Micron’s Trading Volume Surpasses Nvidia as AI Memory Supply Squeeze Meets a Geopolitical Correction

Micron overtakes Nvidia in trading volume, then drops 9% on China AI fears. Yet fundamentals shine: 196% revenue growth, HBM capacity locked through 2026, and forward PE under 8x.

For nine straight trading sessions, investors poured more money into Micron Technology than into Nvidia – a stark signal that the market’s focus is shifting from processors to the memory chips that feed artificial intelligence systems. During that stretch, roughly $47 billion worth of Micron shares changed hands, compared with $34 billion for Nvidia. The smaller memory specialist first topped its rival in daily volume on March 19 and held the lead for six of the nine sessions.

Yet the week that began with a fresh 52-week peak of €685.40 ended in a rout. By Friday’s close in Frankfurt, the stock had surrendered about 9% of that high to land at €624.00, dropping more than 8% on the day alone. The trigger was a wave of selling across the semiconductor space after reports surfaced that Chinese technology companies had failed to place anticipated orders for advanced AI chips. For a stock that has more than doubled year-to-date and risen over 630% in the past twelve months, such a sharp pullback was almost predictable: the relative strength index printed 77, signalling technical overbought conditions, while Friday’s elevated volume pointed squarely to institutional profit-taking.

Behind the headline volatility, Micron’s operating performance tells a fundamentally different story. In the second fiscal quarter the company posted revenue of $23.86 billion, a 196% jump from a year earlier, and earnings per share of $12.07 – an eye-popping 756% increase. Management has guided for third-quarter revenue of roughly $33.5 billion and a non-GAAP gross margin of about 81%, which would push earnings per share to $18.90, representing growth of over 1,000%. Perhaps more telling, Micron’s entire high-bandwidth memory (HBM) production capacity for the remainder of 2026 is already locked in under binding contracts, underscoring the structural shortage that has turned cloud providers into long-term contract buyers rather than spot-market traders.

The supply picture only tightens from here. SK Group, the parent company of rival HBM maker SK Hynix, has warned that a wafer shortage could persist until 2030, and meaningful new DRAM capacity is not expected to hit the market before late 2027. The AI data-centre market that drives Micron’s growth is now projected to reach $1.7 trillion by 2030, up from a prior estimate of $1.4 trillion – a revision that prompted Bank of America to double its price target on the stock from $500 to $950 while maintaining a buy rating.

Despite the narrative of runaway demand, Micron’s valuation remains curiously modest. The forward price-to-earnings multiple stands below eight times, an unusual level for a company generating a net margin of 41%. That disconnect may partly reflect the geopolitical risk baked into the stock: any China-related headline can whip the share price in either direction, as Friday’s selloff demonstrated. Yet the long-term infrastructure buildout that requires Micron’s HBM is global in scale, and short-term jitters have done nothing to alter the fundamental gap between supply and demand.

The next major inflection point arrives on May 20, when Micron’s management speaks at the J.P. Morgan Technology Conference. Investors will be listening for confirmation of the aggressive Q3 guidance and any update on the supply outlook. Should the company reaffirm its trajectory, the institutional profit-taking of the past week is likely to be remembered as a brief pause in a much longer climb.