Micron shares climb as analysts raise targets on DRAM supply squeeze

Micron shares rose about 2.7% premarket after Deutsche Bank lifted its price target to $1,500 from $1,000 and Citi raised its target to $1,200, both citing tighter DRAM supply-demand conditions. Deutsche Bank sees DRAM supply tightness persisting into 2028 and beyond, while Citi expects DRAM ASPs to rise 200% in 2026 on strong data center demand and constrained supply. Spot DRAM prices are already up 52% since the start of January and 21% above contract pricing, pointing to further contract price increases ahead.

Analysis

The setup is no longer just a Micron story; it is a memory-capacity bottleneck story with second-order beneficiaries across the AI infrastructure stack. If DRAM pricing really keeps compounding into 2027-2028, the marginal winner is not only the largest merchant supplier, but also equipment vendors, HBM-adjacent ecosystem names, and server OEMs with better pricing power and inventory discipline. The key implication is that supply additions are being absorbed faster than the market expected, which usually extends the cycle and punishes any buyer of the "peak margins" narrative too early.

For semis, the more important read-through is that memory is becoming a structural tax on AI compute economics. That can slow budget reallocation toward non-memory-heavy AI workloads, while reinforcing demand for high-end accelerators that justify the memory bill; in other words, compute vendors with bundled platforms should be relatively insulated, while commoditized server and storage buyers face margin compression. The spillover risk is a late-cycle inventory flush in enterprise hardware if customers over-order to secure allocation, which could create a lagged air pocket 2-3 quarters from now even if pricing stays firm.

The consensus may be underestimating the duration of pricing power because it is still anchored to prior semiconductor cycles that were demand-led rather than structurally supply-constrained. The contrary risk is that announced capacity, if it comes online in a synchronized way, can flatten the curve faster than bullish estimates imply; but the more immediate reversal trigger would be spot-to-contract convergence stalling, which would signal buyers have finished front-running the next leg higher. On the trade side, the asymmetry favors being long the memory complex on pullbacks rather than chasing the first spike, because the rerating can persist for quarters even if the next earnings print is only a confirmation event.

Key Decisions for Investors

• Add MU on any post-print pullback, targeting a 6-12 month horizon; prefer scale-in over momentum chasing because the multiple can rerate for several quarters if pricing stays above contract.
• Pair long MU / short a semiconductor equipment basket over 3-6 months; if memory tightness extends, suppliers with direct pricing exposure should outpace capex-levered names, but use tight stops if capex accelerates faster than expected.
• Buy 6-9 month call spreads in MU into the June earnings catalyst to express upside on guide raise with defined downside; structure for ~2:1 to 3:1 payoff rather than outright premium.
• Reduce exposure to low-margin server/OEM names with weak inventory control for the next 2-4 quarters; they are the most likely to absorb pass-through cost inflation and suffer margin squeeze.
• Watch for confirmation in spot-vs-contract spread and lead times over the next 30-60 days; if the spread stops widening, trim longs because the market will start pricing in supply normalization sooner.