pretty helpful for doing so and there’s an argument to be made that’s there’s no better position for a company to be in than being a trusted public company.

I should zoom out for a second again - one of the goals with Inevitability Research is to dispel the false dichotomy of business vs tech. Or at least blur the lines. If you start a company you need to build and sell. Can you be exceptional by focusing on one aspect vs the other? Probably. But I think the ideal model of a founder is Jeff Bezos (clearly I stan,) who seems to have harnessed both business and tech and used this combined knowledge as a force that drove Amazon to be one of the most successful and incredible companies of all time. Reading his early shareholder letters probably has a higher ROI than getting an MBA, especially for a tech founder looking to grow a durable business.

The other way to generate liquidity is via being acquired. It could be a strategic acquisition by a company looking to expand into the market you’ve captured, or by a private equity fund that specializes in companies like yours. There are various outcomes and transaction structures - ranging from all cash, no earn outs or vesting (so if it’s a $1B acquisition, the billion dollars is paid in straight cash to the current shareholders) to deals where more of the purchase price is paid in stock of the acquiring company and may require the founder to stay on and “earn” some of the compensation via achieving milestones set in the transaction. It may be the case that the acquisition is the most beneficial option for you as a founder, and you get to continue to run your business almost as if it was an independent startup but with the resources of a bigger company. The current trend to note is that there is a considerable amount of antitrust scrutiny aimed at Big Tech and this has made it pretty tough for companies with lots of cash and valuable stock to use it for acquisitions.

Another aspect of the alignment goal is your employees. Chances are you can’t pay market rate salaries for early employees, so you compensate them with equity. This is a big risk on their part given the average probability of success of any given startup and the time to exit for the ones that do get to a liquidity event. Astronomical valuations give the appearance of momentum and success but the number really doesn’t mean anything to an employee unless they’re able to trade their equity for cash. Just like you can’t eat IRR, you can’t buy a house with paper wealth.

Conclusion

At the end of the day, your task as a founder is balancing the long term vision with the countless decisions you have to make everyday to make your company successful. You have to be able to see things from a 30,000 foot view and put boots on the ground so you know what’s happening in the trenches. You have to know someday you’ll trek through victory road to face the elite four but still take it one battle at a time. One level at a time.

Understanding your startup’s win conditions doesn’t mean obsessing over the endgame. It means staying aware of the terrain: what capital you take, who you align with, and how those choices shape your eventual outcomes. Whether you IPO, get acquired, or chart some less common route, the founders who win tend to be the ones who know what game they’re playing. And why. Both a live player, and a student of the game.

As always, thanks for reading and I look forward to hearing your thoughts! See you again really soon.

Appendix: Bezos Shareholder Letter, April 2005

Below is one of my favorite letters from Jeff Bezos (included in the linked tweet but also wanted to share here) where he fluently demonstrates his understanding of the importance of growing FCF/share and how it relates to Amazon’s strategy. An understanding of key financial metrics and the mechanics that drive them are essential for any founder looking to build a great business that has a model durable enough to succeed as a public company.

1. Any successful investment is by definition a refutation of the efficient market hypothesis for which you have experimental evidence.

2. Every decision at a company, at any stage, is one of resource allocation (and since a lot of it is how you spend time and we all know time ~= money, it’s essentially capital allocation too.) Thus, investing and operating, while obviously not exactly similar, *can* be viewed as two sides of the same coin. And someone successful in one domain *can* be successful in the other, though it requires a successful mapping of one side to the other and it’s by no means a guarantee (i.e. just because you’re a successful investor doesn’t mean you can step in and run a company, but a perspective from one side of the table can be an advantage for someone who wants to do both. There are many examples of people who have been good at both, would love to hear their thoughts.) The operator is making investment decisions with the company’s capital and the investor needs to understand how those drive value (ideally in a unique, differentiated way that will drive above market returns over a long time horizon.)

3. Since decision making at any stage company is one of capital and resource allocation (time ~= money) the intangibles such as - the people you hire and the velocity at which you iterate, are the biggest force multipliers (certainty at early stage and probable at any stage) that can make founders/operators successful and give the investor a lens through which to understand the company’s unique advantage (defensibility, durability, general ability to succeed and deliver above market returns)

Starting a company is an investment decision. The Efficient Marker Hypothesis says the best way to allocate your money is in a portfolio of securities that delivers the best possible returns for your given risk profile. You’re saying “No, i have information that will help me get better returns. Thus, I will allocate my time and maybe even my money towards proving this hypothesis right.” Now, the EMH is often written off as overly academic. And while I agree with that general viewpoint, I do think it’s worth understanding why it’s the prominent theory of finance for the sake of understanding what it takes to be a statistical outlier. Point 1 from above suggests that a founder starting a company believes that have information that the market either doesn’t or hasn’t fully priced in. It is a statistically insane choice to start a company, yet when you look at the Forbes list of richest people, most have attained their wealth through their equity in one company. So, the founder had information sufficient to make them believe they could beat the market by concentrating their time and money into one bet. This information asymmetry is far larger in private markets, and thus the variance of outcomes is far wider.

In public markets, companies have standard quarterly disclosures and there are laws about what information is “material non public,” upon which it is illegal for investors to trade on. In private markets, capital is allocated differently, often based on storytelling and vibes (vibe rounds or excel rounds.) Thus, the founder must convince investors they have an information advantage. And to attempt to continue bridging the gap between the academic world of EMH and the power law world of early stage venture capital, this information asymmetry + the ability to execute on it is what creates the necessary conditions for a company to be successful. Otherwise, by definition, there would be no opportunity for this kind of alpha.

The second point should naturally follow from the first. An investor able to generate alpha over a long time horizon should have a unique understanding of what differentiates successful companies from average ones and a founder able to start a successful company should have both an information asymmetry advantage and the ability to act and execute upon their unique vision. This means that every decision (for the sake of simplicity, I will use capital allocation to refer to both literally spending money and how companies spend their time) is one of capital allocation. Startups are not just “good founders, good team, good idea, big market, consistent execution and iteration” well ok they are, but they’re the cumulative result of thousands and thousands of decisions made every day. There is probably no way to formalize early stage strategy - it is probably closer to entirely based on vibes than a formal field like Decision Science, but this is the lens I believe can be helpful to people on both sides of the table (and hopefully for those who want to both invest and start/operate companies.) You need the asymmetrical information (Peter Thiel talks about how “All great companies have secrets) and the ability to execute. It is probably overdoing it to calculate return on capital for each decision you make, but eventually there is a real tangible value that can be ascribed to decision making. And these decisions certainly have a magnified impact in the early stages (key hires, go to market strategies, product decisions etc.) This can perhaps explain why so many successful venture capitalists talk about the bet in the early stage being so much about the founders. You are betting on both their unique vision based on what is hopefully an information advantage, and their ability to make the thousands of decisions necessary every day that will eventually drive outsized value and returns. That is something that’s probably impossible to put in a spreadsheet in the early days, though I really believe it shows up later on, once the value of their decision making has compounded and the company raises a spreadsheet round / goes public (guess an IPO is just another spreadsheet round lol.) One relevant metric might be revenue per employee, something that seems to be high across the best companies, even though they have vastly different business models, strategies and founders.

The third point, which is by no means a unique insight (in fact it’s probably so well known in startup culture that it’s almost a truism,) should be now quite obvious to a founder thinking of how to build an outlier business or an investor looking for ways to pattern match traits of outlier companies. Your unique vision that’s based upon what you believe to be information asymmetry drives the “what?” the “why?” and the “when?” of the business you’re starting, and your decision of how to best allocate resources drives the “how?” There are many ways to map the story of a company to their numbers - this is essentially the job of a fundamental public market investor (and a later stage private growth investor) but for a founder, the decision of who to hire and how fast to iterate on feedback are the intangibles that provide the largest possibility for positive force multipliers, especially at the early stage. While teams comprised of exceptional people that iterate quickly do not necessarily guarantee success, an investor looking for the best way to “de-risk” an early stage investment may consider the degree to which the founder has a valuable “secret” (information asymmetry they’re choosing to act upon) and then also the degree to which they can build an organization that iterates rapidly and can attract the best and brightest people. This would suggest then, that almost by definition, outlier companies (in terms of their returns) are built upon outlier cultures. And so, outlier founders are ones who consistently allocate resources in such a way that their company is able to compound its unique advantage in a way that proves their hypothesis (that their unique information + their ability can deliver above market returns over a long time horizon, the bet they made when they started their company.)

In summary, whether starting, joining or investing in a company; you need to look for information asymmetry and exceptional decision making in terms of capital allocation, especially in terms of the intangible force multipliers (iteration speed and hiring being two of the most essential/highest leverage.) Someone who understands these aspects - information asymmetry and intangible force multipliers, would give themselves a better chance as a founder/investor and could likely fluently navigate between the two roles.

There is probably a lot more to write on the matter, and this would likely have been better delivered as a well thought out coherent essay with concrete examples but best I can do is a one take apple note stream of consciousness. Do not hesitate to reply with your opinion, always curious to hear what people have to say.

TLDR:

AirBnB, Booking and Expedia are all, in their own ways, exceptional companies. They are inevitably linked to overall travel demand and thus will experience a degree of cyclicality in their earnings. When travel was ground to a halt during Covid, their businesses suffered but have since rebounded strongly. This has been great for their companies but has perhaps made it a bit harder to estimate normalized earnings. If you are looking for a way to play travel demand from a capital returns perspective, these are definitely companies worth taking a look at as they are highly levered to the overall travel environment. If you are looking for quality businesses to buy for the longer term, these are also worth your time. AirBnB perhaps contains the most upside potential as they have been innovative in unlocking new supply and new paradigm for travel for both travelers and hosts and there is still a fairly long runway for their potential growth. They do face various risks, including regulation and a slowdown in overall travel or the short term rental market, as well as a fairly rich valuation at the present moment so their potential upside comes with a fair degree of uncertainty. In terms of the modern era of tech startups though, they are a standout for many reasons - exceptional leadership (founder is still the CEO and routinely takes to Twitter to address user feedback and new product features) as well as a healthy balance sheet and fairly significant free cash flow generation. Booking is a solid balance of healthy growth, a strong balance sheet and solid operational metrics. If they are able to execute on their strategic plans over the next 3-7 years they do seem positioned to earn above market returns. Expedia lags AirBnB and Booking in operational metrics (lower total sales and worse margins) but the story seems to lie in driving more profitable growth, leveraging AI and creating a more comprehensive experience for travelers planning their trips, expanding margins via more customers starting and ending their experience on Expedia’s platform and returning capital to shareholders in the form of buybacks.

AirBnB (on a reverse DCF) is fairly valued for a 10% IRR if you believe they will continue growing pre-tax profit at 32% for the next 5 years and can sustain a ~20% EBIT margin during that time. The company has a strong balance sheet with $6.1B in net cash ($8.1B in cash against $2B in debt) another $2.8B in short term investments and $6B in funds held on behalf of guests. We believe Brian Chesky to be an exceptionally strong leader, focused both on the operating metrics of the business and the granular nuances of the company’s products that drive them. He has outlined the product strategy for the company and continues to take to social media to address user concerns and feedback. The questions surrounding the story with AirBnB are:

• To what degree will regulation dampen their revenue growth?

• How serious are some of the common concerns? These include but are not limited to: the company being increasingly levered to the short term rental market as management companies buy properties to rent them out on AirBnB. The lack of transparency surrounding fees and the final booking cost of rooms on the platform.

• Can the company continue to grow into the current valuation? There is no doubt they are an exceptional company, and perhaps one of the best of the recent era of new tech companies, but they are valued to continue growing sales over 30% per year and stabilize operating margins at around ~20%. This may cause some investors to exercise caution and potentially wait for a pullback (that could possibly coincide with a weakling travel environment.)

• How serious is the competition? Booking and Expedia both are developing alternative accommodation offerings and are looking to compete for guests with AirBnB. This may be a question of are these companies taking share from a static pie? Or is the overall market growing at such a clip that Booking and Expedia can gain share while AirBnB remains the leader and still experiences healthy growth.

With exceptional leadership, an innovative business model, a strong balance sheet and free cash flow margin, AirBnB appears poised to be the market leader in its niche in online travel for the foreseeable future. The company currently trades at 15x earnings(maybe a bit distorted due to a one time net income boost due to a tax break), 31x EV/EBITDA and 17x EV/FCF. The company has a very health balance sheet, with a strong net cash portion and is using a stock buyback program ($2.5B announced in May 2023) to somewhat offset dilution from stock based compensation.

Booking (on a reverse DCF) is fairly valued for a 10% IRR if they can continue growing pre-tax profit at 16.5% per year for the next 5 years and if they can sustain a ~30% EBIT margin over that time frame. The company has a healthy balance sheet with $13.2B in cash and $475mm in net debt. The company has strong management with a history of market leadership and strategic, accretive acquisitions that has all led to Booking becoming the market leading online travel platform that it is today. They are focused on continuing to grow their core business, driving more users to their connected trip platform and integrate generative AI into their products in a way that both delights users and increases profitability. The questions surrounding Booking include:

• How resilient can the company stay in the case of a recession and overall slowing travel demand?

• Can the company continue to maintain superior margins to Expedia? Historically, this was because Booking mainly operated in Europe and could charge higher commissions with their agency business model (whereas Expedia operated more in the US and couldn’t command as high of a take rate through their merchant model because chains like Hilton and Marriott controlled a lot of supply and therefore had more bargaining power.)

• Can Booking deliver on it’s long term vision to “make booking and experiencing travel easier, more personal and more enjoyable, while delivering better value to our traveler customers and supplier partners?” This vision includes making the customer experience better, their launch of their Connected Trip platform strives to help users book their entire trip (hotels, plane tickets and car rentals) on Booking’s platform.

• Can this goal of having users start their trip planning process on Booking’s platforms help reduce ad spend? (Booking and Expedia are heavily reliant on search ads to drive users to their respective platforms.)

• Can the company leverage AI effectively in ways that augment the user experience and truly benefit their customers and supplier partners?

The company currently trades at 20x earnings, 15x EV/EBITDA and 13.5x EV/FCF. Management is focused on profitable growth as well as retiring shares, with almost $20B left in the company’s recently authorized buyback program. The company is a strong mix of high single digit to low double digit growth, 30% operating margins, healthy balance sheet, competent management with a good balance between investing for the future and prioritizing shareholder returns. In terms of investors looking for high quality growth companies at a fairly reasonable price, Booking certainly fits this billing.

Expedia would appear overvalued (on a reverse DCF) as it would need to grow pre-tax profit at 25% over the next 5 years (vs 10 year average of 16%) to justify it’s current valuation and maintain historical EBIT margins of around 7%. The company has $5B of cash on hand with a modest net debt position of $1.2B billion. The company has lower margins than Booking due to lower take rates on their merchant model (where they buy inventory wholesale from hotels and resell it online) due to a large portion of their lodging business being US based, but they are actively expanding into Europe and the rest of the world and increasingly adding agency model business (traditionally, Expedia dominated the US market with a primarily merchant model and Booking was the leader in Europe with a primarily agency model but both businesses have been venturing into one another’s territory both on the business model front and geographically.) The questions surrounding Expedia’s story include:

• Similar to the question about Booking, can the company remain resilient during periods of weakening travel demand? This is a concern to some degree for any company that experiences cyclical trends in their business, as they must capitalize on tailwinds and build cash buffers and other mechanisms to survive the inevitable downturns.

• Can the company’s efforts to combine their various travel platforms under a single loyalty program, OneKey, lead to high customer satisfaction, increased user retention and lead to more customers starting and ending their journey on Expedia’s platforms. This is a key point of focus for investors as the company is able to earn incredible gross margins (82% average over the last 10 years) but a significant portion of operating expenses are due to the company (and the overall industry) being reliant on search ads as the primary means of acquiring customers.

• Will the company’s efforts to leverage AI (something management has highlighted on their recent earnings calls and calls their efforts “industry leading”) pay off both in the form of increased customer satisfaction and increased profits via sales growth, higher margins or (perhaps ideally) a combination fo the two?

• Does the company’s efforts to drive additional vacation rental share via VRBO prove to be materially accretive to the top and bottom lines? While on the surface it may appear this is a direct threat to AirBnB’s business model, it may be the case that the market for alternative accommodations is sufficiently large where these three companies can all have a share of an increasingly big market.

• Will management’s capital allocation efforts be sufficient to drive adequate total shareholder return? Since the company is priced more richly than Booking with similarly growth and lower margins, they will need to be smart allocators of capital in order to deliver reasonable returns for shareholders. Their recently announced $5B buyback program is evidence that management is attuned to this reality.

The company currently trades at 22x earnings, 10x EV/EBITDA and 9x EV/FCF. If they are able to drive more profitable growth through investing in initiatives like OneKey and by leveraging AI to create better, more personalized travel experiences and also create more of a reason for customers to plan their entire trip on Expedia platforms (thus lowering their search ad spend) it may make for an interesting margin expansion story.

Introduction

The online travel industry, marked by the emergence of platforms such as AirBnB, Booking.com, and Expedia, revolutionized the way consumers approach travel. Each company has a distinct history that contributed to the industry's evolution.

Booking and Expedia were both founded in 1996, Booking as a way to book hotels throughout Europe online (which made it popular initially with American consumers who were early users of the consumer internet and because at the time, you had to call hotels to book rooms which presented a challenge for Americans traveling to Europe who didn’t speak the language of the country to which they were traveling) and Expedia as an internal division of Microsoft, with the goal of creating a site to provide bookings for airline tickets, hotel reservations, car rentals, cruises, and vacation packages. The idea was to use the power of the internet to simplify the process of booking travel, providing consumers with direct access to a wide range of travel options. Global Distribution Systems existed (computerized systems, initially pioneered by airline companies, that travel agents could use to book reservations) but there were no consumer friendly front-ends for users to access these systems online. AirBnB emerged later, founded in 2007, the founders (who were broke and decided to rent out air mattresses in their apartment to conference attendees because all the hotels in San Francisco were booked) envisioned a world where anyone could rent their spare guest room or couch to interested travelers.

Fast forward to 2023 and these three companies have come to dominate the online travel industry. 

Over the last 12 months, AirBnB saw 438 million nights and experiences booked on their platform (growing 14% year over year as of last quarter and 32% over 2019, a comp that travel companies use to indicate where their business was pre-covid) with a gross booking volume of $71.3 billion (growing 17% yoy and 89% vs 2019), which translated to $9.6B in revenue and $2.25B in operating profit (23% operating margin). They have averaged 32% revenue growth since becoming a public company and that includes 2020 where sales fell by almost 30%. 

The last 12 months for Booking.com have seen over 1 billion room nights booked (growing 15% as of last quarter, healthy growth in their rental car and airline ticket segments as well, which are smaller by the numbers but becoming more important as Booking seeks to become a platform where consumers can book entire trips) and gross bookings of over $146B (growing 21% year over year on a constant currency basis) from both their agency and merchant segments. The agency business model is similar to the analogue travel agent model - the platform facilitates a trip and Booking earns a commission (this was Booking’s initial business model but recent years have seen them expanding into the merchant model which has become a very meaningful portion of their total revenue) whereas the merchant model is a wholesale transaction where the online site buys room from hotels wholesale and marks them up on their platform and earns the difference (this was Expedia’s initial model, and when you consider that they are not responsible for the inventory (if the room doesn't sell they don't pay) and don’t need to pay the hotel until 30 days after the stay (and with an average booking time of 2 months in advance, Expedia holds customer cash for 90 days, a great negative working capital cycle) you can almost see why the Expedia board passed up up on buying Booking.com in the early 2000s, more on this later.) As of last quarter, merchant and agency bookings represented 56% and 44% of Booking’s revenue respectively. The $146B of gross bookings translated into $20.6B of revenue and $6.3B of operating profit (30% operating margin.) Booking.com has averaged 16% sales growth over the last 10 years, including 2020 where they saw a 54% decrease in revenue due to the pandemic.

For Expedia, the last 12 months have seen 344 million room nights booked across their online platforms and $72.7B of gross booking volume which translated to $12.5B of revenue and $1.3B of operating profit ( ~11% operating margin.) Expedia has averaged around 16% sales growth over the last 10 years including 2020 which saw a 57% decline in revenue due to covid.

These companies are ultimately levered to travel demand which itself is dependent on the business cycle and while this can cause results to be cyclical, these companies all have strong operating models and have proven resilient over the years which has led to them emerging as the leaders in the online travel business. There has been a considerable amount of consolidation in the industry over the years which has led to greater operating leverage and more value to accrue to Booking and Expedia, as their scale creates significant barriers to entry for any would-be competitors. AirBnB’s success is due to many factors but one critical component is that they weren't necessarily directly competing with Booking and Expedia as much as they were unlocking new (and exciting!) supply entirely. Ultimately, a marketplace’s product is matching supply and demand and the companies that provide the best experience for consumers will earn the highest conversion rates.

These companies all face various risks. For AirBnB, these risks include potential changes to the innovation and lodging tax as well as potential headwinds from regulation that may require hosts to apply for a license to list their “alternative accommodation”, restricting the number of nights a unit can be rented and requiring units to meet certain safety standards. The risks for Booking and Expedia have less to do with the potential regulation AirBnB may face (to some degree, a model that is as innovative as AirBnB will inevitably face pressure from regulators, as they have disrupted the status quo) and more to do with overall travel demand and the potential for another covid level exogenous shock that grounds travel to a halt (maybe a bit more of a threat to Expedia in the scenario where they are liquidity constrained as they carry more debt and have lower margins than Booking.) Another thing to note is the degree to which Booking and Expedia need to spend marketing dollars on Google Search ads to be the first result when someone searches “hotels in Rome.” They would love for consumers to begin and end their journey on their own platforms vs starting on Google which costs both companies billions per year in advertising spend (online travel is one of the largest spenders on search ads as a category and both companies spend billions per year on these ads, which eats into their operating margin.) AirBnB has a bit of a relative advantage in this regard as their customers tend to start their searches on the AirBnB app or website. Even in the face of the various risks, if you like businesses where even if you had nearly unlimited capital to put to work, it would be hard to compete with them, these three companies are worth taking a look at. In this piece, we will explore each company’s history, recent commentary by their management teams, take a look at their financials, build a valuation model and try to get a sense of why they are uniquely positioned in the online travel industry.

AirBnB

Overview

AirBnB, Inc., together with its subsidiaries, operates a platform that enables hosts to offer stays and experiences to guests worldwide. The company's marketplace model connects hosts and guests online or through mobile devices to book spaces and experiences. It primarily offers private rooms, primary homes, or vacation homes. The company was formerly known as AirBed & Breakfast, Inc. and changed its name to AirBnB, Inc. in November 2010. AirBnB was born in 2007 when two Hosts welcomed three guests to their San Francisco home, and has since grown to over 4 million Hosts who have welcomed over 1.5 billion guest arrivals in almost every country across the globe. Every day, Hosts offer unique stays and experiences that make it possible for guests to connect with communities in a more authentic way. Whether the available space is a castle for a night, a sailboat for a week, or an apartment for a month, AirBnB is the easiest way for people to showcase these distinctive spaces to an audience of millions. By facilitating bookings and financial transactions, AirBnB makes the process of listing or booking a space effortless and efficient. With 4,500,000 listings in over 65,000 cities in 191 countries, AirBnB offers the widest variety of unique spaces for everyone, at any price point around the globe.

Key Metrics 

Why they are uniquely positioned in the online travel industry 

AirBnB's unique position in the online travel industry can be attributed to several strategic advantages and business model innovations that have allowed it to disrupt traditional accommodation sectors and maintain a competitive edge.

Peer-to-Peer Business Model:

AirBnB's platform capitalizes on the sharing economy, allowing individuals to rent out their private spaces to guests. This peer-to-peer model has fewer barriers to entry compared to traditional hotels, which require significant capital investment. It enables AirBnB to offer a vast and diverse inventory without the costs associated with owning or leasing properties.

Diverse Inventory:

The company offers a broad range of accommodations, from shared rooms to luxury villas, providing options for various customer preferences and price points. This inclusivity caters to budget travelers, experience seekers, and those looking for unique stays, expanding AirBnB's market reach.

Global Reach:

AirBnB operates globally, with listings in more than 220 countries and regions. This extensive reach allows travelers to find accommodations in virtually any destination, including areas underserved by traditional hotels, which can be a significant draw for users.

Local Experiences:

Unlike hotels, many AirBnB properties offer a local and personalized experience, which appeals to travelers seeking authenticity. This community-centric approach can lead to higher customer satisfaction and repeat business.

Technological Innovation:

AirBnB has consistently leveraged technology to improve user experience. Its platform includes features like smart pricing algorithms, search filters, and a robust review system that fosters trust and safety among users. Its investment in AI and machine learning further enhances personalization and recommendation accuracy.

Brand Strength and Community Engagement:

The AirBnB brand is synonymous with alternative lodging and experiences. Strong branding efforts have resulted in high brand equity. The company's community-driven approach, which includes host-guest interactions and a shared sense of belonging, builds loyalty and advocacy among its users.

Cost Structure:

AirBnB has a relatively lean cost structure. It doesn't own the properties listed and has lower overhead costs compared to traditional hotels. This efficiency enables competitive pricing and the ability to scale quickly in response to market demands.

Adaptability and Crisis Management:

AirBnB has demonstrated adaptability in the face of industry disruptions, including the COVID-19 pandemic. It quickly pivoted to promote local and longer-term stays to address the decline in tourism and travel restrictions, showing resilience and a capacity to innovate under pressure.

Regulatory Engagement:

Despite facing regulatory challenges, AirBnB has engaged proactively with policymakers and has worked toward finding mutually beneficial solutions, such as agreeing to collect tourist taxes and enforce rental caps in certain cities.

Sustainable and Responsible Travel:

With an increasing number of travelers prioritizing sustainable and socially responsible travel, AirBnB's model of using existing resources (people's homes) aligns well with these values. The platform encourages economic benefits for local communities by enabling hosts to generate income directly.

AirBnB's innovative business model, technological leadership, global reach, and strong brand presence have uniquely positioned it to succeed in the online travel industry. These attributes allow AirBnB to capture diverse market segments, adapt to changing consumer behaviors, and maintain a competitive edge in a dynamic market.

History 

AirBnB's history is a testament to innovation and rapid expansion within the travel industry.

Initial Concept and Founding (2007-2008):

• The AirBnB concept was born in 2007 when Joe Gebbia and Brian Chesky, then roommates in San Francisco, needed to pay their rent. They decided to rent out three airbeds in their living room and offer breakfast to their guests, which were attendees of a local design conference.

• The official founding of AirBnB occurred in August 2008, with Brian Chesky, Joe Gebbia, and Nathan Blecharczyk (the technical architect behind the site) coming together. The original website was called AirBedandBreakfast.com.

Early Growth and Funding (2009-2011):

• In the early days, the founders struggled to gain traction. They famously sold novelty cereal boxes to fund their operation during a financial crunch.

• By 2009, AirBnB had participated in Y Combinator, a startup accelerator that provided them with much-needed seed funding.

• In 2011, the company saw significant growth, with a $7.2 million Series A funding round from Greylock Partners and Sequoia Capital. In the same year, AirBnB announced its 1 millionth booking.

International Expansion and User Growth (2012-2015):

• AirBnB began expanding internationally, acquiring German clone Accoleo in 2011, and opening several international offices including in London, Paris, Barcelona, Milan, Copenhagen, Moscow, and São Paulo.

• By 2012, AirBnB was offering lodgings in 192 countries and had hit 10 million nights booked.

• The company continued to raise more funding, securing hundreds of millions of dollars, which valued the company in the billions.

Product Expansion and Branding (2016-2019):

• AirBnB diversified its offerings, introducing AirBnB Experiences in 2016, which allowed travelers to book activities hosted by locals.

• The company also underwent a rebranding, adopting a new logo in 2014 that it called the Bélo, symbolizing belonging.

Regulatory Challenges and Adaptation (2010s):

• As the company grew, it faced regulatory hurdles in various cities around the world. Cities like New York, San Francisco, and Berlin struggled with how to regulate the short-term rental market.

• AirBnB responded by cooperating with authorities, agreeing to enforce limits on how many days a year hosts can rent out their properties in certain cities and collecting tourist taxes directly.

IPO and Market Maturation (2020-2021):

• AirBnB went public in December 2020, with one of the most anticipated IPOs of the year. Despite initial concerns due to the COVID-19 pandemic's impact on travel, the company's stock soared on its first day of trading.

• As of September 2021, AirBnB had over 4 million hosts with about 5.6 million listings scattered across the globe, having served more than 1 billion guests

Throughout its history, AirBnB has not only grown in terms of listings and global reach but has also influenced the hospitality industry and local economies, often becoming a point of discussion regarding the impact of short-term rentals on housing markets and community dynamics.

Recent Management Commentary 

Brian Chesky, the CEO & Co-Founder, highlighted a record high net income (even without the tax benefit) and free cash flow during this quarter, as well as buying back $500mm of stock

We had over 113 million Nights and Experiences Booked. Revenue of $3.4 billion grew 18% year-over-year. Net income was $4.4 billion. Now this includes a onetime income tax benefit from the release of a valuation allowance of $2.8 billion. But even excluding this tax benefit, adjusted net income was $1.6 billion, our highest ever and represented an adjusted net income margin of 47%. And free cash flow for the quarter was $1.3 billion. In fact, on a trailing 12-month basis, our free cash flow was $4.2 billion, which is also our highest ever. And because of our strong cash flow and balance sheet, we repurchased over $500 million of our stock. Now during the quarter, we saw a number of positive business highlights. First, we have added nearly 1 million active listings this year. Our supply grew 19% in Q3 compared to a year ago. We once again saw double-digit supply growth across all regions with the highest growth in regions with the highest demand. Urban and nonurban supply increased at nearly the same rate, and we saw relatively similar supply growth among individual professional hosts with the majority of new listings exclusive to AirBnB. 

He also highlighted that it was a record travel season, a possible indication of where overall travel demand is at in terms of the overall business cycle

Second, Q3 was a record travel season on AirBnB. Nights and Experiences Booked grew 14% in Q3 compared to a year ago. We saw an acceleration of nights growth across all geographies, and we are particularly encouraged by the growth of first-time bookers during Q3, and we saw more nights than ever booked in the AirBnB app with 53% of gross nights booked in the app compared to 48% in the same period last year. And finally, international expansion markets are gaining momentum. Cross-border nights book increased 17% in Q3 compared to a year ago. In Asia Pacific, our business has fully recovered to pre-pandemic level. And we're seeing significant growth in Asia Pacific markets such as Taiwan, Thailand and Indonesia, all experiencing year-over-year nights growth above 30% on an origin basis.

He discussed the comapny’s commitment to making the experience better for hosts

Now we've been able to achieve these results by continually making progress on our 3 strategic priorities. First, we're making hosting mainstream. We've been focused on making hosting as popular as traveling and our Q3 results show that our approach is working. We ended the quarter with the highest number of active listing, and we saw strong active listings growth across all regions of the market types. And hosts are benefiting. During Q3 alone, AirBnB host earned more than $19 billion. We'll continue growing supply by raising awareness around hosting, making it easier to get started and improving the overall experience for a host.

He also touched on the company’s response to criticism of the product, something we believed to be evidence of his founder mindset still being present this far into AirBnB’s journey which we believe is important for AirBnB’s overall culture as well as their business goals. There are various complaints about the service, high cleaning fees and less than transparent display pricing being amongst them, and it is at least somewhat comforting for shareholders to see a CEO so involved in the iteration of the company’s product and so focused on making improvements on every level

Second, we're reflecting our core service. We've collected millions of pieces of feedback on how to improve AirBnB. And 2 years ago, we started doing twice a year of product releases to address this feedback. And since then, we've launched more than 350 new features and upgrades across our entire service. And in the past year alone, this has included things such as improved customer service, total price display and new tools to help host set more competitive prices. These upgrades are paying off for both guests and host. For example, we redesigned our tool and we made it easier for hosts to add discounts and promotions. And now almost 2/3 the host offer weekly or monthly discount. We also added a new feature called similar listings that let hosts see listing prices in the area, so they know what to charge. And since we launched the similar-listings tool, nearly 1 million hosts have used this feature. In mid-September, we shared progress we've made to help lower cleaning fees, reduce prices and improved search and reliability. We have even more improvements coming as part of our November 8 winter release next Wednesday where we'll introduce dozens of new features aimed at making AirBnB more reliable.

And finally, our third strategic priority is expand AirBnB beyond their core. Now we made significant progress in the past few years in building a strong and profitable business. And in addition to laying the foundation for new services and offerings, we've been focused on international expansion. We are investing in underpenetrated international markets, and we're seeing great results. Following the success, we've seen in recent quarters in Germany and Brazil, Korea has now become one of our fastest growing countries compared to 2019 with gross nights booked 54% higher than they were in Q3 2019 on origin basis. As international travel continues to recover, we're building greater momentum for AirBnB in underpenetrated markets. 

Financials

Valuation 

AirBnB Income Statement

AirBnB DCF Valuation

Booking.com

Overview

Booking Holdings Inc. provides travel and restaurant online reservation and related services worldwide. The company operates Booking.com, which offers online accommodation reservations; Rentalcars.com that provides online rental car reservation services; and Priceline, which offer online travel reservation services, and consumers hotel, flight, and rental car reservation services, as well as vacation packages, cruises, and hotel distribution services. It also operates Agoda that provides online accommodation reservation services, as well as flight, ground transportation and activities reservation services. In addition, the company operates KAYAK, an online meta-search service that allows consumers to search and compare travel itineraries and prices, comprising airline ticket, accommodation reservation, and rental car reservation information; and OpenTable for booking online restaurant reservations. Further, it offers travel-related insurance products, and restaurant management services to consumers, travel service providers, and restaurants; and advertising services. The company was formerly known as The Priceline Group Inc. and changed its name to Booking Holdings Inc. in February 2018. The company was founded in 1997 and is headquartered in Norwalk, Connecticut. 

