• A record number of fund managers now say US stocks are overvalued. 
• They’ve tried rotating into better value. But recent earnings performance has forced them back into tech.
• That dichotomy creates a vulnerability. If investors don't like the valuations, at the first signs of trouble — for example if inflation reignites as tariffs fully kick in — they’re liable to sell.

Buying stuff you don’t believe in creates downside risk

We’re in the unusual situation where fund managers almost uniformly say US stocks are overvalued, yet everyone is piling in. This disconnect, as the S&P 500 Index sets repeated all-time highs, is the market’s biggest vulnerability going forward.

At the first sign of trouble — for example if tariffs do ultimately lift consumer prices and derail expectations for Federal Reserve interest-rate cuts -- they’re liable to sell. And it’s a risk the market may have to contend with in the next few months.

To get a sense of the latest positioning, look at the most recent Bank of America Global Fund Survey results.

A record 91% of institutional investors say US equities are overvalued. But don’t watch what fund managers say, watch what they do, because they’re fully invested. Cash levels are near the bottom of historical levels over the past quarter-century.

BofA says investors increased allocation to emerging markets, global stocks and utilities given the valuations in the US. However, going long the so-called Magnificent Seven — i.e., megacap US technology shares — is the most-crowded thematic trade, the survey shows.

Basically, institutional investors — cautious after President Donald Trump’s April 2 tariff announcement and slow to buy the dip amid recession fears — have lagged behind mom-and-pop investors who piled in when Trump delayed most of his tariffs in April. Big money managers are now playing catch-up, with BofA saying it just saw a near-record week of buying by institutional clients. That marked not only the 10th-largest flow into stocks in its data since 2008, but the flow into tech shares was in the 99th percentile, led by institutional buyers.

This is extreme. It creates huge risks too if people are buying stuff they consider to be overvalued.

Two things beyond chasing returns are driving this bull market. The first is the lack of an economic hit to the economy from tariffs and tariff uncertainty. The second is the strong earnings performance of megacap tech. In an economy still doing well, and given Big Tech’s weight in the indexes, investors are almost forced to  buy in order to keep up with benchmarks -- especially if they tried to hedge or rotate into areas they considered less expensive.

Take Meta Platforms Inc., for example. It’s not an obvious place to go to invest in artificial intelligence. Yet, of all the Magnificent Seven stocks, its earnings performance was the most surprising, with gains in ad spending offering a sign that tariffs have yet to hurt the US economy significantly.

The stock surged when the company said it would invest heavily in artificial intelligence.

Most of the other Magnificent Seven stocks that reported beat estimates. Tesla Inc. was an exception. Nvidia’s earnings, still ahead this month, should be stellar too.

Why would anyone sell?

It’s a bit discomforting though. Most of the tech behemoths are outperforming, but for how long? The economy is holding up, but can that last?

I see a third risk, what I’d call a “Black Swan’” risk that’s especially potent given the low conviction of investors in the value of what they’re buying.

Wikipedia lays out three elements The Black Swan author Nassim Nicholas Taleb has highlighted:  The risk has to be a surprise and have a big impact, and after the first recorded instance, it will be rationalized by hindsight, as if it actually could have been anticipated.

This is exactly how I would describe the tariff announcement on April 2. It certainly shocked investors. It pushed the S&P 500 Index to the brink of a bear market too. We now act like it wasn’t that big a deal. The Trump administration’s tariff policy is on par with the levels revealed on April 2. Remember, even China, which has entered another 90-day pause with the US on tariffs, still faces a 30% levy on goods imported into the US. Yet, stocks that swooned after April 2 on recession fears are at a record high.

This episode shows the vulnerability. Any unexpected policy action or event that could precipitate a US recession will shock investors.

BofA says 68% of investors predict a soft landing for the US, with 22% even calling for no landing, where growth doesn’t decelerate. Just 5% say we go into a recession here. That’s back to the thinking we had pre-April 2. And that risk is only heightened by the 91% of fund managers who say US equities are overvalued.

Let’s get back to the risks around megacap tech’s outperformance and the durability of the economy’s resilience. Last week, I expressed my stagflation fears. That scenario is a lot closer than we think. Though Trump mocks the firm for expecting inflation, Goldman Sachs says US consumers have only borne 22% of the tariff costs so far. And that’s simply because it takes time to negotiate lower import prices or raise consumer prices.  They say by the autumn consumers will absorb 67% of tariff costs and US companies less than 10%. So, the inflation hasn’t hit but it’s coming. And that’s bad for stocks.

Then there’s AI investment. The large language models behind Big Tech’s artificial intelligence are going to be game changers regarding internet search and productivity. But as a heavy AI user, I know that the output is very error-ridden and unlikely to yield enough revenue yet to justify the spend. Just ask your chatbot to list all the US states with the letter ‘R’ in the name and you’ll see what I mean. The output is never reliable without extensive verification of the details. It’s still only good mostly for drafts, background information and first cuts.

While AI will yield cost benefits to companies employing it, I don’t think these savings will be large enough to prevent an AI investment pause when growth turns down. My expectation is that we will see at least one AI investment cutback before large language models reach maturity — much as we saw after the first burst of internet euphoria. I remember buying this book in 1999 on residential broadband and looking forward to the future of streaming media that dominates media consumption today. The ideas were solid but it took almost two decades to get there. AI will be no different.

Buying overvalued stocks is working for investors for now. If Goldman is right about inflation, that won’t be the case a few months down the road.

Things on my radar

• Tariff-induced inflation hasn’t hit hard yet. It’s there as core CPI is the highest since January. But inflation numbers are low enough to induce traders to expect rate cuts.
• Without that impact, small businesses are becoming more optimistic.
• US budget deficits are really high.  That’s even after tariff revenue and a huge reason the economy has held up.
• The 15% cut the US government is taking out of chip sales to China is a dangerous precedent seeded by the lack of economic or market impact of tariffs.
• Leveraged loan pricing is extremely favorable to borrowers right now. It’s another sign of looser financial conditions (and a potential vulnerability too)

Like getting The Everything Risk? Check out these newsletters:

• Markets Daily for what’s moving in stocks, bonds, FX and commodities
• Odd Lots for Joe Weisenthal and Tracy Alloway’s newsletter on the newest market crazes
• Economics Daily for what the changing landscape means for policymakers, investors and you
• CFO Briefing for what finance leaders need to know

You have exclusive access to other subscriber-only newsletters. Explore all newsletters here to get most out of your Bloomberg subscription.