The most accurate analyst speaks out—this is the most dangerous moment

—It’s time to move away from the most crowded areas to places that can better withstand volatility.

Michael Hartnett, the so-called “Wall Street’s most accurate analyst,” has released another report this week. Over the past year, people who traded based on his advice basically made a fortune. Last December, he issued a “sell signal” exclusive to Bank of America; after two months of U.S. stock market turbulence, the market then crashed. On March 29, he loudly proclaimed that the “sell signal is over,” and that very day was the S&P 500’s phase-cycle bottom.

This week’s report is a warning: the environment is no longer one for “buying the dip.”

1）The biggest problem in the market is that there are no short sellers. Bank of America’s proprietary bull-bear indicator has risen to a historical high of 9.6. The market is crowded, but investors have not truly pulled back. Over the past three weeks, technology stocks have attracted nearly $50B in inflows, setting a historical record, and even though semiconductor prices have already pulled back noticeably, capital has continued to flow into semiconductor ETFs.

2）The most dangerous thing isn’t AI—it’s that AI has become a “consensus.” If everyone believes that AI capital expenditures won’t decline, then the market has already priced this positive development entirely into the stock price. Even if AI investment only slows slightly, the stock market could still see sharp pullbacks.

3）South Korea could be the first global signal. South Korea is one of the markets globally that loves trading AI, chips, and leverage. After the Bank of Korea unexpectedly raised interest rates, the Korean stock market underwent a major adjustment, and many highly leveraged AI trades began cooling off rapidly. The author believes that if even the most reckless capital starts withdrawing, it indicates that global AI trading sentiment may be changing. Of course, South Korea can’t represent the entire U.S. market, but it acts like an early warning system.

4）Hartnett is watching three trades:

· Buy long-term Treasuries: Everyone currently believes the economy is doing well. If the economy cools even slightly in the future, long-term Treasuries are actually the most likely to rise.

· Buy the U.S. dollar: If the market suddenly starts re-pricing rate hikes, the dollar could become the winner again.

· Stay bullish on Mag7, but be cautious on chips: Many people think chips and Mag7 are the same thing, but they aren’t. Hartnett believes that if AI capital expenditures truly slow in the future, the first things to be hurt will be chips, servers, industrial manufacturing, and banks. Meanwhile, platform companies with stable cash flows—like Microsoft, Apple, and Meta—may not be the worst off.

5）In the author’s final hypothesis—if AI giants really cut capital expenditures in the future, the market would of course fall. But the danger is that even if AI giants cut capex, if Mag7 doesn’t continue to rise, that would indicate that money has started a broad exit. At that point, the downside may spread beyond chips—potentially to banks, industrial stocks, cyclicals, and even the entire U.S. stock market.

When everyone believes “the economy won’t slump, the Fed won’t hike rates, and AI won’t slow,” the market fears not bad news, but the good story no longer being perfect.

Risk warning: The above content is for general information purposes only for readers.
The above content is provided solely for general informational purposes for readers.