The Memory Collapse JUST Started | Here's WHY.

Samsung stock cratered 8.9% in Korean trading after missing earnings and, more importantly, disappointing on buyback expectations by favoring dividends over the buybacks markets had priced in. Kevin argues this is an expectations-versus-reality issue rather than a fundamental collapse in the memory shortage thesis.

Kevin frames Micron within a broader memory 'shortage cycle' that Nomura research suggests could persist through 2027 and beyond due to structural high-bandwidth-memory constraints, but he cautions that markets already price in an eventual decline in memory prices and that the key variable is how long the 'plateau' lasts. He flags downside risk if the catalysts needed to extend the cycle (robotics, AGI, healthcare AI) arrive later than markets expect.

SanDisk fell about 11% as part of the broader memory sector selloff triggered by Samsung's disappointing buyback guidance, even though Kevin notes SanDisk is technically more of a storage play than a pure memory chipmaker. He treats it as swept up in sector-wide sentiment rather than having a distinct thesis of its own in this segment.

Kevin notes Nvidia recently raised server chip prices about 15%, which markets read as bullish for margins, but attributes much of that pricing power to memory cost inflation rather than Nvidia-specific strength. He floats the idea that Nvidia could actually benefit if memory prices eventually fall, since it could keep GPU prices elevated while lowering its own cost of goods sold, but frames this as speculative.

Kevin says the Nasdaq-100 is being dragged down alongside the memory stock selloff, compounded by other macro factors like Treasury Secretary credibility concerns, Jackson Hole, and trade war tensions with Canada. He references a prior call for further downside but the specific numeric levels cited in the transcript are unclear and inconsistent.