The Trainwrecks

Back when I was Risk Manager for a large global hedge fund, my quarterly review always began with a category I called “the trainwrecks.” These were the parts of the market that had performed especially poorly over the recent past. Identifying struggling industry segments early helped me map key risks to the market and broader economy, avoid potholes, spot accumulating stress, and detect signs of more systemic risk.

I revisited that process recently while dissecting performance across different market segments. Despite a fairly benign environment at the headline equity-index level, several areas are not doing well at all. Below are some examples.

The Anti-Craving Trend

Unit sales of snacks are down 4% over the past four years and 17% for sweet snacks in particular. Traffic growth at fast-food chains has turned negative: using Chipotle as one example, traffic growth has averaged -1.8% over the past five quarters. Total beverage alcohol volume fell 5% in 2025, according to IWSR, with declines across nearly every major category: beer down 6%, wine down 6%, and spirits down 4%. McKinsey’s State of Grocery North America 2026 survey showed grocery sales rising 1.2%, but that growth was more than explained by price increases, which contributed 2.2%, while volumes actually declined by 1.0%. The broader story is a decline in consumption across several categories within food and beverage, somewhat masked by companies lowering the serving size and increasing prices in order to maintain revenue growth.

It is difficult to draw a direct line between consumption patterns and any single underlying cause, but obesity drugs appear to be having at least some effect. According to Gallup, 12.4% of U.S. adults were taking a GLP-1 drug for weight loss last year, and that figure is projected by some analysts to reach 20% in the near term. That is a large number of people, 32 million and growing. A peer-reviewed Cornell/Numerator study published in the Journal of Marketing Research found that households using these drugs clearly reduced grocery spending, with the largest reductions in calorie-dense, processed categories. They also drink less: alcohol spending as a share of overall expenditures now sits near a 40-year low, although that’s partly due to the younger consumer cohort drinking materially less than older generations and by some substitution toward cannabis.

The 12-month stock price returns for many names in those sub-industry groups are deeply negative despite a positive overall market: Chipotle -39%, Domino’s Pizza -34%, General Mills -33%, Constellation Brands -15%, Albertson’s -37%, McCormick -34% to name just a few.

The Private Equity Logjam

The traditional private-equity playbook is straightforward: buy a high-potential business using an optimized mix of equity and, ideally, cheap corporate debt; streamline operating costs; pursue bolt-on acquisitions to expand across product verticals and geographies; and then wait for a more favorable market environment to exit at a higher multiple than the one originally paid.

The past fifteen years have been a golden era for private equity. The cost of debt was exceptionally low, and financing was abundant, partly due to the rapid growth of private credit funds that were eager to provide deal financing. Low inflation made it easier to control operating costs, while rising equity markets made exits profitable even when the underlying companies had not created much internal value. Private equity firms and their fund managers profited handsomely, as did the private credit firms financing their deals. High profits drew new entrants, causing both the number and average size of private equity and credit funds to balloon.

The current backdrop is quite different. Both inflation and interest rates are now sitting on a much higher plateau than they were before 2020. At the same time, the sheer number of companies owned by private equity, 13,500 as of June 30, according to PitchBook, including almost 4,000 that have been held for six or more years, means that even a healthy IPO market could not realistically absorb that much supply. Fundraising has also been muted and concentrated among the largest managers, making sponsor-to-sponsor exits more difficult. As we discussed in our last letter (CAM Insights Q1 2026), another challenge is that a significant portion of private equity acquisitions occurred in the software sector. Many were bought at the elevated multiples of the 2020–2021 valuation era, which look increasingly difficult to justify as the threat from AI has contributed to a sharp contraction in public software valuations.

The market is slowly coming to terms with two realities: first, it may take many years to clear the logjam of unsold portfolio companies; and second, many of these underlying businesses could become “zombie” companies or ultimately default as financing options narrow and fundamental growth becomes harder to generate.

One-year returns for the publicly listed names in the space are deeply negative: TPG -23%, KKR -31%, Carlyle -18%, Blue Owl -54%, and Ares -36%.

The Agent Substitution

The basic concept is that artificial intelligence models are making it possible to execute certain tasks directly, without the need for a traditional middleman such as a software service, information and data provider, IT consultant, or call center. Although AI companies often describe themselves as partners to these enterprises, helping make their products and services better, it is clear to anyone who has done the math that they will eventually try to replace many of them.

The reason is straightforward. AI is currently being heavily subsidized, and the hundreds of billions of dollars being spent on compute will eventually need to be funded through increased monetization. That will likely force AI companies to move deeper into the application layer. Their valuations also depend on it because their current market value is difficult to justify unless they can capture the revenue pools currently being spent on the products and services they claim to augment.