Key Metrics 

Why they are uniquely positioned in the online travel industry 

Booking.com's unique positioning in the online travel industry is driven by a combination of strategic business practices, extensive market penetration, and customer-centric features.

Broad and Diverse Accommodation Listings:

Booking.com is known for offering one of the most extensive ranges of accommodation options among all online travel agencies. This includes hotels, hostels, apartments, vacation homes, and even more unique properties like treehouses and igloos, catering to a wide variety of traveler preferences and budgets.

Strong Global Presence:

The platform's extensive global presence, with a vast array of properties in almost every country, makes it a go-to option for international travelers. This global reach is crucial for capturing market share in both popular and emerging travel destinations.

User-Friendly Platform:

Booking.com has invested heavily in creating a user-friendly interface that simplifies the booking process. Features such as free cancellation, no booking fees, and a best-price guarantee have been key to attracting and retaining customers.

Strategic Marketing and Branding:

The company's aggressive marketing strategies, including performance-based and brand advertising, have significantly increased its visibility and brand recognition. Booking.com has effectively utilized online marketing channels to drive traffic and conversions, making it one of the most recognized names in online travel.

Data-Driven Personalization:

Booking.com uses data analytics and machine learning algorithms to offer personalized recommendations and dynamic pricing. These technologies enhance the user experience by tailoring options to individual user preferences and willingness to pay.

Customer-Centric Services:

The platform's focus on customer service, with 24/7 support in multiple languages, builds trust and loyalty among users. Their robust review system further aids customers in making informed decisions, contributing to repeat bookings.

Investment in Technology and Innovation:

Booking.com continues to invest in technological advancements, such as artificial intelligence and augmented reality, to improve search functionality, customer engagement, and the overall booking experience.

Strong Supply Chain Relationships:

The company has established strong relationships with property owners and managers, offering them tools and services to maximize their bookings and revenue. This supply-side focus ensures a steady inventory of accommodation options for customers.

Agile Response to Market Trends:

Booking.com has shown agility in responding to emerging market trends, such as the rise in demand for alternative accommodations and sustainable travel options, by diversifying its listings and implementing sustainability initiatives.

Competitive Pricing Strategy:

Through a competitive pricing strategy that often includes the option for free cancellation and pay-later services, Booking.com has positioned itself as a customer-friendly platform that offers flexibility and value.

Booking.com's combination of an extensive inventory, global reach, customer-focused features, and technological prowess uniquely positions it to remain a leader in the online travel industry. These strengths align well with the evolving needs and behaviors of modern travelers, providing Booking.com with a robust platform for growth and success in a competitive market.

History

Booking.com's journey to becoming one of the leading figures in the online travel industry is characterized by strategic growth, technological innovation, and market adaptation. 

Early Days (1996-1999):

• Booking.com has its roots in a small startup named Bookings.nl, founded in 1996 by Geert-Jan Bruinsma, a student from the Netherlands who wanted to connect travelers with hotels.

• The initial concept was to provide a platform that enabled Dutch hotels to offer their rooms to an online audience, and for Dutch customers to find accommodations easily.

Expansion and Rebranding (2000-2005):

• In 2000, Bookings.nl merged with another similar service, Bookings Online, which operated as Bookings.org. After this merger, the company rebranded to Booking.com.

• The newly formed Booking.com rapidly expanded its inventory and started including properties outside the Netherlands, covering a more significant part of Europe.

• In 2003, the company launched its first mobile application, an early move into the mobile space which would later become vital for the travel industry.

Acquisition by Priceline and Global Growth (2005-2010):

• In 2005, Booking.com was acquired by Priceline Group (now known as Booking Holdings) for $133 million. This acquisition provided the resources for further expansion. One could argue this is amongst the most valuable acquisitions of all time.

• Post-acquisition, Booking.com accelerated its growth, adding more properties worldwide and becoming a go-to platform for international travelers.

Technological Advancements and Market Leadership (2011-2015):

• Booking.com made significant investments in technology, improving user experience with features like instant confirmations and customer reviews.

• The platform's user-friendly interface and extensive property listings helped solidify its position as a market leader in the online accommodation booking sector.

• By the early 2010s, Booking.com had become one of the world’s largest e-commerce companies in the travel sector.

Continued Innovation and Challenges (2016-Present):

• Booking.com has continued to innovate by integrating new types of accommodations like homes, apartments, and unique stays like castles and igloos.

• The company faced and adapted to various challenges, including regulatory changes in Europe and competition from other platforms and direct hotel bookings.

• Despite these challenges, Booking.com has maintained a robust presence in the market, aided by its extensive inventory and strong customer service.

Corporate Evolution:

• Booking.com is part of Booking Holdings Inc., which also owns other travel industry giants like Priceline, Agoda, Kayak, and OpenTable.

• The company's mission has evolved to not just connect travelers with hotels but to make experiencing the world easier for everyone

Recent Management Commentary 

Glenn Fogel, President, CEO & Director of the company, spoke about a strong environment for travel demand and how Booking’s business benefitted during the third quarter:

I am encouraged by the strong results we are reporting today and by the strong leisure travel demand environment that we continue to see. In the third quarter, our traveler customers booked 276 million or more than 0.25 billion room nights. Which was an increase of 15% year-over-year and we had gross bookings of $40 billion, which was an increase of 24% year-over-year. Room night growth versus 2019 was 24% in Q3. Both room nights and gross bookings were record quarterly amounts for the company and both came in ahead of our previous expectations. Third quarter revenue of $7.3 billion grew 21% and adjusted EBITDA of $3.3 million increased 24%, both versus Q3 last year and both exceeded our prior expectations. Finally, our non-GAAP earnings per share in the quarter grew 36% year-over-year and was nearly 60 that's 6-0, 60% higher than in the third quarter of 2019. Our earnings per share growth benefited from our improved profit levels, as well as our strong capital return program, which reduced our end-of-quarter share count by 10% versus the third quarter 2022. Now turning to October. We estimate that room night growth was about 8% year-over-year and about 20% versus 2019. Excluding Israel, we estimate these growth rates would have been about 9% and 22%, respectively. We saw a significant negative impact on our business in Israel, and there was some impact on travel trends outside of Israel. 

Nevertheless, we were encouraged to see global room night growth improve towards the end of the month. Overall, we continue to see resiliency in global leisure travel demand. And as we take a very early look ahead to 2024, we see strong growth on the books for travel that will take place in the first quarter of next year. Though a high percentage of these bookings are cancelable. Given current trends, we expect customers and consumers will continue to prioritize travel over other discretionary spend in 2024.

He highlighted important initiatives that the company is investing in, including integrating AI into their product offerings (maybe relevant, Booking was mentioned on Amazon’s earnings call as a company that is building generative AI experiences on AWS)

I firmly believe we are well positioned to continue our work attracting customers and partners to our platform, while making progress on several important initiatives, which will help strengthen our business over the long term. These initiatives include: one, advancing our Connected Trip vision; two, further integrating AI technology into our offerings; three, continuing to grow alternative accommodations and four, building more direct relationships with our traveler customers. Starting with the Connected Trip. This is our long-term vision to make booking and experiencing travel easier, more personal and more enjoyable, while delivering better value to our traveler customers and supplier partners. In the third quarter, we saw an increase in the percentage of transactions, which we count as connected trips, meaning 2 or more travel components within a trip. But still a small percentage of our total transactions today, it is encouraging to see an increasing number of our travelers booking more elements of their travel with us. 

Glenn spoke about their efforts to continue growing their airline ticket business and the company’s plans to appeal the European Commission blocking their acquisition of Etraveli

Outside of accommodations, one of the most important elements of travel is Flex, and we continue to focus on further developing our flight offering on Booking.com. In the third quarter, air tickets booked increased 57% year-over-year, driven by the growth of Booking.com's flight offering. To provide some context on how this has developed over the last few years, the 9 million tickets booked on our platforms during the third quarter, were more than 5x the number of air tickets booked through us in Q3 2019. This significant growth of our flight offering at Booking.com over the last 4 years was achieved through our successful partnership with Etraveli. As previously announced, our proposed acquisition of Etraveli was blocked by the European Commission in September, a decision we will appeal. While we strongly disagree with the EC's decision to block the deal, our commitment to building the flight vertical at Booking.com has not changed. In fact, we have extended our partnership agreement with Etraveli through at least the end of 2028. Which means we anticipate continuing to work with them on improving Booking.com's flight offering over the coming years. We believe offering a compelling flight product alongside our accommodation, ground transportation and attraction offering, helps to create a better, easier and more comprehensive travel booking experience for our travelers and more opportunities for our partners. 

He also discusses the company’s efforts to build out Connected Trip, which ideally would be an AI companion that helps travelers at every step of their trip planning process. He noted ways that the company is already seeing success in their efforts to integrate Generative AI

We will continue to build out our Connected Trip vision, which we believe will ultimately result in increased customer and supplier engagement with our platform. As we discussed last quarter, we have always envisioned AI technology at the center of the Connected Trip. We have a long history of investing in AI technology and incorporating it in our platforms across our company. I previously spoke about the hard work, our team has been doing to integrate Generative AI into our offerings in innovative ways, including Priceline Generative AI travel assistant, named Penny and Booking.com's AI Tripplanner. It is still very early days. But both teams are gaining valuable insights on booker questions, concerns and behavior, as the tools continue to interact with customers. At Priceline, we're seeing some encouraging signs of lower customer service contact rates. And we're exploring other areas across our business, where we believe we can use Generative AI tools to increase productivity. For example, our brands are running projects using Generative AI to enhance the productivity of our software developers with encouraging results so far. And we look forward to using these tools more widely in the future. I remain confident in our company's ability to benefit from AI developments by improving our products for our customers and operating more efficiently over time. Turning to our supply partners. We strive to be a trusted and valuable partner for all accommodation types on our platform. 

He also noted Booking’s efforts to increase alternative accommodations on their platform, a space that is currently led by AirBnB but one that both Booking and Expedia are entering and trying to capture market share

We look to add value for our partners by delivering incremental demand and developing products and features to help support their businesses. During the quarter, some of our partners at Booking.com experienced delayed payments due to a planned upgrade to our finance and payment platforms in early July. We've now cleared the backlog of outstanding payment issues related to the system upgrade. We plan to provide compensation to partners who experienced an extended delay, and we recorded this in our Q3 results. We plan to communicate to all partners who were impacted by these payment delays within the next few days. We continue to focus on strengthening our alternative accommodation offering at Booking.com by increasing supply and raising awareness among travelers. In the third quarter, alternative accomodation room nights grew at about 24% year-over-year, which was faster than our traditional hotel category. Alternative accommodations represented about 33% of Booking.com's total likes, which is about 3 percentage points higher than Q3 2022. We are seeing continued momentum in terms of alternative accommodation, supply growth, both globally and in the U.S. with global listings reaching about 7.2 million by the end of the third quarter. Which is about 9% higher than Q3 last year. We aim to build on this progress by continuing to improve the product for our supply partners and travelers, particularly in the United States. For our travelers, we remain focused on building a better experience that leads to increasing loyalty, frequency, spend and direct relationships over time.

Financials 

Valuation 

Booking.com Income Statement

Booking.com DCF Valuation

Expedia 

Overview

Expedia Group, Inc. operates as an online travel company in the United States and internationally. The company operates through Retail, B2B, and trivago segments. Its brand portfolio includes Brand Expedia, a full-service online travel brand with localized websites; Hotels.com for marketing and distributing lodging accommodations; Vrbo, an online marketplace for the alternative accommodations; Orbitz; Travelocity; Wotif Group; CheapTickets; ebookers; Expedia; Hotwire; CarRentals.com; Classic Vacations; and Expedia Cruise. The company's brand portfolio also comprises Expedia Partner Solutions, that offers private label and co-branded products through third-party websites; and Egencia that provides travel services to businesses and corporate customers. In addition, its brand portfolio consists of Trivago, a hotel metasearch website, which send referrals to online travel companies and travel service providers from hotel metasearch websites. Further, the company provides loyalty programs, hotel accommodations and alternative accommodations, and advertising and media services. It serves leisure and corporate travelers, that includes travel agencies, tour operators, travel supplier direct websites and call centers, consolidators and wholesalers of travel products and services, online portals and search websites, travel metasearch websites, mobile travel applications, and social media websites, as well as traditional consumer ecommerce and group buying websites. The company was formerly known as Expedia, Inc. and changed its name to Expedia Group, Inc. in March 2018. Expedia Group, Inc. was founded in 1996 and is headquartered in Seattle, Washington. 

Key Metrics 

Why they are uniquely positioned in the online travel industry 

Expedia Group's unique positioning within the online travel industry can be attributed to a strategic confluence of comprehensive service offerings, technological innovation, and a diversified business model.

Diversified Portfolio of Brands:

Expedia Group encompasses a variety of well-known travel brands, including Expedia.com, Hotels.com, Hotwire, Trivago, Orbitz, Travelocity, and Vrbo, among others. This diversified portfolio allows the company to capture a wide range of consumer segments across different geographies and market niches, from budget to luxury travel, and from hotels to vacation rentals.

End-to-End Travel Services:

Unlike some of its competitors that focus solely on accommodations, Expedia offers an extensive array of travel services, including flights, car rentals, cruises, vacation packages, and activities. This full-service approach enables cross-selling and bundling strategies, creating value for customers and driving higher transaction volumes.

Advanced Technology and Data Analytics:

Expedia has made significant investments in technology to facilitate efficient and user-friendly search and booking experiences. Utilizing data analytics, the company can personalize offerings and dynamically price services, enhancing customer satisfaction and loyalty.

Strong Supplier Relationships:

Expedia has fostered strong relationships with airlines, hotels, and other travel service providers, securing competitive rates and inventory. These partnerships are crucial for providing customers with a wide range of options and ensuring pricing competitiveness.

Global Reach and Local Expertise:

With localized websites in over 70 countries, Expedia combines global scale with local presence, understanding and catering to regional travel preferences and behaviors. This localized approach is key to capturing domestic travel markets, which are substantial in many regions.

Customer Loyalty Programs:

The company's loyalty programs, such as Expedia Rewards, incentivize repeat bookings through points accumulation that can be redeemed for travel discounts. This enhances customer retention and lifetime value.

Strategic Acquisitions and Integrations:

Expedia's history of strategic acquisitions has expanded its capabilities and market reach. The integrations of acquired companies have allowed Expedia to leverage new technologies and inventory, maintain a growth trajectory, and diversify revenue streams.

Innovative Mobile Platform:

With a strong emphasis on mobile user experience, Expedia provides a seamless booking experience across devices. Given the increasing prevalence of mobile bookings in the travel industry, this is a significant competitive advantage.

Brand Recognition and Marketing Expertise:

Expedia Group has invested heavily in marketing and brand-building efforts across its portfolio, resulting in high brand recognition and trust among consumers. Effective marketing campaigns have driven user acquisition and engagement.

Financial Flexibility:

Expedia Group's financial strength affords it the flexibility to invest in growth opportunities, technology, marketing, and respond to market disruptions, as evidenced by its ability to navigate through the impacts of the COVID-19 pandemic.

Crisis Management and Adaptability:

The company's response to the COVID-19 pandemic highlighted its ability to adapt quickly to changing market conditions, by offering flexible cancellation policies and focusing on domestic travel and alternative accommodation options when international travel was limited.

Expedia's broad brand portfolio, comprehensive service offerings, technological capabilities, and global footprint uniquely position it to succeed in the online travel industry. These factors enable the company to meet diverse customer needs, adapt to market changes, and continue to expand its market share.

History 

Expedia's history is a story of technological innovation, strategic acquisitions, and industry leadership in the online travel sector. 

Founding and Microsoft Era (1996-1999):

• Expedia was created as a division of Microsoft in October 1996, marking one of the first forays of a major tech company into online travel services. The service was launched on Microsoft Network (MSN) to provide online bookings for airline tickets, hotel reservations, car rentals, cruises, and vacation packages.

• The idea was to use the power of the internet to simplify the process of booking travel, providing consumers with direct access to a wide range of travel options.

Spin-Off and Public Offering (1999-2001):

• In 1999, Expedia was spun off from Microsoft in a successful public offering, allowing it to operate as an independent company and focus more intensely on the travel industry.

• Microsoft retained a controlling interest after the spin-off but later divested its shares, giving Expedia the independence to grow its business.

Expansion and Acquisitions (2001-2005):

• In 2001, USA Networks Inc. (later renamed IAC/InterActiveCorp) acquired a controlling stake in Expedia. Under IAC's ownership, Expedia began a series of acquisitions, including Travelscape and VacationSpot, which helped it to expand its service offerings.

• Expedia continued to grow its platform and user base, becoming one of the largest online travel agencies (OTAs) in the world.

Growth as a Travel Giant (2005-Present):

• In August 2005, IAC spun off Expedia as a separate entity, which included Tripadvisor in its portfolio at the time.

• Expedia Group expanded its influence in the travel industry by acquiring a range of other travel sites, including Hotels.com, Hotwire, Travelocity, Orbitz, Trivago, and HomeAway, among others.

• The company developed a comprehensive travel platform that provided everything a traveler might need, from reviews and advice to bookings for all aspects of travel.

Technological Innovation and Market Leadership (2010s):

• Expedia invested heavily in technology to refine its search capabilities, making it easier for consumers to find and book the travel options that best suited their preferences.

• The company also embraced mobile technology, developing apps and mobile-optimized services that catered to the growing number of travelers using smartphones and tablets to plan and book their travels.

Recent Developments and COVID-19 Impact (2020-2021):

• Like many travel companies, Expedia faced challenges due to the COVID-19 pandemic, which caused a significant downturn in travel. However, the company has been working on strategies to adapt to the changing landscape and prepare for the recovery of the travel industry.

• Expedia has continued to invest in its platform, focusing on flexibility and user experience to remain competitive and meet the evolving needs of travelers.

Throughout its history, Expedia has been at the forefront of the online travel industry, continuously evolving to meet the changing demands of the market. As an investment analyst, understanding Expedia's adaptability, technological prowess, and strategic acquisitions provides valuable insight into its market position and potential for future growth.

Recent Management Commentary 

Peter Kern, Vice Chairman & CEO, highlighted a strong overall demand for travel during the third quarter, consistent with the tone in both AirBnB and Booking’s earnings calls. He also talked about finishing migrating VRBO (Expedia’s alternative accommodation platform) to their single front end stack which leaves the company strongly positioned to focus on accelerating growth vs rebuilding their tech.

Our third quarter results came in ahead of our expectations with record revenue and EBITDA. This was particularly gratifying, considering the fires in Maui, which had a disproportionate impact on our Vrbo business and put pressure on the top line overall. Travel demand otherwise remained solid with broad trends consistent over the last few months. North America and Europe demand remained stable with more pronounced growth in APAC and Latin America. Prices also remained stable by and large. Hotel and Vrbo ADRs are holding up in each region, but mix effects are leading to a slight year-over-year decline in overall lodging ADRs. Conversely, we have seen some modest price pressure in air and car. We are also keeping a close eye on the escalating violence in the Middle East, which appeared to have some impact on global travel in early October. More relevant to our specific performance, I'm happy to share that we've just completed the final leg of our Vrbo migration onto our single front-end stack with the conclusion of our global launch of the new Vrbo app this past Monday in the U.S. This marks the last of our major migrations associated with our multiyear transformation. It has been a long, complex journey, but well worth the effort as we are now in a position to accelerate into the future without the drag of transformation work that forced us to go backwards in order to go forwards. 

He also spoke about Expedia’s AI efforts and their OneKey rewards program, which is a way for users to earn points across all of Expedia’s online platforms

We are now in a position to dramatically increase our test-and-learn capacity and feature release velocity while also providing a scalable and efficient base to operate upon. We continue to utilize our industry-leading AI and ML capabilities to radically improve all aspects of our traveler experience. And with the launch of One Key and our increasing ability to understand the long-term value of our travelers, we can now begin to drive faster, more profitable growth. Moving on to the key pillars of our performance, starting with our category-leading B2B business, which remains on track for a strong year with Q3 revenue growing 26% versus last year. We're winning new deals, increasing wallet share with existing partners, and launching new products and features to support our growth. Demand from China, in particular, continues to pick up with Q3 bookings from China Partners up over 150% year-on-year. We anticipate continued strength from B2B going forward, driven by our continuing push into the addressable market along with the advantages that our platform improvements will bring to the B2B business, whether in core technology, the application of AI and machine learning or in service and payments. 

As we unify stacks, this will also further enhance the capabilities on offer for our B2B partners. But as pleased as I am with the continued growth of our B2B business, I'm even happier to see our B2C business picking up momentum with year-over-year revenue growth in Q3, accelerating over 400 basis points sequentially. This is what we've all been working so hard for, so it's very gratifying to see these results beginning to improve. Another major milestone for us was the U.S. launch in July of One Key, our new loyalty program. One Key unifies our major brands of Expedia, Hotels.com and Vrbo, allowing our members to earn and burn one simple currency, OneKeyCash across our vast marketplace. As I've mentioned before, getting dividends from a program like this will take some time, given the frequency with which most consumers travel. That said, we are very happy with the early results and traction One Key has with our members. We have already migrated over 82 million members to the program. And with the addition of Vrbo to the mix, we have seen 34% growth in new members over the last year. We have already seen many members using OneKeyCash across brands, including on Vrbo and have been pleased that Hotels.com members have not been unduly impacted and have already been using OneKeyCash to shop for other products on Expedia. 

He discussed the company’s efforts to increase the percentage of bookings that occur within the company’s apps (which is a long term goal to reduce google ad spend and improve operating margins.) Expedia has a goal of making the experience of planning group travel more seamless and also believes that the launch of their OneKey platform and how they are leveraging AI will lead to higher consumer satisfaction, more return customers and therefore improved operating metrics for years to come.

Overall, these promising results have given us solid learnings that will be useful as we launch One Key in other countries next year. And with the Vrbo migration complete, we can now more fully lean into the core differentiation that One Key gives Vrbo in the vacation rental space. One Key, along with our ongoing efforts to more generally attract higher lifetime value customers, is accelerating our mix of loyalty members, app users and app members. And the percentage of bookings coming through our apps continues to grow and was up approximately 300 basis points sequentially in the third quarter, which ultimately contributed to our year-over-year marketing leverage in our B2C business for the third quarter. We have also been releasing exciting products and features that remove more and more friction from the planning and booking process. In September, we announced our Fall Release, showcasing a series of new features and products, squarely aimed at solving complex traveler problems and enhancing engagement. In case you missed it, I'll give you a few of the highlights. We have simplified group travel planning, providing for the first time one place for friends and family to collaborate on a group trip to see one another selections and add saved options across air lodging, car rentals and activities, significantly using the group planning process and creating a better and more successful trip for everyone. 

Products like these not only enhance the consumer experience, but allow us to utilize a consumer's own network of friends and family to expand our reach. We've also launched tools to aid research into a given destination, hotel prices, weather, best times to visit, crowd levels and even generative AI-powered tools to determine the best neighborhoods to stay, giving the traveler one place to research where and when to stay in their dream location. We now also leverage generative AI to scrape reviews to answer traveler questions about amenities and property details, so no more sorting through hundreds of reviews to find out how strong the WiFi is, the quality of the pool or whether you're going to like the breakfast. We also launched the first project of its kind, EG Labs, which allows interested customers to test our AI-powered beta products, allowing them to play a hand in how we shape the future of travel. I could literally go on and on, but the takeaway is no one in travel is innovating faster than us and with so much important platform work behind us and with our leading capabilities and machine learning and AI, we will out-innovate in the space for many years to come. 

The company is focused on faster and more profitable growth, as well as their high level strategy of providing the best possible product for their users.

We have literally worked for years and given up many short-term opportunities to get to this place, and I don't believe anyone is in a better spot technologically, which is not only exciting for our existing business but sets us up to go back on offense in many more markets next year. In closing, I'm pleased to see our solid execution in Q3 and through 2023 so far. While the geographic mix of business distorts global growth rates, we believe that we have held up or grown hotel gross booking share in virtually all of our key markets and with Vrbo finally completing its migration and the key differentiator of One Key, we expect our vacation rental share to improve going forward. In addition, we are finalizing our plans for '24, where we expect to drive faster and more profitable growth. Our high-level strategy is not going to change, best product, best loyalty program, best marketplace and best service. But instead of spending most of the year doing surgery on our own business, we will be focused on growth, innovation and efficiency. I'm excited for 2024 and beyond and for what we will bring to our travelers, partners and shareholders

Financials 

Valuation 

Expedia Income Statement

Expedia DCF Valuation

Conclusion

In conclusion, the online travel industry has come to be dominated by a small group of companies that can leverage their massive scale to provide the best experiences for customers who are looking to book travel plans online. AirBnB unlocked new supply with their innovative marketplace model for guests and hosts and has created and captured a lot of value as a company. They appear well positioned for the future but are also priced to continue growing at very high rates over the coming years. With exceptional leadership, a strong balance sheet and a combination of improving operational metrics and a thoughtful capital allocation strategy, the company is certainly one that investors looking for quality growth should place on their watchlist. Booking is perhaps the most solid of the three - healthy growth, a good balance sheet and strong margins. If they can execute on managements vision for the near to mid term future, shareholders can likely expect above market returns from the current price. Expedia appears ok on growth, but lags on margins vs AirBnB and Booking and will need to increase profitability or return capital to shareholders via buybacks for the current valuation to make sense. If Expedia can drive more users to their platforms (vs Google) and leverage AI (something AirBnB and Booking are also actively investing in) there may be a possibility for a margin expansion story to unfold. If you are looking for a way to play travel demand (you expect travel to reaccelerate, for instance) these companies are the way to go as they are all inevitably levered to the overall level of travel demand.

These three companies are all great examples of businesses that leveraged the internet to provide better services to consumers while also creating significant value for shareholders in the process. In terms of companies where it would be difficult to compete with even if you had unlimited capital to do so, these three are definitely in that category as they leverage the network effects from their massive scales to continue growing year over year. While AI may enable new ways of startups attempting to enter the online travel market, these companies are actively investing in the technology as well and in general, we believe it would be it would be extremely difficult for a new entrant to compete in the online travel business (there may be a larger threat from existing large internet companies making a move into the space but it would be tough for a new startup to directly compete.)

Thats all for now. If you enjoyed this piece, consider sharing it and subscribing and please feel free to reach out on Twitter (@netcapgirl) with any thoughts or feedback.

He talks about how important it is to be mentored by a great investor. His professor at UC Santa Barbara (who received a PhD from Stanford for their work on the business models of public companies in the post great depression era) took him on as a research assistant & had him reading S-1s and talking to CEOs of companies the professor was considering investing in. He joked that although he was technically a research assistant, he was essentially the professors investment analyst.

He stresses the importance of focusing on business model for understanding durability of the business. This is a recurring theme in his strategy and makes sense when you consider 1. how concentrated the IGSB portfolio is 2. his work for the UCSB professor who had done extensive work on business models and the companies that successfully navigated the era in the aftermath of the great depression.

He talks about how there's synergy between investing in public & private companies (something he did early in his career) when viewed through the lens of evaluating business models and trying to understand "is this an exceptional company?" & "what are some of barriers to competing with them?" Which are questions he’s always asking and doing lots of deep work in his effort to understand if the company clears the high bar to be included in the portfolio.

"If you wanted to focus on reducing risk, you diversified. if you wanted to focus on knowing your investments better than anyone else, you have a limited amount of time and thus need to focus on certain business models and certain companies" (The fund usually has 70-80% of assets in 3-4 companies which is certainly a highly concentrated strategy, but it's sort of more akin to thinking of portfolio construction as a collection of operating businesses and your company or portfolio is going to be the aggregate result of your stake in each one you own.)

"It’s hard to be an expert in 20 companies, it became obvious that concentration was going to be our key to returns" the pushback against this is that you take on unnecessary risk because if one company blows up it's a big hit but he never took outside money so that was an advantage and allowed him to focus on doing what he felt was best to generate returns. This has worked for Reece as the fund has been in business over 54 years.

He discusses the difference between being an investor & being in the investment business.

He never took outside money and talks about how the downside of not having outside capital was needing to pay expenses, salaries, taxes etc. all out of pocket but upside was being able to sculpt the strategy in what they felt was the best way to maximize returns (concentration

investing in both public & private companies etc.)

He realized a lot of outsider investors would not want the "risk" or volatility that comes with being so concentrated & investing in a mix of public & private but the way that Reece defines risk is "not knowing your companies" as opposed to not being diversified

How they've built a fund that's now in its 6th decade of operations. Reece talks about how there's lot of ways to make it work in investing business and how they refined what worked for them over the years.

By end of their 2nd decade in business they started to focus on software and eventually came to really focus on B2B vertical market SaaS.

They decided to take a concentrated approach to their private investments as well & wanted to own around 30% of the private businesses they backed. They also were willing to hold the company for a long time because they were so close to it, understood it and felt it was "de risked.” This is opposed to the portfolio strategy of traditional venture capital firms which may have 25-30 companies or more in one of their funds.

In fact, they would often buy shares at each round when the company was private, invest in the IPO & continue building their stake after the company went public (a recent example of this was their continued investment in AppFolio.)

They felt it was better to let money continue compounding in great companies that they knew well rather than sell somewhat arbitrarily after 10 years, have to pay taxes and then need to redeploy the capital.

There are only so many exceptional companies. He refers to a study done on companies from 1926-2016 that looked at 26,000 companies and only 1000 of them produced returns in excess of treasury bills

He describes the companies that are exceptional & stay that way for decades as having customers that are fanatically reliant on what you deliver to them, whether it's a consumer product or business product. He talks about they spend time talking to customers help understand the qualitative advantages the company has, especially in terms product market fit & go to market strategies.

“You’re in control of your own destiny.• Reece talks about how his father passed away when he was 14 & how much the time he spent with him during last his 6 months shaped his life.

He wants to do everything he can to keep everything as simple as possible. Most people needlessly add complexity to investing & to life. This is a core tenet of his philosophy. He also talks about the importance of mentors.

His goal was to wake up in the morning & love what he was doing and who he was working with more than it was ever about money. He spends a good portion of the year in Rome, where he is able to focus & read in peace.

His key metrics were "am I learning?" & "am I having fun?" and he figured if he could say yes to both of those that eyentually the IRR would take care of itself.

He talks about having a loving family, a simple life & amazing mentors and how they all helped him succeed. The interview is absolute gem and definitely worth a watch. I believe it’s one of if not his first public interview ever. There’s an incredible amount of wisdom in these 60 minutes and I’m happy he decided to share his thoughts with the world and think that any investor could take something valuable away from watching it.

At a high level, there are three buckets of work one might be doing at any given time:

• Monitoring and Reviewing Longs

• Reviewing Watchlist

• Finding New Ideas

Monitoring and Reviewing Longs

These are companies you already own so ideally you have a good sense of the story and the key drivers of the business and the stock. This bucket involve monitoring news, reviewing earnings report and transcripts from the quarterly management calls, as well as transcripts from any conferences that management attends. In terms of the decision tree for a core long (and this is different for every investor as everyone has different time horizons) it boils down to:

1. The story is pretty much the same, I’m comfortable with the numbers and managements vision so doing nothing is the appropriate course of action at the present time.

2. The story is deteriorating, I’m not comfortable with the direction that management is taking the company or the guidance they’re providing for the future. Maybe the company’s moat is not as strong as evidenced by contracting margins or slower than guided revenue growth. Maybe management is demonstrating poor capital allocation decisions by spending recklessly on M&A, overpaying for non core assets or companies with little synergy to the existing business. Maybe growth or margins or some other KPI you measured in your initial work is just not going to hit your targets and this is the case where you need to take a sober look at the position and reassess how comfortable you are with the story as it stands today. A stock with a deteriorating underlying business or one that generally doesn’t match your initial assumptions for negative reasons is a good candidate to sell. Whether you sell out of the entire position or reduce the exposure depends on the specific company, but in the context of monitoring your longs, these are the sort of things you want to look out for in terms of making the decision to continue holding the stock.

3. The story is improving or exceeding expectations. This is usually good news because it means the company is outperforming the expectations in your original thesis. What will ultimately determine your actions is your reassessment of the story now that the market has had time to digest to the company’s outperformance. If the market is slow to react to the news, it may be the case that the stock remains undervalued even in the wake of the news.

An example might be a company that is experiencing accelerating revenue growth because of a secular trend. The company reports earnings and the stock goes up as a result of the market learning the positive news. The company rerates from a 10x multiple to a 20x multiple in a day. If you’re able to look at the results and say well it might not be as cheap as before, but i’m still confident this is 15x ntm earnings and maybe 9-10x in two years, it might be a position you want to add to.

This is psychologically difficult.

You’re averaging up. You’re increasing your cost basis and diminishing the return you’ve already earned. It is of the utmost importance to take a cold and sober look at the business’ trajectory and how it lines up with your initial thesis and assumptions. In my experience I have found it more difficult to average up winners than to cut losers, but it can often be the case that good companies tend to compound their competitive advantage and that is why it is so important to not let the psychological difficulty of adding to a position that’s going up get in the way of building a successful position in a great company by averaging up on a winner.

It’s not to say that every time a company exceeds expectations it is a time to buy more stock. The second possibility in this scenario is that while yes that the business exceeded expectations this quarter, on a normalized basis the performance is still just above the upper end of your assumptions and based on the market reaction, you believe the stock to be approximately fairly valued and so you’re comfortable holding the position as is, not adding or reducing.

The third possible case is the market drastically over reacts to the better than expected performance and possibly pulls forward a lot of your expected future returns in a much shorter time frame. I can choose to sell out of the entire position because of course there’s an opportunity cost to keeping money with any one company and now that I’ve achieved my expected return (albeit over a shorter time horizon) I need to find a better use for that capital. Or it can be the case that one can reduce the position to a smaller percentage of their portfolio and see the position as “house money” if it was able to return a portion or even all of your initial investment. One other possible variation of this scenario is that the market barely reacts to a company exceeding expectations. This may be the opportunity to continue building a position in a company I view as a winner but it may also be a sign that the market views the company as more cyclical and is assigning the stock a lower multiple because it believes eventually earnings will come back to a more normalized state.

