AI has rewarded investors. It may now pose their greatest risk.

Near the midpoint of the year, stocks and bonds both report good returns. But the global stock market has become highly concentrated, our columnist says. (Peter and Maria Hoey/The New York Times)

By JEFF SOMMER

Caterpillar is known for its yellow construction machinery, not for artificial intelligence. But it also makes power-generation turbines, and as AI data centers consume increasingly large amounts of electricity for their operations, Caterpillar’s stock has soared.

Like scores of other companies, it has become a backdoor bet on AI. The Caterpillar stock surge is just one of myriad ways in which AI has been fueling the stock market, propelling the retirement accounts of millions of investors to their best quarterly returns in years.

Huge sums being funneled into AI infrastructure have bolstered returns in a vast range of stocks around the world. Capturing those gaudy AI returns while diversifying sufficiently to minimize the pain of potentially severe downturns is becoming a significant challenge. If the enormous expectations for AI aren’t met, those remarkable portfolio returns could suddenly vanish.

AI is still at an early stage of development, but enthusiasm for the technology is already a dominant force in the markets, and not just in the United States. In much of the world, “AI is the only story in town,” said Indrani De, head of global investment research for stock market index provider FTSE Russell. Building a broader portfolio is a major challenge.

US stock funds

Mutual funds and exchange-traded funds — as well as the workplace-based retirement plan trusts based on them — are how most people in the United States invest. Morningstar, a financial services company, has been providing quarterly fund returns to The New York Times for decades. The latest report provides a window on how the stock and bond markets — and the AI trade — have affected real people so far.

For the three months ending June 30, the average fund investor in domestic stocks gained 14.8%. That was the best quarterly return since 2020, when the stock market rebounded after a fierce decline set off by the COVID-19 pandemic and recession, Morningstar data shows. In the second quarter of that year, domestic equity funds gained 22.4%.

This time around, the market staged a second-quarter recovery as well — from a market decline set off when the United States and Israel attacked Iran on Feb. 28. The month of March was particularly fraught in the markets, with the S&P 500 falling 5.1% and the price of Brent crude oil soaring 30%, according to FactSet. A series of temporary ceasefires set off a strong rally, though trading has been choppy in recent weeks during periodic bursts of fighting around the Persian Gulf.

Viewed as a whole, the U.S. stock market has experienced a rolling AI rally, shifting focus on different groups of stocks but retaining its power. First to ascend in the market were Big Tech stocks in the so-called Magnificent Seven, comprising Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla. As the rally spread, traders bid up stocks that designed, produced or were in other ways connected to semiconductors — the silicon chips needed to run AI. A variety of other shares gained, too, including equipment and generator companies like Caterpillar, utility companies and even fossil fuel companies like Exxon Mobil, which help to sate the voracious appetite of AI data centers for electric power.

In fact, practically wherever you look, AI is a factor — though not always a favorable one. Software stocks, like Oracle, Salesforce and even Microsoft (also a major developer of AI data centers), were pummeled on fears that AI agents may make many of their profit centers irrelevant. That interpretation of the implications of AI is open to question. The point is that for better and for worse, you can’t easily avoid AI in the stock market.

A global view

Take international stock funds. They returned 12.8% in the three months through June, according to Morningstar. Over 12 months, international stock funds did even better, with a 26.8% return — 3.6 percentage points ahead of domestic stock funds.

You might think that this is the result of diversification away from AI, but you would be wrong. Instead, it reflects the global strength of the AI trade. For example, if you hold an index fund, like the Vanguard Total International Stock Fund (as I do, in a workplace retirement account), you will find that its top five holdings are all AI-adjacent. These include Taiwan Semiconductor Manufacturing, Samsung Electronics, SK Hynix, ASML and Tencent.

Global stock markets have tended to move like synchronized swimmers lately, with their higher correlations connected to the Middle East war — and, increasingly, by the cascading effects of the AI trade.

Fixed income

Bond fund returns were not nearly as spectacular as those generated by stocks. That’s the proper order of things in traditional investing. Bonds are generally far less riskier than stocks and therefore produce more modest rewards. On the other hand, they often, but not always, provide a buffer when the stock market declines.

One major exception to the role of bonds as a solid form of portfolio insurance tends to crop up in periods of high inflation. That was the case in 2022. The problem is that bond yields, or interest rates, move in the opposite direction of bond prices. In 2022, interest rates rose, bond prices fell, and bond fund investors lost money. That’s worth keeping in mind now.

Inflation has been running higher than most central banks — and many consumers — find comfortable, and it’s quite possible that interest rates will be rising in the months ahead. Bond buyers who intend to stick with their holdings for decades should be fine, but it still makes sense to be careful. If you think that you may need to cash in your bonds, or bond funds, over the next few years, you might be better off with short-term holdings, like Treasury bills, bank certificates of deposits or money market funds.

For the quarter through June, most bond funds shrugged off a modest rise in interest rates. Taxable domestic bond funds rose 1.8%. And they gained 5.1% over 12 months — compared with an average annualized return of 3.2% for a major investment-grade bond benchmark, the Bloomberg U.S. Aggregate Bond Index.

Target-date 2035 retirement funds — aimed at those planning to leave the workplace in that year — returned 9.4% for the quarter and 16.5% over 12 months. Retirement income funds, aimed at those already in retirement and presumably holding a high proportion of bonds, returned 5.1% for the quarter and 10% for the year.

If you can, diversify

Those are all high returns, historically speaking. Whether they can be sustained is a big question. There are reasons to doubt it.

On the positive side, corporate earnings are sizzling. Unlike in the dot-com era, when the market assigned lofty values to loss-making firms, many of the tech giants in the current market are generating enormous profits.

Nonetheless, stock prices are already high, and signs of irrational exuberance abound. The SpaceX initial public offering, for example, commanded a $2 trillion-plus market value that can scarcely be justified by traditional earnings metrics, as I have noted. Geopolitical stresses, natural disasters and new leadership at the Federal Reserve amid a surge in inflation all increase the risk of market declines.

Focusing mainly on the AI-driven rise in stock prices, Bank of America’s stock strategists, led by Savita Subramanian, project a decline of about 4% for the S&P 500 for the remainder of the year. “It’s a matter of valuations,” she said on Monday. That would be a tiny decline, given the gains of recent years, but it’s a decidedly bearish view for a major firm on Wall Street, where bullish takes on the stock market prevail.

Betting against AI in the stock market has been costly in recent years, and momentum still seems to be strong. Even in an environment of rich share prices, abandoning the stock market because it’s risen too high would be a difficult move to make. It’s not one that I’d advocate because it’s impossible to predict short-term market movements accurately and because I think it’s wise to stick with the stock market for the long haul.

Nonetheless, I’m spreading my bets as widely as I can, investing in global markets that are, admittedly, too closely tied to the U.S. market for comfort, and holding plenty of cash and bonds, too.

AI has driven global markets relentlessly higher but there will inevitably be setbacks. The eventual faltering of the AI trade may now be the market’s greatest risk.