Burry’s Micron Short Isn’t Anti-AI — It’s A Warning The Memory Cycle Is About To Snap

(SeaPRwire) –

By: Reginald Vance

Michael Burry’s latest short book is jolting every semiconductor trading desk. The investor who called the 2008 crash holds no fear of contrarian positions. His newest targets cut straight to the core of the AI trade. That trade has carried public tech valuations for three straight years. Most retail commentary frames his bets as a rejection of AI’s potential. That take misses the point entirely. Burry is not betting AI will turn out to be a passing fad. He is betting the market has priced in a decade of perfect execution. It has priced out every memory cycle pullback of the last 40 years. It assumes endless supply deficits that have never held for the sector. The real panic is not about imminent corporate collapse. It is about the gap between priced-in perfection and cyclical reality. That gap is wide enough to erase 30%+ of shareholder value in months. Micron alone has seen 34 such drawdowns since 1984.

Let’s track the hard, documented facts on both sides of the trade. Burry rolled out his disclosures in a string of Substack posts starting June 30. His initial short book covered Nvidia, Tesla, Applied Materials, Caterpillar. It also included the broad iShares Semiconductor ETF. He added a Micron position on July 1, entered at $1,051.87 per share.

He framed the dominant AI narrative as little more than mass addiction. He paired that take with a line from the 1989 Batman Joker. The quote warned “the end is nigh” for overstretched market positions. He posted two Bloomberg charts to back his thesis. The first showed AI chip stocks drastically outperforming cloud providers. Those same cloud providers foot the bill for most AI infrastructure buildout. The chips also outran every other category of AI beneficiary. The second chart placed the Philadelphia Semiconductor Index near the top of its 15-year valuation band. His harshest criticism landed directly on Micron. He flagged 34 separate drawdowns of 30% or more for the stock over 42 years. He called Micron more cyclical than any other peer in the sector. He calculated its median return on invested capital at just 4%. Median return on equity landed at 7%, figures he called frankly terrible. He noted the company destroys capital roughly one quarter out of every three. He added the stock trades further above its 200-day moving average than any point since 1984. He wrote off its AI-tied high-bandwidth memory chips as no special advantage. Those chips, he argued, are just another entry in a long line of incremental products. He pinned the stock’s recent run on FOMO, greater fool dynamics, and public commitment bias. Micron’s latest reported numbers paint a far rosier operational picture. For the quarter ending May 2026, revenue hit $41.5 billion, up 346% year over year. Gross margins reached 84.6%, up from 37.7% in the year-ago quarter. Net income landed at $28.2 billion, compared to $1.9 billion the prior year. Free cash flow hit $17.6 billion, a sharp turnaround from $1.7 billion 12 months prior. On the June 24 earnings call, Chief Business Officer Sumit Sadana shared guidance. He said customer demand for memory chips outstrips supply across nearly all categories. That supply crunch is set to last through 2028, per his comments. Micron stock has returned close to 1,000% over the past three years. It is up roughly 260% in 2026 alone. It trades near $976 per share as of the latest disclosures, around 22 times earnings. Management guided for roughly $50 billion in revenue for the coming quarter.

Cut through the noise of partisan hot takes on both sides. Burry is not shorting Micron over imminent business failure. His bet hinges entirely on timing and valuation stretch. He sees a stock that has run too far, too fast. It has outpaced even the rosiest plausible operational outcomes. He is betting the memory cycle will reverse, as it always has. No wave of AI demand has ever repealed memory’s core cyclical nature. Memory companies ramp capacity at full tilt when margins hit records. Those new production lines come online with multi-year lead times. They often hit the market just as initial demand surges cool. That dynamic sends prices and margins tumbling fast. Right now, Micron is posting its highest gross margins in company history. Its free cash flow is at record levels, hitting $17.6 billion last quarter. That cash is being earmarked for expanded production to meet demand. Management is guiding for multi-year supply deficits across product lines. Every prior memory cycle has featured identical guidance at the peak. Investors currently pay 22 times earnings for the stock. That multiple prices in years of unbroken, record-setting growth. It ignores Micron’s long track record of single-digit median returns on capital. It writes off 42 years of 30%+ drawdowns as irrelevant to the AI era. It treats high-bandwidth memory as a permanent, uncopyable moat. It dismisses the fact that every prior “revolutionary” memory product commoditized quickly. Capital is flooding into HBM production lines across every major memory vendor. All players are racing to capture a slice of AI-driven demand. That capacity buildout will not stop the second supply matches demand. It will overshoot, as it has in every single prior cycle. The endgame here is not a collapse of AI adoption. It is a sharp, painful valuation reset for semiconductor stocks. Those stocks have priced out every ounce of cyclical risk. The reset will hit the most cyclical names first. Micron sits at the very top of that list. The memory sector never delivers permanent, high-margin plateaus. It delivers booms. It delivers busts. Anyone buying Micron at these levels is betting 42 years of market history no longer applies.

Author bio: Reginald Vance, a venture partner specializing in semiconductor valuation and advanced materials, with 18 years tracking global memory market cycles.