“We find whole communities suddenly fix their minds upon one object, and go mad in its pursuit.” - Charles Mackay, Extraordinary Popular Delusions and the Madness of Crowds
As we’ve pointed out quite a few times (some say we are perseverating) over the past few years, we believe the markets are showing classic signs of a bubble. At the risk of continued underperformance, we have taken profits from many of our positions over the past twelve months. Many of you will see that we have reached double-digit cash positions. We have decided to break this article into two pieces. The first will be a brief explanation of what we think are the characteristics of today’s market bubble. An upcoming second article will discuss several key measurement tools we use to ascertain when/how the market data support our thesis.
In 1929, at the height of the post-World War I economic boom in America, Joseph P. Kennedy Sr. was working as a stockbroker on Wall Street. As the story goes, Kennedy was heading back to his office when he decided to get a shoeshine. While polishing his shoes, the young worker shared some of his favorite stock picks. As he listened, Kennedy realized that when shoeshine boys give their patrons stock picks, it’s time to get out of the market. Kennedy exited his market positions and took short positions that bet on the market going down. Was the story true? While probably apocryphal, it certainly proved a point. Market bubbles and their associated tops generally have similar attributes. One of them is that even the most unwise of investors feel like geniuses as their portfolios keep increasing in value with each move.
Having said that, it’s nearly impossible to call the exact moment when bubbles reach their apex. It can be one day, one month, or three years after you become convinced the markets are in one that the bubble finally pops.
So how does one predict when a bubble might implode and take action to protect their portfolio? We think there are several attributes common to all bubbles. When markets begin to exhibit these simultaneously, we begin to reduce our positions and raise cash at Nintai. The major signs we look for are a.) valuations at sky-high levels, b.) a constant refrain of “it’s different this time”, c.) the bubble's theme becomes an inordinately large percentage of market share, and d.) companies beginning to spend like drunken sailors on the new-fangled toy. As we look out today, we believe the market exhibits all of these, and then some.
We should be very clear: we have no idea when or how this bubble will collapse. But it will. Looking back at history, the devastation in the marketplace is usually broad and deep. In this instance, we aim to prevent the one thing that can devastate long-term returns: the permanent loss of capital in our investment partners’ portfolios.
Valuations Have Reached Nosebleed Levels
Overall market valuation metrics, such as the cyclically adjusted price-to-earnings (CAPE) ratio, are at or near historical highs, indicating that investors have extremely high expectations for future growth. The Shiller PE ratio has reached levels comparable to those seen during the 1999-2000 Internet Bubble. In December 1999, the market’s CAPE ratio reached an all-time high of 44.2. In January 2026, the market reached a second-highest CAPE ratio ever of 40.6. The historical mean is roughly 17 - 18. We believe the levels we are seeing in today’s market are certainly indicative of a market bubble.
As we’ve pointed out quite a few times (some say we are perseverating) over the past few years, we believe the markets are showing classic signs of a bubble. At the risk of continued underperformance, we have taken profits from many of our positions over the past twelve months. Many of you will see that we have reached double-digit cash positions. We have decided to break this article into two pieces. The first will be a brief explanation of what we think are the characteristics of today’s market bubble. An upcoming second article will discuss several key measurement tools we use to ascertain when/how the market data support our thesis.
In 1929, at the height of the post-World War I economic boom in America, Joseph P. Kennedy Sr. was working as a stockbroker on Wall Street. As the story goes, Kennedy was heading back to his office when he decided to get a shoeshine. While polishing his shoes, the young worker shared some of his favorite stock picks. As he listened, Kennedy realized that when shoeshine boys give their patrons stock picks, it’s time to get out of the market. Kennedy exited his market positions and took short positions that bet on the market going down. Was the story true? While probably apocryphal, it certainly proved a point. Market bubbles and their associated tops generally have similar attributes. One of them is that even the most unwise of investors feel like geniuses as their portfolios keep increasing in value with each move.
Having said that, it’s nearly impossible to call the exact moment when bubbles reach their apex. It can be one day, one month, or three years after you become convinced the markets are in one that the bubble finally pops.
