The S&P 500 Index returned 8.12% and the S&P Global Broad Market Index returned 7.83% in the quarter ended September 30, 2025.1
The third quarter was a continuation of a trend we discussed in our December 2024 letter.2 In short, indexes are increasingly concentrated in a small group of companies, all of which are betting heavily on artificial intelligence (AI). In fact, since that letter, the aggregate weighting of the top 8 companies in the S&P 500 index (all of which are AI-related) has increased by another 2% to 37.5%.3 Additionally, the equity research team at Citi recently completed a comprehensive analysis of the S&P 500 index’s exposure to the AI investment theme, identifying 84 AI-related stocks that currently account for roughly 35% of the index’s earnings and 50% of its market capitalization.4 In some ways, this concentration makes sense. AI is undoubtedly a transformative technology that has huge potential to drive global growth and productivity over the coming decades. Furthermore, AI could create massive, durable new wealth for some of the investors currently funding the trillions of dollars that are flowing into energy, commercial real estate, cooling, semiconductors, and the other technologies that are required to make AI’s potential a reality.
We are hopeful that some of the companies we own will durably capture some of this new wealth. Early indications are promising. So far, our holdings in Microsoft, Amazon, and Google have benefited the most from AI. As the three largest cloud services providers, they are the digital infrastructure layer on which the majority of AI compute runs, and there is such overwhelming demand for their services from start-ups and enterprises that they can’t even build capacity fast enough to fulfill it. Moreover, all three companies are already using AI tools in their other businesses to both increase revenue (for instance, through improved ad targeting) and reduce costs (by, for example, incorporating AI tools that increase employee productivity in analysis, coding, and customer service). In addition, our holdings in commercial real estate (CBRE and Costar Group) and construction aggregates (Martin Marietta Materials, Vulcan Materials, and CRH Group) are currently benefiting from exponential growth in data center construction, which often requires commercial real estate transactions and management services and almost always requires a great deal of aggregates and cement.
Furthermore, if AI and its associated advancements such as robotics continue to progress, we think even more of our companies have the potential to increase their revenue and profitability. Intelligent robots could increasingly dominate Amazon’s fulfillment centers over time, dramatically reducing costs. Apple could eventually use its combination of privacy and edge computing to further monetize its valuable position as the personal operating system for over a billion customers. For example, if Apple’s AI agent becomes the consumer’s trusted assistant, the company would gain even more leverage over the other players in its ecosystem. It could also potentially use further algorithmic advances to manufacture wearables that more accurately track health and even predict future medical risks and needs. As a last example, we think Moody’s and S&P Global could eventually benefit from AI solutions that increase the number of companies each financial analyst can rate, simultaneously generating additional revenue and cost savings.
However, all these growth opportunities from new technology also come with increased risk. First, investors are, on average, avaricious, impatient, and overconfident. As a result, any category of opportunity that promises quick riches tends to be overpriced. We believe this phenomenon occurs for two reasons: 1) Because they so strongly desire quick riches, investors convince themselves that this time is different, and, therefore, overestimate the likelihood that the whole category will generate attractive returns. 2) Even if they acknowledge that most investments in the category of opportunity will likely result in poor returns, they overestimate their own skill in picking the winners. In our view, AI clearly checks both these boxes.
Second, millions of talented investors and operators are pouring trillions of dollars and hours into their pursuit of AI profits. As a result, we believe innovations that massively increase the supply of compute are quite likely. If these occur, the current scarcity of compute could quickly become a glut, crushing prices and profits in the process. The following quote from a retrospective on the fiber buildout during the late ‘90s internet boom and subsequent bust is the best we’ve ever found on this clear and present danger:
A powerful factor contributing to excess capacity in long-distance telecommunications was the unexpected degree of improvement in dense wave division multiplexing. At first, a given strand of fiber was split into two channels. The technology rapidly advanced to where a given strand of fiber now has over 100 channels, with the possibility of over 1000 channels in the future. Thus, as companies installed new long-distance networks, technology improved so dramatically that capacity outpaced growth in demand, even with the Internet’s rapid growth.5
In our view, this quote should terrify investors who are all-in on AI.
Lastly, because missing out on a new technology paradigm can be fatal, existing technology leaders often feel forced to invest even when they know prospective returns are highly uncertain. But don’t take our word for it. Take theirs.
Bill Gates (Microsoft)
[AI] Agents will affect how we use software as well as how it’s written. They’ll replace search sites because they’ll be better at finding information and summarizing it for you. They’ll replace many e-commerce sites because they’ll find the best price for you and won’t be restricted to just a few vendors. They’ll replace word processors, spreadsheets, and other productivity apps. Businesses that are separate today—search advertising, social networking with advertising, shopping, productivity software—will become one business.
