Best Protection Against an AI Bubble? Index Funds

“A sentiment common among Wall Street professionals is that the stock market is dangerously concentrated.

 

The market value of the 10 largest stocks represents almost 40% of the total value of the S&P 500 index.

 

Market indexes, the argument goes, are insufficiently diverse, exposing investors to dangerous risk and making index-fund investing an unjustifiable investment strategy.

 

The stock market is highly concentrated. But broad low-cost indexing remains the optimal strategy for investors, and staying invested in equities remains the best vehicle for earning generous risk-adjusted returns.

 

Worries about market concentration have reached a fever pitch. Well-known professional investors have issued dire warnings. Jeremy Grantham writes that we're experiencing a speculative bubble and that the narrow stock-market indexes will produce unique underperformance. Howard Marks opines that index investors are, unwisely, simply pouring money into the largest stocks regardless of value. Bill Gross declares that indexing isn't buying the market -- it is owning just a few stocks, all betting on a single AI theme. Investment advisory firms like JPMorgan and Goldman Sachs are telling investors that the narrow market makes actual stock picking essential.

 

But is it really unprecedented for a handful of stocks to dominate the market and produce the highest returns? Research by Hendrik Bessembinder indicates that almost all the wealth creation from equity investing was produced by a tiny fraction of individual stocks. For the 100 years since 1926, the U.S. stock market has returned about 10% annually. But only 3.4% of the stocks were responsible for all the wealth creation. The rest of the stocks combined didn't produce net gains that exceeded the amount that could be earned from investing in 30-day U.S. Treasury bills. And more than half the stocks in the market lost money. The generous average return from the market came from a tiny handful of stocks. An indexing strategy works because it captures these rare winners.

 

Those investment advisers who worry about today's market structure would argue that the current unique concentration requires a different approach. If the concentration reflects an expanding market bubble spawned by overoptimism about AI, then continuing to invest in index funds is too risky. They point to experience from the early 2000s, when the stock market was nearly as concentrated with internet-related stocks. Stock prices declined sharply until mid-2002, with index funds losing almost half their value. Had investors been able to time the decline correctly and shift to "value" stocks, they would have limited their losses and beat the index. The problem is that no one can time the market perfectly, and index funds overperformed the average actively managed fund even during the market decline.

 

Standard & Poor's began publishing its S&P Indices Versus Active reports, comparing market indexes with the performance of active portfolio managers, in mid-2002, at the bottom of the dot-com bust. It found that a majority of actively managed portfolios underperformed the market indexes even during a crushing bear market. Fifty-nine percent of active funds underperformed the S&P 500 in 2001, and 72% did in 2002. From mid-1999 through June 2002 (the period that began when the bubble was reaching its peak to the bottom of the subsequent decline), 62% of active funds failed to beat the market. If the calculations are carried out over the next 10 or 20 years, more than 90% of actively managed funds will have failed to beat the market. Indexing gives you superior performance in all markets.

 

It is certainly correct that today's stock market presents substantial risks. It is possible, perhaps even likely, that we are experiencing an AI bubble. Stock markets always go to extremes. Stock price bubbles have expanded during all the major technological revolutions. The railroad boom in stock prices was followed by rail stocks' falling by three-quarters of their value during the panic of 1873. The electrification revolution ended in the crash of 1929. The dot-com bubble led to a crushing bear market ending in 2002 with some internet-related stocks losing 90% of their value.

 

But we don't know if we are in the top of the third inning or the bottom of the ninth of the AI revolution. And we know neither when the stock market will fall nor what the best assets are to buy if we remain invested. Remember that Alan Greenspan used the term "irrational exuberance" to describe the Internet bubble in December 1996. The market continued to rally and peaked in March 2000.

 

While staying the course with your investment program is likely the best strategy, you can take steps to lessen the risks inherent in today's "narrow" market.

 

 First, reduce the concentration of AI-related stocks in your portfolio by using the broadest stock index possible.

 

Use a mutual fund or ETF indexed to the total stock market, including small- and midcap stocks as well as value stocks. Consider adding foreign stocks from developed and emerging markets. This won't eliminate instability, but it will reduce concentration and increase diversification.

 

Having a substantial share of your portfolio in broadly diversified low-cost index funds remains the most reliable method for building wealth in the stock market.

 

---

 

Mr. Malkiel is author of "A Random Walk Down Wall Street."” [1]

 

1. Best Protection Against an AI Bubble? Index Funds. Malkiel, Burton G.  Wall Street Journal, Eastern edition; New York, N.Y.. 23 Mar 2026: A17.