So how does one predict when a bubble might implode and take action to protect their portfolio? We think there are several attributes common to all bubbles. When markets begin to exhibit these simultaneously, we begin to reduce our positions and raise cash at Nintai. The major signs we look for are a.) valuations at sky-high levels, b.) a constant refrain of “it’s different this time”, c.) the bubble's theme becomes an inordinately large percentage of market share, and d.) companies beginning to spend like drunken sailors on the new-fangled toy. As we look out today, we believe the market exhibits all of these, and then some.
We should be very clear: we have no idea when or how this bubble will collapse. But it will. Looking back at history, the devastation in the marketplace is usually broad and deep. In this instance, we aim to prevent the one thing that can devastate long-term returns: the permanent loss of capital in our investment partners’ portfolios.
Valuations Have Reached Nosebleed Levels
Overall market valuation metrics, such as the cyclically adjusted price-to-earnings (CAPE) ratio, are at or near historical highs, indicating that investors have extremely high expectations for future growth. The Shiller PE ratio has reached levels comparable to those seen during the 1999-2000 Internet Bubble. In December 1999, the market’s CAPE ratio reached an all-time high of 44.2. In January 2026, the market reached a second-highest CAPE ratio ever of 40.6. The historical mean is roughly 17 - 18. We believe the levels we are seeing in today’s market are certainly indicative of a market bubble.
Figure 1: Historical CAPE Ratio: 1870 - Present, Source: https://www.multpl.com/shiller-pe
No, really….it’s different this time.
During bubbles, investors are often confronted with questions about their willingness to pay what seem like incredibly high prices for what they’re investing in. The almost invariable response is the assertion that “this time is different.”[1] In recent years, as prices and valuations have slowly risen to nosebleed levels, we have begun to hear this phrase used all too often to explain such conditions. Here are some recent examples:
“It’s Different This Time”: Debt
One of the things cited as the major difference between the last technology or real estate bubbles is that AI generates lots of revenue, unlike the 1999-2000 Internet hype. That may be true, but it also matters a great deal about the underlying costs of generating this revenue. For many years, large tech could fund initiatives with strong free cash flows and cash on the balance sheet. No more. In 2025 alone, large tech issued over $121B in debt. In the first three months of 2026, an additional $42B was raised. In Oracle's (ORCL) case, the company has gone from $37B in cash/short-term instruments on the balance sheet in 2020 to less than $10B in 2025. Conversely, total debt has gone from $71B in 2020 to $123B in 2025. The company expects to raise an additional $20-$40B for AI investments over the next five years. Meta's AI spending frenzy has taken the company from a net cash position of nearly $30 billion in 2023 to minus $7 billion today. Some relatively large assumptions are being built into the ability to service such debt.
“It’s Different This Time”: Investors are wiser
We like to think that after the technology bubble of 1999 - 2000 and the real estate/credit crisis of 2007 - 2009, it would be painfully evident to most investors that remarkable returns, inflated valuations, and accounting shenanigans are too good to be true. Unfortunately, this simply isn’t the case. Two instances in particular show that investors never lose their desire to make big profits quickly. Both were fanned by rates close to zero percent, leaving savers with little choice but to seek investment gains in the bond market.
In 2020 and 2021, Wall Street saw an explosion in Special Purpose Acquisition Companies (SPACs). SPACs offered a faster, less regulated path for private companies (especially in tech, EV, and space) to go public than traditional IPOs. SPACs would raise billions of dollars as “blank check” companies, then use these funds to acquire a private company and merge it into their publicly traded SPAC. Over 600 SPAC-based acquisitions received billions of dollars in less than two years. Inevitably, the market collapsed for several reasons. First, SPAC owners made their profit on the acquisition, not on the acquired company’s performance. Second, the explosion in SPACs created a race to acquire with a relatively small pool of quality companies. With both factors at work, most SPACs began to fail within months, as their acquired companies simply weren’t good investments. As usual, Wall Street movers and shakers made their profits, while individual SPAC investors were left holding the bag.
After the SPAC model imploded, they were quickly replaced by the MEMO stock craze.
