The AI Bubble, Market Concentration, and the Cost of Overconfidence

Key Takeaways: Is the AI Bubble a Risk to Your Portfolio?

• Historic Market Concentration: The U.S. market is at its most concentrated level since 1927. Currently, just seven companies make up 36% of the S&P 500, meaning the market is heavily reliant on the performance of a handful of tech giants.‍
• Valuation vs. Concentration: While concentration itself doesn't guarantee a crash, high valuations do. U.S. equity prices are approaching 1999 dot-com levels. History shows that paying premium prices for earnings usually leads to lower long-term returns.‍
• Lessons from Nortel: The 2000 collapse of Nortel (which was 36% of the Canadian market) proves that real technology can still result in investment losses if the price is too high. However, investors who stayed diversified recovered quickly.‍
• The "Productive Bubble" Theory: AI is a real technological revolution, much like railways or the internet. These sectors often experience bubbles that burst, yet the infrastructure remains valuable to the economy long-term.‍
• How to Protect Your Wealth: Since timing a bubble is impossible, the best strategy is broad global diversification. Reducing reliance on U.S. mega-cap tech and tilting portfolios toward value and high-profitability stocks reduces downside risk while capturing future growth.

Recently, there’s been growing discussion about the risk of a stock market crash, fueled by sky-high valuations in a small group of stocks and increasing market concentration. One of the core promises of index investing is diversification: own the market, spread your risk, and avoid reliance on any single company or sector. But today, many investors are understandably asking whether that promise still holds.

As of now, roughly 36% of the S&P 500 is concentrated in just seven stocks. Even when we zoom out to the total U.S. stock market, those same companies represent about 32% of total market capitalization. That is the highest level of index concentration in U.S. market history, going back to 1927.

At the same time, U.S. equity valuations are approaching levels last seen in 1999, just before the dot-com crash. A period that ultimately delivered a decade of flat to disappointing returns for investors who bought at the peak.

It’s understandable that this feels uncomfortable.

If a small handful of companies dominate returns, earnings growth, and capital investment, and those companies falter, the impact on portfolios could be substantial. This is not a hypothetical concern. It’s something we’ve seen before.

A Canadian Cautionary Tale: Nortel

In July 2000, one company made up approximately 36% of the entire Canadian stock market. That company was Nortel Networks.

Nortel was not a scam. It was not hype with no substance. It was a real business, building real infrastructure, at the heart of what many believed would be the future of the internet. Investors piled in. Valuations exploded. Employees accumulated massive stock-based wealth. At its peak, the Canadian market reached a Shiller CAPE ratio above 60, far exceeding even the U.S. market at the height of the dot-com era.

In plain English, the Shiller CAPE ratio compares stock prices to the average earnings companies have generated over the past 10 years.

The idea is simple: earnings can be volatile year to year, so looking at a longer average gives a more stable picture of what investors are paying for corporate profits.

A very high CAPE ratio means investors are paying a lot for each dollar of earnings, which historically has been associated with lower future returns, unless earnings grow far more than expected.

In Nortel’s case, we saw how high levels of hype around a real company that led to over investment and speculation. And then reality set in. Nortel collapsed. The Canadian TSE 300 fell 43% between 2000 and 2002. Eventually, Nortel became worthless.

Nortel delivered us two important lessons:

• The Canadian market recovered relatively quickly, returning to its prior level by 2005 and going on to deliver strong long-term returns, outperforming the U.S. market over the following decade.
• Canadian value stocks did not crash, and in fact outperformed during the recovery.

Concentration hurt, but it didn’t permanently cripple a disciplined, diversified investor, who avoided over concentration in highly valued stock.

Technology Bubbles Are a Feature, Not a Bug

‍

‍

Periods of technological revolution often come with bubbles. This isn’t new. Railways in the 1840s, electricity, telecommunications, and the internet are all prime examples. And now, we’re in the age of artificial intelligence.

In each case, the pattern is remarkably consistent:

• Excitement about transformational potential
• Massive capital investment
• Soaring stock prices
• Eventual disappointment for investors who paid too much
• Long-term economic benefits from the infrastructure that remains

Bubbles feel irrational in hindsight, but they often accelerate real innovation. The excess fiber-optic cables laid in the 1990s were wasteful at the time, yet they formed the backbone of today’s internet. The overbuilding of railways enabled modern commerce. From an economic standpoint, these “productive bubbles” may be beneficial overall, even if they are painful for investors caught at the wrong time.