I don’t think any single company will dominate the agents business–there will be many different AI engines available. Today, agents are embedded in other software like word processors and spreadsheets, but eventually they’ll operate on their own. Although some agents will be free to use (and supported by ads), I think you’ll pay for most of them, which means companies will have an incentive to make agents work on your behalf and not an advertiser’s. If the number of companies that have started working on AI just this year is any indication, there will be an exceptional amount of competition, which will make agents very inexpensive.6
Satya Nadella (Microsoft)
And by the way, one of the things is that there will be overbuild. To your point about what happened in the dotcom era, the memo has gone out that, hey, you know, you need more energy, and you need more compute. Thank God for it. So, everybody’s going to race.
In fact, it’s not just companies deploying, countries are going to deploy capital, and there will be clearly… I’m so excited to be a leaser, because, by the way; I build a lot, I lease a lot. I am thrilled that I’m going to be leasing a lot of capacity in ’27, ’28 because I look at the builds, and I’m saying, “This is fantastic.” The only thing that’s going to happen with all the compute builds is the prices are going to come down.7
Joe Tsai (Alibaba)
I’m still astounded by the type of numbers that are being thrown around in the United States about investing into AI.
In the same talk
People are talking, literally talking about $500 billion, several 100 billion dollars. I don’t think that’s entirely necessary. I think in a way, people are investing ahead of the demand that they’re seeing today, but they are projecting much bigger demand.8
Sam Altman (OpenAI)
When bubbles happen, smart people get overexcited about a kernel of truth.
In the same interview
Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes. Is AI the most important thing to happen in a very long time? My opinion is also yes.9
Jeff Bezos (Amazon)
Now, what the stock market does, which is when we think of bubbles, we think of valuations and market caps and things like this, and how many billions of dollars are being invested in these six people at a $20 billion valuation, even though they just started yesterday. That’s very unusual behavior. Investors don’t usually give a team of six people a couple of billion dollars with no product. It’s rare and that’s happening today.
But the great thing about industrial bubbles, this is a kind of industrial bubble as opposed to financial bubbles. And I’ll tell you what I mean by that. If you go back like the ‘90s had a biotech bubble and there were a bunch of pharma startup companies that were designing drugs and using new techniques and the world got very excited, the investment world got very excited. As a group they all lost money. But we did get a couple of life-saving drugs.
A bubble like a banking bubble, a crisis in the banking system, that’s just bad. That’s like 2008. And so those bubbles society wants to avoid, the ones that are industrial are not nearly as bad. It could even be good because when the dust settles and you see who are the winners, society benefits from those inventions. They still get those life-saving drugs. And that’s what’s going to happen here too. This is real. The benefits to society from AI are going to be gigantic.10
Mark Zuckerberg (Meta)
I think it’s quite possible [there’s an AI bubble]. I mean, I think basic if you look at most other major infrastructure buildups and history, whether it’s railroads or fiber for the internet, in the .com bubble, these things were all chasing something that ended up being fundamentally very valuable. In most cases, it ended up being even more valuable than the people who were kind of pushing the bubble thought it was going to be.
But in at least all of these past cases, the infrastructure gets built out, people take on too much debt, and then you hit some blip, whether it’s some macroeconomic thing or maybe you just have like a couple of years where the demand for the product doesn’t quite materialize, and then a lot of the companies end up going out of business, and then the assets get distressed and then it’s great opportunity to go buy more.
So, I think that that, it’s obviously impossible to predict what will happen here.
Later in the interview
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. If you build too slowly, and then superintelligence is possible in three years, but you built it out assuming it would be there in five years, then you’re just out of position on what I think is going to be the most important technology that enables the most new products and innovation and value creation in history.11
Sundar Pichai (Alphabet)
[AI] is one of the most profound technologies we are working on, as important or more than fire and electricity.12
The biggest risk could be missing out.13
The risk of under‑investing is dramatically greater than the risk of over‑investing for us here.14
Larry Page (Alphabet)
I am willing to go bankrupt rather than lose this race.15
Concluding Thoughts
As stewards of your capital, we believe it’s our job to construct a portfolio that can maintain and grow your purchasing power through many different future scenarios. Today, given the S&P 500 index’s concentration in perceived AI beneficiaries, we worry that the typical investor’s portfolio is too reliant on the future scenario where AI continues its rapid growth and where companies successfully monetize this growth. Even if we thought this scenario were likely to occur, we would be nervous about making such a concentrated bet. However, given the frequency with which past capital-intensive technology booms have turned into busts and given how closely the AI boom mirrors these past ones (with investor’s get-rich gold rush mentality, with boatloads of money and talent chasing innovations to create more supply of the scarce assets, and with existing technology players viewing potential AI competition as existential), we fear the odds of this rosy scenario could be quite low. Therefore, we view such a concentrated AI bet as completely untenable.