This is an overly simplistic description of managing core longs and every portfolio manager will have different approaches to constructing their portfolio. Some favor a more concentrated approach (8-10 stocks, even 15 is pretty concentrated) while others will be more diversified. Some may have less constraints around position sizing while others may have a strict 5% or 10% of nav limit meaning that any time a position exceeds that limit they’ll reduce it to be in line with their risk management protocol. Tax considerations are of course relevant when thinking of selling a position as short term gains are taxed differently than long term gains.

In summary, it’s important to stay on top of the longs in your portfolio and be monitoring for inflection points where you may want to add to the position, reduce your exposure or sell out if it entirely. Although it’s painful to watch a stock you sold continue to go up, one is likely to regret averaging down on a loser more than selling out of a future winner. As always, investing is as much about the in the moment psychology of making decisions with limited information as it is doing the research and modeling to inform the decision making process.

Reviewing Watchlist and Starter Positions

This part varies from fund to fund and from pm to pm but at a high level it is about monitoring and reviewing the companies on your watchlist, which you may or may not have built a “starter position” in (a position that is probably significantly smaller than a core long, bought because of your interest during the research process but sized smaller perhaps because of reservations about short term performance or current valuation. Or because you want to be Druck and you’re following his “invest first, research after” approach.)

Whatever the reason for starting a position in a company, the decision tree is somewhat similar to monitoring and reviewing your core longs. If you only did a bit of research before buying a starter it may make sense to spend some time really digging in and getting to know the company. Again, everyone will have different parameters for what would constitute the appropriate conditions for adding to a starter so it becomes a core long or selling the position but what is important is that you understand yourself and you know when you know enough to make that decision one way or another.

In the second case, where you have added a company to your watchlist but haven’t started a position, the work to be done will vary depending on why you added the stock to your watchlist. If you are someone who only builds a watchlist when you have done a reasonable amount of work on a company, you might have a price target on the stock at which you’d be more comfortable buying. You might also be concerned about cyclicality or seasonality in the business and want to get more clarity around their trajectory by waiting to hear their next earnings call.

This is another thing that can be psychologically challenging.

It can be tempting to watch a stock on your watchlist go up and say “well I missed that one” but as previously discussed, it’s important to remain detached psychologically and assess the situation from a sober point of view. It may be the case that the underlying business is inflecting and the recent run up is only the start of a secular acceleration for the company. And if that’s the case, you want to do the work to try to understand that so you can get comfortable buying something on your watchlist after it’s already ran up.

This is definitely not easy in practice as for some reason the mind seems to anchor to the price at which you added the stock to your watchlist.

At the same time, if you have not done any real work on a company, seeing a stock go up in the short term isn’t necessarily a reason to just ape into it without any real work. This brings me to perhaps one of the most important points of this piece: it’s important to have your own criteria for decision making. You have to know yourself, your biases and how you operate and understand the criteria by which you’ll feel comfortable making a decision one way or another. Because at the end of the day, your process is one of your most important assets and building one that is repeatable (and one you can trust) is essential to longevity in investing. That’s why it’s crucial to understand what works best for you and what information you need in order to make the decision. Doing post mortems can be helpful to assess if you stayed true to your process. For instance, say you sold meta after their high capex guide, massive headcount ramp through covid and stated metaverse ambitions. Seeing Meta’s run up in 2023 can be hard to stomach but ultimately it comes down to did you make the decision for the right reasons? If so, you can trust your process moving forward and understand that there’s going to be times where you have losers. Ultimately limiting them is your long term goal and understanding yourself well enough to have a repeatable process that you trust is as well. If the findings of your post mortem is “well I sold Meta because everyone on Twitter was saying Zuck lost the plot” that’s maybe not as strong a reason in terms of developing your own unique process.

What works for some people might not work for you and that’s fine, I think the most important thing is iterating and understating the core tenets of your decision making process to set yourself up for success in the future. Some people can spend a few days to a week looking at a company and get comfortable enough to add it to their watchlist and buy it after good results. Others will need more time after the initial decision to add it to your watchlist. Personally, I have 3 sections of my watchlist:

1. I have done the work, like the company but want to see the next quarter or two results before making it a core long. This is when I’ll buy a starter position.

2. I have done the work but can’t get comfortable with valuation or some other aspect of the story but the company is interesting, so add it to watchlist A. In some cases this may also merit a starter position, as there is always the possibility that even though a stock appears “expensive” on current valuation metrics, it’s actually relatively fairly valued or even cheap based on its projected growth and margins over the next few years.

3. I haven’t done enough work but for whatever reason came across the company and something about the story sounds intriguing. These get added to watchlist B where I still need to ramp up on the name in order to have any sense if I’d ever want to own it in some capacity.

This is of course different for every fund, but having sections of your watchlist like this can help when you’re looking to prioritize work for the day or when you have some free time because you know the next company you need to look into.

Finding New Ideas

This is probably the most vast and open ended of the three, as inevitably the companies you own and have on your watchlist are the result of your idea generation process. There are a number of ways to go about this and as is true with investing, it will be different for every fund and may even differ from person to person at the same fund.

In some cases, your scope may be narrowed by your investment mandate. For instance, a growth fund may screen based on sales or earnings growth and derive their investable universe that way or a fund with a specific sector focus such as tech may limit its coverage universe to tech stocks. In cases such as these, it makes sense to follow Druck’s wisdom of “the best way to learn about an industry is to study every company in it.” This might not be quite as possible for those who are more generalist investors but it still isn’t a bad strategy, even if you examine each company at a relatively high level. This is also where having a strong network can be beneficial as talking to someone with experience covering the industry can be helpful during the process of ramping up.

But let’s say you’re not sure where to start finding new companies, where are places you can look for potential ideas? One that is viable if you already have a portfolio of longs and a watchlist is to look into the supply chain of some of your best longs. The simple reason being that companies focused on growth are going to be investing in the future, whether through the income statement or the balance sheet, and the companies that supply them are going to be downstream beneficiaries of their increased investment.

For instance, if you owned AMD or Nvidia and were interested in exploring who puts together servers from their chips and then ultimately sells it to data center customers, this process might have led you to Super Micro Computer. Or if you owned Taiwan Semiconductor and were interested in what equipment they buy in order to make semiconductors, this process may have led you to ASML or other providers of semiconductor capital equipment.

This isn’t to say it’s a sure thing to find companies that are downstream of ones you are already long and it’s often the case that a supplier or vendor has economics that vary to different degrees from the company they are working with, but it’s at least somewhere to start. Companies are required to disclose if one customer is 10% or more of their revenue so this can make an interesting place for research, and while they often won’t disclose who their big customers are, with some additional research there is a possibility that one can narrow it down.

One of the main takeaways here is to spend a lot of time reading and keeping notes. In my own process, I start a google doc for a company or a trend that I’m researching and add notes from what I find as I’m doing the research. This allows me to come back to the doc to review things and also can serve as inspiration when Im looking for a new idea. Again, management won’t always give away exactly who their big customers or suppliers are, but between 10ks, earnings calls, investor presentations and conference talks one can get a decent sense of who it might be. I once tweeted something that was along the lines of “one person’s capex is another person’s revenue” in reference to Mark Zuckerberg increasing Meta’s capex and buying GPUs with the capital, and this idea is along those lines. There’s really no substitute to doing the work yourself and reading 10Ks, earnings call transcripts and investor presentations released by the company and though it might sound surprising, there can be alpha found in taking the time to actually read all the relevant documents and materials.

The essence here is that there’s often breadcrumbs that are worth following when a company is growing and investing in the future, and diligent research here can be a fruitful source of idea generation.

The next suggestion is not necessarily easy to navigate because of the complexity and potential difficulties in determining the real from the hype, but doing your best to understand near to mid term thematic trends can be profitable if you are indeed able to parse the companies that are truly structural beneficiaries of whatever the trend is from the companies that attempt to build hype when they don’t benefit or may even be a loser as their industry undergoes a secular shift in the future. Longer term trends are great too but by their nature will take more time to play out and it is more difficult to determine who will benefit and how the market will price the likelihood of certain companies being winners or losers.

The most obvious examples from recent memory are AI and GLP-1 drugs. Both of these themes are still in the early stages of actually playing out in reality but based on moves in the market, investors feel these will inevitably create new winners and losers in their respective industries. In information technology more specifically, investors are familiar with the disruption caused by secular trends such as cloud and mobile and even the internet in general. While it’s become a bit of a meme, “secular tailwinds” can truly create new business models and even new industries entirely. I think there’s some hindsight bias in the sense that investors look back and think “wow all you had to do was get mobile right and you would’ve massively outperformed.” Which is not exactly true, companies such as Ericsson and Nokia may have looked like promising beneficiaries but have lagged indices in recent years and Blackberry and Palm Pilot, once industry leaders, are pivoting and out of business respectively. Microsoft missed mobile but now cloud is a major aspect of their business.

It is not enough to be early in noticing a trend, you have to really dig in and understand the businesses within the industry to get a sense of who may emerge as the winners and losers. There is of course the meta game of trying to predict who the rest of the market will crown the winners or losers over the next 6-18/24 months but since this piece is written from the perspective of someone looking to invest in quality businesses I’ll stay mostly focused on that.

Being able to parse reality from hype is certainly a crucial skill for any investor but it is especially relevant when you’re looking at a larger thematic trend and trying to find companies poised to benefit. An example of this research process can be found in these two pieces from (I helped with the AI losers piece and wrote a small part about luxury retail being a potential beneficiary in the GLP-1 piece.)

Citrini Research

Less Deus, More Machina

Preface As discussed in my piece on how to play AI beneficiaries (which you should read first here), I laid out the case that the best risk reward in my triphasic model came in the form of data center…

Read more

3 years ago · 44 likes · 16 comments · Citrini and Sophie

Citrini Research

Upgrading from Overweight: The Effects of GLP-1 Drugs on the Investment Landscape

Upgrading from Overweight: The Effects of GLP-1 Drugs on the Investment Landscape The Questions: GLP-1 Drugs and Obesity If you haven’t, please read my article on Thematic Investing & Megatrends here fi…

Read more

3 years ago · 61 likes · 22 comments · Citrini

These were useful exercises in trying to understand a thematic trend and the companies poised to benefit from the potential secular tailwinds . There are countless other ways of finding new ideas and investing is a practice that rewards the diligent and curious mind. An investor who truly loves the game will enjoy the opportunity to learn about new ideas, regardless of if it leads to an actionable investment. Also, the benefit of constantly remaining curious is building a larger mental library of analogies from which to analyze future investments.

The more ideas and stories you’ve looked at, the better positioned you’ll be to evaluate future investments because you have a larger repository from which to draw analogies.

You can look at something and quickly understand ok that’s an accelerating revenue story or that’s an incremental margin story, similar to the ones you’ve looked at in the past. While no two companies will ever be exactly alike, the larger your bag of experiences the better equipped you’ll be to evaluate the next one you find.

Two other ways of potentially finding new ideas are through stock screeners and reading quarterly letters of funds whom you respect. Stock screeners are a great way to look for companies who fit some criteria but you may have not come across otherwise and fund letter ideas have the advantage of being vetted by someone whom you already respect, though it is of course not sufficient to buy something because someone wrote about it in their quarterly letter, it’s of vital importance that you do your own work and to reiterate an earlier point, you develop and firmly understand the criteria that is necessary for your own decision making process. But with the main goal of the idea generation work being to generate new ideas (lol) these are both good places to start and may occasionally lead you to venture outside of your comfortable universe and learn about something new. This is also something that investors can do well to varying degrees and it’s important to know yourself when approaching this kind of idea generation. Warren Buffett is on record numerous times recommending that investors stick within their “circle of competence” which is perhaps ironic because when one looks at the variety of investments the Berkshire chairman has made over his storied career it’s hard to fit them all within a niche, but the advice is certainly sound as a good precaution. Every industry has its own nuances and while it may be possible for a tech investor to get up to speed on a consumer or retail idea, wading into territories like energy or financials may prove exceptionally difficult for someone who is more of a generalist. That’s not to say one shouldn’t be open to ideas in any industry, there would be no way of learning about them if you kept a close mind, it’s only to continue to reiterate the point of knowing yourself and reinforce the idea that’s central to this piece: understand your own decision making criteria and when you are comfortable making a decision. Over the long run, you are only going to be as good as your repeatable process.

My Own Examples

I have tried to create somewhat of a formal process for my idea write ups by way of a simple template, I am still working on a similar template for my thematic write ups.

An example of a single name write up that follows this template can be found here with my piece about Nintendo.

An example where I looked at an overall industry through the lens of some of the best companies can be found here with my piece about luxury retail

I am always working to improve my process. For single name idea write ups, I’m always trying my best to understand the story and be able to distill in as few words and numbers as possible. Understanding narrative is crucial in the idea generation process and being able to understand how the numbers match up to the story (and vice versa) is one of the most important intangible skills for an investor to build and something I am always actively working on.

The basic template (found in the Nintendo piece and below) centers around the story and all the work being done is an attempt of finding the “inevitable” underlying objective reality through available information.

• Introduction

• TLDR on the Story (Bull and Bear cases)

• Key Metrics & KPIs

• Financials

• Commentary from Management

• Valuation

• How numbers are key drivers of the story

• Assumptions

• Interesting points/Considerations

• Risks/Bear Case

• Conclusion

This is the TLDR from the Nintendo piece, to give an example of what I’m talking about

Nintendo appears cheap on valuation metrics because the market views them as having a hit-driven, cyclical business model where they need to reestablish their user base every time they launch new hardware. The market seems to think we’ve reached “peak Switch” meaning that sales of the popular gaming console, the Nintendo Switch, have peaked and revenue is likely to decline in coming years, which will put pressure on margins and depress earnings. The market also doesn’t place much value on the ~$13B of cash & short term investments on their balance sheet, as management has conveyed their intent to build a cash buffer as a potential safeguard against things going wrong in the future (a failed hardware launch, for instance, could be expensive and hurt the company economically, which is why - in management’s opinion, it is prudent to hold a large amount of cash). If you had to pay a multiple of earnings for a given earnings stream, would you prefer one that is stable and experiences low but steady growth? Or one that is dependent on major product launches, and will have rather “lumpy” sales cycles where the first few years after a successful launch see dramatic revenue growth, only for that growth to taper off as everyone who wants a device eventually wants one? The market, as it stands today, does not seem willing to award Nintendo a high multiple of earnings, which seems to reflect a view that the company will always be chasing their next hardware launch and it’s financial success will depend it’s ability to reestablish its user base with the next “hit console”. The bull case is that Nintendo is evolving as a company (and Japan is simultaneously evolving its corporate governance laws to be more aligned with shareholders) and shifting towards a stabler revenue mix where a greater percentage of their sales comes from higher margin opportunities like software games (selling the games digitally through their eshop vs a physical retailer) and subscription revenue from their online service offering, Nintendo Store Online.

If you’re interested, I also wrote up thematic pieces about Electric Vehicles and The Future of Networking & Telecom as well as a piece about a trend I called “Hyper Individualization.”

I also wrote a piece about Arista Networks and did a piece with about Celestica and an earnings review where I looked at the earnings reports of some companies in my coverage universe.

I’m always trying to get a sense of the story behind the current numbers and valuation and how to best interpret management’s commentary and guidance as it relates to their vision for the company. As a great account once pointed out on Twitter, business analysis and modeling are table stakes, your job is ultimately to synthesize the story in a way that makes you money. This is of course obvious to some of my readers but hopefully helpful in the overall context of this piece about process, understanding the story and building a somewhat repeatable process that you can trust to make money in the long run

Conclusion

I hope this piece was helpful as it pertains to your own process for monitoring the companies you already own and finding new ideas. In summary, this piece covered some core ideas related to -

Monitoring and Reviewing Longs: This involves continuous assessment of existing investments, evaluating whether the company's story is consistent, and deciding to hold, reduce, or sell based on the evolving situation. This process includes analyzing earnings reports, management calls, and market reactions to performance. As well as updating your model to reflect new information.

Reviewing Watchlist and Starter Positions: There are various challenges in managing a watchlist, especially when stocks on the list start performing well. I wanted to emphasize the importance of understanding your own personal decision-making criteria, being patient, and avoiding impulsive decisions.

Finding New Ideas: There are many different ways to generate new investment ideas, including exploring supply chains of current holdings, understanding thematic trends, reading fund letters, and utilizing stock screeners. I wanted to again stress the importance of staying curious, building a mental library of analogies, and continuously learning. There is no perfect way to generate new ideas, so keeping an open mind and following your curiosity can lead you to ideas in places you may have never looked otherwise.

Hopefully the examples of my own research process provided some insight into how I put these ideas into practice and highlighted the significance of distilling complex narratives, understanding key metrics, and assessing management commentary and guidance but above all else, knowing yourself well and building your own (at least somewhat) repeatable process based on individual decision-making criteria.

I hope this piece was helpful. If you found any useful insights, consider sharing and subscribing and don’t hesitate to reach out on Twitter X @netcapgirl with any thoughts or feedback. I will try to get more pieces out as I’m doing research into new companies and industries and also updates about companies I have covered in the past.

Continuing with our series of looking at industries that should inevitably experience healthy growth well into the foreseeable future, this piece will take a look at the market for luxury goods. While we will provide a general overview of the industry at large, we think the luxury market is best understood through not only three of the best companies in the sector, but perhaps three of the best companies in the world; LVMH, Hermès and Ferrari. This piece will take a look at the story of each of the three companies, why they are considered the epitome of luxury, look at their current investments and strategies, and also look at their financial statements and provide a valuation model for each of them. 

LVMH

LVMH Moët Hennessy Louis Vuitton, commonly known as LVMH, is a world-renowned French multinational luxury goods conglomerate. Founded in 1987, it has emerged as a global leader in luxury fashion, cosmetics, wines, spirits, and jewelry. LVMH's portfolio includes iconic brands like Louis Vuitton, Dior, Moët & Chandon, Hennessy, and many others. With a commitment to innovation and craftsmanship, LVMH has consistently set industry standards for quality and style. The company's success is a testament to its ability to blend tradition with modernity, offering customers exceptional experiences and products that define luxury across the globe.

Over the past decade, LVMH has exhibited a strong upward trend in its stock performance. The luxury conglomerate, known for its premium brands in fashion, cosmetics, and spirits, has experienced consistent growth. LVMH's stock benefited from expanding global luxury markets, particularly in Asia, and its ability to adapt to changing consumer preferences. High demand for iconic brands like Louis Vuitton and Christian Dior contributed to its financial success. While there were occasional market fluctuations and economic uncertainties, LVMH's overall trajectory has been positive, solidifying its position as a leader in the luxury goods industry.

Bernald Arnault, on the inevitability of LVMH (he has bought $215mm worth of LVMH stock since the shares dropped on a disappointing earnings report in July):

Arnault tells the story of coming to the US for the first time and having a cab driver who was talking about Richard Nixon. Arnault asked him if he knew who the French president was.

“No, but I know Christian Dior” replied the cab driver.

This inspired Arnault, working for his family’s construction business at the time, to pursue an acquisition of the struggling group that owned Dior. He is similar to Warren Buffett in a lot of ways, but one key similarity that provides intangible value for LVMH (and the corporation’s strategy for M&A driven growth) is being the desirable buyer for companies that may have not sold otherwise. In France and across Europe, many luxury houses are family businesses that have been passed down through generations. They are often reluctant to part with the business that their ancestors have built from the ground up. LVMH, much like Berkshire, has become a trusted home for companies like this, providing operating leverage for LVMH in the form of having access to potential acquisitions that would not otherwise be on the market, but the owners will come to the table because they have the utmost respect for Arnault, LVMH, and the way he has managed the acquisition of other family owned luxury houses previously. This is similar to how Berkshire has become the preferred buyer for similar businesses in the US, as otherwise skeptical buyers trust Buffett with their company, which is often their life’s work.  

Mr. Arnault bought Dior and bought back the licenses that the company was selling, as he felt it was diluting the brand. The very idea of creating a luxury conglomerate was new at the time, but during the merger of Moet Hennessy and Louis Vuitton, both sides wanted Arnault to side with them, as they wanted a sort of shield from an external takeover. Mr. Arnault eventually sided with Moët Hennessy and fired the head of Louis Vuitton. When the people in charge of Moet left, he found himself in charge of LVMH and Dior, laying the foundation for the company he controls today, with over $80B in sales across 75 brands. The company’s origin certainly shares traits with Buffett’s purchase of Berkshire Hathaway, a struggling textile mill at the time, and the two men share a mindset for running their companies. They’re generally fiscally conservative, they look to invest counter-cyclically and think in terms of very long term horizons, where ultimately the performance of the company’s stock should match the performance of the underlying operating businesses owned by the company. 

When trying to get a sense of the story, it’s always important to understand both the bull and bear case for any potential investment. It’s helpful to think of yourself like a scientist, searching for an inevitable underlying objective reality amidst lots of information and uncertainty. In the case of LVMH, aside from the natural cyclical risks associated with the consumer discretionary sector (the Chinese market especially, though even amidst economic downturns and recession, their business proves to be resilient due to the nature of their customer base being wealthy individuals and their ability to raise prices and maintain strong gross and operating margins) the biggest risk to LVMH’s continued success is probably the key man risk associated with Mr. Arnault’s planned succession. With LVMH being the product of his life’s work (he was trained as an engineer but has the rare ability to balance the calculated mindset with an artistic one. Both a ruthless business operator and someone who can appreciate the craftsmanship and passion that goes into producing the products his company sells) there is a natural risk that his successor won’t live up to his legacy. For us, that’s ok. The same is true of Berkshire. What is true of both businesses is they would be almost hard to mess up. And furthermore, because of the principled nature of their leadership, we expect the principles and core ethos and values that Mr. Arnault has established during his tenure as CEO to be firmly embedded in LVMH’s DNA. If you can imagine the worst case for the inevitable succession that looms over the company - say an economic downturn and the loss of key leaders of their operating companies, it’s hard to imagine a future where the company isn’t resilient and continues to overcome challenges as it has in the past. We believe that is the inevitable value of compounding heritage, and why LVMH equity should continue to compound at low to mid double digits over the next 10+ years. 

In the context of making an investment in a company like LVMH, the question “what is luxury?” inevitably comes to the mind of the investor. One can of course look through the lens of the consumer and experience the desire for the company’s products (to reiterate Mr. Arnault, luxury is “how do we create desire ?”) There are certainly some similarities with how the consumer views the presentation of luxury with how the investor evaluates the investment opportunity. The customer is not just buying a product but buying the timelessness of the brand. The heritage that has compounded across so many years that has created the brand equity that is so valuable today. The customer is also usually relatively inelastic to price increases (and is often the case, higher price creates more demand as it is seen as even more desirable and exclusive.) When viewed through the lens of an investor, this timelessness is an intangible asset that is hard to precisely value. Warren Buffett often looks for businesses that would be hard to compete with even if you could invest the same amount of capital as them. (For instance, if you had the capital to match all that’s been invested in apple, coca cola or american express, would you want to compete with those businesses?) That’s one way to quantify a competitive advantage and that’s certainly the case with Luxury Houses such as LVMH, Hermès and Ferrari. The heritage and timelessness of these companies has created brand equity that would be extremely difficult to replace. Even if a startup had unlimited capital, how long would it take to catch up to these iconic brands that are known all around the world as the epitome of luxury? These intangible values such as heritage, timelessness and brand equity show up in the financial statements as healthy gross margins, strong operating margins and fairly robust free cash flow growth, with what seems like a relatively ample runway for reinvestment at solid rates in the near to mid term future. These attributes are also why the market is comfortable assigning a relatively higher multiple on these businesses than peers within the luxury industry and certainly higher than other consumer goods businesses.  

Composition of revenue and earnings by geography

Composition of revenue and earnings by segment and brand

In terms of corporate structure, all the brand CEOs report to Mr. Arnault. He is known to be demanding and detail oriented, while also being known for having a deep appreciation for creative vision and is the rare CEO who can create a culture that balances a passion for the highest level of artistic craftsmanship with a laser-like focus on business performance and optimization. This is core to LVMH’s DNA. There is a bit of a push and pull between brands being fully autonomous and running their own P&L, but at the end of the day, everyone is reporting to a CEO & Founder who is obsessed with quality and operating performance. Similarly to Berkshire, all the individual businesses (of which there are 75+) send capital to management where they make capital allocation decisions. From the financial crisis until now, LVMH deployed $30B+ of capital and earned a return of around ~20% on the incremental invested capital. (Capex averages around 6% of the group’s revenue each year)

LVMH is also focused on investing through the income statement, Mr. Arnault believes (and is evidenced by the continued growth in earnings with consistent margins) this compounds brand equity vs focus on expanding operating margins (which have been stable around ~20%.)

Why it's the epitome of luxury 

LVMH is often considered the epitome of luxury for several key reasons:

1. Iconic Brands: LVMH owns and manages some of the world's most prestigious and iconic luxury brands, including Louis Vuitton, Christian Dior, Givenchy, Fendi, Celine, Tiffany, Loro Piana and more. These brands have a rich heritage and are synonymous with luxury, elegance, and exclusivity.

2. Craftsmanship: LVMH brands are known for their exceptional craftsmanship and attention to detail. Whether it's Louis Vuitton's handcrafted leather goods, Dior's haute couture fashion, or Dom Pérignon champagne, the products are often created using traditional techniques and the highest-quality materials.

3. Innovation: While rooted in tradition, LVMH constantly pushes the boundaries of luxury through innovation. For example, Louis Vuitton has introduced high-tech features in its luggage, and Dior has embraced sustainable fashion practices. This blend of tradition and innovation appeals to a wide range of luxury consumers.

4. Exclusivity: LVMH carefully manages the scarcity of its products, creating an aura of exclusivity. Limited editions, bespoke services, and the occasional scarcity of certain products drive desire and demand among luxury consumers.

5. Global Presence: LVMH has a strong global presence, with stores and boutiques in some of the world's most prestigious shopping districts. This global reach ensures that its luxury products are accessible to consumers around the world.

6. Marketing and Branding: LVMH is renowned for its effective marketing and branding strategies. The company invests heavily in advertising and celebrity endorsements, further enhancing the desirability of its brands.

7. Customer Experience: LVMH places a strong emphasis on providing a luxurious and personalized customer experience. From elegant store designs to attentive staff, customers often feel pampered when shopping at LVMH boutiques.

8. Heritage and Tradition: Many of LVMH's brands have a long and storied history, dating back to the 19th century or earlier. This heritage adds to the perception of authenticity and timelessness.

9. Sustainability Initiatives: LVMH has also taken steps to align with modern consumer values by implementing sustainability initiatives. Brands like Louis Vuitton and Dior have embraced eco-friendly practices, demonstrating a commitment to environmental responsibility.

10. Cultural Influence: LVMH and its brands have a significant cultural influence. They are often seen on red carpets, in films, and at high-profile events. This visibility reinforces their status as symbols of luxury and sophistication.

In summary, LVMH's status as the epitome of luxury is the result of a combination of factors, including its portfolio of iconic brands, commitment to craftsmanship and innovation, global reach, marketing prowess, and the ability to create an exclusive and aspirational experience for consumers. This combination of factors makes LVMH a dominant force in the luxury goods industry and a symbol of luxury worldwide.

History / Story

1987-1990: The Birth of LVMH

LVMH was officially created in 1987 when Moët Hennessy, a company known for its champagne and cognac, merged with Louis Vuitton, a renowned French fashion house specializing in luxury luggage and accessories.

In its first year of operation, LVMH reported sales of approximately €4.6 billion, with diverse revenue streams from fashion, spirits, and wines.

The Visionary Leader:

The driving force behind the creation of LVMH was Bernard Arnault, a visionary businessman with a keen interest in luxury goods. Arnault, known for his shrewd acquisitions and brand management, became the Chairman and CEO of LVMH.

Early 1990s: Expansion and Acquisitions

In the early 1990s, LVMH embarked on an ambitious expansion plan, acquiring several luxury brands, including Christian Dior, Givenchy, and Céline.

By 1993, LVMH's sales had grown to over €6.2 billion, reflecting its successful strategy of diversifying its brand portfolio.

Mid to Late 1990s: Global Reach and Profitability

Throughout the mid to late 1990s, LVMH continued to expand its luxury empire, venturing into new markets like Asia.

By 1998, the company's sales had surpassed €10 billion, with a growing global presence and strong profitability.

Early 2000s: Challenges and Recovery

The early 2000s presented challenges for LVMH due to the global economic downturn and a decline in luxury consumption.

Despite these challenges, LVMH's diversified brand portfolio and resilience helped it weather the storm. By 2005, sales had rebounded to approximately €12.6 billion.

Mid-2000s to Late 2000s: Continued Growth

In the mid to late 2000s, LVMH's sales and profitability continued to climb steadily. The company made strategic acquisitions, including the purchase of Bulgari in 2011.

By 2010, LVMH's sales exceeded €20 billion, reflecting its growing presence in emerging markets and the enduring appeal of its luxury brands.

Early 2010s: Emerging Markets and Digital Expansion

LVMH capitalized on the growing middle and upper classes in emerging markets like China and expanded its digital presence.

By 2015, LVMH's sales had reached approximately €35 billion, cementing its status as the world's leading luxury goods conglomerate.

2016-2020: Resilience and Adaptation

LVMH faced various challenges during this period, including changing consumer preferences and geopolitical tensions.

However, the company adapted by investing in e-commerce, sustainability initiatives, and enhancing customer experiences.

By 2020, LVMH's sales exceeded €53 billion, demonstrating its ability to evolve with the times.

2021-2023: Navigating the Pandemic and Beyond

The COVID-19 pandemic posed significant challenges to the luxury industry, with travel restrictions and economic uncertainty affecting sales.

LVMH responded by focusing on digital innovation, online sales, and sustainable practices.

Financials

Valuation

LVMH Income Statement

LVMH DCF Valuation

Hermès

What began as a saddle company in the 1850s has grown into a company that epitomizes luxury at every step of their process. Hermès did nearly $13B of sales last year at a ~70% gross margin, which is incredibly impressive and perhaps even more so considering they deliberately restrict the production volume of the bags and other products that they make.

Hermès approach to luxury is perhaps best understood through their flagship offerings, the Birkin and the Kelly bag. The eponymous bags, named after Jane Birkin and Grace Kelly, respectively, are not only not available online, but customers who want to purchase them must put their names on a waiting list at one of the company’s stores. Once they have been approved to purchase the bags (keep in mind, these bags start anywhere from close to $10k usd with prices going much higher, and can be resold on the secondary market for often 3-4x the purchase price (or more, depending on the year and the specific bag, making them an attractive investment which is an interesting overlap for us as we look at luxury houses from both the consumer and investor lens)) there is an entire ceremony that lasts roughly an hour and a half at one of their flagship stores in major cities across the globe. They serve champagne and cake and strive to make the experience as special as possible for the customer (a sort of hyper individualized experience, a theme we have explored previously at inevitability research.) Usually customers will start with an entry level purchase, such as a scarf, shoes, perfume or a tie. It is customary to begin with one of these items when you put your name on the waitlist for the privilege of purchasing a Birkin or a Kelly. What other brands that you know of put their customers on a waitlist and screen them before permitting them to make a $8k (and often much more) purchase? This is another aspect of how Hermès approaches the entire experience differently and defines luxury in their own way. 

Axel Dumas, Executive Chairman of Hermès, on their latest earnings call said: “The 2023 first half results reflect the strength of the pillars of the artisanal model of the house: quality of materials, exceptional know-how and abundant creativity. To support this growth, we continue to invest in our production capacities, in the expansion of our network, while accelerating job creation and training in all of the group’s métiers.”

Financially, Hermès looks maybe more like a software business than a consumer goods or manufacturing business with 70% gross margins and 40% operating margins. They are profitable through all their business segments and the family is careful with reinvestment and focused on being extremely good stewards of capital for the business. In other industries, investors may look at the margin structure and think “why can’t another company come in at a discount and compete some of that away from Hermès?” And the difficulty of doing so is perhaps illustrative of the competitive advantage that Hermès has built up after the many years of compounding heritage. The 100+ years of exceptional decision making and an unrelenting commitment to quality at every step of their supply chain and distribution is reflected in the company’s intangible brand equity and would make it exceedingly difficult for another company to come along and compete. Similar to the case with LVMH, economic theory would suggest that a competitor could create a reasonable substitute and compete away some of Hermès’ best in class margins, but as many companies have found out over the years, this is nearly impossible. Such is the inevitable advantage of the many years that the company has spent compounding its incredible heritage, and creating a world class brand and company in the process. 

When it comes to luxury, Hermès is the pinnacle of heritage, craftsmanship, image, and being careful with how they produce and allow supply into the market, all important components of the intangible brand equity of a luxury company and also strong competitive advantages in their own right. When you buy a Birkin or a Kelly, not buying the utility of a bag, you’re buying a piece of history. 

How big can Hermès get? What are the risks?

20 years ago, the luxury market was more centralized in the coastal US, capital cities in Europe and Japan. Now Asia is a huge market and the Middle East is experiencing healthy growth. A looming question for all luxury companies is just how big is the size of the overall market? This is a risk factor we will address in the conclusion of this piece. As noted in the graphic below, all geographic segments are experiencing 20%+ growth in the first half of 2023

An example of Hermès' commitment to exclusivity and deliberately restricting supply volume can be seen some years back, when they had an extremely popular beach bag that was flying off the shelves in Japan. While many consumer goods companies would be thrilled with the success and make plans to ramp production volume, Hermès killed the SKU. This example is illustrative of their approach to exclusivity and luxury. 

While they have been reluctant to embrace digital sales to the degree of other luxury brands, Hermès’ online presence serves as a sort of digital storefront with the goal of driving customers to their physical retail stores (you can’t buy a new Birkin or Kelly online, for instance.) The average Hermès store is 525 square meters, which means they do about $50k of revenue per square meter. On the supply chain side, Hermès owns their entire manufacturing process and assures quality at every step so that customers know their bag was crafted from a single piece of the finest leather in the hands of France’s most skilled craftsmen at Hermès owned ateliers.

