Baird downgrades Rambus stock rating on DRAM shortage concerns

Baird downgraded Rambus to Neutral from Outperform and cut its price target to $120 from the current $141.31, implying downside. The firm cited slower registered DIMM growth prospects as DRAM shortages persist through 2027, partly offset by stronger x86 demand tied to agentic AI. Separately, Rambus reported Q1 2026 EPS of $0.63 versus $0.64 expected and revenue of $180 million versus $189.71 million, a miss on both lines.

Analysis

This reads as a classic “good business, bad setup” rerating rather than a thesis break. The key second-order issue is that Rambus is levered to unit growth in a market where constrained DRAM supply can actually help memory vendors more than it helps an IP / controller monetization model; if memory makers keep diverting capacity to HBM, the scarcity premium accrues upstream, while RMBS is left exposed to weaker module volumes and delayed design wins. That makes the stock more vulnerable to multiple compression than to a clean earnings reset, especially after a miss that undermines the market’s willingness to pay up for AI adjacency.

The market is likely underestimating how long this can persist: shortages extending into 2027 implies this is not a one-quarter cyclical wobble, but a multi-cycle allocation problem. In that environment, any near-term x86 AI demand uplift can be offset by ASP inflation in components that customers may simply not absorb, pushing procurement into later periods. The result is a “phantom tailwind” where end-demand improves on paper but realized revenue growth lags because shipment timing and mix become the bottleneck.

The contrarian angle is that the market may already be pricing in too much of the bad news if RMBS is being treated like a structurally ex-growth semiconductor story. With very high gross margins, the business can absorb some volume pressure better than lower-quality analog names, so the right short is not necessarily outright size but timing around estimate revisions and guidance language. The inflection to watch is whether management starts to frame 2027 as a normalization window rather than a continued shortage regime; that would shift the debate from cyclical to durable and can stabilize the multiple quickly.

Key Decisions for Investors

• Short RMBS on strength over the next 1-3 weeks into analyst downgrade follow-through; target a 10-15% drawdown toward the new price target area, with a stop if management reiterates acceleration in AI-related demand on the next call.
• Buy RMBS put spreads 1-2 quarters out, struck around current levels down to ~15% out-of-the-money; this expresses downside from multiple compression while limiting theta if the stock stalls.
• Pair trade: short RMBS / long a memory beneficiary with clearer pricing power such as MU over the next 3-6 months; the trade captures the spread between a unit-sensitive IP model and a direct pricing beneficiary of DRAM tightness.
• If RMBS retraces sharply on further guidance cuts, cover into the first post-earnings selloff and wait for evidence of order stabilization; this name can mean-revert violently if the market decides the miss was timing, not demand.
• Do not add on dips until there is evidence that DRAM allocation is easing or that RMBS can offset volume pressure with mix expansion; absent that, the risk/reward favors patience over averaging down.