A meme stock is one whose company’s financials and business fundamentals appear very weak, yet whose stock price suddenly surges, driven by social media hype (such as Reddit communities or subreddits) and massive retail buying. Frequently, these stocks are heavily shorted by Wall Street players, including private equity firms. This model was a way to profit from short squeezes against these players. It should be noted that this model has nothing to do with investment models utilizing valuation metrics or business fundamentals. This is wholly market gambling, with little interest on investing for the long term. Originating in 2021 with GameStop, the phenomenon has shown resilience, with resurgences occurring as recently as 2024 and 2025. In its later stages, the MEMO stocks began to create their own groupings, such as the “DORK” stocks - "DNUT (Krispy Kreme), OPEN (Opendoor Technologies), RKLB (Rocket Lab), and KSS (Kohl's).
It’s Different This Time: Profits
One factor that led to the bursting of the Internet bubble in 1999 was the proliferation of companies that went public with no reasonable way to generate profits, short-term or long-term. We are beginning to see the first ripples of concern as the AI model is questioned about how companies expect to generate a profit. In Q4 2025, Harvard Business Review published a report[2] that questioned this very issue.
There’s been a push to have investors focus only on gross revenue, assuming that many of the costs that go into generating net revenue will go away over time. This seems an awful lot like the move to using EBITA as the valuation criterion or the old days of not expensing stock options. As Warren Buffett stated: "If options aren't a form of compensation, what are they? If compensation isn't an expense, what is it? And, if expenses shouldn't go into the calculation of earnings, where in the world should they go?" Or as Charlie Munger so niftily stated, “EBITDA is bullshit earnings.”
Clearly, we haven’t figured out how AI will be profitable. In 2024, OpenAI lost $5B, and Anthropic lost $5.3B. In 2025, OpenAI was expected to lose roughly $8B, and in 2026, roughly $14B. Anthropic, after initially estimating a $3B loss in 2025, saw the actual number come in at $5.2B. Last year, OpenAI committed to $1.4T (yes, that’s trillion) in capital spending on data centers and related infrastructure before reaching profitability.
To become profitable, OpenAI will need to generate roughly $200B in revenue (15x current revenue in 2025) by 2030. Sam Altman, CEO of OpenAI, stated that investors are overexcited and "someone is going to lose a phenomenal amount of money." Not what one wants to hear as an AI investor.
“It’s Different This Time”: AI Lifts All Boats
In their recent post, “Valuing AI: Extreme Bubble, New Golden Era, or Both?”,
Jeremy Grantham and Edward Chancellor discussed how AI has become the central focus for market capitalization growth. They write:
“The United States, whose companies are at the center of the AI boom, now makes up more than 70% of the MSCI World index. J. P. Morgan estimated last September that 44% of S&P 500 capitalization—over a quarter of total world market capitalization—was accounted for by just 30 AI-linked names. In January 2000, 7 of the top 10 companies by market cap came from the technology, media, and telecoms (TMT) sectors. At the time of writing, the eight largest companies in the world were Nvidia, Apple, Google, Microsoft, Amazon, Meta, Broadcom, and TSMC. Of those eight, Apple is the only one
that is not a highly active participant in or apparent beneficiary of the AI race (although doubtless there will be synergies between AI and Apple's products). Saudi Aramco, the producer of 10% of all the world's oil, is down in ninth place.”
It is easy to see why the Magnificent Seven (or is it Ten now?) has generated most of the equity gains over the past few years. This is where almost all the profit growth across the S&P 500 has come from. This is evident in the chart below. The M7 has delivered consistent earnings growth of more than 20% annually since 2023. Profit growth from the S&P 493 has been much lower, averaging 3% over the same period.
Conclusions
The past five years have been extraordinarily difficult for those who haven’t invested in AI/Magnificent 7/Meme stocks. At Nintai, we’re among those who missed out on this cycle. Not investing in any of these stocks has crushed our returns relative to the S&P 500. That said, we aren’t disappointed with the business returns of our portfolio holdings. On average, our companies increased free cash flow by 11.9% annually over the past 5 years. Our average return on invested capital is 31.6%. While our investment returns have been difficult, our underlying businesses are doing just fine. When will things turn around from an investment perspective? We have no idea, but the case we’ve laid out in this (far too) brief article gives us some comfort that we will avoid the inevitable return to the mean. We didn’t boast when we outperformed in the first four years of our investment management services, and we aren’t going to cry in our cups about missing out on bubble-like returns. We will continue to do what we know best – find outstanding businesses with great capital allocators at the helm and let long-term compounding do its work. Until then, we thank our investment partners for their patience and trust in our investment process.
