What’s Troubling Investors

In a change of trend, stocks closed lower in today’s session. Not sharply lower, you understand, but just enough to remind us that we can’t have a record high every day. This has been a very good year for investors. Broad stock market indices are now up somewhere between 8% and 12%. Even the bond market seems to be finding some footing with the Bloomberg Aggregate Bond Index up around 3%.

Investors are naturally getting a bit nervous about this rally, which continues to be highly concentrated in technology, and which may be a little extended. The S&P 500 Index’s P/E ratio, a common measure of value, is up around 23 vs. its long-term average range of 16-18. The Nasdaq’s P/E is north of 30. That’s a bit pricey. August and September haven’t yet lived up to their typical seasonal weakness. But we could be staring a correction in the face.

We’re hearing the word “froth” more often these days, referring to overconfident traders pushing up speculative asset prices quickly. And some segments of the stock market are increasingly untethered to reality. Think SPACs or profit-less IPO stocks. Certainly FOMO, rather than fundamentals, is driving some of that.

But a much more important issue is beginning to be felt as investors worry about the sustainability of the AI boom. How closely is it mirroring the run-up to the dotcom crash 25 years ago? Back then companies were racing to invest in the internet in order to attract eyeballs. Now, tech companies are racing to spend as much as they possibly can in order to build credible AI product offerings. They’ve got huge FOMO. OpenAI just announced a commitment to spend $850bil to build data centers. Oracle’s (ORCL) increased spending just pushed its cashflow into negative territory and the company will have to take on debt. Meta CEO Mark Zuckerberg just announced a $600bil plan to build out data centers though 2028. These are just a few examples.

Deepwater Asset Management’s Gene Munster says investors are a little skeptical that all this capital spending can be monetized. He senses “unbelief about where we’re ultimately going” with the AI boom. Over the long run, he thinks it will lead to much higher sales and profits for the “hyperscalers.” Nearly everyone seems to agree with him. But at the moment, he believes “the market is heavily discounting this type of [expected] growth” because it requires such a massive up-front investment. And if we’re honest, no one can really predict the future. There is a growing realization that we may have to put up with significant market volatility in the near-term even if AI ultimately ends up being the pot of gold at the end of the rainbow.

Calling for a correction seems like the sensible thing to do. After all, it would reset valuations and knock down some of the runaway speculative fervor among traders. The trouble is, we see two reasons why a correction is unlikely before year-end. First, if we learned anything from second quarter earnings announcements it’s that profit expectations are rising, not faltering, for the balance of 2025 and into 2026. And the sector that looks most expensive on a P/E ratio basis—technology—is expected to see the highest profit growth next year (+20%). So the valuation problem clearly hasn’t reached dotcom bust levels.

Second, the economy is holding up well despite tariffs. In fact, it looks as though economists & Federal Reserve officials were completely wrong to call for much slower economic growth this year. Since “Liberation Day,” they’ve been projecting a mere 1.5% Gross Domestic Product (GDP) growth for 2025. But after a wonky first quarter, growth has accelerated far beyond that figure. Second quarter GDP growth was just revised up to a very healthy 2.5% from the initial estimate of 1.5%. The boost was driven by better than expected consumer spending and business investment. In addition, the Federal Reserve Bank of Atlanta’s GDP tracker projects third quarter growth around 3%.

The market has had an extraordinary run, and while the rally has left valuations stretched and speculative behavior in pockets of the market, it isn’t yet showing the hallmarks of a full-blown bubble. Technology spending on AI is sparking déjà vu for some, but unlike the dotcom era, profits and demand remain pretty strong across most parts of the economy. Still, investors should brace for volatility in 2026, particularly as AI spending stories evolve and sentiment swings between enthusiasm and skepticism.