JPMorgan Asset Management chair could hardly have timed his semiconductors' commentary any better

One investor correctly called the top of the semiconductor stocks trade.

On June 23, Michael Cembalest rang the alarm on semiconductor stocks. June 22 turned out to be the all-time high on both the iShares semiconductor exchange-traded fund and SK Hynix, perhaps the sector bellwether after its meteoric rally of the previous year.

Michael Cembalest examines what he got right and what he got wrong in his latest note.

Within five weeks, the SOXX ETF SOXX tumbled almost 30%, SK Hynix (KR:000660) more than halved and other chip companies like Taiwan Semiconductor Manufacturing Company (TSM) and Micron Technology (MU) dropped sharply in value.

How did Cembalest, the chairman of market and investment strategy at JPMorgan Asset Management, manage to time the peak of the memory trade so presciently? In his "Eye on the Market" note published at the time, Cembalest highlighted how "the technicals of soaring semiconductor stocks had breached levels not seen since the dotcom boom," and also the crowded positioning of hedge funds in chip stocks and what investors were calling 'the picks and shovels trade' of the AI investment bonanza.

Cembalest was also alarmed by the rapidly mounting margin debt accumulated by retail traders that was turbocharging the rally, and the market risks from the multiplication of leveraged ETFs on the sector.

MSCI Korea and semiconductors plus outstanding stock margin loans

This was far from the only call Cembalest got right, as he reviewed both his hits and misses in a new research note published on Tuesday. Among other predictions he reviewed that proved prophetic in 2026 were his bets on a buying opportunity in healthcare XLV, software stocks being oversold in the aftermath of the Citrini research piece back in March , a reassuring call that the dollar would remain stable when many expected its 2025 decline to be replicated and a wager that the Fed's next move would be to hike and not ease interest rates.

Cembalest also felt the headwinds coming for hyperscaler stocks in a note titled "Smothering Heights" back on Jan. 1. Declining free cash flow, a reliance on debt to fund capex and questions about returns on capital invested persuaded him to expect underperformance and that proved accurate.

S&P 500 vs Hyperscalers

Cembalest did get some forecasts wrong, however, and did miss some developments. He didn't see the Asian surge chalked up by Taiwan, Korea and Japan, while he forecast China AI stocks to rally and they have underperformed American rivals.

He failed to foresee the return to form of small-cap stocks in the U.S. IWM while the Western critical mineral stocks he thought would rally have gone nowhere. The outperformance of Asian and emerging markets fixed-income assets versus U.S. Treasurys and investment-grade bonds also wrong-footed him.

The success or failure of one remaining recommendation has yet to be determined. On the first day of the year Cembalest encouraged investors to reduce the overweight on U.S. equities versus the rest of the world, motivated primarily by the latter's heavily discounted valuations. So far in 2026 these two asset classes are in a dead heat.

"The jury is out," notes Cembalest.

-Jules Rimmer