Composition of revenue and earnings by geography

Composition of revenue and earnings by segment

10 year CAGRs

Why it’s the epitome of luxury

Hermès is often considered the epitome of luxury for several compelling reasons:

1. Heritage and Tradition: Hermès was founded in 1837, and its long history is synonymous with artisanal craftsmanship and timeless elegance. This heritage adds an aura of authenticity and tradition to the brand, appealing to those who appreciate a storied legacy.

2. Artisanal Craftsmanship: Hermès is renowned for its commitment to exceptional craftsmanship. The brand's artisans meticulously handcrafted products using the finest materials, from the renowned Birkin and Kelly bags to its silk scarves and leather goods. Each item is a testament to the skill and dedication of its craftsmen and women.

3. Exclusivity and Scarcity: Hermès carefully manages the scarcity of its products, particularly its iconic Birkin and Kelly bags. These items are often in high demand and have long waiting lists, creating an air of exclusivity and desirability among luxury consumers.

4. Quality and Materials: Hermès is uncompromising when it comes to the quality of its materials. The brand sources the finest leathers, silks, and other materials for its products, ensuring durability and a luxurious feel.

5. Timelessness: Hermès products are known for their timeless design. Unlike trendy or seasonal items, Hermès pieces remain stylish and relevant for decades, making them valuable investments for those who seek enduring luxury.

6. Customization: Hermès offers customization and bespoke services for many of its products, allowing customers to personalize their items to their exact preferences. This level of customization enhances the sense of exclusivity and individuality.

7. Limited Editions and Artistic Collaborations: Hermès frequently releases limited-edition collections and collaborates with artists and designers, creating unique and collectible pieces. These collaborations add an element of artistry and creativity to the brand's offerings.

8. Retail Experience: Hermès boutiques are known for their elegant and inviting design, creating a luxurious and personalized shopping experience. The brand's attentive and knowledgeable staff further enhance the customer experience.

9. Sustainability: Hermès has demonstrated a commitment to sustainability by implementing eco-friendly practices and sourcing materials responsibly. This aligns with the values of modern luxury consumers who value ethical and sustainable practices.

10. Cultural Influence: Hermès products are often seen on influential individuals, celebrities, and in popular culture. This cultural visibility reinforces the brand's status as a symbol of luxury and sophistication.

11. Scarves and Silk: Hermès' silk scarves are iconic and have become collectible items in their own right. The intricate designs and quality of these scarves have made them highly sought after by collectors and fashion enthusiasts.

In summary, Hermès' status as the epitome of luxury is the result of a combination of factors, including its rich heritage, commitment to craftsmanship and quality, exclusivity, timelessness, and the ability to create a unique and personalized experience for its customers. These qualities make Hermès a highly regarded and aspirational brand in the world of luxury.

History / Story

1837-1900: Early Beginnings and Saddlery

Hermès started as a harness workshop in Paris, specializing in crafting high-quality leather goods for horse riders. Their attention to detail and fine craftsmanship quickly gained recognition.

1900-1920s: Expansion and Diversification

In the early 20th century, under the leadership of Thierry's grandson Emile-Maurice Hermès, the company expanded its product line beyond saddlery to include leather bags and accessories. 

The iconic Hermès Haut à Courroies bag, designed for riders to carry saddles, was created in this period.

1930s-1940s: Silk Scarves and International Expansion

In the 1930s, Hermès introduced silk scarves, which became a global sensation. 

The brand also expanded its presence internationally, opening stores in locations like New York, London, and Tokyo.

1950s-1970s: The Kelly and Birkin Bags

In 1956, Grace Kelly famously used an Hermès bag to shield her pregnancy from paparazzi, leading to the bag being named the "Kelly Bag."

In the 1970s, the Birkin Bag was introduced, named after actress and singer Jane Birkin. It quickly became one of the most coveted and iconic luxury handbags in the world.

1980s-1990s: Continued Growth and Succession

Jean-Louis Dumas, a member of the Hermès family, served as the company's artistic director and CEO during this period. 

Hermès continued to expand its product range, including the introduction of fragrances and ready-to-wear fashion.

2000s-Present: Resilience and Global Expansion

In the 21st century, Hermès demonstrated resilience during economic downturns and maintained its commitment to artisanal craftsmanship.

The company expanded its global footprint and continued to launch successful products, such as the Hermès Apple Watch collaboration.

Hermès' financial performance has generally shown steady growth over the years, driven by its strong brand reputation and consistent demand for its luxury products.

The company is known for its high margins due to its focus on craftsmanship and exclusivity.

Hermès has earned a reputation as one of the world's leading luxury brands, known for its unwavering commitment to quality, craftsmanship, and timeless design. It's safe to say that Hermès' success is rooted in its ability to maintain these core values while adapting to the changing landscape of the luxury industry.

Financials

Valuation

Hermès Income Statement 

Hermès DCF Valuation

Ferrari

Ferrari is among the world’s leading luxury brands focused on the design, engineering, production and sale of the world’s most recognizable luxury performance sports cars. The Ferrari brand symbolizes exclusivity, innovation, state-of-the-art sporting performance and Italian design. Its history and the image enjoyed by its cars are closely associated with its Formula 1 racing team, Scuderia Ferrari, the most successful team in Formula 1 history. From the inaugural year of Formula 1 World Championship in 1950 through the present, Scuderia Ferrari has won 242 Grand Prix races, 16 Constructors’ World titles and 15 Drivers’ World titles. Ferrari designs, engineers and produces its cars in Maranello, Italy, and sells them in over 60 markets worldwide.

Why it's the epitome of luxury

Ferrari is often considered the epitome of luxury for several reasons:

1. Exclusivity: Ferrari maintains a strict limit on the number of vehicles it produces each year, creating an aura of exclusivity and rarity. This limited production approach ensures that owning a Ferrari is a privilege reserved for a select few.

2. Heritage and Racing Pedigree: Ferrari has a storied history in motorsports, particularly in Formula 1 racing. This racing pedigree is deeply ingrained in the brand's DNA, and it adds an extra layer of prestige to its road cars.

3. Design Excellence: Ferrari's design philosophy emphasizes both form and function. The sleek, aerodynamic, and iconic design of Ferrari cars not only enhances their performance but also makes them works of art. The brand's prancing horse logo is instantly recognizable worldwide.

4. Engineering Excellence: Ferrari is renowned for its cutting-edge engineering and innovation. The brand continually pushes the boundaries of automotive technology, producing high-performance vehicles that are celebrated for their precision and handling.

5. Performance: Ferrari cars are known for their blistering speed and exceptional handling. They consistently set benchmarks for performance in the automotive industry, making them aspirational for auto enthusiasts.

6. Customization: Ferrari offers an extensive range of customization options, allowing buyers to tailor their vehicles to their exact preferences. This bespoke approach enhances the sense of individuality and luxury.

7. Legacy of Luxury: Ferrari's brand legacy is synonymous with luxury, sophistication, and an appreciation for the finer things in life. Owning a Ferrari signifies a certain level of affluence and discernment.

8. Resale Value: Ferrari vehicles often maintain strong resale values, thanks to their exclusivity and enduring appeal. This factor further emphasizes their status as luxury assets.

9. Cultural Icon: Ferrari is a cultural icon that transcends the automotive world. It frequently appears in films, music, and popular culture, contributing to its image as a symbol of luxury and success.

10. Ownership Experience: Ferrari provides an exceptional ownership experience, including access to exclusive events, track days, and a global network of Ferrari owners' clubs. This sense of community and prestige enhances the overall luxury ownership experience.

11. Limited Editions and Collector's Items: Ferrari releases limited-edition models and collector's items, such as the LaFerrari and the Enzo, which are highly sought after by collectors and enthusiasts.

12. Brand Loyalty: Many Ferrari owners are fiercely loyal to the brand, often owning multiple Ferraris throughout their lives. This loyalty and passion contribute to Ferrari's status as an epitome of luxury.

In summary, Ferrari is considered the epitome of luxury due to its combination of exclusivity, heritage, design and engineering excellence, performance, customization, and the overall experience it offers to its discerning customers. Owning a Ferrari is not just about having a fast car; it's about being part of a prestigious legacy and embodying a symbol of luxury and high achievement.

History

The story of Ferrari is a captivating tale of passion, racing excellence, and the creation of one of the most iconic automotive brands in the world. 

Early Years (1898-1939): 

1. Enzo Ferrari, the founder of Ferrari, was born on February 20, 1898, in Modena, Italy.

2. Enzo developed a passion for cars and racing from an early age, and his interest led him to work for Alfa Romeo as a test driver and racing driver in the 1920s.

3. In 1939, Enzo left Alfa Romeo due to a disagreement with the company, and he began Auto Avio Costruzioni, a company specializing in manufacturing aircraft parts. This company later played a pivotal role in Ferrari's history.

Post-World War II (1940s-1950s):

4. After World War II, Enzo Ferrari shifted his focus to racing cars and founded Scuderia Ferrari in 1947. Initially, Scuderia Ferrari was primarily a racing team that used Alfa Romeo cars.

5. The first car to bear the Ferrari name, the 125 S, was built in 1947. It featured a 1.5-liter V12 engine designed by Gioachino Colombo.

6. In 1950, Ferrari entered the Formula One World Championship, and Alberto Ascari drove a Ferrari car to victory in the Silverstone Grand Prix.

7. The 1950s were a period of racing dominance for Ferrari, with wins in prestigious races like the Mille Miglia, Le Mans, and the Formula One World Championship.

1960s-1970s:

8. The 1960s saw Ferrari continue its success in motorsports, with drivers like John Surtees, Phil Hill, and Niki Lauda bringing home titles.

9. Ferrari introduced the iconic 250 GTO in 1962, widely regarded as one of the greatest sports cars ever built.

10. Tragedy struck in 1967 when Enzo's son, Dino Ferrari, passed away at a young age. The Dino brand was subsequently used for a series of V6 and V8 sports cars.

11. In 1973, Ferrari launched the Berlinetta Boxer, the company's first mid-engine road car.

1980s-1990s:

12. The 1980s saw Ferrari continue its success in Formula One, particularly with the partnership between the team and driver Michael Schumacher in the 1990s.

13. Ferrari introduced the iconic F40 in 1987, which became a symbol of supercar excellence.

14. The 1990s also marked a period of financial success, thanks to the booming demand for high-performance sports cars.

21st Century:

15. In 2002, Ferrari launched the Enzo, a hypercar named after the founder, which showcased the brand's technological prowess.

16. Ferrari's Formula One team enjoyed significant success in the 2000s and 2010s, with multiple championships won under the leadership of Michael Schumacher and later drivers like Kimi Räikkönen and Sebastian Vettel.

17. Ferrari expanded its product line to include GT cars like the 458 Italia, 488 GTB, and more recently, the F8 Tributo.

Present Day:

18. Ferrari remains one of the most prestigious and sought-after automotive brands globally. The Prancing Horse logo is synonymous with luxury, speed, and Italian craftsmanship.

19. The brand continues to produce high-performance road cars and is known for its limited-edition models like the LaFerrari and the SF90 Stradale.

20. While staying true to its racing heritage, Ferrari is also exploring electric and hybrid technology for future models to adapt to changing industry trends while maintaining its performance standards.

Ferrari's journey from a racing team to a globally recognized luxury brand is a testament to Enzo Ferrari's vision and passion for automotive excellence. Today, Ferrari represents the pinnacle of automotive craftsmanship, performance, and style, and its iconic cars are celebrated by enthusiasts and collectors worldwide.

Financials

Valuation 

Ferrari Income Statement

Ferrari DCF Valuation

Conclusion 

What is luxury?

As is evidenced by the sales figures in this piece, certain consumers want something that is meticulously crafted to the highest degree and also want a sense of belonging.There's an element of showing off, sure, but it’s more about being in a sort of “if you know you know” club. Luxury is owning the upstream and downstream aspects of the business to ensure the highest quality at each step. From the supply chain to the distribution down to the minutiae of the customer expended in the stores. Luxury companies follows laws of anti marketing - don’t pander to your customer, don’t increase supply or volume of popular products, cultivate an air of desire and exclusivity, keep the ratio of people who know about your products to people who own them as high as possible (many aspects of this strategy run contrary to how a traditional consumer goods business would think about their marketing efforts.) Perhaps amongst the most important - true luxury brands are selling you something timeless. By buying something from LVMH, Hermès or Ferrari you are buying a piece of the heritage. Heritage that has been preserved over centuries by holding their companies to the highest standards and never compromising on their values. From the company’s perspective, luxury is about, to quote Bernard Arnault, “creating desire.” From an investor’s perspective, luxury is about intangible brand equity, best in class pricing power, world class gross and operating margins and years of above average shareholder returns. From a research perspective, we believe luxury is best understood through these three shining examples of timeless companies that epitomize luxury.

Risks

From an investment perspective, while there are a number of attractive qualities about these businesses and their position in their respective markets (for instance, the luxury industry should experience tailwinds during good economic times and the wealthy, relatively price inelastic consumers should, in theory, prove to be  more resilient during economic downturns) there are a number of risks to consider. Valuation is always a consideration and the market certainly places a premium multiple on the earnings of LVMH, Hermès and Ferrari. They trade above their respective indices average multiples and above their peer group as well (LVMH is at 24x earnings, Hermes is at 50x earnings and Ferrari is at 48x earnings, at the time of writing this.) This is somewhat to be expected, as these businesses have proven over many years why they are exceptional, their ability to grow and reinvest at healthy margins has earned the premium multiple that investors pay for shares today. As we look towards the future, one wonders just how much runway there is for the luxury market. If the overall market was to eventually stall out, there is a risk that growth at these companies would do the same. Though unlikely based on their past (these businesses, while required as public companies to release quarterly earnings and forecasts, are always focused on long term time horizons. They think in decades, and this makes sense when they have been around for centuries) a slowdown in the overall growth of the market could cause the companies to ramp production and cut prices, perhaps diluting the tremendous value of their brand equity. We don’t see this being a likely scenario but it is possible nonetheless. A possibly more likely scenario for investors to consider is that growth stalls and the multiples that investors are willing to pay contract. Even with modest growth, a moderate to drastic rerate of earnings or cash flow multiples could severely dampen returns for shareholders. Though to the contrary, the management teams at these respective companies are savvy and perhaps exceptional capital allocators, and with healthy balance sheets, could buy back shares at depressed levels, offsetting some of the damage of the multiple contracting. A perhaps less tangible risk is the risk that overall consumer sentiment shifts harshly against luxury. While probably a low percentage in terms of near term possibility, it is worth consideration nonetheless. If there was some sort of mass movement that was able to shift the consciousness of a generation, it is possible that an entire generation of consumers would not see luxury goods as desirable as generations past. 

Bain & Company, however, presents an opposing view, finding in their research that GenZ is spending on luxury earlier than millennials did and predicts solid growth throughout most of the next decade.

“The prospects for personal luxury goods out to 2030 are positive. Solid fundamentals are set to boost the market’s value to between €540 billion and €580 billion by the end of the present decade, from an estimated €353 billion in 2022—a rise of 50% or more.

A powerful factor for sector growth this decade will be generational trends. Generation Y (millennials) and Generation Z accounted for all of the market’s growth in 2022. The spending of Gen Z and the even younger Generation Alpha is set to grow three times faster than other generations’ through 2030, making up a third of the market. This reflects a more precocious attitude toward luxury, with Gen Z consumers starting to buy luxury items some three to five years earlier than millennials did (at 15 vs. at 18–20); Gen Alpha is expected to behave in a similar way.

New types of activities, often powered by technology, should also spark an additional €60 billion to €120 billion in sales by 2030, from sources such as the metaverse and brand-related media content.

The luxury market’s consumer base will expand from some 400 million people in 2022 to 500 million by 2030. The share of top customers has been expanding and accounted for some 40% of market value in 2022, compared with 35% last year. These consumers are hungry for unique products and experiences, putting brands’ VIC (very important client) strategies into overdrive.”

• “Luxury brands have faced three years of tremendous turbulence and uncertainty, but the industry shows more strength, resilience, and ability to innovate than before . Profit levels that had quickly recovered post-Covid to an average 21% in 2021 have slightly eroded in 2022, down to 19%–21% . While the industry has benefited from increased prices and a continued shift to higher-margin direct channels, the lower profit levels reflect luxury brands’ investment in future growth, particularly through increased marketing spending and ambitious transformation programs . However, the profit erosion also reflects higher energy prices and increased labor costs . None of this has stopped brands from investing in modernizing their operations, especially through more robust information technology infrastructure to support the ongoing digitalization of the industry, and througha reconfiguration of their store networks (primarily through renovation and relocation projects) .

• The robust performance in 2022 suggests that growth should stay healthy for the personal luxury goods market in the medium term . We expect that solid market fundamentals will result in annual growth rates between 5% and 7% until 2030 . We therefore forecast that the market value of personal luxury goods will rise to between €540 billion and €580 billion by the end of the present decade, from an estimated €353 billion in 2022—an increase of more than 50% .

• Four growth engines will profoundly reshape the luxury market by 2030:

  • –  Chinese consumers should regain their pre-Covid status as the dominant nationality for luxury,growing to represent 38%–40% of global purchases .

  • –  Mainland China should overcome the Americas and Europe to become the biggest luxury market globally (25%–27% of global purchases) .

  • –  Younger generations (Generations Y, Z, and Alpha) will become the biggest buyers of luxury by far, representing 80% of global purchases .

  • –  Online should become the leading channel for luxury purchases with an estimated 32%–34% market share, followed by monobrand stores (30%–32% market share) .”

In Summary

From a financial perspective, these three companies have strong metrics across the board.

LVMH has grown revenues at a 9% CAGR the past 10 years, with 65% gross and 20% EBIT margins, while compounding free cash flow at 15% per year and earning a 13.5% return on invested capital.

Hermes has grown revenues at a 10.5% CAGR the past 10 years, with 69% gross and 34% EBIT margins, while compounding free cash flow at 20% per year and earning a 25% return on invested capital.

Ferrari has grown revenues at a 6% CAGR the past 10 years, with 50% gross and 22% EBIT margins, while compounding free cash flow at 11% per year and earning a 18% return on invested capital.

These numbers are attractive to investors on the surface level, and perhaps even more so when one considers the strength of their respective competitive advantages. Perhaps deservingly so, all these companies are valued at relatively expensive multiples, and conservative investors may want to add these to their watchlists while they wait for a pullback. There are certainly risks - whether the key man risk and succession worries with Mr. Arnault at LVMH or the market growth concerns relevant to Hermès and Ferrari, but investors can sleep well at night knowing they own shares in brands that have successfully compounded their heritage over many years to deliver attractive returns for shareholders. As always, one should do their own due diligence when researching an investment idea and the information in this piece is for research purposes only and is not investment advice. We believe that inevitably, the luxury industry will continue to experience healthy growth through the resilience of their target demographic amidst varying economic conditions and because of the various competitive advantages that have been built over the decades. We recognize the premium multiples that investors are currently paying for the earnings of these companies and while it’s impossible to time the market, we would feel comfortable building positions in these shares at a price that is a bit closer to the fair value in our valuation models. Nonetheless, they are certainly companies that belong on most investors watchlists and are undoubtedly amongst the best companies in the world. Please don’t hesitate to reach out on Twitter with any thoughts or feedback and if you enjoyed this piece, let me know and I will do my best to follow up with more in depth coverage on each company and review their quarterly earnings results.

• Introduction 

• Overview 

• Supply Chain 

• Policy Support 

• Conclusion

• Relevant Companies 

If you have been noticing more and more electric vehicles on the roads these days, you’re not alone. 14% of all cars sold in 2022 were EVs, up from 9% in 2021 and 5% in 2020. While a large share (60%) of all sales come from China, sales in the US rose 55% year over year, bringing the US’ share of global EV sales to around 8% (the third highest in terms of sales with Europe being the second.) With a combination of fiscal support in the form of tax credits and other incentive programs, growing demand from consumers (data from the International Energy Agency suggests that 14 million EVs will be sold in 2023, a 35% year over year increase) and increased investment efforts from automakers and their suppliers to support this demand, a future with more and more electric vehicles feels inevitable. How can investors best understand this electrified road ahead of us and profit from this inevitable future? This piece is written with the goal of providing an overview of what is happening in the Electric Vehicle industry with a focus on highlighting the many steps involved in the complex supply chain. We will discuss 8 key stages of the EV supply chain and highlight relevant companies along the way.

From the International Energy Agency:

The increase in demand for electric vehicles is driving demand for batteries and related critical minerals. Automotive lithium-ion (Li-ion) battery demand increased by about 65% to 550 GWh in 2022, from about 330 GWh in 2021, primarily as a result of growth in electric passenger car sales. In 2022, about 80% of lithium, 30% of cobalt and 10% of nickel demand was for EV batteries. Only five years prior, these shares were around 15%, 10% and 2%, respectively. Reducing the need for critical materials will be important for supply chain sustainability, resilience and security, especially given recent price developments for battery material.

New alternatives to conventional lithium-ion are on the rise. The share of lithium-iron-phosphate (LFP) chemistries reached its highest point ever, driven primarily by China: around 95% of the LFP batteries for electric LDVs went into vehicles produced in China. Supply chains for (lithium-free) sodium-ion batteries are also being established, with over 100 GWh of manufacturing capacity either currently operating or announced, almost all in China.

The EV supply chain is expanding, but manufacturing remains highly concentrated in certain regions, with China being the main player in battery and EV component trade. In 2022, 35% of exported electric cars came from China, compared with 25% in 2021. Europe is China’s largest trade partner for both electric cars and their batteries. In 2022, the share of electric cars manufactured in China and sold in the European market increased to 16%, up from about 11% in 2021.

EV supply chains are increasingly at the forefront of EV-related policy- making to build resilience through diversification. The Net Zero Industry Act, proposed by the European Union in March 2023, aims for nearly 90% of the European Union’s annual battery demand to be met by EU battery manufacturers, with a manufacturing capacity of at least 550 GWh in 2030. Similarly, India aims to boost domestic manufacturing of electric vehicles and batteries through Production Linked Incentive (PLI) schemes. In the United States, the Inflation Reduction Act emphasizes the strengthening of domestic supply chains for EVs, EV batteries and battery minerals, laid out in the criteria to qualify for clean vehicle tax credits. As a result, between August 2022 and March 2023, major EV and battery makers announced cumulative post-IRA investments of at least USD 52 billion in North American EV supply chains – of which 50% is for battery manufacturing, and about 20% each for battery components and EV manufacturing.

The supply chain for electric vehicles (EVs) is a complex network that involves various stages and components. Here’s a high-level overview:

1. Raw Materials: The supply chain begins with the extraction and processing of raw materials like lithium, cobalt, nickel, and other minerals used in battery production. These materials are often sourced from different parts of the world.

2. Battery Manufacturing: Battery cells are produced by assembling the processed materials in specialized facilities. These battery cells are a crucial component of EVs, as they store and supply electrical energy.

3. Vehicle Assembly: After manufacturing the battery cells, the vehicle assembly process takes place. This involves integrating the battery packs, electric motors, and other EV-specific components into the car’s chassis.

4. Tier 1 Suppliers: Numerous tier 1 suppliers provide specialized components like electric drivetrains, charging systems, electronics, and other parts necessary for the EV’s operation.

5. Tier 2 and Tier 3 Suppliers: These suppliers provide a wide range of components and parts that go into making the various systems and sub-assemblies for the electric vehicles.

6. Distribution and Logistics: Once the vehicles are assembled, they are distributed to dealerships and customers. Logistics play a significant role in transporting parts and vehicles efficiently throughout the supply chain.

7. Charging Infrastructure: The EV supply chain also involves the establishment and maintenance of charging infrastructure, including public charging stations, fast-charging networks, and home charging units.

8. After-Sales Support: After-sales support and service networks are vital for maintaining and repairing EVs, which includes spare parts supply and service centers.

Raw Materials 

The transformation from conventional combustion engines to electric propulsion hinges on a range of critical raw materials, each playing a pivotal role in the EV ecosystem. These elements, often referred to as battery minerals, include lithium, cobalt, nickel, and graphite. The journey of these minerals begins in resource-rich areas scattered across continents.

1. Lithium: The "White Gold"  

Lithium is the cornerstone of EV batteries, offering high energy density and light weight. The lithium triangle formed by Argentina, Bolivia, and Chile holds the largest known reserves. Extracting lithium involves solar evaporation in salt flats, followed by chemical processing. 

“Lithium demand has almost doubled since 2017 to 130 kt (kilotons) in 2022, of which demand for EV batteries accounts for 80%, up from 47% in 2021, 36% in 2020 and only 20% in 2017. Lithium is also used in the production of ceramics, glass and lubricants. Batteries are now the dominant driver of demand for lithium and therefore set the price. The availability of lithium supply is of particular concern because it is irreplaceable for Li-ion batteries and there are no commercial alternative battery chemistries available at scale today that meet the performance of Li- ion batteries.” - IEA

2. Cobalt: The Complex Dilemma  

Cobalt, though essential, poses ethical and supply challenges due to its concentration in politically unstable regions. The Democratic Republic of Congo dominates cobalt production. Efforts are underway to reduce cobalt usage in favor of more sustainable chemistries.

“Cobalt demand was 187 kt in 2022, of which the EV battery share was 30%, up from 24% in 2021 and 18% in 2020. Cobalt is also used in superalloys, hard metals and catalysts. The cobalt intensity of Li-ion batteries has decreased significantly over recent years as battery makers moved to higher nickel content chemistries to achieve higher energy densities and lower costs (cobalt is the most expensive constituent per kg of Li-ion battery metal).” - IEA

3. Nickel: The Powerhouse  

Nickel-rich cathodes enhance battery energy density. Indonesia and the Philippines are key suppliers of nickel, with Tesla striking deals to secure long-term supplies. As EV adoption surges, nickel demand is set to soar.

“Nickel demand is dominated by stainless steel production. Total demand was 2,640 kt in 2021, of which the share of EV-related demand was 10%, up from 7% in 2021 and 4%2020. Batteries require Class 1 nickel (>99.8% purity), while Class 2 nickel (<99.8% purity) cannot be used without further significant processing. Nickel-based cathodes are the dominant EV battery chemistries today and are expected to remain so in the future due to the demand for longer driving range EVs particularly in Europe and the United States. There is almost seven times more nickel than lithium by weight in an NC811 battery, therefore, EV Li-ion battery prices are most sensitive to nickel prices. This is of significant current concern given the war in Ukraine because Russia is the world’s largest supplier of Class 1 battery-grade nickel, producing around 20% of global supply.” - IEA

“The increase in battery demand drives the demand for critical materials. In 2022, lithium demand exceeded supply (as in 2021) despite the 180% increase in production since 2017. In 2022, about 80% of lithium, 30% of cobalt and 10% of nickel demand was for EV batteries. Just five years earlier, in 2017, these shares were around 15%, 10% and 2%, respectively. As has already been seen for lithium, mining and processing of these critical minerals will need to increase rapidly to support the energy transition, not only for EVs but more broadly to keep up with the pace of demand for clean energy technologies.15 Reducing the need for critical materials will also be important for supply chain sustainability, resilience and security. Accelerating innovation can help, such as through advanced battery technologies requiring smaller quantities of critical minerals, as well as measures to support uptake of vehicle models with optimized battery size and the development of battery recycling.” - IEA

4. Graphite: The Conductive Element  

Graphite serves as an anode material in batteries, facilitating the movement of lithium ions. Natural graphite production centers on China, followed by Brazil and Canada. Synthetic graphite is also gaining prominence.

5. Manganese is relevant to electric vehicles primarily due to its use in the production of lithium-ion batteries, which are the most common type of batteries used in EVs. Here are a few reasons that manganese is used in the EV supply chain:

Battery Cathodes: Manganese is a key component of the cathode material used in lithium-ion batteries. In particular, manganese is often used in lithium manganese oxide (LiMn2O4) or lithium nickel manganese cobalt oxide (NMC) cathodes. These cathode materials are chosen for their stability, high energy density, and cost-effectiveness, which are crucial factors in the design and performance of EV batteries.

Improved Safety: Lithium manganese oxide cathodes, in particular, are known for their thermal stability and improved safety characteristics compared to some other cathode materials. This safety feature is highly desirable in EV batteries to reduce the risk of thermal runaway or battery fires in the event of a malfunction or accident.

High Energy Density: Manganese-based cathodes can provide a relatively high energy density, which is essential for extending the driving range of electric vehicles. High energy density batteries can store more energy in a given volume or weight, allowing for longer trips between charges.

Cost Efficiency: Manganese is more abundant and less expensive than some other cathode materials, such as cobalt. This can contribute to cost savings in battery production, which is a significant factor in making EVs more affordable for consumers.

Hybrid and Plug-in Hybrid Vehicles: Manganese-based cathodes are also used in hybrid and plug-in hybrid vehicles, which combine internal combustion engines with electric propulsion. These vehicles benefit from the energy storage capabilities of manganese-based batteries to support electric driving modes.

Recycling: As the EV industry grows, there is a growing interest in recycling materials from end-of-life batteries to reduce waste and environmental impact. Manganese can be recovered and reused from recycled lithium-ion batteries, contributing to sustainability efforts in the EV sector.

It's important to note that not all lithium-ion batteries for EVs use manganese, as different manufacturers may use various cathode chemistries based on factors like performance requirements, cost considerations, and environmental goals. Some batteries may use cathode materials that contain nickel, cobalt, or iron in addition to or instead of manganese, depending on the specific battery design and intended application.

6. Copper is also relevant to electric vehicles in several ways:

Electric Wiring and Conductivity: Copper is an excellent conductor of electricity, which makes it essential for the wiring and electrical components of electric vehicles. Copper wiring is used to transmit electric power from the battery to the various components of an EV, including the electric motor, lights, sensors, and the charging system. Its high conductivity minimizes energy losses during transmission.

Electric Motors: Electric vehicles rely on electric motors for propulsion. These motors often use copper windings to create magnetic fields and generate motion. Copper's high electrical conductivity and thermal conductivity are advantageous for optimizing the efficiency and performance of electric motors in EVs.

Batteries: While the main components of lithium-ion batteries, which are commonly used in electric vehicles, include lithium, cobalt, nickel, and other materials, copper is also used in the battery's conductive components. Copper is used in the current collectors, connectors, and bus bars within the battery pack, ensuring efficient electricity flow within the cells.

Charging Infrastructure: Charging infrastructure for electric vehicles also relies on copper wiring and connectors. Charging stations and cables are made with copper components to efficiently transmit electricity from the grid to the EV's battery.

Thermal Management: Copper is known for its high thermal conductivity. In EVs, copper is used in various thermal management systems to dissipate heat generated by the electric motor and the battery pack. This helps maintain optimal operating temperatures, which is critical for the efficiency and longevity of the vehicle's components.

Electric Vehicle Production: Copper is used in the manufacturing process of electric vehicles. It's used in various parts and components, including wiring harnesses, connectors, and other electrical elements that are integral to the assembly of an electric vehicle.

Recycling: As the electric vehicle industry grows, there is an increasing focus on recycling materials to reduce environmental impact. Copper is a recyclable material, and recycling practices are becoming more critical in the EV industry to recover and reuse copper from end-of-life EV components.

The journey from extraction to EV battery assembly involves a complex supply chain that transcends borders and industries. It's an orchestra of miners, processors, manufacturers, and recyclers working in harmony.

1. Mining and Extraction: 

Miners extract raw materials from mines, often using environmentally sensitive methods. Cobalt mining in the DRC has faced scrutiny due to concerns about child labor and unsafe conditions.

“The five key battery materials are lithium, nickel, cobalt, graphite and manganese. Lithium is extracted from two very different sources: brine or hard rock. Lithium brines are concentrated salt water containing high lithium contents and are typically located in the high elevation areas of Bolivia, Argentina and Chile in South America with Chile being the largest producer. Brine deposits often contain large quantities of other useful elements such as sodium, potassium, magnesium and boron which offsets some of the cost of pumping and processing brine. Lithium hard rock (spodumene) is primarily mined in Australia. Novel processes are being developed to extract lithium from unconventional resources such as geothermal brine. Currently, the top-five lithium suppliers account for about half of global lithium production. Major lithium suppliers include a mixture of large chemical and mining companies including: Sociedad Química y Minera de Chile SA (Chile); Pilbara Minerals (Australia); Allkem (Australia); Livent Corporation (United States); and Ganfeng Lithium Co. (China). Unlike for other battery metals, lithium extraction companies tend to be specialized in lithium mining and chemical companies. Nickel is found primarily in two types of deposit – sulfide and laterite. Sulfide deposits are mainly located in Russia, Canada and Australia and tend to contain higher grade nickel. It is more easily processed into Class 1 battery-grade nickel. Laterite, however, tends to contain lower grade nickel and is mainly found in Indonesia, Philippines and New Caledonia. Laterite requires additional energy intensive processing to become battery-grade nickel. Nickel production is less concentrated than lithium with about nine companies supplying half of global nickel production. Key nickel suppliers include: Jinchuan Group (China); BHP Group (Australia); Vale SA (Brazil); Tsingshan (China); Nickel Asia Corporation (Philippines); and Glencore (Switzerland).

Cobalt is predominantly mined as a by-product of copper or nickel mining. Over 70% of cobalt is produced in the Democratic Republic of Congo (DRC) and Glencore (Switzerland) is the largest global producer. Other key cobalt suppliers include: Jinchuan Group (China); CN Molybdenum (China); and Chemaf (DRC). Artisanal and small-scale mining is responsible for 10 – 20% of cobalt production in the DRC.

Graphite is the dominant anode material and can be found naturally or produced synthetically. Natural graphite mining is dominated by China (80%), though global production is becoming more diversified, with many greenfield graphite mining projects being developed including in Tanzania, Mozambique, Canada and Madagascar.