Fortunately, we employ an index-agnostic approach that we believe can avoid a heavy reliance on any single future scenario. To accomplish this goal, we first identify what we believe to be the most durable, predictable businesses in the world. We’ve found that these businesses have seven main characteristics: 1) They own a dominant network, giving them both pricing power and a check on competing supply since the nature of network economics means that large networks provide far more value to their customers than smaller networks; 2) They possess additional checks on competing supply such as NIMBY (not-in-my-backyard) zoning restrictions or institutional risk aversion; 3) They serve large and broad end markets that we believe are highly likely to grow at least as fast as GDP; 4) They have built up significant untapped pricing power because the value to customers of their products and services has grown far faster than the price they charge; 5) Their pricing power and returns on capital are unlikely to be capped over time because the government will find it either too difficult or not important enough to permanently cap these metrics through either customer concentration power (as in the case of defense or healthcare) or regulatory changes (as in the case of utilities or, possibly in the future, healthcare); 6) They are run by ownership-minded management teams that treat minority shareholders fairly and tend to think in decades much more than quarters; and 7) They possess conservative balance sheets that enable equity shareholders to survive even deep recessions with their ownership intact.
Then, so long as we believe they are attractively priced, we diversify among these businesses as much as possible. This process results in a portfolio of companies that perform mission-critical services in large and varied end markets all over the world.
As you can see from the graphic above, some of these businesses are AI-related, but many are not. This is by design. We’re not interested in creating a portfolio that could work out spectacularly but also possibly end in ruin. Rather, what we’re interested in, and what we believe we’ve created, is a portfolio that acts like a diversified, recession-resistant toll collector on global GDP. While this portfolio is perhaps less likely to perform spectacularly over the short term, we believe it will perform much more durably. And the interesting thing about durability is that its importance has almost always been underrated. Steady progress often beats short term flash, as the well-studied phenomenon of high-quality stock outperformance has consistently demonstrated. In fact, this tendency to misprice durability is so timeless and core to human nature that Aesop even made a fable about it over two thousand years ago. As a result of this remarkable universality, we believe your portfolio of boring toll collectors will prove, much like the tortoise, to be not only more durable than many of today’s flashier portfolios but also more remunerative in the end.
As always, thank you so much for your trust, know that we continue to be invested right alongside you, and please always reach out to us if you have any questions or concerns. We’re here to help!
Sincerely,
The YCG Team
Disclaimer: The specific securities identified and discussed should not be considered a recommendation to purchase or sell any particular security, nor were they selected based on profitability. Nothing said in this piece may be considered to be an offer to buy or sell any security. Rather, this commentary is presented solely for the purpose of illustrating YCG’s investment approach. These commentaries contain our views and opinions at the time such commentaries were written and are subject to change thereafter. The securities discussed do not represent an account’s entire portfolio and, in the aggregate, may represent only a small percentage of an account’s portfolio holdings. These commentaries may include “forward looking statements” which may or may not be accurate in the long-term. It should not be assumed that any of the securities transactions or holdings discussed were or will prove to be profitable. Data presented was obtained from sources deemed to be reliable, but no guarantee is made as to its accuracy. S&P stands for Standard & Poor’s. All S&P data is provided “as is”. In no event, shall S&P, its affiliates or any S&P data provider have any liability of any kind in connection with the S&P data. No further distribution or dissemination of the S&P data is permitted without S&P’s prior express written consent. All MSCI data is provided “as is.” In no event, shall MSCI, its affiliates or any MSCI data provider have any liability of any kind in connection with the MSCI data. Past performance is no guarantee of future results.
1 For information on the performance of our separate account composite strategies, please visit www.ycginvestments.com/performance. For information about your specific account performance, please contact us at (512) 505-2347 or email [email protected]. All returns are in USD unless otherwise stated.
2 See https://ycginvestments.com/2024-q4-investment-letter-prepared-for-any-outcome/.
3 See https://www.ssga.com/us/en/intermediary/etfs/spdr-sp-500-etf-trust-spy. Data as of October 13, 2025.
4 See https://www.marketwatch.com/story/its-not-an-ai-bubble-its-a-longer-term-bull-market-says-this-wall-st-firm-909b7fe2.
5 See https://law.yale.edu/sites/default/files/documents/pdf/sidak.pdf.
6 See https://www.gatesnotes.com/ai-agents.
7 See https://www.dwarkesh.com/p/satya-nadella.
8 See https://www.datacenterdynamics.com/en/news/alibabas-joe-tsai-voices-concern-about-ai-data-center-construction-bubble/.
9 See https://www.cnbc.com/2025/08/18/openai-sam-altman-warns-ai-market-is-in-a-bubble.html.
10 See https://singjupost.com/transcript-amazon-founder-jeff-bezos-on-tech-future-ai-space-revolution-italian-tech-week-2025/.
11 See https://podcasts.apple.com/us/podcast/mark-zuckerberg-on-the-ai-bubble-and-metas-new/id1073226719?i=1000731779809.
12 See https://mitsloan.mit.edu/ideas-made-to-matter/googles-sundar-pichai-says-tech-a-powerful-agent-change.
13 See https://observer.com/2025/02/biggest-risk-ai-is-missing-out-google-ceo-sundar-pichai/.
14 See https://finance.yahoo.com/news/google-shows-investor-patience-with-big-techs-ai-spending-might-be-running-short-154439569.html.
15 See https://progressive-research.com/insights/ais-god-delusion/.