Even though there is still little evidence of companies canceling their QuickBooks, Salesforce, or Office subscriptions, the core earnings algorithm for these businesses has been disrupted. They have traditionally relied on customer stickiness, upselling, and annual price increases. AI threatens all three. Even if clients do not cancel outright, they now have the option to build internal AI tools that require some upfront investment but then replace steadily growing recurring expenses. That creates long-term uncertainty for these businesses and increases client negotiating leverage.

The market has started to price this in the rising uncertainty: the 12-month return of Salesforce is -43%, Workday -49%, Intuit -67%, Cognizant -50%, Accenture -58%, FactSet -49%.

The Concentration Problem

So how do we interpret this destruction of value?

In a recent presentation on the state of private markets for technology companies, Thomas Laffont, the co-founder of Coatue, included several data points I found interesting. Somewhat unsurprisingly, AI is dominating fundraising, accounting for 78% of venture dollars deployed. But those dollars are also accruing to only a handful of companies: the top 10 fundraisers accounted for 77% of overall capital raised. In the ten years through 2024, that figure had averaged only 16%. That represents an extraordinary concentration of capital allocation.

Investors tend to overstate the value of management and understate the importance of access to capital when assessing the drivers of a company’s success. Having an innovative idea is important, but having the financial resources to implement that idea is often even more critical. Laffont’s research found that the likelihood of a $1 billion company becoming a $10 billion company is only 8%, while the likelihood of a $100 billion company becoming a $1 trillion company is 31%.

His data set may be too narrow to be statistically significant, yet it illustrates what I believe has been happening: the overall pie is growing, but value is accruing to a small number of companies with massive capital resources at the expense of many smaller competitors.

The market capitalization of Eli Lilly, the leader in obesity drugs, has increased by just under $1 trillion since 2021. That is more than the combined market-cap loss of many of the food and beverage companies being pressured by the rise of those drugs. On a net basis, the economic pie has grown, but the value creation has been highly asymmetric: one major winner and many losers. A similar dynamic is visible in AI and semiconductors. Three companies that were private for most of this period (OpenAI, Anthropic, and SpaceX, which went public in June), plus three semiconductor companies (Nvidia, Micron, and SK Hynix) have generated approximately $7.5 trillion in combined market-cap gains over the past three years. Not only does that dwarf the market-cap losses seen across the software and services sector during that period, but as a standalone figure, it would rank as the third-largest economy in the world, behind only the United States and China. A small number of very large companies are creating so much incremental value from already massive bases that they can overwhelm the negative growth experienced by many smaller companies that are being disrupted.

This is why portfolio diversification, while useful in limiting downside, can also dilute returns in a market dominated by a small number of outsized winners. A highly diversified portfolio may still generate little return unless it owns some of those winners. And outperforming requires more than simply owning them; it requires allowing them to compound and naturally become a larger share of the portfolio, even if that reduces diversification over time.

Of course, we can only spot the winners with the benefit of hindsight. How do you identify them before the parabolic return? And how do you avoid coming in too late, right before the music stops? It is not easy, and that is where an analytical process becomes critical. But there are some simple rules.

First, invest in companies that at least have the potential to compound at exceptionally high rates of return. If a company is not addressing a very large market, benefiting from a multi-year industry tailwind, operating in an oligopolistic competitive environment, or offering a product that is in high demand, it is unlikely to ever deliver truly massive outcomes.

Second, do not shy away from companies that are already large or have already been strong performers, provided that performance has been supported by corresponding earnings growth. As we noted above, it is precisely those companies that often have the ability to continue generating high returns on invested capital, build broader ecosystems, and capture additional market share.

Third, volatility does not equal risk. Volatility is simply the degree of variation in a stock’s trading price over time. In fact, it is often a prerequisite for the positive price spikes that create unusually high return outcomes. True safety comes from the predictability of an underlying investment’s fundamental progress, not from controlling the variability of its share price.

Finding Value in the Wreckage

When do the trainwrecks become value opportunities?

Value and momentum are symptoms rather than drivers of fundamental developments. Momentum works in the short to medium term because success tends to reinforce itself. A company that is growing quickly attracts talent and capital, which can further accelerate growth. Value works over a longer horizon because a large decline in valuation can sharpen focus and trigger change. A company may enter survival mode by cutting unnecessary costs or selling non-core assets. It may shift its product offering to better align with the changing environment, replace ineffective management, or attract activist capital that accelerates the pace of change. There is significant execution risk, which is why value investing is typically much harder than momentum trading. But when it works, the upside can be substantial.

Big Food needs to reinvent itself. The old model of driving growth through expensive advertising, prominent placement on Walmart shelves, and steady price increases without volume pressure needs to change. The broader food and beverage category lost 20 basis points of share of personal expenditures from 2019 to 2025. Competition from new brands without legacy portfolios of unhealthy, high-sugar products is intensifying, as is competition from lower-priced private-label offerings from large grocers. A handful of brands will likely reinvent themselves successfully and gain share as weaker competitors fail. But I sense we are still early in that journey, making the space less attractive as a hunting ground for now.