Manganese resources are more widely distributed around the world than the other battery metals and remain available at relatively low cost. There is a general expectation that there will not be an ore shortage in the near term. The leading producers of manganese ore include South Africa, Australia, Gabon and China.” - IEA

2. Refining and Processing: 

Raw materials undergo refining and processing to achieve the chemical composition required for batteries. This stage is crucial for ensuring the quality and safety of EV batteries.

“Batteries require high purity materials and therefore high-grade sources, as well as significant refining, is required to reach sufficient quality battery chemical precursors. These refining processes typically involve heavy industrial processes based on heat or chemical treatment (typically pyrometallurgy and/or hydrometallurgy) to refine the raw ore into the usual required chemicals, lithium carbonate or hydroxide, or cobalt and nickel sulfate. Adding complexity, certain raw materials are more or only suitable for the production of battery precursors. For instance, lithium carbonate is produced from lithium brine, which is useful for wider lithium demand, however, unsuitable for use in leading high-nickel Li-ion batteries. Lithium hydroxide is more suitable for high-nickel chemistries and is more easily produced from spodumene hard rock sources. Similarly, battery production typically requires nickel sulfate, typically only synthesized from Class1nickel, which is most economically produced from nickel sulfides. Class 2 nickel can be processed into Class 1 nickel but requires significant additional processing.

New processing technologies, however, are increasing the flexibility of nickel processing routes. These include:

• High-pressure acid leaching (HPAL) is a process which is able to produce Class 1 nickel from lower grade laterite resources.

• Mixed hydroxide precipitate (MHP), an intermediate product in nickel refining, can be further refined into nickel sulfates at low cost from laterite resources.

• Nickel matte (a battery-grade nickel precursor) can be produced from laterite resources, but is more emissions- intensive than conventional production routes.

Raw material processing is highly concentrated. For example, in lithium carbonate and hydroxide production, five major companies are responsible for three-quarters of global production capacity. Often the refining is done by the mining company together with the extraction. For example, Ganfeng, a Chinese mining company, has evolved to include processing and refining lithium and is increasingly focussed on boosting their lithium hydroxide production. In other cases, it is exported to third parties to do the refining, with many processing companies in China, such as Chengxin Lithium Group or Zhejiang Huayou Cobalt. This is particularly the case for Australian spodumene as almost no miners yet produce integrated lithium chemical supply.

While manganese resources are widely distributed, production of high-purity manganese sulfate raises concerns around geographical concentration of supply. China currently accounts for around 90% of the global production capacity, raising the need for new, diversified manganese refining capacity. New manganese sulfate projects are starting to come online in Australia, Europe, Indonesia and the United States.” -IEA

3. Manufacturing Battery Components:

Battery manufacturers like CATL, LG Chem, and Panasonic produce cells using refined materials. The choice of cathode chemistry (NMC, NCA, LFP) influences the mineral mix.

“Batteries are composed of several highly specialized components including cathode and anode materials, electrolytes and separators. These components require advanced materials chemistry and engineering for their production. The most complex processing is required to form the battery active materials from the high purity chemicals produced from raw material processing, such as lithium hydroxide and nickel sulfate. These materials are further processed using specialized syntheses to produce active materials for the cathode and anode. The leading cathode active materials for Li-ion are transition metal oxides including lithium cobalt oxide, lithium nickel manganese cobalt oxide (NMC), lithium nickel cobalt aluminum oxide (NCA) and lithium iron phosphate (LFP). Seven companies are responsible for 55% of global cathode material production capacity. Key players include: Sumitomo (Japan); Tianjin B&M Science and Technology (China); Shenzhen Dynanonic (China); and Ningbo Shanshan (China).

The dominant anode active material is graphite which can be natural or synthetic. Producing graphite anode materials is more mature and established than cathode material production given graphite has been the dominant anode for a long time, though both graphite types require sophisticated processing. Flake natural graphite is used in batteries and is processed into spherical graphite to be more homogenous for use as anode material. Synthetic graphite is produced from refining hydrocarbon materials such as coke. To improve graphite anode performance, small and increasing fractions of silicon are being added to the graphite anode to increase energy density. Anode material production is even more highly concentrated with four companies responsible for half of global production capacity. The largest players include: Ningbo Shanshan (China); BTR New Energy Materials (China); and Shanghai Putailai New Energy Technology (China). The top-six companies are all Chinese and account for two-thirds of global production capacity.

Separators are engineered microporous membranes, typically made of polyethylene or polypropylene and often ceramic coated for improved safety for EVs. Separator production is also concentrated with five companies responsible for half of the global production capacity. Key players include: Zhuhai Enjie New Material Technology (China); Shanghai Putailai New Energy Technology (China); and SK IE Technology (Korea). Electrolytes are made of a salt and solvent and both require synthesis and then mixing. Jiangxi Tinci Central Advanced Materials in China alone produces 35% of global electrolyte salt. The top electrolyte producing companies include: Zhangjiagang Guotai-Huarong New Chemical Materials (China); Shenzhen Capchem Technology (China); and Ningbo Shanshan (China). Most companies that engage in cell component manufacturing are highly specialized and only produce those components.” - IEA

4. Automotive Assembly:  

Automakers incorporate batteries into vehicles, with EV manufacturing differing significantly from traditional assembly lines. Battery packs influence vehicle design, safety, and performance.

5. End-of-Life Recycling: 

As EVs proliferate, end-of-life recycling becomes vital. Efficient recycling can recover valuable materials like cobalt, nickel, and lithium, reducing the industry's environmental footprint.

Supply Chain Challenges and Future Trends

The EV supply chain isn't without challenges. Geopolitical tensions, ethical concerns, and supply constraints can disrupt material availability. However, the industry is actively addressing these issues through innovation, sustainable practices, and diversification of supply sources.

1. Sustainable Sourcing: 

Industry stakeholders are focusing on ethical sourcing and minimizing the environmental impact of mining and extraction processes.

2. Technology Advancements:  

The pursuit of solid-state batteries, reduced cobalt chemistries, and improved recycling methods is reshaping the materials landscape.

3. Diversification:  

Strategic partnerships and investments in resource-rich regions aim to diversify material sources and reduce dependency on single suppliers.

The EV revolution rests on a complex yet interconnected network of raw material suppliers, processors, manufacturers, and innovators. Understanding this intricate web is essential for investors, policymakers, and enthusiasts alike as we navigate the electrified future.

Relevant Companies:

• Lithium: Albemarle Corporation, SQM (Sociedad Química y Minera de Chile)

• Cobalt: Glencore plc, China Molybdenum Co., Ltd.

• Nickel: PT Vale Indonesia Tbk, Norilsk Nickel

• Graphite: Syrah Resources Limited, China National Building Material Group Corporation (Sinosteel Corporation)

Battery Manufacturers:

• Contemporary Amperex Technology Co. Ltd. (CATL)

• LG Chem

• Panasonic Corporation

• Samsung SDI Co., Ltd.

• SK Innovation

Automakers:

• Tesla, Inc.

• General Motors

• Volkswagen Group

• BMW Group

• Nissan Motor Corporation

• Ford Motor Company

• Hyundai Motor Group

• Rivian Automotive, Inc

Battery Manufacturing 

At the core of every electric vehicle lies a sophisticated energy storage system designed to provide the required range and performance. Lithium-ion batteries, the dominant technology, have transformed the automotive landscape due to their high energy density, longevity, and relatively lightweight nature. 

1. Raw Material Sourcing and Refinement:

The journey commences with sourcing crucial materials like lithium, cobalt, nickel, and graphite from mines across the world. These raw materials undergo refining processes to attain the required purity and composition, ensuring the battery's optimal performance and safety.

2. Electrode Preparation:

The battery's electrodes are prepared by coating thin sheets of copper or aluminum foil with active materials. The anode is coated with a layer of graphite, while the cathode is coated with materials like lithium cobalt oxide (LiCoO2), lithium iron phosphate (LiFePO4), or lithium nickel cobalt manganese oxide (NMC).

3. Cell Assembly:

Cell assembly involves layering the anode, separator, and cathode, followed by winding or stacking these layers to form the cell. A separator prevents direct contact between the anode and cathode, preventing short circuits. An electrolyte is added to enable ion movement.

“Producing the battery cells is a multi-step process with two broad stages: electrode manufacturing and cell fabrication. Though cell manufacturers have different cell designs, the cell manufacturing processes are similar, use mature technologies and are well established. These processes are energy intensive, being conducted in highly controlled clean and dry room conditions to avoid any impurities and moisture. Using low-carbon sources of electricity is key to reducing emissions in cell production. First electrodes are produced by mixing cathode or anode active materials with a binder, solvent and additives before coating on aluminum (cathode) or copper (anode) foil current collectors. The electrodes are rolled (calendared) and subsequently dried. The cell is then created by stacking the electrodes with a separator in between.

Manufacture of the battery pack may be completed either by the cell manufacturer or by the automaker. Cells are first housed together in module frames, then the battery pack is assembled through integration of modules, the battery management system, electronics and sensors, all encased in a final housing structure.

Battery cell production is a capital-intensive process and production is highly concentrated, with the top-three producers in 2021, CATL (China), LG Energy Solution (Korea), and Panasonic (Japan), accounting for 65% of global production. Cell manufacturers from Japan and Korea tend to be established conglomerates having decades of experience making batteries for consumer electronics. There are also Chinese companies that began producing batteries for consumer electronics in the 1990s and then specialized in batteries for EVs such as CATL and BYD. A third wave of new battery makers is taking shape in Europe and North America, but today they are mostly in planning or upscaling stages. With recent supply chain strains many battery makers and automakers are becoming increasingly involved in the mining and processing of critical minerals to ensure access to production; Tesla, CATL and LG Energy Solution have all become directly involved in upstream stages.” -IEA

4. Formation and Testing:

The cells undergo a formation process to stabilize their performance. Rigorous testing ensures quality and safety standards are met. Faulty cells are discarded, maintaining the integrity of the battery pack.

5. Battery Pack Assembly:

Battery packs are assembled by connecting multiple cells in series and parallel configurations. The pack includes cooling systems, electronics for monitoring and balancing, and safety mechanisms to prevent overcharging or overheating.

6. Integration into EVs:

Automakers incorporate battery packs into vehicle architectures, impacting vehicle design, weight distribution, and overall performance. The pack's placement and integration are crucial for safety and efficiency.

Battery manufacturing isn't without challenges, but innovative solutions are driving progress:

1. Scale and Efficiency:  

Meeting the growing demand for EVs requires scaling up production while maintaining efficiency and quality.

2. Sustainable Sourcing:  

Industry stakeholders are actively seeking sustainable and ethical sourcing of battery materials, aiming to reduce environmental and social impacts.

3. Solid-State Batteries:  

Solid-state batteries promise improved energy density, safety, and longevity, with potential to reshape the battery landscape.

4. Recycling and Second Life:  

Developments in recycling technologies aim to recover valuable materials from used batteries, reducing waste and environmental impact.

5. Innovations in Chemistry:  

Advancements in battery chemistries, such as high-nickel cathodes, are pushing the boundaries of energy density and cycle life. Solid State batteries are also an interesting trend that we explore a bit more below.

Global Players in Battery Manufacturing

Several companies lead the charge in battery manufacturing, each contributing to the evolution of EV technology:

• Contemporary Amperex Technology Co. Ltd. (CATL): A Chinese giant renowned for its lithium-ion battery prowess and collaborations with major automakers.

• LG Chem: A South Korean company with a significant presence in the EV battery market, supplying to global automakers.

• Panasonic Corporation: Known for its partnership with Tesla, Panasonic is a key player in supplying batteries for electric vehicles.

• Samsung SDI Co., Ltd.: Samsung's battery division is actively shaping battery technology with innovations and collaborations.

• SK Innovation: A South Korean company involved in the production of EV batteries and development of next-generation battery technologies.

Looking to the Future of Batteries

Over the summer, Samsung announced that they had completed construction of a pilot plant for solid state batteries and plans to complete a prototype in the second half of the year. Solid state batteries are sort of a “dream battery” because of their high energy density and safety. Samsung has a goal for these batteries to be able to go almost 500 miles on a single charge and to be able to be charged over 1000 times in their life cycle, they have a target to commercialize these solid state batteries by 2027. Toyota has also been working on similar technology, hoping to to reduce the number of processes that is required to make battery materials so that they can make solid state batteries at a similar or even lower cost than liquid based lithium ion batteries. Inevitably, the winners in this space will be able to harness chemistry to optimize the energy density, power, number of cycles and cost of the battery as well as its form factor. This is a space that is certainly worth watching and a possible topic for us to explore in a future in depth piece. The intersection of physical inevitabilities and financial market realities is where we hope to provide value from a research perspective.

Silicon Carbide

Earlier this year, Tesla sent shockwaves through the silicon carbide industry by announcing they planned to use 75% less silicon carbide in future power modules. While it sounds severe (and sends the stocks of SiC companies tumbling), it is likely that silicon carbide will remain relevant to the automotive market in at least the short to mid term, due to it being the best material for high power and high voltage devices.

Why is Silicon Carbide relevant to electric vehicles?

1. Power Electronics: SiC is a wide-bandgap semiconductor material that has superior electrical properties compared to traditional silicon (Si) used in power electronics. SiC-based power electronics components, such as diodes and transistors, can operate at higher temperatures, voltages, and frequencies while minimizing energy losses. This leads to more efficient power conversion in EVs, which can increase overall vehicle efficiency and extend battery range.

2. Energy Efficiency: The use of SiC power electronics can significantly improve the energy efficiency of EVs. They reduce switching losses and conduction losses in electronic components, resulting in less heat generation and more power being delivered to the vehicle's electric motor. As a result, the vehicle requires less energy to perform the same tasks, which can increase the driving range on a single battery charge.

3. Faster Charging: SiC-based power electronics can enable faster charging of EVs. These components can handle higher power levels and operate efficiently at higher frequencies, allowing for rapid battery charging without excessive heat generation. Fast-charging stations using SiC technology can deliver more power to the EV's battery, reducing charging times.

4. Lighter and Smaller Components: SiC devices are smaller and lighter than their silicon counterparts for the same power rating. This is especially beneficial for EVs, as it can lead to weight reduction and space savings, which can be used for other components or increase passenger and cargo space.

5. Extended Battery Life: SiC power electronics can help extend the life of EV batteries. The improved efficiency and reduced heat generation mean that the battery operates under less stress during charging and discharging cycles, which can slow down degradation and increase the overall lifespan of the battery pack.

6. Regenerative Braking: SiC-based power electronics can enhance regenerative braking systems in EVs. These systems capture and store energy when the vehicle decelerates or brakes, which can then be used to recharge the battery. SiC components can improve the efficiency of this energy conversion process, maximizing the energy recovered during braking.

7. Thermal Management: SiC's ability to operate at higher temperatures with lower losses reduces the demand for complex and heavy thermal management systems. This can lead to more compact and lightweight EV designs, as well as cost savings in the production of cooling systems.

In summary, SiC plays a crucial role in improving the performance, efficiency, and range of electric vehicles by enabling more efficient power electronics and faster charging, reducing heat generation, and contributing to overall system optimization.

Tesla in particular has been laser focused on making its entry level vehicles more accessible and for all it’s benefits, SiC is an expensive material. Tesla may be exploring ways “to stretch the performance of the SiC substrate through more advanced system designs and higher integration. While it is unlikely that a single technology could decrease SiC by 75%, a variety of advances in packaging, cooling (i.e. double sided and liquid cooling), and trench device structure could enable more compact, better-performing devices. Tesla will no doubt be exploring such opportunities, and the 75% figure could refer to a highly integrated inverter design that reduces its use of die from 48 to 12. However, if this were to be the case, it would not equate to such an aggressive reduction in SiC material as has been suggested, because each die would need to be larger in size to handle higher powers.

Meanwhile, other OEMs who are releasing 800V vehicles in 2023-24 will still rely on SiC, which is the best candidate in this space for high-power and high voltage-rating devices. Therefore, there will likely be no short-term impacts on SiC penetration for OEMs.” (Yole Intelligence)

Also from Yole -

“While the push towards higher integration will result in minimal impact on the device market, there could be an impact on wafer shipments. Despite not being as dramatic as many initially thought, each scenario predicts a decrease in SiC demand, which could impact semiconductor companies.

However, this could increase the supply of material to other markets that have been growing alongside the automotive market in the last five years. The automotive market was and is the main driver of SiC and will retain its large market share going forwards (Yole Intelligence predicts the sector to reach $8 billion by 2028), but the industrial, energy (driven by renewables) and transport (high voltage) sectors are all expected to grow significantly in the next few years – which will all but be helped by lower costs and more access to supply of material.”

Vehicle Assembly 

Next we’ll discuss the many steps involved in assembling electric vehicles, focusing on the integration of battery packs, electric motors, and other EV-specific components that define the driving experience of the future.

1. Battery Pack Integration:

The journey begins with the integration of the heart of the EV – the battery pack. The battery, comprising numerous cells, is placed within a structurally designed enclosure that ensures safety, cooling, and optimal weight distribution. This step requires meticulous engineering to determine the ideal placement, protecting the battery while minimizing its impact on the vehicle's center of gravity.

“The battery pack is integrated into the EV by the automakers, where it is connected with the electric motor, on-board charger module, high voltage distribution box, electric transmission and thermal systems, depending on the vehicle architecture. Automakers focussing only on EVs must develop greenfield factories, while for incumbent automakers pre-existing vehicle assembly factories can be retooled and repurposed for EV production. EV manufacturing is currently concentrated in a small number of OEMs, with the top-six companies responsible for 52% of production in 2021. This is a slight decrease from 2020 where the top-six were responsible for 55%. The three largest producers, Tesla (United States), VW Group (Germany) and BYD (China), accounted for a third of EV production in 2021. The rapid growth of BYD has been particularly impressive, it was not even among the top-six producers in 2020, but ranked as the third-largest producer of EVs in 2021.” - IEA

2. Electric Motor Integration:

Electric vehicles derive their power from electric motors, which replace traditional internal combustion engines. The integration of the electric motor involves precise positioning within the vehicle's chassis to ensure efficient power delivery and drivability. Depending on the vehicle's design, the motor might be placed in the front, rear, or even on each axle for all-wheel drive configurations.

3. Power Electronics and Inverters:

Power electronics play a crucial role in controlling and optimizing the flow of electricity between the battery, motor, and other vehicle systems. Inverters convert direct current (DC) from the battery to alternating current (AC) for the motor. These components are carefully integrated to ensure seamless communication and energy management.

4. Thermal Management Systems:

Electric vehicles rely on effective thermal management systems to regulate temperatures within the battery, motor, and power electronics. Liquid or air cooling systems prevent overheating, ensuring the longevity and performance of critical components.

5. Chassis Integration:

The vehicle's chassis, or structural frame, provides the foundation for the entire assembly. EV-specific chassis designs accommodate the unique weight distribution, incorporating reinforced sections to support the battery and ensure rigidity.

6. Wiring and Electronics:

The intricate web of wiring and electronics connects all components, facilitating communication between the battery, motor, sensors, and various vehicle systems. Electric vehicles require sophisticated electronic architectures to manage power flow, charging, regenerative braking, and other functions.

7. Interior and User Experience:

Interior assembly involves incorporating EV-specific features such as advanced infotainment systems, digital instrument clusters, and energy usage displays. Attention to detail is crucial to enhance the user experience and align with the vehicle's eco-friendly ethos.

8. Safety and Quality Checks:

Rigorous safety and quality checks are conducted at various stages of assembly to ensure the vehicle meets regulatory standards and performance expectations. These checks encompass everything from crash tests to functional assessments of the powertrain and electronics.

9. Final Testing and Quality Assurance:

Before an electric vehicle leaves the assembly line, it undergoes comprehensive testing to verify its performance, safety features, and battery health. Functional and performance tests assess acceleration, braking, energy consumption, and more.

The assembly of electric vehicles presents unique opportunities and challenges:

1. Integration Expertise:  

Integrating complex electric powertrains requires specialized engineering expertise, particularly in terms of thermal management and electronic architecture.

2. Battery Safety:  

Ensuring the safe integration and protection of high-voltage battery systems is paramount for EVs' success.

3. Lightweight Materials:  

Electric vehicles often employ lightweight materials to offset the weight of the battery, further requiring advanced assembly techniques.

4. Enhanced User Experience:  

EVs offer an array of digital features that require seamless integration for an intuitive and enjoyable user experience.

Tier 1 Suppliers 

Tier 1 suppliers hold a pivotal role in the EV ecosystem, responsible for delivering advanced solutions that define the performance, efficiency, and user experience of electric vehicles. These suppliers possess the expertise to design, engineer, and manufacture complex components that propel EVs into the future.

1. Electric Drivetrains:

The heart of an electric vehicle, the electric drivetrain, includes components like electric motors, power electronics, and transmission systems. Tier 1 suppliers design and produce these integrated systems that govern power delivery, torque, and energy efficiency, shaping the driving experience.

Leading Tier 1 Suppliers:

• BorgWarner Inc.

• Aisin Seiki Co., Ltd.

• Magna International Inc.

• Siemens AG

2. Battery Systems:

Tier 1 suppliers play a critical role in crafting battery systems that define the range, charging speed, and longevity of electric vehicles. These systems include battery packs, thermal management solutions, and energy management systems.

Leading Tier 1 Suppliers:

• LG Chem

• CATL (Contemporary Amperex Technology Co. Ltd.)

• Panasonic Corporation

• Samsung SDI Co., Ltd.

3. Charging Systems and Infrastructure:

As EV adoption rises, charging systems and infrastructure become essential. Tier 1 suppliers contribute to charging solutions, including hardware, software, and network management systems that enable efficient and rapid charging.

Leading Tier 1 Suppliers:

• ABB Ltd.

• Siemens AG

• Schneider Electric SE

• ChargePoint, Inc.

4. Electronics and Connectivity:

Electric vehicles rely heavily on advanced electronics and connectivity solutions to manage power distribution, energy flow, and communication between various vehicle systems.

Leading Tier 1 Suppliers:

• Delphi Technologies (Aptiv PLC)

• Continental AG

• Infineon Technologies AG

• Bosch Automotive Solutions

5. Infotainment and User Experience:

Tier 1 suppliers contribute to the user experience by developing infotainment systems, touchscreen interfaces, advanced driver assistance systems (ADAS), and connected vehicle technologies.

Leading Tier 1 Suppliers:

• Harman International Industries, Inc. (a Samsung company)

• Panasonic Corporation

• Visteon Corporation

• Denso Corporation

6. Lightweight Materials and Components:

Given the focus on efficiency, Tier 1 suppliers contribute lightweight materials, such as advanced composites and aluminum, to reduce vehicle weight and improve energy efficiency.

Leading Tier 1 Suppliers:

• Magna International Inc.

• Novelis Inc.

• Toray Industries, Inc.

• SGL Carbon SE

Tier 1 suppliers encounter several challenges while striving for excellence:

1. Innovation Pace:  

Keeping up with rapid advancements in EV technology demands continuous innovation in design, materials, and manufacturing processes.

2. Scalability and Supply Chain:  

As demand for electric vehicles increases, Tier 1 suppliers must ensure the scalability and resilience of their supply chains to meet production targets.

3. Collaboration and Compatibility:  

Components from different Tier 1 suppliers must seamlessly integrate to ensure the vehicle's overall functionality and performance.

4. Regulatory Compliance:  

Meeting global safety and emissions regulations poses challenges in the design and production of electric vehicle components.

Tier 1 suppliers are the architects of the electric vehicle renaissance, driving innovation through their specialized components that power, connect, and enhance electric vehicles. As the EV landscape evolves, these suppliers will continue to redefine the driving experience, pushing the boundaries of technology and sustainability.

Tier 2 & 3 Suppliers 

While Tier 1 suppliers take the spotlight, Tier 2 and Tier 3 suppliers are the backbone of the electric vehicle ecosystem. They specialize in delivering a wide range of components, each contributing to the seamless functioning of EVs' complex systems. From intricate sensors to resilient materials, these suppliers leave an indelible mark on electric mobility.

1. Components for Electric Drivetrains:

Tier 2 and Tier 3 suppliers contribute to the drivetrain ecosystem by providing motors, inverters, and power electronics components. These players craft specialized parts that optimize energy conversion, enhancing overall efficiency and performance.

2. Battery Components and Materials:

Beyond the battery pack itself, suppliers at these tiers provide materials for anode and cathode electrodes, separators, electrolytes, and thermal management solutions. Their innovations shape battery chemistry and longevity.

3. Wiring and Connectors:

The complex network of wiring and connectors that interconnects various systems relies on components supplied by Tier 2 and Tier 3 suppliers. Their contributions ensure reliable communication and power distribution throughout the vehicle.

4. Thermal Management and Cooling Solutions:

Efficient thermal management is crucial for the safe and optimal operation of EVs. These suppliers deliver cooling systems, heat exchangers, and advanced materials to regulate temperature and prevent overheating.

5. Interior and Comfort Systems:

Tier 2 and Tier 3 suppliers provide interior components such as seats, dashboard components, infotainment systems, climate control systems, and interior lighting that enhance the driver and passenger experience.

6. Safety and Driver Assistance Systems:

Suppliers at these tiers contribute to advanced driver assistance systems (ADAS), sensors, cameras, LiDAR, and radar components that enable features like adaptive cruise control and lane-keeping assistance.

7. Suspension and Chassis Components:

Steering systems, suspension components, shock absorbers, and chassis components are sourced from Tier 2 and Tier 3 suppliers. These components influence ride comfort and handling.

8. Exterior Body Components:

From panels to lighting systems, Tier 2 and Tier 3 suppliers provide the exterior components that shape the vehicle's aesthetics, aerodynamics, and visibility.

Tier 2 and Tier 3 suppliers encounter distinct challenges and innovations:

1. Customization and Diversification:  

The diverse needs of EV manufacturers demand flexible and customizable solutions, requiring suppliers to offer a wide range of components.

2. Quality Control and Standards:  

Ensuring high-quality components that meet stringent automotive standards is paramount for maintaining safety and performance.

3. Innovation and Collaboration:  

Tier 2 and Tier 3 suppliers need to stay abreast of technological advancements and collaborate closely with Tier 1 suppliers to ensure seamless integration.

4. Sustainability and Efficiency:  

Suppliers are increasingly focusing on sustainable materials and manufacturing processes to align with the environmental ethos of electric mobility.

The evolution of electric vehicles relies heavily on the intricate contributions of Tier 2 and Tier 3 suppliers. Their specialized components, systems, and sub-assemblies are the unsung heroes that enable the electric revolution to flourish.

Acknowledging the pivotal role played by these suppliers in creating a cohesive ecosystem unveils the complex interplay between innovation, collaboration, and dedication. As the electric mobility landscape advances, Tier 2 and Tier 3 suppliers remain the silent architects of the transformation, shaping the vehicles that are propelling us into a sustainable and electrifying future.

Distribution and Logistics 

The journey of an electric vehicle doesn't conclude on the assembly floor; it's a continuation, orchestrated by the intricate web of distribution and logistics. The seamless movement of components, sub-assemblies, and completed vehicles from suppliers to dealerships is a harmonious symphony that demands precision, adaptability, and sustainability.

1. Parts and Component Logistics:

The EV supply chain begins with the movement of components and parts from Tier 1, Tier 2, and Tier 3 suppliers to the assembly plants. This process involves efficient sourcing, inventory management, and just-in-time delivery to minimize storage costs and ensure timely availability.

2. Vehicle Distribution:

Once assembled, electric vehicles are transported to dealerships or directly to customers. The distribution process requires meticulous planning to optimize routes, ensure on-time delivery, and maintain the quality and integrity of the vehicles during transit.

3. Challenges and Innovations:

a. Charging Infrastructure for Transport:

EVs often need to be transported over long distances, necessitating charging infrastructure along routes to maintain optimal battery levels. The establishment of charging stations at distribution centers and along transportation routes is a key innovation.

b. Range Considerations:

Balancing the range capabilities of electric vehicles with transportation distances is crucial. Innovations include route optimization to minimize stops and maximize distance covered within a single charge.

c. Vehicle Loading and Unloading:

Ensuring safe loading and unloading of EVs requires specialized equipment and training due to the unique characteristics of electric vehicles, such as high-voltage systems.

d. Supply Chain Visibility:

Real-time tracking and monitoring technologies provide supply chain stakeholders with visibility into the status and location of vehicles during transportation, enhancing efficiency and accountability.

e. Last-Mile Delivery:

For urban areas, the challenge lies in last-mile delivery to dealerships or customers' doorsteps. Electric delivery vehicles are emerging as an eco-friendly solution, aligning with the ethos of electric mobility.

f. Regulatory Compliance and Customs:  

Navigating international borders requires compliance with various regulations and customs procedures. Innovations in digital documentation and tracking streamline these processes.

4. Sustainable Logistics:

a. Green Transport Solutions:

Promoting sustainability in logistics involves adopting electric or hydrogen-powered trucks for transportation, reducing emissions and the carbon footprint.

b. Multimodal Transport:  

Combining various modes of transportation, such as rail and sea, can reduce the environmental impact of long-distance vehicle transport.

c. Collaboration and Consolidation:  

Collaboration among manufacturers, logistics providers, and transport companies can lead to consolidated shipments and optimized routes, reducing waste and emissions.

d. Packaging and Recycling:  

Sustainable packaging and recycling practices minimize waste and contribute to the circular economy by repurposing materials.

Charging Infrastructure 

The electrification of transportation hinges on the availability and reliability of charging infrastructure. This network serves as the lifeblood of electric mobility, enabling EV owners to recharge their vehicles conveniently and efficiently. The landscape of EV charging infrastructure encompasses various types and levels of chargers, each catering to diverse user needs.

1. Types of Charging Infrastructure:

a. Level 1 Charging (Residential Charging):

Level 1 charging involves plugging an EV into a standard household electrical outlet. While it's the slowest charging option, it's ideal for overnight charging at home, providing about 3-5 miles of range per hour of charging.

b. Level 2 Charging (Residential and Public Charging):

Level 2 charging stations deliver power at a faster rate, typically through a dedicated charging unit installed at homes or public locations. This option provides approximately 10-20 miles of range per hour of charging.

c. DC Fast Charging (Public Charging):

DC fast charging, or Level 3 charging, is the most rapid option available. It's commonly found at public charging stations and offers high-power charging, delivering 60-80 miles of range in just 20 minutes.

2. Importance of Charging Infrastructure:

a. Range Anxiety Mitigation:

A widespread charging network alleviates range anxiety by assuring drivers of available charging options during long journeys.

b. Encouraging Adoption:

An extensive and reliable charging infrastructure incentivizes potential EV buyers, knowing that charging won't be a hindrance to their daily routines.

c. Sustainable Transportation:

Charging infrastructure complements the environmental benefits of EVs by relying on renewable energy sources, reducing emissions, and supporting the transition to clean energy.

3. Challenges and Innovations:

a. Infrastructure Investment:

Setting up charging infrastructure demands significant investment in terms of capital and installation.

b. Charging Speed and Convenience:

Innovations are focused on faster charging technologies, battery technology advancements, and enhancing user experience through app-based charging information and remote management.

c. Grid Capacity and Stability:

Rapid adoption of EVs could strain the power grid. Smart charging solutions that manage load distribution and time-of-use pricing are being explored to optimize grid stability.

d. Urban and Rural Accessibility:

Balancing the distribution of charging stations in urban centers and rural areas is essential for comprehensive EV adoption.

e. Interoperability and Standardization:

Creating uniform standards for charging connectors, payment systems, and access methods ensures a seamless experience for all EV owners.

4. Innovations and Future Prospects:

a. Ultra-Fast Charging:

Research is ongoing to develop ultra-fast charging technologies that could provide hundreds of miles of range within minutes.

b. Wireless Charging:

Inductive and wireless charging technologies aim to eliminate the need for physical cables, enabling effortless charging.

c. V2G (Vehicle-to-Grid) Integration:

V2G technology enables EVs to return energy to the grid, contributing to grid stability and providing an additional revenue stream for EV owners.

d. Charging Integration with Renewable Energy:

Combining charging stations with solar panels and energy storage systems can offer sustainable and grid-independent charging solutions.

The electrified future hinges on the transformative power of charging infrastructure. This intricate network has the potential to reshape transportation, making EVs a viable choice for drivers across the globe. From residential to public charging stations, and from fast chargers to wireless solutions, the innovation in EV charging infrastructure underscores the dynamic journey towards sustainable mobility.

After Sales Support 

The shift to electric vehicles demands a new paradigm of after-sales support that caters to the unique needs of EV owners. This support covers a spectrum of services, from routine maintenance to addressing technical concerns, ensuring a seamless and satisfying ownership journey.  This stage is crucial for many reasons but especially for inspiring customer satisfaction and loyalty. 

1. The Significance of After-Sales Support:

a. Range of Services:

After-sales support encompasses a wide array of services, including regular maintenance, battery health checks, software updates, repairs, and technical assistance.

b. Building Consumer Confidence:

Comprehensive after-sales support instills confidence in consumers, reassuring them that their EV ownership experience will be well-supported.

c. Customer Loyalty and Brand Reputation:

Providing excellent after-sales support fosters customer loyalty, contributing to positive word-of-mouth and enhancing the reputation of EV manufacturers.

2. Challenges and Innovations:

a. Specialized Training:

Service technicians need specialized training to diagnose and repair electric vehicle components, including high-voltage systems.

b. Battery Health Management:

Developing effective strategies for battery health management is crucial to prolong the lifespan of EVs and maintain optimal performance.

c. Charging and Infrastructure Support:

Helping EV owners navigate charging options, install home charging stations, and troubleshoot charging-related issues is essential.

d. Remote Diagnostics and Telematics:

Innovations in remote diagnostics and telematics enable proactive monitoring of vehicle health and identifying potential issues before they escalate.

e. Spare Parts Availability:

Ensuring the availability of genuine spare parts is vital for timely repairs and maintenance.

