Prediction: These Will Be Micron's and Sandisk's Stock Prices by the End of 2027

Micron and Sandisk are benefiting from a severe memory chip supply shortage, with demand tied to AI data center build-outs expected to keep pricing elevated through at least 2027. Wall Street estimates cited in the article imply FY2027 EPS of $113.81 for Micron and $183.05 for Sandisk, which at 20x earnings would value them at about $2,276 and $3,661 per share, respectively. The piece argues upside could extend to $2,500+ for Micron and $4,000+ for Sandisk if AI capex continues accelerating.

Analysis

The market is now pricing memory as if the current shortage is a multi-year secular re-rating rather than a cyclical spike, and that matters for positioning. The first-order winners are obvious, but the second-order winner is the capex ecosystem: equipment, process-control, and advanced packaging names should see a longer-than-expected order tail because foundry expansions in memory are lumpy, tool-intensive, and difficult to accelerate without destroying yields. The key competitive edge is not just volume, but mix: vendors that can qualify higher-density, higher-margin products into AI infrastructure will capture disproportionate economics even if unit demand normalizes.

The bigger risk is not demand disappearing; it is earnings inflecting faster than the market can sustain on multiple. If pricing stays elevated into 2027, consensus numbers may still prove conservative, but once supply additions begin to show up, the marginal move in share prices will likely depend more on forward capacity utilization than on headline EPS. That creates a classic “good news becomes less incremental” setup over the next 6-12 months: the stocks can keep running, but the path becomes more violent as investors start discounting the post-shortage phase before it appears in reported results.

The contrarian miss is that the market may be underestimating how much of the AI capex boom is concentrated in a handful of hyperscalers, which increases duration risk. If one or two customers pull back even modestly, memory demand can reprice quickly because procurement is often scheduled far in advance and inventories are intentionally lean. At the same time, if AI capex broadens beyond the current cohort, the current valuation framework is still too low, because memory leverage to every incremental dollar of data-center spend is much higher than most investors assume.

Key Decisions for Investors

• Stay long MU and SNDK into the next 1-2 quarters, but use a staggered approach: add on 10-15% pullbacks rather than chasing strength. Risk/reward remains favorable while supply is constrained, but upside becomes more path-dependent once the market starts discounting 2027 capacity additions.
• Pair long MU/SNDK against a basket of near-term beneficiaries with less duration to the cycle, or simply trim after sharp rallies and rotate part of the gain into memory equipment and process tool names that benefit from the same capex wave with lower earnings volatility.
• Consider a call spread on MU or SNDK with 6-12 month tenor instead of outright stock for new capital. This preserves upside if shortage tightness persists while limiting downside if multiples compress before supply normalizes.
• Watch for any signs of inventory rebuild or customer concentration risk at hyperscalers; if procurement lead times shorten or capex guidance rolls over, reduce exposure quickly. The setup is strong, but the reversal could be abrupt once the market believes supply relief is within 6-9 months.
• Avoid treating NVDA as the direct trade on this theme; it is only an indirect beneficiary. If you want pure memory beta, keep exposure centered on MU/SNDK and use NVDA as a hedge only if you want to offset broader AI hardware sentiment risk.