BlockBeats News, May 30th, BCA Research's Chief Global Strategist Peter Berezin analyzed that the current AI bubble is primarily a profit bubble rather than a traditional valuation bubble. Unlike many past stock market bubbles characterized by a rapid increase in price-to-earnings (PE) ratio, the current AI-related sectors, especially the semiconductor sector, have relatively reasonable valuations, but profit expectations are overly optimistic and unsustainable. Similar situations have occurred in history with real estate developers and banks before the 2008 financial crisis: they had seemingly low PEs, but actually relied on unsustainable profit surges. Once the profits cannot be realized, the bubble bursts. Peter pointed out that semiconductor sales have been growing parabolically, and currently, AI demand indicators do not yet show that the bubble is about to burst, but all bubbles will eventually come to an end.

Peter emphasized that investors should not overly rely on Wall Street analysts' profit forecasts because stock prices often plummet significantly before profit expectations begin to decline. In past cycles, stock prices often peaked months before the forward EPS started to decline, and waiting until the EPS is lowered to sell will result in severe losses. The key to current investment is to monitor changes in AI demand indicators in advance.