3. Innovations and Future Prospects:

a. Predictive Maintenance:

Using data analytics and AI, manufacturers can predict maintenance needs, optimizing scheduling and reducing downtime.

b. Mobile Service Units:

Mobile service units equipped to handle routine maintenance and minor repairs at the customer's location enhance convenience.

c. Augmented Reality Support:

AR technology can assist technicians in diagnosing and repairing complex issues by providing real-time visual guidance.

d. Over-the-Air Updates:

Software updates sent directly to the vehicle enable continuous improvement and new feature introduction without requiring a visit to the service center.

e. Enhanced Customer Engagement:

Manufacturers are using digital platforms to engage with customers, providing them with information, tips, and support throughout their ownership journey.

After-sales support is the bridge between innovation and ownership in the electric vehicle realm. The seamless integration of maintenance, technical assistance, and innovative solutions plays a pivotal role in driving the EV revolution forward.

Government Incentives for EVs

Increasingly, policy across the world is aiming to boost the manufacturing of electric vehicles, not just their deployment.  Governments are finding various ways to support the entire supply chain for EVs, as the efforts are seen as a “win win” by furthering sustainability goals in line with net zero by 2050 (or a comparable emissions target) and boosting domestic economic production.

In China (the largest market for electric cars) support for the EV industry has been going on for some time. Over the past ten years the Chinese Government has made various efforts on both the supply and demand side to support domestic companies as well as engaging in joint ventures with international automobile companies. Support has been particularly stimulative at the local level which has led to the development of some major EV companies growing and succeeding in China. Individual regions have also set targets for EV goals, for instance Chongqing has a goal of producing and selling more than 10% of China’s new energy vehicles and Jilin is aiming for an annual production capacity of 1 million EVs a year by 2025. Goals that are supported by the Chinese Government’s investment plans.

In the United States, the Inflation Reduction Act (which was passed in August 2022) included tax incentives and various funding programs for clean energy initiatives. Overall, there was $369B designated for climate investments. The Clean Vehicle Tax Credit makes some changes to current EV incentive programs. It states that starting in 2023, the final assembly of the electric vehicle must occur in North America, and that vehicles must have a 7 kWh battery or greater (to exclude low-range plug-in hybrid electric vehicles [PHEVs]), be under 6.35 t gross vehicle weight (GVW), and have a suggested retail price of less than $80k for vans, SUVs and pickup trucks, or $55k for other vehicles. In order to qualify for the incentive, the EV buyer’s household income must be below the limit set by the US Internal Revenue Service. These conditions open eligibility for an incentive of up to $7500 per vehicle: $3750 if the battery meets the critical mineral requirement, and another $3750 if it meets the component requirement. Furthermore, from 2025, vehicles with any critical minerals from “foreign entities of concern” will not be eligible for the credit, and vehicles with battery components from such entities will be ineligible from 2024.This critical mineral requirement stipulates that: 1) in 2023, 40% or more of the battery critical minerals must be extracted or processed in the United States or a US free trade country, or have been recycled in North America, gradually increasing to 80% in 2027 and beyond; and 2) in 2023, 50% of the value of the components in the battery must be manufactured or assembled in North America, gradually increasing to 100% in 2029 and beyond. (Data from the IEA)

The update also removed the sales cap of 200k, which allows General Motors and Tesla to participate in the program. The average listed price of the models registered for the program is just below the $55k and $80k limits for cars and SUVs respectively, indicating the manufacturers intent to participate in the program even if it means cutting prices. There is also an additional $1B for heavy duty vehicles. The Inflation Reduction Act also contains supply side tax credits for advanced manufacturing, providing subsidies for domestic battery manufacturing of up to $35 per kWh and another $10/kWh for model assembly. With average battery prices in 2022 around $150/kWh, these subsidies may account for almost a third of battery prices. Also, in January of this year, the United States signed a memorandum of understanding with The Democratic Republic of Congo and Zambia, with a commitment to develop a productive supply chain for electric vehicles from mining to assembly.

As far as efforts in Europe go, in February 2023, “the European Union presented the Green Deal Industrial Plan, which has four pillars related to progress on net zero-related projects: faster permitting, financial support, enhanced skills, and open trade. The plan also includes provision for the creation of a Critical Raw Materials Act, the proposal for which was issued in March 2023, with a focus on security of supply, extraction and environmental standards, as well as recycling.

The faster permitting for facilities, including battery production, will be formalized via the proposed Net Zero Industry Act, providing simplified and predictable planning approvals. As well as loosening rules on state aid until 2025, the financial support package of the plan attempts to allow faster access to essential subsidies and loans, to compensate businesses for high energy prices, to help ensure liquidity, and to reduce electricity demand. The plan also aims to reskill workers affected by the green transition, and to establish Net Zero Industry Academies. Lastly, the trade element focuses on improving the resilience of the EU’s supply chains, opening trade with new partners and attracting private investment.

​​In March 2023, the European Union proposed the Net Zero Industry Act, which aims to meet 40% of the European Union’s needs for strategic net zero technologies with EU manufacturing capacity by 2030. These technologies explicitly include battery and storage technologies, and for batteries the aim is for nearly 90% of the European Union’s annual battery demand to be met by EU battery manufacturers, with a combined manufacturing capacity of at least 550 GWh in 2030, in line with the objectives of the European Battery Alliance. These announcements came just as CO2 standards for car sales over 2030-2035 tightened under the Fit for 55 package.” - IEA

There are efforts from countries all across the world to put forth similar government programs to increasingly incentivize the development and production of electric vehicles, as well as secure their supply chains for the critical minerals required for production. For brevity sake we only highlighted the efforts of China, Europe and the US here, but lots of other countries across the globe are creating their own fiscal programs to stimulate electric vehicle initiatives. 

Conclusion: Navigating the Electric Road Ahead

In the search for the inevitable, in both technology and financial markets, it’s hard to find a trend that has as many concurrent tailwinds as electric vehicles. If you were to say “inevitably, there will be a much larger number of EVs on the road in 5-10 years” it would be difficult to debate that statement. There are of course concerns, a main one for consumers (particularly in the US) is the range that EVs can travel on one charge and how long it takes to charge the vehicle.  However, with investment from the entire automotive industry (as they attempt to catch up to Tesla’s massive lead in the space) and serious fiscal programs designed to support the industry with subsidies and tax incentives amongst other benefit, it is hard to imagine the industry not experiencing strong secular growth, at least in the near to mid term horizons. This overview was meant as a general primer to the electric vehicle supply chain, and was written as I conducted research in my own “search for the inevitable.” From the mining of the critical materials (this process seems to be inherently inflationary and is evident by looking at price charts of the minerals) to their refining and eventual assembly, the supply chain for electric vehicles is certainly complex, but is also a battleground that is sure to contain some great investments for those who can do the work to uncover what companies stand to be the largest beneficiaries of all the committed investment capacity, both from industry and the government. There is also the question of some second order effects in the future that are definitely worth exploring - what % of EVs vs ICE (internal combustion engine) vehicles wi there be when EVs reach 100% in some markets? What will the long term impact of self driving EVs be? Right now, utilization of cars is relatively low but this may change as autonomous driving becomes increasingly common. This may have an impact on oil markets and also things like parking garages. Also, how does the electric grid need to evolve in order to handle EV charging at scale? What is the resale value of EVs? What does the distribution look like in terms of consumers buying an EV for their first car vs their second or third? Is there an ample supply of mechanics and parts for servicing electric vehicles? These are all topics that merit further exploration and discussion in a future piece, as we look to find the best way to invest in these larger trends. I hope you found this overview somewhat informative, I’ll try to explore some of these themes more in depth in a future post where I focus more on the company level and which companies might be good investments on certain time horizons (I’ll try to do the same with the networking and telecom piece, as it reflects how I do research when I am attempting to first identify the large scale trend and then zoom in and try to understand stand which companies will inevitably be the winners).  If you enjoyed this piece, please consider subscribing and don’t hesitate to reach out on Twitter @netcapgirl.

Relevant Companies

Battery Manufacturers:

• CATL (Contemporary Amperex Technology Co. Ltd.)

• LG Chem

• Panasonic

• SK Innovation

• BYD Company Limited

Automakers:

• Tesla, Inc.

• General Motors

• Volkswagen Group

• BMW Group

• Nissan Motor Corporation

• Ford Motor Company

• Hyundai Motor Group

• Rivian Automotive, Inc

Semiconductor Providers:

• NVIDIA Corporation

• Qualcomm Incorporated

• Infineon Technologies AG

• Texas Instruments Incorporated

• Broadcom Inc.

Charging Infrastructure:

• ChargePoint, Inc.

• Tesla Superchargers Network

• EVgo Services LLC

• ABB Ltd.

• Siemens AG

Materials and Mining:

• Albemarle Corporation

• BHP Group Limited

• Rio Tinto Group

• Glencore plc

Energy Storage Solutions:

• Enphase Energy, Inc.

• Sonnen GmbH

• AES Corporation

• Stem, Inc.

• Fluence Energy, LLC

Utilities and Grid Infrastructure:

• Duke Energy Corporation

• E.ON SE

• Enel S.p.A.

• Dominion Energy, Inc.

• National Grid plc

Automotive Technology and Software:

• Mobileye (Intel Corporation)

• Aptiv PLC

• BlackBerry Limited

• Daimler AG (Mercedes-Benz)

• Rivian Automotive, Inc.

Geopolitics and Critical Minerals:

• China Molybdenum Co., Ltd.

• Albemarle Corporation

• Ganfeng Lithium Co., Ltd.

• SQM (Sociedad Química y Minera de Chile)

• Rio Tinto Group

Renewable Energy Providers:

• NextEra Energy, Inc.

• Vestas Wind Systems A/S

• Orsted A/S

• Enel Green Power S.p.A.

• First Solar, Inc.

• Overview of Article

• Introduction

• History 

• AI Bottlenecks - Inter GPU Bandwidth 

• NVLink & Infiniband, Google’s Data Center Efforts

• ISPs / Carriers, Starlink

• Edge Computing, Dojo

• Networking Companies & Component Providers 

• Data Centers

• Conclusion

One goal of these thematic pieces is to provide sufficient context for a generalist to get up to speed with what’s happening in an industry or with a trend while also hopefully providing some useful information that someone close to the industry would appreciate. These longform pieces are the result of my own personal research into industries and trends with the goal of approaching them like a scientist might, hoping to form hypotheses and test them in the market in hopes that we may get ever closer to discovering what may be “inevitable” in the future.  Even then we must conduct careful analysis because as history has shown, just discovering a prominent trend isn't enough, you need to understand the underlying mechanics in order to grasp which companies are truly poised to benefit. For instance, if someone had told you in the 1990s the number of cellphones that would be in use in 2023, you might have bought Ericsson and Nokia. The purpose of this research is to arm investors with a framework for thinking about companies, industries and trends and hopes to over time, develop a process for searching for what might be inevitable. I aim to write the piece I wish that I had when doing research and ramping up on a new industry and while this is generally a high level overview of the different industries (and companies) within the broader networking/telecom sector, I plan on exploring these topics individually, in a more in depth manner in the future.

Introduction

Before computers, the internet or even telephones, sending information in various forms has been a core part of human existence. In this piece, we will take a look at the current state of various industries and companies within them that traffic in the transfer of data.  We’ll explore some of the history of the networking and telecommunications industry, examine the current state of the industry and try to get a sense of where companies are investing their capital with the hope of discovering what might be “inevitable” in the future.  We’ll take a look at AI bottlenecks (inter GPU bandwidth) and other current networking challenges and investments ,Edge Computing, Internet Service Providers & Carriers, Networking Companies, Data Centers, and Component Providers amongst other topics and highlight some relevant companies in each sector.

History 

These are some of the major scientific and technological breakthroughs that paved the way for the modern computer networking and telecommunications industries:

1. 1837 - Telegraph and Morse Code:

Samuel Morse and Alfred Vail developed the telegraph system, which allowed messages to be transmitted over long distances using electrical signals. Morse code, a system of dots and dashes, was used to encode and decode messages.

2. 1876 - Telephone Invention:

Alexander Graham Bell and Elisha Gray independently developed the telephone, enabling voice communication over electrical wires. Bell is widely credited with being the first to patent the telephone.

3. 1895 - Wireless Telegraphy (Radio):

Guglielmo Marconi conducted experiments in wireless telegraphy, which led to the development of radio communication. He sent the first radio signal across the English Channel in 1899.

4. 1920s - Early Broadcasting:

The 1920s saw the beginning of radio broadcasting, allowing people to receive news, music, and entertainment over the airwaves.

5. 1960s - ARPANET and Packet Switching:

The Advanced Research Projects Agency Network (ARPANET) was created by the U.S. Department of Defense. It used packet switching, a method of breaking down data into packets for efficient transmission. ARPANET laid the foundation for the modern internet.

6. 1973 - Ethernet Invention:

Robert Metcalfe and his team at Xerox PARC invented Ethernet, a technology for connecting computers in local networks. Ethernet became the standard for wired local area networks (LANs).

7. 1983 - TCP/IP Standardization:

The Transmission Control Protocol (TCP) and Internet Protocol (IP) were standardized, providing a common language for computers to communicate over networks. This was crucial for the growth of the internet.

8. Late 1980s - World Wide Web:

Tim Berners-Lee developed the World Wide Web, a system of interconnected hypertext documents accessed via the internet. This marked the beginning of user-friendly internet browsing.

9. 1990s - Commercialization of the Internet:

The 1990s saw a surge in internet adoption and the rise of commercial internet service providers (ISPs). This period marked the transition of the internet from a research tool to a global communication and commerce platform.

10. Late 1990s - Mobile Communication and GSM:

The Global System for Mobile Communications (GSM) standard enabled digital mobile communication. This paved the way for the widespread use of cell phones and text messaging.

11. 2000s - Broadband Internet and Wi-Fi:

Broadband internet, offering high-speed connectivity, became widely available. Wi-Fi technology allowed wireless internet access within short ranges, transforming how people connected to the internet.

12. 2010s - 4G and Smartphones:

The deployment of 4G (fourth-generation) mobile networks enabled faster data speeds and facilitated the widespread use of smartphones, which became essential devices for communication and accessing online services.

13. 2020s - 5G and IoT:

The rollout of 5G (fifth-generation) networks began, promising even higher data speeds and lower latency. This is expected to support innovations like the Internet of Things (IoT), where various devices are interconnected for data sharing and automation.

These breakthroughs, spanning over centuries, collectively shaped the computer networking and telecommunications industries into what we know today, enabling seamless global communication, information sharing, and technological advancements.

Inevitably, AI training is constrained by inter GPU bandwidth 

One area of interest in modern networking and telecommunications is how the training of large AI models is inevitably constrained by inter GPU bandwidth.  Say you’re a new AI startup who has just been funded and you plan on training a foundation model. You spin up some ND H100 virtual machine instances on Azure to train your mode, which is 8 H100s initially but can scale up to thousands with an upper limit of 3.2 Tb/s of interconnect bandwidth per virtual machine (each GPU in a v5 instance has a 400 Gb/s Infiniband connection.)  As one of Microsoft’s flagship cloud offerings for AI, we can understand how the upper limit of bandwidth in these cloud instances represents an inevitable bottleneck for training AI models and an area of interest for the companies participating in this arms race to invest in.

AI training involves training large neural networks on vast amounts of data to learn patterns and make accurate predictions or classifications. This process is computationally intensive and requires significant processing power. Graphics Processing Units (GPUs) are commonly used for this purpose due to their parallel processing capabilities, which allow them to handle the massive amounts of matrix calculations involved in neural network training.

When training neural networks, data is divided into batches, and these batches are processed simultaneously on the GPU's many cores. However, during this process, there is a need for frequent communication between the cores or GPUs. This communication involves exchanging intermediate results, gradients, weights, and other information to update the model effectively. This communication happens through what's called the inter-GPU communication or inter-GPU bandwidth.

Inter-GPU bandwidth refers to the speed at which data can be transferred between different GPUs within a system. If this bandwidth is low, it can lead to performance bottlenecks and slower training times. Here's why inter-GPU bandwidth is a constraint in AI training:

1. Data Parallelism: In many AI training scenarios, a technique called data parallelism is used. This involves splitting the dataset into multiple parts and processing each part on a separate GPU simultaneously. The GPUs then need to communicate their updates to the model's parameters and gradients. If the inter-GPU bandwidth is limited, this communication can slow down the overall training process.

2. Gradient Descent: Neural networks are typically trained using optimization algorithms like stochastic gradient descent. In each iteration, gradients are computed with respect to the loss function for each batch of data. These gradients need to be synchronized across GPUs to update the model's parameters properly. A slow inter-GPU communication can delay this synchronization and increase the time it takes to reach convergence.

3. Model Parallelism: In certain cases, especially when dealing with very large models, model parallelism might be employed. This involves dividing the neural network itself into segments and distributing these segments across different GPUs. Again, efficient communication between these GPUs is essential for synchronization and proper functioning of the model.

4. Large Models: Deep learning models are becoming increasingly large and complex, with billions of parameters. Training these models involves even more data exchange between GPUs, putting additional strain on inter-GPU bandwidth.

5. Multi-Node Training: In distributed training scenarios, where multiple machines are connected to work together, communication between nodes is crucial. Inter-GPU bandwidth becomes inter-node bandwidth in this case. Slow communication between nodes can lead to poor scaling and reduced efficiency.

Efficient communication between GPUs is essential for maintaining parallel processing and avoiding idle times where GPUs are waiting for data from others. Researchers and engineers often strive to optimize inter-GPU communication, either through hardware improvements or software techniques, to minimize the impact of this constraint on AI training speed and efficiency.  

Considering the arms race that is AI training, companies providing the “picks and shovels” are investing heavily in innovating in the space.  Nvidia, for instance, is investing in further developing its Infiniband and NVlink technologies. 

NVIDIA's InfiniBand and NVLink are two different technologies that are used to improve interconnectivity and communication between GPUs and other components in high-performance computing (HPC) and data-intensive applications. Both technologies aim to provide high-speed and low-latency connections, but they have different design principles and use cases.

1. InfiniBand:

InfiniBand is a high-speed networking technology designed for connecting servers and high-performance computing systems. It's not exclusive to NVIDIA and is used across the industry. InfiniBand provides a hh-speed, low-latency interconnect solution that is commonly used in cluster computing and supercomputing environments.

   Key features of InfiniBand:

• High Bandwidth: InfiniBand supports very high data transfer rates, ranging from gigabits per second (Gbps) to terabits per second (Tbps), depending on the specific implementation.

• Low Latency: InfiniBand offers low communication latency, which is crucial for applications that require rapid data exchange, such as scientific simulations and real-time analytics.

• Message Passing Interface (MPI) Support: Many high-performance applications, especially those used in scientific research, utilize the MPI standard for distributed computing. InfiniBand provides efficient support for MPI communication.

• Clustering and Supercomputing: InfiniBand is often used in cluster computing and supercomputing environments to connect multiple nodes and GPUs for high-performance computations.

2. NVLink:

NVLink is a proprietary high-speed interconnect technology developed by NVIDIA specifically for connecting their GPUs together. It's designed to provide faster and more efficient communication between GPUs than traditional PCIe (Peripheral Component Interconnect Express) connections.

   Key features of NVLink:

• High Bandwidth and Low Latency: NVLink offers higher bandwidth and lower latency compared to traditional PCIe connections. This is particularly beneficial for scenarios where GPUs need to exchange a significant amount of data, such as in deep learning model training.

• Direct GPU-to-GPU Communication: NVLink allows GPUs to communicate directly with each other without passing through the CPU or host system, reducing communication bottlenecks.

• Unified Memory Space: NVLink enables GPUs to share memory more seamlessly, making it easier to distribute and manage data across multiple GPUs.

• Multi-GPU Scalability: NVLink facilitates building large-scale multi-GPU systems for tasks like AI training, simulation, and scientific computing.

• Data-Intensive Workloads: NVLink is particularly advantageous for workloads that involve heavy data exchange between GPUs, such as in deep learning training and complex simulations.

Both InfiniBand and NVLink technologies are designed to improve interconnectivity and communication between components in high-performance computing environments. InfiniBand is a general-purpose networking technology used for connecting servers and clusters, while NVLink is NVIDIA's proprietary technology optimized for connecting GPUs together with a focus on high-speed and low-latency communication, making it especially valuable for data-intensive tasks like deep learning and scientific simulations. Infiniband (a Mellanox product, a company that Nvidia acquired in 2019 for $6.9B) is designed for somewhat longer distances than NVLink as in any physical medium, there’s a natural and inevitable trade off between bandwidth and length. Whether to use copper or fiber is a critical point that researchers in the industry are focused on and being able to put optics actually “on the board” is a sort of holy grail because not only would it create incredibly fast speeds compared to copper, but it would also reduce power consumption per bit (saving energy in data centers) and expand bandwidth by achieving higher channel densities. On board optics is still extremely expensive to produce but it is certainly an area worth watching (an perhaps a topic for a follow up piece), especially as the AI arms race accelerates investment in the space, as it would be an extraordinary (and potentially revolutionary) leap forward for optical technology.

Google’s Project Apollo

Google has been investing heavily in building custom data center networking infrastructure (Project Apollo) to support its massive scale of operations and to provide the necessary performance, reliability, and efficiency for its services. One of the key components of Google's data center networking approach is its use of software-defined networking (SDN) principles (something that Arista Networks is also using to win market share). Here's an overview of Google's efforts:

1. Jupiter Network Architecture:

Google's custom networking infrastructure is based on a design called the "Jupiter" network architecture. The Jupiter architecture is designed to handle the enormous amount of data traffic that Google's services generate. It employs a two-tier Clos network topology that provides high bisection bandwidth and minimizes the number of hops data needs to travel between servers.

2. Software-Defined Networking (SDN):

Google heavily relies on SDN principles to manage and control its data center networks. SDN separates the network's control plane from the data plane, enabling centralized control and dynamic management of network resources. Google uses an SDN controller to dynamically allocate and manage network resources based on traffic patterns and demand.

3. B4 Network Infrastructure:

B4 is Google's global backbone network that connects its data centers around the world. It uses high-capacity links to provide low-latency and high-speed connectivity between data centers. The B4 network is designed to handle both user-facing traffic and backend communication between Google's services.

4. Andromeda Virtual Network Stack:

Google developed the Andromeda network virtualization stack to provide networking capabilities to virtual machines (VMs) running on Google Cloud Platform. Andromeda allows Google to offer features like software-defined firewalls, load balancing, and network performance monitoring to its cloud customers.

5. Custom Networking Hardware:

Google designs its own networking hardware, including switches and routers, to match its specific requirements. By designing its hardware, Google can optimize for performance, energy efficiency, and cost-effectiveness based on its data center workloads. Specifically, they are designing optical circuit switches to replace the traditional electronic packet switches they would buy from Broadcom.

6. OpenFlow and OpenConfig:

Google has been involved in initiatives like OpenFlow and OpenConfig, which aim to standardize and open up network control and configuration protocols. OpenConfig, for instance, provides vendor-neutral network configuration models that align with Google's SDN approach.

7. Data Center Interconnect (DCI) Innovations:

Google has explored innovative ways to interconnect its data centers, including using technologies like optical networking for long-distance, high-capacity connections between data centers. This helps Google maintain high-speed, low-latency communication between its data centers.

8. Network Resilience and Redundancy:

Google's network architecture is designed for redundancy and fault tolerance. It incorporates mechanisms to handle hardware failures, reroute traffic, and maintain service availability even in the face of network disruptions.

Google's efforts to build custom data center networking infrastructure align with its need to deliver reliable and performant services to its users, whether it's search, cloud computing, or other services. These efforts allow Google to have more control over its network, optimize it for its workloads, and ensure that its services are delivered with the quality and scale required by its global user base. Google claims they are able to save $3B vs comparable data center infrastructures and also that their custom network improves throughput by 30%, uses 40% less power, incurs 30% less Capex, reduces flow completion by 10%, and delivers 50x less downtime across their network. It also eliminates the need for them to purchase Broadcom switches for the spine of their networking stack. Google is using this network infrastructure to train its advanced large language models. Google is using internally developed optical circuit switches to replace the traditional electronic packet switches as they offer lower latency (due to not needing to decode packets), ease of upgrading (you no longer need to replace the entire spine) and interoperability between optical switches of different speeds as well as lower power consumption.

There is a fantastic semianalysis article that covers this in more detail.

Relevant Companies

• Nvidia

• Microsoft

• Google

• Amazon

• Meta

• Arista 

• Broadcom

• Cisco

• Juniper 

• Marvell 

• OpenAI (Private)

ISPs & Carriers - 5G, Fiber, FWA, Streaming 

Most people are probably familiar with the process of signing up for a mobile plan and home internet.  They have essentially become requirements for participating in the modern world (a trend that was accelerated by work from home policies during Covid.)  Historically, there was greater delineation between the companies that provided your mobile internet and the companies that provide your cable and home internet. However, as consumer preferences shift, the competition for providing internet access and video entertainment services has heated up and companies are scrambling to innovate to not be left behind. There are some key trends happening in the telecom space 

• Fixed wireless access vs fiber 

• Bundling wireless and home internet (somewhat of a race to zero)

• Streaming and video becoming the main traffic over the internet 

• Telecom companies divesting media assets (AT&T spinning out Warner Brothers) and a sort of war between the cable providers and the media companies (some of whom are also entering the streaming game, Charter vs Disney, for instance. Also perhaps relevant from a few years back was SK Telecom vs Netflix, when Netflix traffic surged due to people watching Squid Game, SK Telecom was on the hook for the increase in cost. Netflix (Media/Tech Company) benefitted at the expense of SK Telecom (Internet Service Provider.)

• Alternative Internet Access solutions such as SpaceX’s Starlink 

The story of the current telecommunications industry seems to be one where consumers are benefitting at the expense of companies margins as cable companies and wireless carriers are engaged in somewhat of a “race to zero,” as they fight one another to offer lower bundled services to consumers in the face of declining average revenue per user (ARPUs.) There are battles on many fronts - cable companies and wireless providers are trying to win customers from each other by bundling home and mobile internet services, cable providers are battling with media companies (most of whom have created their own streaming service, the profitability of which remains to be seen as at the current moment Netflix appears to be the leader in running a profitable streaming company. Charter vs Disney is the most recent example, SK Telecom vs Netflix is another relevant example.) In the Chartr vs Disney Battle, Disney pulled its programming from Charter’s Spectrum TV service the day before college football started. While this causes consumers to get mad at the cable provider (Charter) all the cable company does is provide the programming and pass through the costs. Historically this has been profitable for both entities but as media owners have raised prices, it has eaten away at the cable companies margins. Charter finally had enough and doesn’t want to negotiate with Disney on their terms, which is potentially bad timing for Disney as they are also looking to divest their major media asset (ESPN) rumored to be worth as much as $40B (Disney’s market cap is ~$150B).  It will be interesting to see how this plays out as it may come to be influential in the future of both the cable and media industries.

Also, with video now representing 65% of internet traffic (Streaming, TikTok, Youtube etc.) and the rise of mixed & virtual reality (Meta Quest, Apple Vision Pro) how will the demand for ultra fast internet speeds inevitably shape how telecom companies choose to invest in their infrastructure. One of the interesting battles on the technology side of things is the race to build more internet access capacity.  Some companies are going the route of building and delivering Fixed Wireless Access while others are investing in Fiber Optics to deliver internet services.  After that we’ll also briefly explore SpaceX’s ambitious attempt to deliver the internet via Starlink, their low earth orbit satellite constellation.

FWA vs Fiber

Fixed Wireless Access:

Fixed wireless access (FWA) delivers internet connectivity using wireless signals, typically utilizing radio frequencies. A fixed antenna on your property communicates with a nearby base station to establish a connection.

Pros:

1. Rapid Deployment: FWA can be set up relatively quickly compared to laying physical cables, making it an attractive option for areas where infrastructure development might be challenging or time-consuming.

2. Cost-Effective: Installing wireless infrastructure can be more affordable than laying extensive fiber optic cables, especially in rural or remote areas.

3. Scalability: FWA networks can be expanded easily by adding more base stations to cover larger areas or more customers.

4. No Physical Cables: No need to dig up streets or lay cables, which can be disruptive and costly in urban environments.

5. Decent Speeds: FWA can offer respectable speeds, particularly in areas with good signal quality and lower user density.

Cons:

1. Signal Interference: Wireless signals can be affected by weather conditions, obstructions, and other electronic devices, leading to potential fluctuations in performance.

2. Limited Bandwidth: Shared wireless spectrum can result in reduced performance during peak usage times, causing slower speeds.

3. Lower Speed Potential: FWA speeds might not match the blazing fast speeds achievable with fiber optics, especially over longer distances.

4. Line-of-Sight Required: In some cases, a clear line-of-sight between the antenna and base station is needed for optimal performance.

5. Security Concerns: Wireless signals are susceptible to interception, making them potentially less secure than fiber optic connections.

Fiber Optics:

Fiber optics uses thin strands of glass or plastic to transmit data as pulses of light. It offers extremely high-speed internet connectivity through dedicated physical cables.

Pros:

1. High Speeds: Fiber optics provides some of the fastest internet speeds available, making it ideal for activities like streaming, online gaming, and large file transfers.

2. Stability and Reliability: Fiber is less susceptible to environmental factors like interference or electromagnetic interference, leading to consistent and reliable performance.

3. Symmetrical Speeds: Fiber often offers symmetrical upload and download speeds, which is crucial for applications like video conferencing and cloud-based services.

4. Unlimited Bandwidth: Fiber networks have high bandwidth capacity, making them well-suited for households with multiple devices and high data demands.

5. Future-Proof: Fiber's capacity for very high speeds positions it well for future technological advancements and increasing data demands.

Cons:

1. Deployment Challenges: Laying fiber optic cables requires significant infrastructure development, making it expensive and time-consuming, especially in areas with existing urban infrastructure.

2. Installation Complexity: Fiber optic cables need careful installation, which might involve digging and disruptions to existing infrastructure.

3. Cost: Fiber optic internet can be more expensive due to the high upfront costs of building the infrastructure.

4. Limited Availability: Fiber optic networks might not be available in all areas, particularly in remote or less densely populated regions.

5. Vulnerability to Physical Damage: Since fiber cables are physical, they can be damaged by construction work or accidents, potentially leading to service disruptions.

In summary, fixed wireless access is advantageous for its quick deployment and cost-effectiveness, while fiber optics excels in offering incredibly high speeds, stability, and future-proof capabilities. The choice between the two depends on factors like your location, desired speed, budget, and availability of infrastructure. While Fixed Wireless Access has gained popularity as an option for accessing the internet during the era of 5G rollouts, it appears at the moment that customers are benefitting at the expense of the internet service providers (over $100B has been spent on FWA deployments with very little of the costs passed through to the end customers.) It remains to be seen if the unit economics of FWA are durable in the long run and this is definitely a theme that is worth watching and will be important to listen to what various management teams have to say with respect to the economic viability of continuing to build fixed wireless access internet services. As demand increases for video games with higher quality graphics (as well as trends like cloud based gaming and richer multiplayer experiences) as well as the rise of immersive mixed and virtual reality devices like the Meta Quest and Apple Vision Pro, it will be interesting to observe how consumer preferences shift and how internet service providers invest in order to provide the best experience at the most competitive prices. An interesting point to note when it comes to reviewing history with the goal of better understanding the future - over the last five years, it has been somewhat more profitable to own the companies that own the cell towers and other telecom infrastructure (American Tower, SBA Communications) than it has been to own the Telecom Companies themselves (AT&T & Verizon) with the notable exception of TMobile. (Comcast being somewhere in the middle.)

Starlink

SpaceX's "Starlink" project is an audacious endeavor to revolutionize global internet access. As we discuss this project, we'll also briefly explore the history of satellite internet attempts and the physical challenges inherent in delivering signals via satellites.

The dream of satellite-based internet connectivity has fascinated innovators for decades. Some background, key milestones and challenges that define this industry:

1. Early Experiments:

In the 1970s, experiments like the ALOHAnet demonstrated the feasibility of satellite-based data communication.

In the 1990s, Teledesic, a project backed by Bill Gates and Craig McCaw, aimed to create a global satellite-based internet network but faced funding and technical challenges.

2. Geostationary Satellites:

Traditional geostationary satellites orbit at 36,000 kilometers above the equator, providing global coverage but suffering from latency issues due to signal travel time.

3. Low Earth Orbit (LEO) Satellites:

LEO satellites, at altitudes of 160-2,000 kilometers, offer lower latency but require a vast constellation due to their lower coverage area.

4. Regulatory Hurdles:

Licensing, frequency allocation, and international coordination are complex challenges for global satellite internet projects.

5. Financial Viability:

High launch and satellite production costs have hindered satellite internet projects' economic feasibility.

SpaceX's Starlink project, launched in 2018, aims to address these historical challenges with a new approach:

1. LEO Satellite Constellation:

Starlink deploys thousands of small satellites in LEO, reducing latency and enabling faster data transmission. The system relies on satellite-to-satellite laser links for rapid data transfer within the constellation.

2. Cost Efficiency:

SpaceX manufactures satellites in-house, reducing production costs and reusable Falcon 9 rockets significantly cut launch expenses.

3. Regulatory Progress:

Starlink has secured regulatory approvals in numerous countries, streamlining the path to global coverage.

4. User Terminals:

Starlink user terminals (antennas) are compact and designed for easy installation. The phased-array antenna automatically aligns with the satellites, simplifying user experience.

5. Rapid Deployment:

Starlink's "Better Than Nothing Beta" began offering services in late 2020 to select regions, expanding rapidly to users worldwide.

6. Expanding Coverage:

The goal is to provide high-speed internet access to underserved and remote areas globally.

7. Constellation Growth:

SpaceX continues to launch batches of satellites, with plans to deploy thousands more in the coming years.

While Starlink represents a groundbreaking approach to satellite internet, it grapples with physical challenges:

1. Space Debris:

 LEO satellites risk collisions with space debris, necessitating constant monitoring and collision avoidance maneuvers.

2. Satellite Lifespan:

LEO satellites have a limited lifespan due to atmospheric drag, requiring ongoing replacement.

3. Latency:

Although lower than geostationary satellites, LEO-based systems still face latency challenges compared to terrestrial networks.

