Memory Semiconductors July 2026: The 89% Ceiling on Earnings Revisions
Morgan Stanley’s latest memory note doesn’t argue the trade is over — it argues the easy part is. $MU $SNDK $WDC $STX have run on a combination of pricing surprise, inventory drawdown, and relentless upward earnings revisions. The firm now says all three of those engines are running out of runway at the same time, even as it keeps the structural bull case for 2027 fully intact.
The Thesis
The setup Morgan Stanley describes is “peak rate of change,” not “peak cycle.” DRAM price growth is still positive but decelerating year-over-year, the inventory improvement curve has flattened rather than continuing to steepen, and earnings estimate revision breadth for DRAM has reportedly reached roughly 89% — a level that leaves almost no room for further upgrades to surprise the market. None of that requires demand to actually weaken. It just means the stocks have been pricing in acceleration, and acceleration is the one thing that’s statistically hard to repeat once revision breadth is already near-universal. The next real catalyst, in MS’s framing, is Q2 2026 hyperscaler earnings and capex commentary — a beat-and-raise on cloud capex extends the story, a cautious tone on AI spend, token monetization, or “chipflation” passthrough gives the correction thesis its opening.
Capital Structure of the Trade
This isn’t a single-name story, it’s a capital allocation story across the sector. Hyperscalers are directing spend toward AI infrastructure at a pace that’s pulled DRAM and NAND capacity into HBM and server-grade product, which is precisely why legacy and consumer memory has seen shortages emerge as a byproduct of AI demand rather than organic recovery. Morgan Stanley’s preference ordering reflects where that capital is actually flowing: DRAM over NAND, because DRAM’s long-term agreement terms are more favorable and customers are demonstrably willing to pay up to secure supply, versus NAND where consumer-market resistance and rising module-maker inventory are starting to show up as order cancellations.
Margin Dynamics
The margin story is where the “structural divergence” language matters most. Original manufacturers — Micron, SanDisk, Western Digital on the flash side — are being framed as structurally advantaged over module makers, Seagate included in the broader storage complex, because they control supply and pricing rather than just repackaging it. Module makers have historically shown a repeatable pattern: strong earnings elasticity on the way up as they sell through low-cost inventory bought at the bottom, followed by margin compression once that inventory is depleted and they’re forced to buy at prevailing, now-elevated prices. That’s the mechanical reason Morgan Stanley stays least constructive on module names within this basket even while liking the sector overall.
Stock Trajectory
Near term, expect chop rather than a clean trend. With EPS revisions this stretched, the stocks are more exposed to disappointment risk on any hyperscaler capex wobble, any signs of “token minimization” substituting cheaper open-source inference for premium compute, or any hint that chipflation is compressing customer margins enough to slow orders. None of that changes the multi-quarter shape of the chart, which MS still expects to trend higher into 2027 on 35–40% projected earnings growth and the agentic AI ramp — it just means the path there likely includes a sideways-to-lower stretch rather than a straight line.
The Position
The way to read this note isn’t “sell memory,” it’s “stop treating every name in the basket the same way.” Morgan Stanley’s own positioning — DRAM over NAND, original manufacturers over module makers, structural bull over near-term momentum — is a selectivity call, not an exit call. For anyone holding across $MU $SNDK $WDC $STX, the practical takeaway is that the next few prints matter more for separating winners from laggards within the group than for validating or invalidating the broader AI-capex-driven memory supercycle thesis.