Disclosures: None
[1] We, of course, would be remiss not to thank Sir John Templeton for his classic rejoinder, “The four most dangerous words in investing are ‘it’s different this time.’”
[2] “AI Companies Don’t Have a Profitable Business Model. Does That Matter?”, Danielle Kost, Harvard Business Review, November 12, 2025
No, really….it’s different this time.
During bubbles, investors are often confronted with questions about their willingness to pay what seem like incredibly high prices for what they’re investing in. The almost invariable response is the assertion that “this time is different.”[1] In recent years, as prices and valuations have slowly risen to nosebleed levels, we have begun to hear this phrase used all too often to explain such conditions. Here are some recent examples:
“It’s Different This Time”: Debt
One of the things cited as the major difference between the last technology or real estate bubbles is that AI generates lots of revenue, unlike the 1999-2000 Internet hype. That may be true, but it also matters a great deal about the underlying costs of generating this revenue. For many years, large tech could fund initiatives with strong free cash flows and cash on the balance sheet. No more. In 2025 alone, large tech issued over $121B in debt. In the first three months of 2026, an additional $42B was raised. In Oracle's (ORCL) case, the company has gone from $37B in cash/short-term instruments on the balance sheet in 2020 to less than $10B in 2025. Conversely, total debt has gone from $71B in 2020 to $123B in 2025. The company expects to raise an additional $20-$40B for AI investments over the next five years. Meta's AI spending frenzy has taken the company from a net cash position of nearly $30 billion in 2023 to minus $7 billion today. Some relatively large assumptions are being built into the ability to service such debt.
“It’s Different This Time”: Investors are wiser
We like to think that after the technology bubble of 1999 - 2000 and the real estate/credit crisis of 2007 - 2009, it would be painfully evident to most investors that remarkable returns, inflated valuations, and accounting shenanigans are too good to be true. Unfortunately, this simply isn’t the case. Two instances in particular show that investors never lose their desire to make big profits quickly. Both were fanned by rates close to zero percent, leaving savers with little choice but to seek investment gains in the bond market.
In 2020 and 2021, Wall Street saw an explosion in Special Purpose Acquisition Companies (SPACs). SPACs offered a faster, less regulated path for private companies (especially in tech, EV, and space) to go public than traditional IPOs. SPACs would raise billions of dollars as “blank check” companies, then use these funds to acquire a private company and merge it into their publicly traded SPAC. Over 600 SPAC-based acquisitions received billions of dollars in less than two years. Inevitably, the market collapsed for several reasons. First, SPAC owners made their profit on the acquisition, not on the acquired company’s performance. Second, the explosion in SPACs created a race to acquire with a relatively small pool of quality companies. With both factors at work, most SPACs began to fail within months, as their acquired companies simply weren’t good investments. As usual, Wall Street movers and shakers made their profits, while individual SPAC investors were left holding the bag.
After the SPAC model imploded, they were quickly replaced by the MEMO stock craze.
A meme stock is one whose company’s financials and business fundamentals appear very weak, yet whose stock price suddenly surges, driven by social media hype (such as Reddit communities or subreddits) and massive retail buying. Frequently, these stocks are heavily shorted by Wall Street players, including private equity firms. This model was a way to profit from short squeezes against these players. It should be noted that this model has nothing to do with investment models utilizing valuation metrics or business fundamentals. This is wholly market gambling, with little interest on investing for the long term. Originating in 2021 with GameStop, the phenomenon has shown resilience, with resurgences occurring as recently as 2024 and 2025. In its later stages, the MEMO stocks began to create their own groupings, such as the “DORK” stocks - "DNUT (Krispy Kreme), OPEN (Opendoor Technologies), RKLB (Rocket Lab), and KSS (Kohl's).