4. Regulatory Compliance:

Navigating international regulations and spectrum allocation can be complex and time-consuming.

5. Competition:

Starlink competes with other satellite internet ventures and terrestrial broadband providers.

6. Environmental Impact:

The sheer number of satellite launches raises concerns about space debris and environmental effects.

SpaceX's Starlink project represents a remarkable leap forward in satellite-based internet access. By deploying a vast constellation of LEO satellites, addressing latency issues, and focusing on cost efficiency, Starlink aims to connect the unconnected and reshape global internet accessibility. Despite the physical challenges and regulatory hurdles, Starlink's progress promises to usher in a new era of connectivity, potentially bridging the digital divide for millions worldwide. As history has shown, the path to satellite internet innovation is marked by persistence and innovation, and Starlink is a compelling testament to this journey.

Relevant Companies 

• AT&T

• Verizon

• TMobile

• Charter

• Comcast

• Netflix

• Disney

• SpaceX(Private)

• American Tower

• Crown Castle

• SBA Communications 

• SK Telecom 

• China Mobile

• Nippon Telegraph & Telephone

• Vodafone

• Deutsche Telekom

• American Movil

• China Telecom

• Orange

• Swisscom

• Saudi Telecom

• Singel

• BCE

• Telstra

• (most countries have one or more big telecom providers)

Edge Computing

The “Edge” is where all these various companies meet - Hyperscalers, Communication Service Providers (carriers/teclos), infrastructure equipment providers and Edge Cloud Management Platforms amongst others.  This confluence is driving lots of corporate spending and innovation around computing at the edge. 

Imagine you're considering investing in the real estate market. You have two options: you can either invest in a massive, centralized skyscraper in the heart of the city, or you can invest in multiple strategically located townhouses across the city. Let's relate this analogy to the data center world and explain edge computing in a similar manner:

Edge computing is like investing in those strategically located townhouses. In the world of data centers and networking, it refers to a decentralized approach to processing and analyzing data. Instead of sending all the data to a central data center (the skyscraper), edge computing brings the data processing closer to where it's generated, right at the "edge" of the network – in this case, the townhouses.

Just as these townhouses are placed in key neighborhoods to serve the local residents efficiently, edge computing involves placing smaller computing resources – like servers and data processing devices – in proximity to where the data is being produced. This can be at the site of an industrial machine, a retail store, a connected vehicle, or even in smart home devices.

From an investor's perspective, edge computing has several advantages:

1. Reduced Latency: Since data doesn't need to travel all the way to a central data center and back, processing happens faster. This is crucial for applications that require real-time responses, like autonomous vehicles or industrial automation.

2. Bandwidth Efficiency: By processing data closer to the source, edge computing reduces the need to transmit large amounts of data over long distances. This can lead to cost savings on network infrastructure.

3. Data Privacy: Some data might be sensitive or subject to regulations, making it more secure to process and analyze at the edge rather than sending it to a distant central location.

4. Reliability: If there's a network outage, edge computing can still function locally, ensuring uninterrupted operation for critical systems.

5. Scalability: Edge computing allows for modular expansion. New "townhouses" can be added to the network easily as demand increases.

6. Diverse Applications: Edge computing benefits a wide range of industries, from manufacturing and healthcare to retail and entertainment, making it a versatile market to invest in.

Edge computing is significant in our research as it represents the convergence of networking/telecom/hyperscalers/internet companies etc. and is a key area to watch in the coming years.

“Billions of devices connect to the internet: smartphones, computers, security cameras, machine sensors, and many more. Devices like these generate massive amounts of data, most of which travels over the internet to applications running in the cloud. The cloud, in turn, is powered by enormous, centralized data centers and platforms operated and offered by a few organizations.The problem with this is that, as the number of connection points explodes to 150 billion devices generating 175 zettabytes of data by 2025, sending all that data to faraway clouds for processing will become increasingly inefficient and expensive. Moreover, this model may not be able to deliver the real-time data and response times demanded by newer applications. Consequently, more organizations are considering a hybrid cloud model that augments existing cloud strategies with edge computing.  Edge computing distributes the cloud’s scalable and elastic computing capabilities closer to where devices generate and consume data. These locations can be as varied as an enterprise’s on-premise server, a communication service provider’s central office or cell tower, a hyperscaler’s regional data center, an end-user device, or any point in between.Since data doesn’t have to travel as far, using edge computing can help reduce network resources, cut transit costs, improve reliability, reduce latency, and, perhaps most importantly, enhance enterprise control over data and applications. For example, edge computing can help organizations meet increasingly stringent data sovereignty, privacy, and security requirements by keeping sensitive data on premise. What’s more, when edge computing is combined with advanced connectivity options—especially 5G—it can deliver flexible, near real-time response times for data-heavy, artificial intelligence–driven, time-sensitive, or mission-critical applications. The combination of low latency, advanced connectivity, and enhanced data control makes many IoT use cases, such as the video analytics and computer vision used in security and quality control, immersive mixed reality training, autonomous vehicles, and precision robotics, much more feasible.The developing edge computing ecosystem is highly diverse. While chipset makers, device manufacturers, application developers, security specialists, and system integrators also feature prominently, we’ll focus on four categories of companies that are active in the edge computing market: public cloud hyperscalers, communications service providers (CSPs), infrastructure equipment vendors, and cloud management platforms.” - Deloitte TMT Predictions 2023

Dojo: Custom Supercomputer Infrastructure Built For the Edge

Tesla's "Dojo" supercomputer project represents a pivotal step towards achieving the company's ambitious goals in autonomous driving and artificial intelligence.

Autonomous driving relies on the processing of vast amounts of data from sensors, cameras, and radar systems in real-time. The complexity of analyzing this data, making split-second decisions, and ensuring safety demands computing power beyond the capabilities of conventional computers. Tesla recognized this need and initiated the Dojo project to address it.

Understanding Tesla's Dojo Supercomputer

1. Purpose:

Dojo is designed to accelerate the training of Tesla's neural networks, which are the core of the company's Full Self-Driving (FSD) technology.

2. Impressive Specs:

Tesla aims to make Dojo one of the most powerful AI supercomputers in the world. It's expected to deliver exaflop-level computing, which translates to a quintillion calculations per second.

3. Custom Hardware:

Tesla is developing custom hardware, including AI accelerators and training chips, tailored for Dojo's specific requirements. This hardware is designed to optimize neural network training while minimizing power consumption. In a world where the demand for Nvidia GPUs greatly exceeds supply, this may prove to be a valuable investment over the long term.

4. Data Efficiency:

Dojo focuses on data efficiency, aiming to train neural networks with less labeled data, reducing the need for extensive manual labeling.

5. Neural Network Training:

The supercomputer is intended to significantly accelerate the training of neural networks, allowing for rapid iterations and improvements in Tesla's self-driving algorithms.

6. Potential Applications Beyond Autonomous Driving:

While Dojo's primary focus is on autonomous driving, its capabilities can extend to other AI-intensive applications, such as natural language processing and computer vision.

Significance in the Autonomous Driving Industry

Tesla's Dojo project holds immense significance for the automotive industry and beyond:

1. Competitive Advantage:

Dojo can potentially give Tesla a competitive edge in the race for autonomous driving supremacy by accelerating its development and deployment of FSD features.

2. Enhanced Safety:

Faster neural network training can contribute to safer autonomous vehicles by improving real-time decision-making and reaction times.

3. Cost-Efficiency:

Dojo's data efficiency could reduce the need for extensive data labeling, which can be a costly and time-consuming process.

4. Potential Revenue Streams:

Tesla could potentially leverage Dojo's capabilities for AI-related projects beyond automotive, creating additional revenue streams.

5. Industry Influence:

The success of Dojo could set industry standards and influence the development of AI supercomputers in autonomous driving.

Tesla's Dojo project faces several challenges, including the development of custom hardware, software optimization, and ensuring data privacy and security. Additionally, the practical implementation and integration of Dojo into Tesla's vehicles and infrastructure remain key milestones.

The future of the Dojo project holds the promise of faster advancements in autonomous driving technology. As Tesla continues to refine its neural networks and expands its FSD capabilities, Dojo's role as a critical enabler of self-driving cars is likely to become increasingly apparent.

Tesla's Dojo supercomputer project is an ambitious effort to reshape the landscape of autonomous driving and AI. By harnessing the power of exaflop-level computing, custom hardware, and data efficiency, Tesla aims to accelerate the development and deployment of Full Self-Driving technology. As the project matures and overcomes its challenges, it has the potential to not only elevate Tesla's position in the automotive industry but also influence the broader AI and autonomous driving landscape.

Relevant Companies 

• Amazon

• Google

• Microsoft

• Tesla

• Verizon

• TMobile

• Orange

• SK Telecom

• KDDI

• Telenor

• Telefonica

• AT&T

• Vodafone

• Telstra

• Dell

• Nokia

• Cisco

• Ericsson

• JMA Wireless

• Mavenir

• Hewlett Packard

• Red Hat

• VMWare

• Nutanix

• MobiledgeX

• Amdocs

Component Providers 

As is true in any industry, “where this is a gold rush, sell the picks and shovels” or to frame the quote from the lens of an investor - buy the companies that are providing the picks and shovels for the modern “gold rushes.” In terms of networking and telecommunications, there are many companies that provide the mission critical components that companies need in order to build the most advanced and capable network infrastructure.

• Routers (core, edge, wireless, etc)

• Switches

• Wireless Access Points

• Network Products 

• Transceivers

• Optical Components 

1. Routers:

Routers direct data packets between networks. They manage traffic, determine the best path for data, and enable communication between devices.

Companies: Cisco, Juniper Networks, Huawei, Arista Networks, NETGEAR.

2. Switches:

Switches connect devices within a local network (LAN). They forward data to the intended recipient, improving efficiency and reducing network congestion. 

Companies: Cisco, Hewlett Packard, Juniper Networks, D-Link, Extreme Networks.

3. Access Points:

Access points (APs) provide wireless connectivity to devices within a certain range. They are used to create Wi-Fi networks.

Companies: Cisco, Aruba Networks (HPE), Ubiquiti Networks, Ruckus Networks (CommScope), TP-Link.

4. Modems:

Modems (modulator-demodulators) convert digital data from computers into analog signals for transmission over phone lines or cable systems, and vice versa.

Companies: ARRIS (CommScope), Motorola Solutions, NETGEAR, Technicolor.

5. Firewalls:

Firewalls protect networks from unauthorized access and potential threats by monitoring and controlling incoming and outgoing traffic.

Companies: Palo Alto Networks, Fortinet, Check Point Software, Cisco (ASA), SonicWall.

6. Load Balancers:

Load balancers distribute network traffic across multiple servers to prevent overloading, optimize resource use, and ensure high availability. 

Companies: F5 Networks, Citrix Systems, Radware, Kemp Technologies.

7. Network Security Appliances:

These appliances provide advanced security functions such as intrusion prevention, malware detection, and content filtering. 

Companies: Fortinet, Palo Alto Networks, Sophos, Cisco (FirePower), WatchGuard.

8. Transceivers:

Transceivers are modules that convert data into a format suitable for optical or electrical transmission over network cables.

Companies: Cisco, Juniper Networks, HPE (Aruba), Finisar (now part of Coherent), Broadcom.

9. Network Cables and Connectors:

Ethernet cables and connectors are used to physically connect devices within a network.

Companies: Belden, Panduit, CommScope, Black Box, Legrand.

10. Network Monitoring and Management Software:

Software solutions that monitor network performance, analyze data, and manage network devices. 

Companies: SolarWinds, Cisco (Prime Infrastructure), PRTG Network Monitor, ManageEngine OpManager.

11. Telecommunication Infrastructure:

This includes equipment for mobile and fixed-line communication, such as base stations, antennas, and switches.  

Companies: Ericsson, Nokia, Huawei, ZTE, Samsung Electronics.

12. Satellite Communication Equipment:

Equipment for communication via satellites, including satellite dishes, transponders, and ground stations.   

Companies: Hughes Network Systems, ViaSat (Viasat), Gilat Satellite Networks, Iridium Communications, SpaceX(Starlink)

13. Optical Components

Optical components are essential parts of networking infrastructure that enable the transmission and reception of optical signals (light) in optical fiber communication systems. They play a crucial role in modern high-speed, long-distance, and high-capacity data transmission. Here are some key optical components used in networking infrastructure:

1. Optical Transceivers: Optical transceivers are integrated devices that transmit and receive optical signals. They convert electrical signals from network devices like switches and routers into optical signals for transmission over optical fibers and vice versa. Common types include SFP (Small Form-Factor Pluggable), QSFP (Quad Small Form-Factor Pluggable), and CFP (C form-factor pluggable) transceivers.

2. Optical Amplifiers: Optical amplifiers boost the power of optical signals without converting them to electrical signals. This is essential for long-distance optical transmission where signals may attenuate (weaken) over the fiber. Erbium-Doped Fiber Amplifiers (EDFAs) are a common type used in long-haul networks.

3. Optical Splitters and Combiners: Optical splitters divide an optical signal into multiple signals, while optical combiners merge multiple signals into one. These components are crucial for distributing signals in passive optical networks (PONs) and for creating redundancy in network architectures.

4. Optical Filters: Optical filters allow specific wavelengths of light to pass through while blocking others. They are used for wavelength division multiplexing (WDM) to combine and separate multiple optical signals on a single fiber.

5. Optical Couplers: Optical couplers are passive devices used to combine or split optical signals. They can distribute optical signals among multiple fibers or combine signals from different sources onto a single fiber.

6. Optical Isolators and Circulators: These components control the direction of optical signals within a fiber. Optical isolators allow light to travel in one direction only, while circulators can route light among multiple ports.

7. Optical Attenuators: Optical attenuators are used to reduce the power of optical signals. They are crucial for adjusting signal strength and avoiding overloading optical receivers.

8. Optical Switches: Optical switches enable the rerouting of optical signals to different paths or destinations without the need for conversion to electrical signals. They are used for network redundancy and protection.

9. Optical Connectors and Adapters: These components provide the physical interface for connecting optical fibers and transceivers. Common types include LC, SC, MTP/MPO, and ST connectors and various adapters to facilitate different fiber connections.

10. Optical Reflectors: Optical reflectors are used to redirect optical signals, often in optical testing and measurement applications.

11. Optical Fiber: While not a component per se, optical fibers are the core medium for transmitting optical signals. Various types of optical fibers with different characteristics are used depending on the specific application.

These optical components are the building blocks of modern optical networks, enabling the high-speed data transmission and connectivity that underpin telecommunications, data centers, and internet services. They continue to evolve and improve to meet the growing demands of data-intensive applications and the need for faster, more efficient, and reliable communication infrastructure.

Companies: Cisco, Coherent, Infinera, Fabrinet, NeoPhotonics, Mellanox (now part of Nvidia), Acacia, Ciena, Juniper, Applied Optoelectronics

Data Centers

Jim Chanos, a renowned hedge fund manager and founder of Kynikos Associates, has been known for his skeptical view of certain sectors and companies, as he has risen to prominence from his success as a short seller. One area where Chanos has recently expressed a bearish view is on data center Real Estate Investment Trusts (REITs). His skepticism centers around the belief that these companies may not generate a return on invested capital (ROIC) that exceeds their cost of capital. 

Data center REITs are companies that own and operate data centers, which are facilities housing computer systems and networking equipment. They provide essential infrastructure for cloud computing, data storage, and internet connectivity. These REITs lease data center space to technology companies, enterprises, and cloud service providers.

Chanos argues that there is essentially three ways for companies to maintain their data:

1. On premise - servers owned by the company that an IT department manages

2. Third Party Colocation Services - you keep your servers at a third party location (this is the area that he believes is in secular decline, earning ROICs lower than their cost of capital.) These are companies like Digital Realty and Equinix.

3. Hyperscalers - you don’t own servers at all but rent them from the major cloud providers (Amazon, Google, Microsoft)

Chanos believes these companies are in secular decline, are earning ROICs lower than their cost of capital and are understating their maintenance capex requirements using somewhat misleading accounting that records some necessary capital expenditures (without which the data centers would lose competitiveness or in some cases cease to operate) as growth capex rather than as maintenance.  He points out that Digital Realty is burning $230mm/month on a $30B market cap and is capitalized at roughly 9x leverage (net debt + preferred of ~$19B vs $2.2B of EBITDA.)  He believes these companies are overvalued and are facing numerous headwinds, both competitive and financially as their growth slows and the cost of their debt rises as interest rates go up.

 Chanos's Key Arguments

1. Intense Competition: Chanos believes that the data center industry is highly competitive, with many players vying for market share. This competition can result in pricing pressures and reduced profitability for individual data center REITs.

2. Capital Intensity: Data centers require significant capital investments to build and maintain. According to Chanos, the capital expenditures required to keep up with technological advancements and security demands may erode the ROIC.

3. Cyclicality: Data center REITs are not immune to economic cycles. During downturns, demand for data center space may soften, impacting occupancy rates and rental income.

4. Lease Structure: Data center leases typically have long durations, often ten years or more. Chanos argues that this long-term commitment can be a double-edged sword, as it may lock in rental rates that become less competitive over time.

5. Technological Risk: The fast-paced nature of technology can pose risks to data center operators. Outdated facilities or infrastructure can lead to obsolescence and reduced competitiveness.

6. ROIC vs. Cost of Capital: The core of Chanos's argument is that data center REITs may not generate an ROIC that exceeds their cost of capital over the long term. This means that the companies might struggle to create value for shareholders.

ROIC vs. Cost of Capital

To understand Chanos's viewpoint, it's essential to grasp the concept of ROIC relative to the cost of capital.  This is of course obvious to many readers but just wanted to explain it briefly for anyone unfamiliar with the concept.

ROIC (Return on Invested Capital): ROIC measures how effectively a company generates profits from its invested capital. It's a critical metric for assessing a company's ability to create value for shareholders.

Cost of Capital: The cost of capital represents the return expected by a company's investors in exchange for providing capital. It includes the cost of debt and the cost of equity. If a company's ROIC consistently falls below its cost of capital, it may not be generating enough return to justify the capital invested in it.

Jim Chanos's skepticism about data center REITs revolves around his belief that the competitive and capital-intensive nature of the industry may prevent these companies from consistently earning an ROIC that exceeds their cost of capital. For instance, he calculates that it costs Digital Realty $11 to generate $1 of revenue and on a 50% EBITDA margin that means they need to invest $11 to earn $0.50 of gross cash flow.  With inflated valuations, growth slowing or even the business shrinking, rising debt service costs and lower returns on investment, Chanos believes that colocation data centers are bad investments and is putting his money where his mouth is by being short these companies in size.

There is a Business Breakdowns podcast where he discusses the thesis in more detail.

This piece is already getting a bit long, but interesting data center developments that are worth mentioning (and perhaps exploring in a future piece) are CoreWeave’s efforts to build a special purpose GPU cloud for AI workloads (a great Odd Lots podcast with one of the founders here) and Nvidia’s DGX, a platform that combines their infrastructure, software and expertise and was built from the ground up to be the best choice for enterprise AI workloads (AI training as a service.)

Conclusion

Putting it all together 

So we have cable companies, wireless network providers, some of these companies do both and also provide internet services, hyperscalers that provide cloud computing, network infrastructure companies and the component manufacturers that make products for them.  Data center owners and operators and major internet companies (who are also the major hyperscalers).  Network capacity and bandwidth are a major bottleneck for AI workloads.  Nvidia and Google are spending massively to innovate in this space.  How do we view this from an investment lens? Companies with large capex (sometimes in excess of their ROIC) and declining ARPUs (the telcos) should be losers while the companies that are best positioned to benefit from these tailwinds (the internet companies/hyperscalers, the networking companies, the component manufacturers (both at the equipment and semiconductor level.) However, it’s hardly as easy as looking at a high level trend and then selecting the winners from companies that have historically been the best investments.  In future articles, we will dive into these industries and the sub sectors in much greater detail and do our best to construct thematic long/short baskets from inevitable the winners and losers of the theme we are exploring. However if I took a shot at a basket after writing this it might look something like 

Thematic L/S Basket

Long 

• TMobile 

• Nvidia

• Google

• Microsoft 

• Amazon

• Arista 

• Broadcom

• Marvell

• Comcast 

• American Tower 

• SBA Communications

• Fabrinet 

• Meta

• Netflix

• Dell 

• HPE

• Palo Alto Networks

• Juniper 

• Cisco 

• Fortinet 

Short 

• AT&TCharter

• Verizon

• Disney 

• Digital Realty

• Equinix

• Cogent 

• Altice 

• Lumen

• Frontier

• Orange 

• Vodafone 

• Nokia 

• Ericsson

• BCE

• Deutsche Telekom

I hope you enjoyed this piece and I will be working on more of these for the future. If you have any thoughts, please feel free to reach out to me on Twitter X @netcapgirl. If you liked this write up, please consider subscribing.

Nvidia’s revenue increased dramatically in the 2nd quarter this year compared to last year. this was largely due to an increase in data center revenue from customers placing orders for H100s. A product that commands a high price point and one that Nvidia earns very high margins by selling.

So while some of this increase in revenue was certainly due to the pricing power that Nvidia is able to wield over the market (because there’s nearly infinite demand for H100s right now) it was mostly due to the incremental margin earned by the mix of revenue shifting towards more H100s, rather than an increase in overall sales volume.

While some have placed the spotlight on Nvidia’s relationship with CoreWeave & the recently announced line of credit that CoreWeave secured that was collateralized by the Nvidia chips they own, there is nothing inherently nefarious about vendor financing (CoreWeave ordered $2.3B worth of H100s to continue building out their special purpose GPU cloud.) Especially as the AI arms race heats out & some of Nvidia’s hyperscaler customers may look to develop their own hardware for customers to run AI workloads. Vendor financing is when a customer pays for your product before it’s delivered.

This brings us to the question of the increase in accounts receivable & revenue without the corresponding increase in costs during this quarter. Nvidia books licensing & services revenue when the order from the customer is placed and product revenue

when the products they make are delivered to their customer. Yes their Accounts Receivable went up by $3.24B last quarter but there was also an increase in Accounts Payable & Accrued Liabilities of $3.47B. These might not all the Cost of Goods Sold associated with the sales but they are expenses

that Nvidia owes that have not yet been paid. With respect to 10b5-1, the share buyback plan, this quarter’s buybacks were small purchases (much bigger ones were done previously at a lower price) that were preplanned & limited by knowing the results of the quarter on 7/31, done to cancel out stock compensation.

It’s also unlikely that ~$1B of purchases (done over 3 weeks) would affect the price of the stock for insiders that much when Nvidia can sometimes trade $30B worth of volume in one day. I’m sure

@ResearchQf

@theKantoarbot

could add some more color and nuance here but I hope this clears up some of the confusion around Nvidia’s last quarter & their accounting practices. We’re witnessing a somewhat unprecedented demand for GPUs & since the market is fighting for them, many customers are willing to prepay in order to secure their allocation of Nvidia products.

Airbnb (ABNB)

EPS $0.98 vs $0.80 EST

Revenue $2.5B vs $2.42B EST

Guidance of $3.3-$3.4B in revenue vs $3.22B EST

Airbnb had a solid quarter, beating estimate for earnings per share (EPS) and revenue, and raising guidance for next quarter, but shares were down after earnings as they only increased nights and experiences booked by 11% which were just under the consensus estimates for those metrics. Brian Chesky, the CEO, was optimistic about the company's performance and cited tough comparisons as to why the company missed the estimates for nights & experiences booked (a key metric that helps analysts understand the performance of the company.). The company announced new tools for hosts and Chesky mentioned the potential for Airbnb to explore an advertising platform as well as it’s efforts to leverage OpenAI’s ChatGPT to help make the customer service department more efficient. Airbnb’s strong performance comes in the wake of rumors about Airbnb rentals slowing (there was a tweet awhile back perpetuating the narrative that demand was slowing and prices were decreasing), online complaints about the many layers of fees that guests must pay (some of which were not previously included in the listed price of the rental, though this does seem to be something that management is aware of and is actively addressing) and worries about overall travel demand slowing down as consumers draw down their savings (the bear case would say that Airbnb is probably over-earning since travel is peaking and with student loan payments resuming, consumers are bound to spend less on travel.). This narrative seems to have made Airbnb a relatively popular short on the buyside, and while the numbers from last quarter are strong, the valuation is probably a little rich (the above model suggests the need for a 25% growth rate for the company’s pretax income in order to achieve a 8% IRR on the stock were you to buy it at $125.). However, the company is doing a lot of things well and has an incredibly strong cash position with about ~$10B of cash & short term investments on hand and another $9B in customer deposits. There is certainly substance to both sides of the story for Airbnb and I believe the company is worth following as they are a rare new era tech company who can generate fairly healthy free cash flow in what is still relatively early in the company’s lifetime (founded in 2008.)

Arista Networks (ANET)

EPS $1.58 vs $1.44 EST

Revenue $1.46B vs $1.38B 

Guidance of $1.45-$1.5B in revenue vs $1.39 EST

Arista Networks (a company I did a write up on) posted a very strong quarter, driven by increased demand for AI networking. The company is strongly positioned for the AI arms race and the demand for their networking products is reflected in their record sales and profitability for the quarter. The company also raised guidance for next quarter and even with the 20% pop after their earnings report, the shares appear to be relatively fairly valued based on 16% projected earnings growth and 62% gross margins.

Axcelis Technologies (ACLS)

EPS $1.86 vs $1.46 EST

Revenue $274mm vs $257mm EST

Guidance of $280mm vs $261mm EST

Axcelis Technologies is a company that (per their website) delivers vital equipment, services and process expertise to the semiconductor manufacturing industry, helping customers reach higher levels of productivity with each new technology generation. Today, chipmakers from around the globe rely on their tools and technology insights to produce the transistors that power all electronics. Axcelis’ equipment portfolio comprises a powerful suite of manufacturing technologies for ion implantation - one of the most critical and enabling steps in the IC manufacturing process. They posted a very strong quarter, beating estimates for sales and EPS and raising guidance for next quarter as well. The CEO noted their is significant demand for their Purion Product family, especially in the silicon carbide market. They believe they are the only company with a product line that covers all implant recipes in mature technology markets. The company revised their revenue estimate for the full year to $1.1B, which is about a 20% year over year increase, and with 44% gross margins in the quarter (up from 41% in Q1) and the overall tailwinds for the entire semiconductor industry, one can certainly make the case that Axcelis is well positioned to capitalize on these trends. On a 5 year DCF, the shares appear to be a bit overvalued at their current level but overall business performance probably warrants putting Axcelis on your watchlist as it’s executing well in an industry (semiconductor capital equipment) that should continue to experience structural tailwinds from the increased demand for chips.

ELF Beauty (ELF)

EPS $0.93 vs $0.56 EST

Revenue $216.3mm vs $184.02mm EST

Guidance for next fiscal year revenue of $792-802mm vs $728.82mm EST

ELF Beauty (one of the companies mentioned in my Hyper Individualization write up) had a monster quarter, crushing estimates for both earnings and sales, and raising guidance for the next fiscal year as well. Earnings grew 182% year over year and sales were up 76.5% year over year. ELF has had a big run up this year, with shares up well over 100% already. It’s low cost cosmetic & skin care products are proving to be a hit with customers and the business is operating well - management cited a 260bps market share increase during the most recent quarter as well as a 280bps gross margin expansion which brings their gross margin to 71%. ELF CEO noted that they are one of only five publicly traded consumer companies out of 274 that has grown for 18 straight quarters and averaged at least 20% sales growth per quarter over that period. They believe they are still in the early innings of their long term growth story. Based on our model above, the shares appear to be overvalued by a fair bit, but this is often the case with high growth companies and DCFs aren’t necessarily the most effective way to model a company that is growing rapidly. We can expect ELF to keep investing in the business, both in e-commerce and physical retail channels and if earnings grow faster than our expectations, there may be an adequate return from current levels. Regardless, the company’s stellar performance merits consideration for a watchlist and if the overall market is to experience a pullback, the stock may reach levels that make it more attractive to a conservative investor.

Match.com (MTCH)

EPS $0.48 vs $0.45 EST

Revenue $829.55mm vs $811.40mm EST

Guidance of $875-885mm in revenue vs $863.37mm EST

Match Group, the parent company of popular dating apps such as Hinge & Tinder (another company included in the Hyper Individualization write up), posted a strong quarter - beating estimates for earnings and sales and raising guidance for the next quarter as well. Though total payers declined slightly year over year, Tinder direct revenue was up 6% and Hinge direct revenue showed some acceleration, up 35% year over year. The company posted a solid 26% operating margin and highlighted the undergoing efforts to find more ways of monetizing their various online dating platforms. Management spoke about their commitment to delivering a refreshed user experience for Tinder and how they are focused on adapting the platform to meet the expectations of Gen-Z users, as well as how they are committed to innovation and integrating AI into their products over the long term. The above model suggests that if management is to be believed, and Match can deliver on it’s ambitious vision of growing it’s various dating platforms (and crucially, increasing the degree to which each is monetized) one could expect an 8% IRR if Match is indeed able to grow pretax profit at about ~10% per year. The company’s strong gross margin (~74%), increasing monetization of popular platforms and overall momentum across their portfolio of brands makes it an interesting company to watch and if they are able to capitalize on their expansion plans and effectively monetize their various platforms, may make the company an attractive long at current levels. The company’s efforts seem to be aligned with long term value creation for shareholders - revamping Tinder and increased focused on Gen-Z, expanding into Asia with the acquisition of Hyperconnect and further integrating AI into their platform, to name a few.

A few other companies that posted beat & raise quarters - Confluent (CFLT), KLA-Tencor (KLAC), Lam Research (LRCX), ServiceNow (NOW) & Uber Technologies (UBER). LVMH (LVMH) and Hermes (RMS) also posted solid quarters (companies that were included in the Hyper Individualization write up). I have models that I previously posted on Twitter X for KLA & Lam Research that I will post below and will write up my thoughts about the other companies (CFLT, NOW & UBER are harder to value with a DCF) in future posts. Thanks for reading, please reach out here or on Twitter X and let me know any thoughts and if you liked this post I will try to do more in the future!

KLA-Tencor Model (KLAC)

Lam Research Model (LCRX)

Executive Summary:

• A convertible note private placement could allow Twitter to reduce debt burden from LBO by raising capital from investors aligned with long-term vision. This would lower interest costs and reduce bankruptcy risk.

• New investors would get equity upside if Twitter grows in value above a premium to current price. This incentivizes supporting growth decisions even if they reduce near-term cash flow.

• Elon would likely need to accept dilution of his equity stake to entice investors, but it would increase the chance the stake would be worth much more by the time the dilution is realized and would preserve control, allowing for a potential longer term vision to be executed.

• Alternatively, a group of investors could purchase the underwritten LBO debt from the banks at a discount to par (e.g., 60 cents on the dollar) and simultaneously negotiate a deal with Twitter to swap the acquired debt into new secured (if permitted) PIK notes (e.g., 80 cents on the dollar), helping to alleviate the cash interest burn on Twitter’s FCF.

• Incentives would be better aligned between investors and Twitter, compared to investment banks focused on near-term returns. This allows product innovation and exploring new monetization models without worrying about short-term cash flow for interest payments.

A few days ago, I tweeted my thoughts about a potential way for Twitter to recapitalize so that the shareholders would be more aligned with the longer term vision of the company.

The thesis is pretty simple if you understand a few basics. Essentially - if a group comes together to buy the debt and Elon agrees to swap the bonds/loans for convertible notes (a fixed income security with some clauses for how they may eventually convert into equity), he (Elon) will essentially be taking dilution to his equity stake in Twitter in exchange for relieving the company of the interest burden that is currently sitting around $1.4bn per year. First, some definitions: 

Convertible Note

A convertible note is a fixed income security, effectively a bond with an embedded call option that allows for investors to participate in the growth of the value of the company with certain conditions. Above a certain price, notes may be “converted” into equity by exercising the option (a call). If conversion doesn't occur by the maturity date (because the value of the company has not grown past the “conversion price”), the company repays the loan. The shares issued in association with the exercise of a convertible note are dilutive to existing holders, but since there is a delay between issuing the convert and the actual conversion  it is less dilutive than an equity offering at the money - especially if the company’s valuation is currently depressed. 

LBO: Leveraged Buyout

To quote an old Bloomberg headline that’s been memed countless times on fintwit, “private equity transactions are usually financed with a mix of debt and equity”. In the case of a leveraged buyout, the buyer utilizes debt in order to pay more than they’d otherwise be able to afford - debt that becomes the obligation of the acquired company. This is why they call it a "leveraged" buyout. That word "leveraged" is about a ratio that looks at a company’s debt compared to their earnings. In LBOs, that ratio is usually pretty high. Around roughly 6 to 8 times the company's earnings.If things go south, the company they bought is left holding the bag, not them. Typically, the acquirer only pays for a portion (anywhere between 20-50%, but it can be more or less depending on the transaction) with equity (called “sponsor equity”) and finances the rest with debt (usually provided by investment banks, who go on to “syndicate” the loan, or sell the debt to investors).

The investors who buy these bonds or loans want a return stream that is more predictable than equity (the yield on the bonds) and forgo the potential upside in exchange for consistent interest payments and (in the case of the most senior loans) claims on the company’s assets in the case of bankruptcy (sort of similar to the structure of how one purchases a house with a mortgage). 

This is why companies with relatively stable cash flows and opportunities for enhancing operating leverage are common targets for private equity companies (the stable cash flows allow them to reliably cover interest payments and eventually earn a higher return than they would’ve had the company been financed solely with equity). 

It’s also why startups are usually financed with equity only (since the profits won’t come until someday in the future, it makes sense for the startup to focus on developing their product and growing the company vs making short term decisions to cover interest payments that may hamper the company’s long term growth prospects). A company sells shares to investors willing to take the risk in exchange for sharing potential upside if a company is successful. 