Is There an AI Bubble?

The honest answer is: nobody knows. That’s only knowable in hindsight.

What we do know is that AI-related companies have dominated market outcomes since the launch of ChatGPT in late 2022. According to a recent JP Morgan report, AI-linked firms have accounted for:

• 75% of S&P 500 returns
• 80% of earnings growth
• 90% of capital expenditure growth

This isn’t pure hype. Earnings have grown rapidly. These companies are building real infrastructure. The concern is not whether AI is useful, it clearly is. The real concern is whether current prices already assume a near-perfect future, leaving little room for disappointment.

Concentration vs. Valuation: What Actually Matters?

Market concentration and market valuation are often discussed together, but they are different issues.

• Market concentration tells us how much of the market’s value is held by a small number of stocks.
• Market valuation tells us how expensive those stocks are relative to their earnings.

Historically:

• The relationship between concentration and future returns is weak
• The relationship between valuation and future returns is meaningfully stronger

Highly concentrated markets around the world, including Canada, Switzerland, Taiwan, and others have still delivered solid long-term returns. Even within the U.S., periods of high concentration have not reliably predicted poor outcomes.

Valuations, however, tell a different story.

When investors pay more for future earnings, expected future returns tend to fall. That doesn’t mean a crash is imminent. Markets can stay expensive for long periods. But it does suggest that expecting recent returns to continue indefinitely is unrealistic.

Japan provides the most extreme example. In 1989, Japanese equities were viewed as unstoppable. Valuations were unprecedented. The market crashed, and in inflation-adjusted terms, it still has not recovered more than 35 years later.

So How Do We Protect Against a Bubble We Can’t Identify or Time Them?

The reality is we don’t know whether we are in a bubble, how long it might last, or what ultimately triggers a correction, if one occurs at all. We do know that trying to time the bubble or going all in on new technology like AI is rarely successful for investors. Instead, we believe the goal should be to build portfolios that don’t require us to need either.

That means:

• Broad global diversification, rather than relying heavily on one country or one theme
• Avoiding excessive concentration in any single sector, style, or group of stocks
• Tilting toward investment factors that have shown to perform well over long periods, such as:
  • Value (companies with lower prices relative to fundamentals)
  • Small-cap stocks
  • Highly profitable companies

At Endeavour, our clients are not overly concentrated in U.S. mega-cap technology or growth stocks. That doesn’t mean they avoid all downside risk, no portfolio can, but it does mean that the downside they may experience is likely reduced compared to investors who are heavily concentrated in a narrow segment of the market.

An evidence-based, diversified philosophy helps reduce the severity of market drawdowns and supports a faster recovery when markets rebound. We have always believed that for our clients, downside protection is a critical ingredient in long-term success, not because it eliminates risk, but because it makes risk survivable.

Proper diversification comes with one uncomfortable truth: you will always own something that’s underperforming. But it also ensures you will never own too much of it and that you will always own what comes next.

History shows that when leadership shifts, as it inevitably does, investors who stayed diversified, rebalanced, and resisted the urge to chase what was already popular were the ones who came out ahead.

The Bottom Line

High valuations and high concentration deserve attention, but not panic. Resisting the fear of missing out, especially when headlines and conversations are dominated by AI-driven returns, is an important part of long-term success.

We may look back and call this an AI bubble. Or we may not. Either way, the lessons remain the same as they have been for centuries:

• Markets reward patience, not prediction
• Diversification across many asset classes and styles is still the only free lunch
• Discipline matters most when emotions run high

This information has been prepared by Brandt Butt who is an Investment Advisor and Portfolio Manager for iA Private Wealth Inc. and does not necessarily reflect the opinion of iA Private Wealth. The information contained in this newsletter comes from sources we believe reliable, but we cannot guarantee its accuracy or reliability. The opinions expressed are based on an analysis and interpretation dating from the date of publication and are subject to change without notice. Furthermore, they do not constitute an offer or solicitation to buy or sell any of the securities mentioned. The information contained herein may not apply to all types of investors. The Investment Advisor and Portfolio Manager can open accounts only in the provinces in which they are registered.

iA Private Wealth Inc. is a member of the Canadian Investor Protection Fund and the Canadian Investment Regulatory Organization. iA Private Wealth is a trademark and a business name under which iA Private Wealth Inc. operates.