It’s Different This Time: Profits
One factor that led to the bursting of the Internet bubble in 1999 was the proliferation of companies that went public with no reasonable way to generate profits, short-term or long-term. We are beginning to see the first ripples of concern as the AI model is questioned about how companies expect to generate a profit. In Q4 2025, Harvard Business Review published a report[2] that questioned this very issue.
There’s been a push to have investors focus only on gross revenue, assuming that many of the costs that go into generating net revenue will go away over time. This seems an awful lot like the move to using EBITA as the valuation criterion or the old days of not expensing stock options. As Warren Buffett stated: "If options aren't a form of compensation, what are they? If compensation isn't an expense, what is it? And, if expenses shouldn't go into the calculation of earnings, where in the world should they go?" Or as Charlie Munger so niftily stated, “EBITDA is bullshit earnings.”
Clearly, we haven’t figured out how AI will be profitable. In 2024, OpenAI lost $5B, and Anthropic lost $5.3B. In 2025, OpenAI was expected to lose roughly $8B, and in 2026, roughly $14B. Anthropic, after initially estimating a $3B loss in 2025, saw the actual number come in at $5.2B. Last year, OpenAI committed to $1.4T (yes, that’s trillion) in capital spending on data centers and related infrastructure before reaching profitability.
To become profitable, OpenAI will need to generate roughly $200B in revenue (15x current revenue in 2025) by 2030. Sam Altman, CEO of OpenAI, stated that investors are overexcited and "someone is going to lose a phenomenal amount of money." Not what one wants to hear as an AI investor.
“It’s Different This Time”: AI Lifts All Boats
In their recent post, “Valuing AI: Extreme Bubble, New Golden Era, or Both?”,
Jeremy Grantham and Edward Chancellor discussed how AI has become the central focus for market capitalization growth. They write:
“The United States, whose companies are at the center of the AI boom, now makes up more than 70% of the MSCI World index. J. P. Morgan estimated last September that 44% of S&P 500 capitalization—over a quarter of total world market capitalization—was accounted for by just 30 AI-linked names. In January 2000, 7 of the top 10 companies by market cap came from the technology, media, and telecoms (TMT) sectors. At the time of writing, the eight largest companies in the world were Nvidia, Apple, Google, Microsoft, Amazon, Meta, Broadcom, and TSMC. Of those eight, Apple is the only one
that is not a highly active participant in or apparent beneficiary of the AI race (although doubtless there will be synergies between AI and Apple's products). Saudi Aramco, the producer of 10% of all the world's oil, is down in ninth place.”
It is easy to see why the Magnificent Seven (or is it Ten now?) has generated most of the equity gains over the past few years. This is where almost all the profit growth across the S&P 500 has come from. This is evident in the chart below. The M7 has delivered consistent earnings growth of more than 20% annually since 2023. Profit growth from the S&P 493 has been much lower, averaging 3% over the same period.
Conclusions
The past five years have been extraordinarily difficult for those who haven’t invested in AI/Magnificent 7/Meme stocks. At Nintai, we’re among those who missed out on this cycle. Not investing in any of these stocks has crushed our returns relative to the S&P 500. That said, we aren’t disappointed with the business returns of our portfolio holdings. On average, our companies increased free cash flow by 11.9% annually over the past 5 years. Our average return on invested capital is 31.6%. While our investment returns have been difficult, our underlying businesses are doing just fine. When will things turn around from an investment perspective? We have no idea, but the case we’ve laid out in this (far too) brief article gives us some comfort that we will avoid the inevitable return to the mean. We didn’t boast when we outperformed in the first four years of our investment management services, and we aren’t going to cry in our cups about missing out on bubble-like returns. We will continue to do what we know best – find outstanding businesses with great capital allocators at the helm and let long-term compounding do its work. Until then, we thank our investment partners for their patience and trust in our investment process.
Disclosures: None
[1] We, of course, would be remiss not to thank Sir John Templeton for his classic rejoinder, “The four most dangerous words in investing are ‘it’s different this time.’”
[2] “AI Companies Don’t Have a Profitable Business Model. Does That Matter?”, Danielle Kost, Harvard Business Review, November 12, 2025
RSS Feed