DRAM ETF’s Losing Streak Tests the Memory Shortage Thesis

The Roundhill Memory ETF has delivered a sharp reminder that strong industry fundamentals do not guarantee a rising share price. DRAM was heading for its eighth consecutive down week on October 15 even as memory producers described tight supply and strong demand tied to artificial intelligence infrastructure.

The fund is a concentrated, actively managed portfolio of global memory and storage companies. Roundhill lists a 0.65% expense ratio and launched DRAM on April 2, 2026, so its trading history is short. That makes an “inception record” less meaningful than it would be for an established fund, but the selloff still captures a real debate about whether the memory cycle is approaching its peak.

Recent operating evidence remains strong. Micron reported fiscal fourth-quarter revenue of $54.23 billion, compared with $11.32 billion a year earlier, while gross margin reached 56%. The company also said HBM4E development was progressing toward volume production in calendar 2027. High-bandwidth memory is a critical component in AI accelerators, and long qualification cycles can support visibility for suppliers.

The risks are equally concrete. Memory remains a cyclical industry, and customers can reduce orders quickly once shortages ease or inventory builds. Reports about China’s CXMT moving into more advanced products raise another concern, although the timing and scale of competitive output remain uncertain. Higher interest rates could also slow data-center construction at the margin.

An ETF adds diversification across producers, but it does not remove industry concentration. DRAM’s returns will still depend heavily on memory pricing, capital spending discipline and expectations for AI server demand. Its 0.65% annual fee is also materially higher than broad semiconductor index funds.

The next useful signals are supplier contract duration, capital expenditure plans, HBM shipment growth and any evidence that conventional DRAM supply is catching up with demand. The recent decline does not disprove the shortage thesis. It does show that investors had already priced in a great deal of scarcity, leaving the fund vulnerable when sentiment shifted before the underlying data did.