Michael Burry’s latest market moves reveal a nuanced bearish stance that defies simple crash prediction. In a recent Substack post, the legendary investor warned of a possible “major top” and a 1987-style collapse, but his disclosed portfolio adjustments suggest a more calculated positioning than an immediate doom call.
While the S&P 500 hits new records, Burry closed his bullish bet on Microsoft and eliminated his bearish Oracle position. However, he has maintained a short position against Palantir Technologies, even after the company delivered a blowout quarter with revenue surging 93% year-over-year to $1.94 billion and raised full-year guidance. US commercial revenue jumped 149%, and government sales grew 90%, crushing Wall Street estimates.
The contrast between his exits and his persistence with Palantir underscores a key argument: the short is less about denying artificial-intelligence demand and more about valuation skepticism. Morgan Stanley analysts called the quarter “a remarkable achievement,” and William Blair noted the “stellar performance defies concerns.” Yet Jefferies analysts, including Brent Thill, warned that Palantir’s valuation “leaves little room for a normalization in growth or execution slippage.” Thill views the risk-reward as unfavorable, as the shares require unusually durable expansion to support their revenue multiple.
Burry’s February analysis had questioned Palantir’s receivables, business model, and long-term competitive position against general-purpose AI systems. The latest results challenge some of those points but do not answer what investors should pay for the business. His willingness to change course when conditions shift — demonstrated by closing Microsoft and Oracle trades — makes the Palantir short a deliberate bet on overvaluation rather than a blanket market crash.
This strategic positioning, combined with a low VIX that may reflect complacency or genuine earnings strength, paints Burry not as a crash prophet but as a trader managing risk across an extended bearish timeline.
