Amid epic returns for chip stocks, Jeremy Grantham says the market risks a 70% retrenchment.
Mr. Grantham, 87, longtime head of Boston’s GMO Capital, is notorious for his “permabear” warnings. He’s been right, depending on how you define “right.”
John Maynard Keynes said being right and being early is the same as being wrong. Mr. Grantham was warning investors of imminent risk in 2021. Here we are, much higher than then.
Mr. Grantham told CNBC last Friday that depending on how one counts, “This is the most expensive market in American history.” He channeled Warren Buffett, who said that when the stock market is twice as large as GDP, investors may be playing with fire.
It’s around 2.4 times now.
In normal distribution, two standard deviations capture more than 95% of all events. When market returns lie outside that range, the likelihood of a mean-reverting event rises, says Mr. Grantham. A two-sigma risk signal.
The CFA Institute notes that over 95% of market returns since 1950 have occurred within one standard deviation of the mean.
Mr. Grantham says we’re now more than two standard deviations from the norm, which is roughly 0.03% daily returns over the data set. (I ran the data on SPY, which has half the volatility of stocks, and daily returns and the close vs midpoint have both averaged – strangely – 0.1% in 2026 YTD – a whopping divergence from the mean.)
To revert, the market would have to give up about 70%.
Mr. Grantham carries cachet despite a dubious record on warnings not just because he’s a distinguished British gentlemen possessing gravitas. As I said, he’s been right.
He says there are 26 incidences of two sigma departures from norms in market history and in 100% of them there’s been mean-reversion.
So he thinks AI is a bubble? As with railroads and the internet, the expectation of societal benefit lived up to the hype in the final analysis despite investment markets that first imploded, he says.
He puts AI in that big category of legitimate, life-changing development where current expectations may be optimistic.
Whatever the case, investors and public companies alike have to chart courses through markets that are demonstrably more highly valued than those of the past.
Now to me, high multiples reflect the only deliberate period of monetary inflation in human history, accompanied by the predominance of Passive investment, which is price-insensitive and simply buys products called stocks.
Understanding risk in a product market requires looking not at how things are valued but at the capacity for consumption.
It’s a bit of horse and cart. Under an inflationary regime, prices rise. Rising prices obscure the view of demand and supply. Massive sums are being spent, generating gigantic profits for makers of the chips that power what we’ll all come to hate at some future point: “Compute.”
I don’t hate computing power. I hate that we’re all calling it “compute.” It sounds like something that rightly should blow up.
Anyway. We’re assured by observers that this time is not like the internet bubble because real companies are driving real profits.
Yeah, well. So did the sellers of tulip bulbs, which were mispriced. Because somebodies make money hand over fist selling legitimate products to real consumers doesn’t mean the products are priced correctly.
Did you know Micron used to be a value stock? Chips are historically cyclical. How do we know it’s not still true?
As Mr. Grantham notes, the trouble isn’t in recognizing the problem. It’s knowing when the problem that you recognize manifests its consequences. He says, “Half the time, you’re making up for losses.”
You could pull your money from stocks, investors. My wife and business partner Karen probably would favor it. Mr. Granthan does too, advocating a zero percent allocation to US equities, and a formidable one to international stocks, precious metals, bonds.
I cut my exposure to US equities to 25%. My own choice.
And public companies, we rely on this market for raising capital, benchmarking valuation, liquidity for holders. Do we abandon it? Well, probably not.
Here’s what I know: The stock market depends for valuation on flows. Not fundamentals. So long as money pours in at a rate of an SPCX IPO per month, stocks rise. Unlike Active managers, which generally must sell to buy, Passives just buy.
When that stops, stocks will drop.
You’d think it would be easy to see. Nay. Between Jun 5-22, eight of the eleven trading days had net selling in the S&P 500. Stocks were basically unchanged. In the last week, investment flows in the S&P 500 fell 4.2% during index-rebalances. Stocks rose.
The other thing you watch for is a breakdown in the “arbitrage mechanism” that prices ETFs versus underlying assets like stocks. Why? Because ETFs dominate flows.
The signal is rising volatility. Volatility in stocks is now over 3% for most of 2026. Last week, SPY and the index it ostensibly tracks diverged sharply.
Slowing flows, rising volatility. Mr. Grantham would likely say it’s a signal for the prudent. Who knows? The second half of 2026 starts today. We’ll find out.