The private equity logjam will be felt most acutely by smaller private equity and private credit firms with greater exposure to the middle market, less diversification, and no public listing. The public alternative asset managers are a small subset of the industry and typically the stronger players, with established fundraising franchises and diversified platforms. Alternative assets still represent a small percentage of overall investment assets and have been steadily gaining share, a trend I expect to continue over time. Despite their negative one-year returns, the stock prices of these firms are generally still up significantly over five years, and I would expect them to  continue their uptrend once visibility improves.

The agent substitution category is where the value opportunities appear most imminent. No one really knows how quickly, or in what ways, increased AI adoption will replace certain tasks. As a result, the market’s current approach seems to be: when in doubt, sell first and ask questions later.

The list of affected subsegments is long: vertical software providers, online marketplaces, brokers and other intermediaries, recruiters, information services, IT consultants, and others. The narrative may differ slightly across each disrupted segment, but the broad theme is the same: a service is replaced by a more intelligent AI-driven version, leading to fewer vendors, consultants, intermediaries, and employees. There is no question that long-term visibility is lower and risk is higher for all of these industry groups, but higher risk does not necessarily mean an adverse outcome.

There are also two AI-related factors that I believe will become more prominent in the coming years.

The first is ownership of information. The quality of any AI system depends heavily on the quality of the material used to train it. AI models are trained by ingesting and processing enormous quantities of books, journalism, artwork, music, photographs, code, and other material. They are also improved through the unpaid intellectual and behavioral contributions of their own users. Satya Nadella recently wrote an excellent piece on this topic, which is a must-read:  Satya Nadella: Reverse Information Paradox. In the era of AI, “in consuming intelligence, you are creating intelligence. And what you create should belong to you.” I expect a looming fight between information providers and the AI models that process, distill, and use that information to extract intelligence. That fight could tilt some power back toward the former group.

The second is the view that open-source models will rise in prominence. Research cited by MIT Sloan found that open models averaged roughly 90% of closed-model performance and historically caught up with leading releases within about 13 weeks. These models can be hosted privately, customized and fine-tuned, and run without sending sensitive data to an outside provider. The argument is that once models become sufficiently capable, foundation-model intelligence becomes a commodity, and competitive advantage shifts to proprietary data, applications, distribution, and customer relationships. As this thesis gains traction, we could see some value move away from frontier models and the expensive infrastructure buildout, and toward data owners and application companies.

Either way, the broader argument about value accruing to a small number of winners still stands. Even within the trainwreck categories, the leaders that capture the lion’s share of capital and talent will be best positioned to create a flywheel of market-share gains and compounding returns.

This material is distributed for informational purposes only. Investment Advisory services offered through Chesapeake Asset Management, an investment adviser registered with the U.S. Securities and Exchange Commission (“SEC”). The views expressed are for informational purposes only and do not take into account any individual’s personal, financial, or tax considerations. Opinions expressed are subject to change without notice and are not intended as investment advice. The information contained herein is not intended to be personal legal, investment or tax advice, or a solicitation to buy or sell any security or engage in a particular investment strategy. Nothing herein should be relied upon as such, and there is no guarantee that any claims made will come to pass.

All returns for the names mentioned herein represent price return only and reflect the change in price between 06/30/2025 and 06/30/2026. The names mentioned were selected as representatives of an industry trend and do not necessarily represent investment positions held by Chesapeake Asset Management or its clients. In addition, certain information contained herein or utilized to draw the conclusions contained herein has been provided by, or obtained from, third party sources. All material presented is compiled from sources believed reliable, but the accuracy and completeness, and the opinions based thereon, are not guaranteed. While Chesapeake Asset Management believes that such sources are reliable, it cannot guarantee the accuracy of any such information and does not represent that such information is accurate or complete.

All materials and information are provided “as is” without any express or implied warranties. Opinions expressed are subject to change without notice and are not intended as investment advice or to predict future performance. Past performance is no guarantee of future results. All investments involve risk including the loss of principal. Any security can experience a loss. Please see Chesapeake Asset Management’s Form ADV Part 2A and Form CRS for additional information.

Registration with the SEC does not imply a certain level of skill or training. Neither the SEC nor any other regulatory authority has approved, passed on or endorsed the merits of any investment opportunity that may be described herein. No regulatory authority or other third party has confirmed the accuracy or determined the adequacy of this presentation. Any representation to the contrary is a criminal offense.

The letter and its contents are proprietary information of Chesapeake Asset Management, LLC, (“CAM”) and any reproduction of this information, in whole or in part, without prior written consent of CAM is prohibited. Additional information is available from CAM upon request. Neither CAM nor its affiliates is acting as your advisor or agent.