Michael Hartnett has been Wall Street’s most prominent bear for so long that some clients stopped reading his notes. In his latest, the Bank of America strategist told them to keep selling. His advice is simple: keep taking profits until the Federal Reserve’s July 29 policy meeting, when chair Kevin Warsh is expected to turn hawkish, and hold off on buying again until tighter financial conditions have peaked.
Hartnett’s reasoning rests on an extreme divergence that he says cannot last. On one side, technology funds just posted the largest single-week inflow on record, a stampede he describes as end-of-cycle greed. On the other, Bank of America’s own bull-bear indicator has climbed to 8.8, a level that has historically triggered a sell signal. Record buying and maximum risk at the same moment, in his telling, is a setup that tends to end badly.
The historical comparison he reaches for is 1994. In years when inflation ran above 4% and the macro environment looked like today’s, the Fed’s tightening delivered sharp drawdowns in stocks, and the market that year needed a brutal bond-market correction before it could resume its climb. Hartnett argues the current cycle rhymes with that one: an economy that looks strong, a central bank that is behind the curve, and investors positioned for the trend to continue forever.
The Fed is the hinge. Warsh, who took over the central bank with inflation still above target, has been the market’s main source of anxiety, and Hartnett’s calendar treats the July meeting as the moment the uncertainty resolves. If Warsh delivers the hawkish surprise the market fears, financial conditions tighten, the rate-sensitive trades unwind and the record tech inflows become the top that everyone will blame someone else for missing. If he does not, the bear case loses its anchor.
The note has landed at a delicate moment for stocks. The tech complex has been the engine of the entire rally, and its valuation now assumes that AI-driven earnings growth continues without interruption. A pullback in that complex would be a market-wide event, which is why Hartnett’s advice to trim applies not to a handful of sectors but to portfolios broadly.
Skeptics have heard this before. Hartnett has called for caution through much of the AI rally, and the market has repeatedly proved him early, rallying past every warning he has issued. His defenders note that he was early in the other direction during past cycles, and that being early is not the same as being wrong, but the strategy of selling into strength has cost investors who followed him during the past year’s gains.
The practical question for investors is what “trimming” means. Hartnett’s notes typically favor raising cash, taking profits in the biggest winners and shifting toward defensive sectors and bonds that benefit from a hawkish Fed. The July 29 meeting gives the advice a date, which makes it easier for clients to follow: sell into the current strength, hold the proceeds, and wait for the Fed to show its hand.
The counterargument is the earnings machine. Companies in the S&P 500 have been reporting profit growth that continues to beat expectations, and bears have been wrong about the economy for two years straight. If the Fed holds steady and earnings keep rising, the record inflows will look rational and the bull-bear indicator will have been a false alarm, as it has been before.
The bull-bear indicator deserves a closer look, because it is the mechanical heart of Hartnett’s argument. The gauge blends positioning, flows and sentiment surveys into a single contrarian reading, and readings at or above 8 have historically preceded short-term pullbacks in stocks. A print of 8.8, in the top of the historical range, is the kind of level that has marked local tops before, and Hartnett’s point is that record inflows into technology are pushing the indicator further into warning territory rather than out of it.
His prescription for the weeks ahead is unglamorous. The notes have favored raising cash, trimming the biggest winners and rotating into defensive sectors and bonds that would benefit from a hawkish Fed, and away from the AI trade that has carried the market. Clients who follow the advice will underperform if the rally continues, a cost Hartnett has paid before; clients who ignore it will face the full drawdown if the July meeting delivers the tightening he expects.
Either way, Hartnett has given the market a date to focus on. The weeks before July 29 will be filled with position squaring, and the meeting itself will be one of the most-watched policy events in years. If Warsh turns hawkish, the strategist’s note will be quoted for months as the call that got it right. If he does not, it will join the long shelf of cautious memos that a bull market left behind.