Here’s an example of what a “typical” LBO looks like:

In general, debt financing is cheaper than equity financing. The expected returns for investors are lower with debt (the “equity risk premium”) & the interest payments are tax-deductible. However, this does not always mean that debt is a better option (especially for a company that is having difficulty generating free cash flow). In the case of a company that’s having trouble meeting their debt obligations, selling equity (provided they can get a decent enough price) instead can mean an environment where it is easier to take risks that pay off further down the line - you cannot “default” on equity and the only times when equity holders have claims on the assets of the company are in a bankruptcy liquidation (after the holders of secured/unsecured debt and other senior claims). 

The problem with this is if the company is selling equity to get out from under a stifling debt burden, the valuation is probably pretty depressed. This means the dilutive effect of issuing shares is magnified (raising 50m at 500m means a dilutive effect of 9.1%, versus only 4.8% at 1bn). 

The Twitter LBO

In the case of Twitter, Elon - with a $7.1bn equity investment from the likes of Larry Ellison and others - was able to finance most of the $44bn acquisition with equity  (including Musk’s shares of Twitter purchased on the open market, which were rolled over, the refinancing of Twitter’s existing debt, Elon’s margin loan on TSLA stock & the cash on hand of Twitter) and the remaining ~$13B or so was financed with debt. Now, usually, when a deal like this takes place, the banks that underwrite the transaction will then go on to sell the debt to interested investors (known as the syndication process). 

However, the deal came at a time when financial markets took a turn for the worse and investors’ appetite for risk dried up, leaving the banks holding the loans on their balance sheets. Loans backing LBOs fell 39% to US$114bn, and bringing this much supply of junk rated loans to market risked pricing that was sub par, if they could even get it all placed. A few months ago, there was talk of investor interest in bidding 50 cents on the dollar for Twitter debt, a very bad break for the banks.

This has created a dilemma for the banks who risk taking a multibillion dollar loss if they are not able to sell the debt at par value or close to it. If this situation had been purely influenced by broader market conditions (i.e. a pullback in markets that decreased the risk appetite of investors and slowed overall transaction volume), the banks could likely wait it out and sell the loans once demand picked back up.  However, in the case of Twitter, the stability of near term earnings is in question which makes debt investors uneasy about how reliably the company will be able to meet its interest payments and therefore reluctant to purchase the bonds and loans near par value.

As part of the acquisition, Elon secured $12.5bn of traditional financing that was provided by investment banks. Specifically, the banks involved were Morgan Stanley, Bank of America, Barclays, MUFG, BNP Paribas, Mizuho and SocGen. In addition, Elon secured a $12.5bn margin loan from Morgan Stanley against his other holdings. Despite the headline price, Elon’s actual cash equity investment as part of takeover was only $22bn, a large but relatively small portion of his ~$255bn net worth. 

The below chart illustrates the pro forma and current capitalization. While loan-to-value (“LTV”) based on the “equity” component is fairly low at <30%, the headline leverage at 8.1x / 10.7x is incredibly high, even for levfin standards. With SOFR at over 5%, the blended interest rate burden is ~11% resulting in a cash interest expense of $1.4bn dollars. This is 118% of the Adj. EBITDA the company was generating prior to acquisition. Without any revenue growth or margin expansion, it is very difficult to generate positive FCF. With the ad market in a recession and largely out of Elon’s control, it’s no wonder that Elon chose to attack the only variable that he could: costs. 

A Solution, Potentially

It appears that Elon Musk is running the company like a startup - taking lots of risks, making lots of changes to the product (he does seem to be actively interacting with folks on the platform who are giving earnest product feedback), and experimenting with different monetization routes. All of these things are fine for a startup whose investors are usually aligned with the founder's long term vision but are risky for either one whose investors are not or a company with a significant debt burden. This is due to the problem of incentivization: if incentivized towards generating free cash flow in order to get out from under a significant amount of debt, the company is more likely to jeopardize long-term vision and potential in the short-term just to make interest payments and avoid bankruptcy.  

My recommendation would involve Elon and other original equity investors taking a degree of dilution to their equity stake in Twitter, contingent on Twitter’s value rising above a certain premium to the current valuation (which back of the napkin math puts at about 40% lower than the acquisition value) in exchange for relieving the company of the interest burden that may restrict the ability to grow Twitter in a way that defers payoff immediately for a more significant reward in the future. 

Given Twitter's current financial situation and market conditions, equity financing is likely more expensive than debt financing for them in terms of cost of capital, but it’s also a point worth considering in the analysis of these costs that equity financing carries the least risk to Twitter of eventually ending up in a fire sale out of bankruptcy lead by Morgan Stanley. Additionally, any new equity issued would substantially dilute existing shareholders (mainly Elon Musk but also those who invested 7.1bn at the acquisition price), especially given Twitter's (likely) low valuation right now. The heavy interest burden makes it incredibly difficult to invest in the business and therefore grow revenues. Below is an illustrative LBO example that highlights that even with 5% revenue growth per annum coupled with 500bps of margin expansion, it is incredibly difficult to make a solid return on equity investment outside of a high multiple at exit. The most likely exit for Elon would be an IPO or merger with a strategic (one of his existing portfolio companies).

This makes the true cost of equity capital quite high.

So, what are the possible logical solutions? 

Right now, banks own debt from the Twitter leveraged buyout that they can't sell for full value. Instead of Twitter buying back that debt directly, a group of investors could purchase the debt from the banks at a discount, let's say paying 60 cents for every dollar of debt.

So the investors would pay $6 billion for $10 billion of Twitter debt owed to banks.

Then, those investors could make a deal with Twitter to exchange the acquired LBO debt for new secured debt called PIK notes. PIK stands for "pay in kind" - meaning interest payments are added to the principal rather than paid in cash.

In this example, the investors would swap the $10 billion of acquired LBO debt for $8 billion in new PIK notes from Twitter. So Twitter gets relief on its debt principal and pays interest by issuing more debt rather than cash.

Overall, this allows Twitter to reduce its debt burden from the LBO by having investors buy the debt at a discount from banks, and then further reduce it by exchanging it for PIK notes. This helps Twitter's cash flow by reducing the cash needed for interest payments.

One can look at the recent Carvana debt restructuring/deal for inspiration.

Convertibles Note Refinancing - Convertible notes sold to a group of growth and tech investors who share Elon’s vision for Twitter, the proceeds of which would be used to purchase back the LBO debt from MS at a discounted price, are the optimal solution here. Convertible bonds oftentimes carry very low coupons ranging from [0-2%]. In return, investors get equity upside above a certain equity price (often 30% premium to today’s value, but variable depending on situation). Twitter would benefit as it would lower the interest burden meaningfully.Why?

•  Equity dilution would be delayed, attracting investors & giving Musk time to execute on his vision and get the value of Twitter closer to the original acquisition price or perhaps even above it (which he has demonstrated to investors he can get markets on board with his visions) 

• Coupon payments would likely still be cheaper than an equity stake, definitely cheaper than a straight bond or the current LBO debt.

• It would not be such a slap in the face to the original investors who’s equity stakes reflect a cost of $54.20 and who are important, potentially, as strategic partners, investors and advisors to the future of Twitter.

• With reduced debt burden, Musk and investors can focus on long-term growth to boost the valuation before any note conversion happens. This aligns incentives.

Clearly, if possible, this is the best path: a reasonable middle ground that provides liquidity, avoids imminent bankruptcy, doesn't massively dilute near-term, and sets up the company for growth execution. Creditors also benefit relative to a messy bankruptcy scenario. As mentioned, the existing debt would need to be reduced given the high interest burden.For these reasons, it is the most palatable solution for multiple stakeholders. What would need to occur is:

1. Twitter negotiates a deal with Morgan Stanley et al to repurchase the secured loans at 60 to 70 cents on the dollar. 

2. Twitter issues convertible notes to raise capital in a private placement Twitter offers new convertible notes to the existing investors as well as new investors. 

3. Twitter uses convertible note proceeds to repurchase LBO debt - Twitter takes the cash raised from the convertible note issuance and uses it to buy back the secured debt portion of the LBO debt it took on at a discount to face value.

4. LBO debt is retired - The LBO debt repurchased by Twitter is retired and taken off the balance sheet. This immediately reduces Twitter's debt principal and future interest costs.

5. Interest expense reduced going forward - With lower LBO debt outstanding, Twitter's interest expense is decreased in future periods as it has less debt accruing interest.

6. Cash flow improves - The reduced interest burden on smaller LBO debt balance helps improve Twitter's future cash flow.

7. Balance sheet strengthens - Paying down debt meaningfully improves Twitter's balance sheet by reducing leverage ratios.

8. Alignment of Incentives - By shifting the lender from investment banks to growth investors with investment horizons that are much longer, Twitter has the room necessary to maneuver in anticipation of 5 year forward earnings rather than one quarter forward free cash flow. 

In another scenario, this same group of investors could be organized to purchase the LBO debt from the Investment Banks and then forgive it in return for convertible notes - perhaps that would result in a more advantageous situation for the investors and entice more demand.

Here’s an illustrative example of what this could look like. If Elon were to issue $5bn of new convertible bonds to take out a mix of secured/unsecured debt, he could potentially lower his interest burden by 45% to $755mm - see below.

Other considerations would be: 

• Conversion Terms - The notes convert to equity at a future fundraising round, typically at a 15-30% discount to the valuation. This rewards the note investors. The lower the conversion price (higher the conversion ratio), the better for investors in the notes and worse for investors in the original acquisition and vice versa. A careful balance will need to be struck to avoid unnecessary dilution but also entice enough investors to meaningfully set Twitter up for success.

• Valuation - The conversion valuation would be determined by business performance. If Twitter grows, the valuation will be higher, rewarding note holders.

• Coupon/Interest Rate - A 2-5% coupon rate could be sufficient to make this issue attractive for investors vs. high yield debt, depending on the conversion terms.

• Maturity - Likely a 5 to 7 year maturity, which gives time for a turnaround and room to execute on Elon’s vision for Twitter before conversion.

• Investors - Deep pocketed individuals like founders/CEOs of tech companies, VC firms who align with Musk's vision and growth focused hedge funds like Coatue or Tiger Global may be incentivized to participate.

• Dilution - Conversion dilutes existing shareholders, but likely significantly less than issuing new equity now while in distress and with a delay that grants breathing room.

• Control - Musk/board will likely need to cede some control and seats to large note investors, which makes it imperative that it is attractive to investors who he believes will be accretive as partners to Twitter’s goals.

Overall, it would provide substantial new capital to handle debt, while incentivizing growth that rewards both note and equity investors. This would also make the LBO math much more palpable. As you could see below, the company would be decently FCF positive. More importantly however, Twitter would have the breathing room to invest in the business and explore new revenue streams without the heavy burden of debt payments and constant need to monitor liquidity. There would be a more significant margin for error, things like lawsuits & settlements (e.g. from Twitter’s landlord in San Francisco, Analysis group, Blueprint Studios, Innisfree M&A and Private Jet Services Group or the 500m in severance pay alleged to be owed) would not be as significant a concern if bankruptcy was not a looming threat. 

Here are the details for how a convertible note could work for Twitter’s recapitalization:

It does appear that Elon appreciates that Twitter is a fundamental part of the internet and, while there seemed to be a wide consensus that Twitter was woefully under monetized as a public company, I’m not sure he’s in it for the money. Twitter represents the freedom of ideas and information that was promised by the pioneers of the internet, and while there are certainly some downsides and negative externalities, there’s really no other place on the internet quite like it.  Where else can you find anonymous folks with cartoon avatars that are far more knowledgeable about a field than the credentialed experts? Where else can you engage with world leaders and CEOs of major companies and see the discussion unfold in real time? Maybe it means that ad revenue won’t scale as well as Meta’s pristine walled gardens but Twitter is a systemically important website. 

This is why I believe that, for the long term success of the company, it would be best if the business in its current state is financed with convertible notes (and later, if successful, equity)  ideally held by shareholders whose incentives are aligned with the long term vision for Twitter. Whether that’s a billion users generating ad revenue on the scale of Meta, payments, subscriptions, an “everything app” (in the style of China’s WeChat), or something entirely different, this would be the optimal way to achieve growth and a  longer term goal. Why do I think that this is a better course of action than continuing to face the risk of bankruptcy head on? If Twitter went bankrupt due to being unable to service the debt obligations taken on during Elon Musk's leveraged buyout, there are a few likely outcomes for the fate of the platform:

- Twitter would undergo Chapter 11 bankruptcy restructuring in order to negotiate with creditors and seek to reorganize the company. This allows continued operation.As part of restructuring, equity shareholders would likely be wiped out, with debt holders and other creditors taking over ownership. Musk would lose control.

- Twitter operations would come under oversight of creditors, an appointed trustee, or a court. Strategic direction could drastically change.

- Parts of the business and assets could be sold off to raise money to repay creditors. This could include selling the platform's data, technology, brand name, or real estate. With the advent of Large Language Models, Twitter’s data is likely the most valuable asset held by the company right now, and in a firesale scenario it would be sold for a fraction of its actual value. If Elon can manage to recapitalize with converts or equity and buy Twitter some time, he could find that the data ends up being the only asset necessary to reach positive cash flow in a few years time. 

- Twitter possesses one of the most valuable and unique data sets in the world - real-time posts covering news, events, opinions, and conversations across every topic imaginable. For training LLMs, it is unparalleled as it provides up-to-date, topical information and discussions on current events. With Elon Musk's ambitions in artificial intelligence through XAI, he is well-positioned to leverage Twitter's data for next-generation natural language processing models. A bankruptcy is likely out of the question if he wishes to retain his equity and thus his claim on the future value of this data.

Specifically, combining Twitter's real-time data feeds with advanced AI and machine learning techniques could lead to innovative applications in areas like search, recommendations, speech recognition, and sentiment analysis. As large language models continue to advance, Twitter's data could provide a competitive edge in developing AI assistants, bots, and other services that understand conversational nuance and context. Given the rapid pace of progress in AI, monetizing Twitter's unique data set through advanced language models could potentially happen within the next 2-3 years. The synergies between Twitter's platform strengths and Musk's plans for AI make this a very compelling growth avenue to explore and one that necessitates longevity.

- If no restructuring path is found, Twitter could face complete liquidation under Chapter 7 bankruptcy. However, a liquidation of the platform itself is unlikely since the user base and data would still hold value. 

Of course, nothing comes for free. This would inevitably mean there's a tradeoff.  In order to convince a group to come together and buy the convertible note issue (or the debt from the banks, which the banks are willing to offer below par per the last report), they would need favorable terms regarding how the debt they purchased will eventually convert into equity.  This would almost certainly mean a lower valuation for Twitter than the $44B purchase price and would result in some potentially pretty heavy dilution for Elon’s equity stake. 

However, if he truly cares about the long-term success of Twitter, and believes in how vital a platform it is for the freedom of information, he may be willing to take what would be a multi-billion dollar loss on his equity position in order to stabilize the capital structure of the company by having shareholders who are aligned with the long-term vision. I’m not sure if Tencent breaks out segment revenue for WeChat but I feel like I remember seeing they were around ~$20B USD of sales at one point in the past couple years, which may be a relevant comp for an “everything app.”, 

Twitter’s management would be free to experiment with product decisions and monetization options without interest payments looming each quarter. It seems unlikely without a strategy similar to this one that Twitter will be able to reach cash flow positive status soon - Musk seems to be implementing strategies that work over longer horizons, similar to his approach to Tesla, but Tesla never had a significant debt burden. For example, the drive to get people to pay for verified Twitter blue seems to be a loss leader strategy aware of the costs of paying the accounts with the most reach as an acceptable expense as long as those accounts are likely to share the fact they were paid - enticing other (less expensive accounts for who revenue sharing is unlikely to outweigh the Blue subscription fee) to sign up for Twitter Blue. This seems like a good strategy, but only if there’s enough breathing room for it to organically spread. 

In my own opinion, the problem with a subscription tier on Twitter is that - because the tweets only go to your subscribers - it is nearly impossible to get a subscription tier tweet to go viral. Because of this, the most likely competitor for Twitter’s subscription tier is adult entertainment like OnlyFans. 

This clashes with the ad revenue sharing model of Twitter Blue which ultimately attempts to compete with Substack or YouTube. Twitter needs a Unified Theory of Growth in which all of these actions converge towards a singular goal - right now even focusing on growth is clashing with the fact that free cash flow is negative. This convertible proposal would fix this, and the market seems pretty hot for new issues right now (with most growth investors, at least in public markets, doing pretty well YTD) and it would result in a group of investors that are excited to grow twitter rather than concerned about the NTM impact of LT decision.

This was the essence of my suggestion, as someone who loves Twitter and believes it to be such a fundamental and important part of the internet, I would prefer it not to fall into the hands of creditors and ideally be a company that has shareholders aligned with management, allowing it to operate as close to its optimal state as possible and optimize the decision making process solely for the long term vision.

• Millennials/GenZ stand to inherit trillions in wealth 

• The line between content creator & consumer is increasingly blurred by the day, the phones in our pockets enabling this convergence  

• Ozempic may make it so anyone can be as thin as they want (don’t cancel me pls) 

• AI seems like it will enable increasingly personalized experiences

• Perhaps until the point where our lives are so tailored to who we are, there’s less of a “shared reality” than at any point in recent memory 

Beauty, Cosmetics, Luxury, Travel, Plastic Surgery, House Refurbishing & Upgrades, Tutors & Custom Education, Financial Services, etc. are all industries that should benefit from this trend in one way or another. 

The idea that our phones are not just a camera, but a mirror as well, is nothing new (ok you’ve imagined Sisyphus happy, now imagine Narcissus with an iPhone), and the recent surge in “influencers” has been around about as long as Instagram has been a popular app. as the veil of celebrity came down, the value of celebrity increased. We felt closer than ever to our favorite artists, athletes, and movie stars as we saw their lives presented by Instagram’s feed. Then came Snapchat stories. Soon enough, our feeds were a mix of the aspirational and the familiar. And at the same time, YouTubers and Twitch streamers were emerging with huge, dedicated followings. TikTok seems to have shown up at the crucial inflection point where the incentives in the game were permanently changing. The veil of celebrity coming down is an exciting turning point because it not only made viewing their lives (through the lens of their online feeds) more accessible, but it made many people yearn for a similar lifestyle. I mean, who wouldn’t want to look hot, post about your life and collect big checks from brands? I believe this to be the genesis of the mimetic desire that has spread like the Sumerian mind virus in Snow Crash. 

TikTok seems to have taken the game to a new level, where building a large following from being creative and funny (ok, and maybe also being attractive) can lead to serious ad dollars and lucrative opportunities to grow your own “personal brand.” I can’t imagine how difficult it must be to concentrate on something like taking an integral when you know there are people with 20 million followers who earn more than most careers' lifetime earnings because they did a funny Tiktok dance to a sped-up Lana Del Rey song. The social media stuff is just one side of this multidimensional trend. But supporting these ideas is the increasing push that folks from all industries are making into creating “content”. Doing things to grow your “audience” and strengthen your “brand.” Venture capitalists have podcasts, hedge fund managers have newsletters. Meanwhile, Kim Kardashian is launching a private equity firm. Perhaps it’s not something you can calculate with the NPV formula in Excel. Still, the world seems to agree that engagement on the internet drives real-world value in the form of real dollars you can use to buy a house or pay taxes or whatever else your heart desires. 

The ability to convert engagement to dollars drives some interesting trends and is at the core of what I mean when discussing the hyper individualization hypothesis. We don’t just have avatars and profiles online; we have a digital self. And managing this digital self can have life changing outcomes for those who do so successfully. One example of shifting incentives is found in college athletics. Previously, NCAA athletes were not allowed to sell their likenesses in exchange for brand sponsorships or other sorts of partnerships where they’d earn money in the same way professional athletes do. Now college athletes with large online followings earn more than some professionals will ever make. Another is how the democratization of creating and distributing content (film it on your smartphone, edit it on your laptop, post it on Youtube) has enabled anyone who wants to create media and share it online, build a following, and monetize it. 

So what does this have to do with being hot, having a personalized life, and Ozempic making people skinny? 

Well, in some respects, we’ve all merged with our online selves. And like any mega trend, those who have “cracked the code,” so to speak, on how to monetize their digital selves can not only enjoy the monetary fruits of their labor but are also empowered by the freedom they find from essentially working for themselves, assuming they can make enough to support their lifestyle. This plays into the same mimetic desire that created influencers who mimicked celebrities. 

In a conversation with a friend over the Fourth of July weekend, she mentioned growing her TikTok to over 50k followers, all while working a job most would consider to be financially rewarding and prestigious. “Escaping the 9-5” is a trend seemingly on all corners of the internet, partly due to the lionization of entrepreneurship post social network era. 

https://twitter.com/netcapgirl/status/1591438391325110272?s=20

Partly because of the freedom and potential leisure afforded to those who can set their own daily schedules. And partly because we can now see so many folks doing it, and as humans, it’s our nature to mimic the objects of our desire. And who can blame anyone for choosing this route? Even if you earn a bit less, what’s your time worth? Especially in the prime of your life.  

So how does this tie into AI, Ozempic and wealth transfer? And what should I be looking to own to play these trends? 

Let’s start with Ozempic. In the Instagram era, being fit and looking good was a status symbol. It certainly still will be, but the difference is now you can (essentially) pay to be skinny. And at the risk of being vain, looking good feels good. You want to go out more. You want to try on new clothes. You may even be more inclined to use dating apps. Going to the gym and eating healthy may be more rewarding as Ozempic helps you see results faster. The part of the trend where everyone wants to look good is accelerated by a miracle drug that can aid in curing cravings and makes losing weight easier than ever. 

So we have everyone wanting to be hot, and there being a pharmaceutical aid to assist in part of the process (getting fit), which may or may not make one more likely to more actively engage online and perhaps cross the threshold and become a creator themselves. That’s one part. For this aspect of the trend, i think it makes sense to watch -

Luxury Companies 

• LVMH

• Hermés  

• Richemont 

• Kering

• Prada 

• Ferrari (not clothes, but truly a case where driving the vehicles are an experience) 

• Restoration Hardware

• Essilor Luxottica

• Moncler 

• Pandora 

• Brunello Cucinelli 

• Farfetch 

Beauty & Cosmetics 

• Estée Lauder 

• L’oréal 

• Coty

• Ulta

• L'Occitane

• Beiersdorf AG

• Olaplex

• ELF Beauty

• Givaudan

• IFF

• InterParfums

• Shiseido

• Kao

• AmorePacific

Travel (does solipsism encourage wanderlust or vice versa)

• Airbnb 

• Booking 

• Expedia 

• Mariott 

• Hilton 

Plastic Surgery 

• AbbVie 

• Bausch Health 

• Merz Pharma

• Galderma SA

• Johnson & Johnson

• Cynosure 

• Korean based plastic surgery stocks

Dating Apps

• Match.com

• Bumble 

Companies that enable people to start and manage their own businesses 

• Godaddy 

• Stripe (whenever they go public)

• Intuit 

• Squarespace (now that they own google domains)

• Shopify  

Big Tech (as I mention later on, I think the only real risk in the near to mid term is antitrust. These companies, for the most part, have pristine balance sheets, healthy margins and are staffed with some of the most brilliant people on earth. In an era of hyper personalization, companies with all the data should thrive, assuming of course they can capitalize on the trend)

• Microsoft

• Meta 

• Apple

• Nvidia 

• Google 

• Amazon  

It’s kind of a solipsism basket lol. Wealthy folks will want the best of everything (as has always been the case) but can see this accelerating as life can increasingly be optimized & personalized down to the minutiae because every moment is a chance to post.

So what about personalization? Well, I believe this trend to be especially true in folks who will inherit wealth that affords them a lifestyle in which they do not have to work if they so choose. Their life, and wow does this sound bad but bear with me, will (from an investment thesis lens) be defined by consumption. And with AI enabling hyper personalization in various niches, why wouldn’t someone with the necessary means strive to live a life that’s, in some senses, hyper optimized and in a world where products and services are fighting to cater a unique experience to each individual, hyper personalized. 

For instance, alongside tracking engagement and social status progress via social media apps, one can now constantly monitor a variety of health variables from their phone (and connected devices if they choose). Is it difficult to imagine an AI service integrated with a personal trainer that helps them reach their fitness goals? The feedback loop of data tracking and iteration based on results may create a plan that is hyper personalized to the individual. It might go so far as to plug in with their grocery apps, ordering a set nutrition plan at predefined intervals each week, further personalizing the clients life. (who may or may not have ambitions of getting fit with the help of ozempic and posting luxury fits from dream destinations on their social media, who knows. It may also become a status signal to be offline. Like yeah I've made it so I can afford to be totally off the grid. Completely analogue. Going for long walks in nature and attending ballets, symphonies and operas dressed head to toe in Victorian garb. Ok got carried away. If you like my book recs though, I highly recommend reading Neal Stephenson’s “The Diamond Age” for sci-fi commentary on a futuristic society divided between “Equity Lords” who (as you might have guessed) own everything and can read, and everyone else (literacy is a luxury status thing in diamond age, as is full Victorian costumes)). anyway 

A couple points the other way -

• It doesn’t have to be like this. Maybe there’s instances where internet communities form irl ones that are just as strong. 

• There’s plenty of other ways to create content on the internet where you can make money where it doesn’t matter what you look like at all. 

• The endless interests and topics that one can explore on the internet offers content opportunities for each one. Whether math, history, music or more niche, obscure topics Pokémon nuzlockes, for instance) - passionate audiences will form around the creators who put out the best work in each one. This is cool and a good thing. For instance, Doug's Fabricated Knowledge and Dylan's SemiAnalysis are far superior to any sell side research on semiconductors. 

The opportunity for capital to flow to creators of the best online content would seem to increase the amount of it created, even if it’s just something that some people do on the side. 

There's probably a whole other post on this about how internet culture has just become the Culture and things happen online like more than they happen in the real world. I don't know the most eloquent way to express that. Covid definitely showed this. We can run most of the world in our pajamas on Zoom. Kids look up to streamers & Youtubers. In the post zero to one era, tech Twitter has minted its own form of celebrity. VCs farm alpha from the best anon posters, who are quite often exceptionally talented individuals. Mr. Beast could probably start his own heaven's gate… ok never mind on that analogy. You get it. The internet is real. And now that the stakes are higher than ever and there’s serious money involved, It’s almost impossible to ignore the possibility of monetizing your digital presence. It's like a call option. 

This may of course take other forms. Bryan Johnson, for instance, is using his wealth to pursue something like bio accelerationism. Spending $2m / year to personalize his life to an incredible degree in the pursuit of longevity (man so much of this shit sounds like greek mythology fr).

One interesting area that’s interesting to me will be around tutoring and the increasing personalization of education. This was prevalent in China before the government cracked down on tutoring businesses but I expect premium services to be available to wealthy folks in the coming years, as AI helps to personalize curriculums and learning materials for the individual student. Which they then review and go over with their tutor. Education in this context is not just limited to schoolwork. Music & sports both come to mind but truly the possibilities are endless. Maybe in an ideal world there will be products akin to the illustrated primer in The Diamond Age, which uses AI to democratize the personalization of education. I think that idea is cool. It's sort of like a holy grail of the information age. 

In a sense, one could take the other side of the argument that actually there’s a greater sense of shared reality since we all regularly use the internet and social media. This is true to an extent. However, with companies racing to create even more precise recommendation engines (I think TikTok is doing some crazy work on this, I remember seeing a tweet that they are training a different model for every user or something but maybe I’m tripping, either way, the TikTok algorithm seems to be an example of a feed that’s already hyper personalized). I also think that this means the only legitimate mid term (5-10 years) threat to big tech is antitrust. 

Maybe a relevant analogy for this overall trend is private banking services. At a certain level of wealth, it is worth the bank’s time to hyper personalize a wealth management strategy for their clients. In the future when nearly everyone is online, wealthy folks (or maybe more academically, folks whose propensity to consume is less cyclical) will have access to hyper personalized offerings from all sorts of companies. If in the past people wanted to wear the same outfit as their favorite actress or model, they now may want to wear a new outfit everyday like their favorite TikTOk influencer. taken further, they may want a wardrobe personalized just for them by DoverStreetMarketGPT. An interesting convergence of this trend (and one probably familiar to internet natives) is the AI Balenciaga Dune videos. Balenciaga has been exploring the connection between digital culture and the real world for years now in lots of interesting and provocative ways and this was a particularly interesting crossover episode. Luxury fashion sells fantasy. Part of what you’re buying is the exclusivity of owning pieces that few others have. Luxury brands sell their image as aspirational to folks who can’t afford it but try to market their true merits (expert couture craftsmanship in the instance of balenciaga) to their actual clientele. 

Dune formerly being like a “nerd” (complimentary) thing but now being mainstream (and Balenciaga understanding this paradigm shift) is one example of how internet culture is becoming just Culture. This is one of the reasons I enjoy Twitter. As corny as it sounds to call something “the marketplace of ideas,” the fact that anons can build audiences that are evaluating them solely on their thoughts (in the form of their tweets and replies) is cool. It’s still a form of mimesis nonetheless. 

This brings me back to the SBF tweet. I have been wanting to write something longform about this trend for a while now. It's something I think a lot of people noticed in the post social network era. When people saw Sorkin's portrayal of how dope it was to be Zuck, entrepreneurship became a status symbol all its own. You could go online and find lots of people who were just “into startups”. There was no shortage of resources for how to start your own startup and how to break into tech. Zero to One emerged as a holy bible for folks who wanted to follow the noble path of Zuck (even though Thiel emphatically tells readers “the next Zuck won’t build a social network” lol). It became funny or quirky to call yourself autistic. The lionization of the “wired in” tech founder led to billions (?) of dollars that went up in flames as VCs chased vaporware during the ZIRP era. The most prominent example imo was SBF. There was the fuck it idc attitude (showing up to Sequoia in your bath rode == playing league on VC pitches). The technical credentials of studying physics at MIT. the domain specific credentials of working at autism capital Jane Street. One of the most powerful ideas/ forces in capital markets is information asymmetry. Sure there’s the illegal version but that’s not what i’m talking about. I'm talking about how, as Thiel described, “the best startups have secrets” idea. And as was shown in the FTX debacle, if you could convey an asymmetrical information advantage (I can arb US/Japan Bitcoin markets, for instance) you might be the next Zuck. And just like people mimic celebrities they love, the mimesis was strong amongst VCs who saw Thiel turn $500k to a billi by making an ambitious bet on a goated founder. (does anyone know if Mr. Thiel has any thoughts on this whole mimetic desires thing? lol lol) 

Everyone wanted to hang out and get their photo taken with SBF. The line between real world economic activity and internet clout was increasingly blurred, especially because crypto was by default, an internet native thing. He did the billionaire press run stuff. The cringe interviews where he set effective altruism back hundreds of years. He hung out with Tom Brady and Gisele. Celebrities wanted to be tech founders. Tech founders wanted to be celebrities. This is still probably true to some degree but I digress. 

So what will the world look like in this hyper individualized future where many seek internet clout and further, as many seek to convert their status online to real world currency? (which will still be mainly the us dollar, sorry doomers).

Well with health, fitness, education, financial services, and recreation (travel shopping etc.) able to be personalized (first for UHNW then prob for all wealthy folks) it could maybe get a little weird. Your life could become so tailored to just you it’s difficult to relate to the realities which others experience (solipsism fr?). As wealth gets transferred to a generation with new mimetic desires, I expect consumption habits and hyper personalization to converge. Companies will have increasing amounts of data for each of their users which they will feed into increasingly powerful AI models to tailor experiences which are increasingly personalized. Whether your desire is to escape the 9-5, become the world's best AI researcher or to get completely off the grid, spend time in nature and go to ballets and shit, those with means will be able to live a life catered to them at such a fine level of detail that conscious experiences of reality might be like snowflakes, with no two being the exact same. And of course, to tie it all together, no matter what people are doing, (most) want to look hot while doing it. Ozempic combined with health tracking enabled by smart devices will probably enable anyone to be as fit as they want (which is why I think long luxury brands, beauty/cosmetics, plastic surgery and travel - all things that can be used to reflect status/taste as a mechanism of differentiation). A large percent of the population will want to own their own business or work for themselves to some degree, and with how necessary it is for business owners to have an online presence, the convergence of internet media & real world dollars is all the more inevitable. And with anybody able to look good w Ozempic + personalized health, money may be even more of a status signifier. Perhaps unique interests & tastes and one’s original thoughts on their interests/tastes will also have significant value in a post Ozempic world. How well read you are and how well you’re able to articulate your thoughts, may be a point of differentiation, but that may be far off in the future. Perhaps there’s even a version of the future where biometric verification becomes prevalent (i.e. i can prove i got this fit without Ozempic) maybe that’s a bit too cyberpunk though. 

What can you do? Well, you can undoubtedly choose to personalize your life in a way that exactly you want. In another plug for what I like about Twitter, it’s been (in my experience) a community organized around exploring ideas. I don’t think it’s going anywhere, but there’s always a chance that it can be replicated elsewhere. You can passionately lean into things that you love. If you want to learn math or music, it’s never been easier. If you want to find online communities about investing or startups, it’s never been easier. You can strive to write in prose that’s just (for now) out of the reach of the latest GPT model. You can strive to align your work with who you are as a person (it’s like that Ikigai chart or something) and lead a life that you truly find fulfilling. You can strive to be a great member of your irl community and take care of your family (whether blood or chosen). You can strive for optimal health for reasons beyond vanity (it’s not a crime to want to look good fr, it’s in fact, very human. But like overall health prob more important in the long run fr). To continue with the prior Stephenson references, I sometimes wonder if, similar to a tower of babel thing with like an innate language stored in our brain stems, there’s goals we all share as humans. Ok some people wanna watch the world burn, I get that. The one I come back to a lot is I think that (whether you personally want children or not) most people want a better world for the next generation. We’re all sort of working on raising the standard of living in our own way. Maybe there’s a way the picture of a hyper individualized future could be painted more optimistically. I’m trying to find a less cringe way to say some shit like “in a hyper personalized world, the next generation needs you to be the best possible version of yourself” because that, amongst all the bleakness and vanity, does sound like a worthy pursuit.