Bank of America hikes Sandisk price target to $2,100. Here's why

Bank of America reiterated a Buy on Sandisk and raised its price target to $2,100, implying about 34% upside from Friday's close. Analyst Wamsi Mohan said AI-driven memory shortages should keep pricing power strong, with over one-third of fiscal 2027 revenue already signed under new business model contracts and more than 60% of NAND supply still available at higher prices. Despite a 15% two-day pullback last week, Sandisk remains up more than 550% in 2026.

Analysis

The key incremental takeaway is not that SNDK’s demand story is intact, but that the company is actively converting a cyclical commodity business into a quasi-contracted supply business. That matters because it reduces the usual reflexive downside in memory: when pricing weakens, producers typically keep output elevated to absorb fixed costs, which deepens the price spiral. If SNDK can actually throttle wafers without blowing up cash generation, the stock earns a structurally higher multiple than legacy NAND peers because earnings volatility should compress over the next 4-8 quarters.

The second-order winner is likely the broader memory ecosystem tied to AI infrastructure, but only the differentiated names. A tighter supply regime tends to favor vendors with pricing discipline and strong customer lock-in, while pressuring lower-quality competitors that rely on spot pricing and volume to preserve utilization. If the market starts believing this new contracting model is sticky, the overhang shifts from "AI demand optionality" to "capacity scarcity," which can extend the cycle even if hyperscaler capex growth normalizes.

The main risk is not an imminent collapse in end demand; it is a sentiment reset once investors realize the stock already discounts a very long duration of perfect execution. In the next 1-4 weeks, the trade is vulnerable to factor rotation out of crowded AI winners, especially if semicap or networking earnings temper near-term expectations. Over a 3-6 month horizon, the real test is whether contract coverage keeps expanding fast enough to offset any deceleration in spot pricing; if not, multiple compression can happen well before fundamentals roll over.

Contrarian view: the market may be underestimating how much optionality SNDK retains from self-supply reduction. The ability to cut production in a slowdown is effectively a call option on margin preservation, and that is rare in memory cycles. The flip side is that any evidence of inventory normalization or delayed AI orders could hit the name harder than the street expects because positioning is likely still crowded after a 550% move.

Key Decisions for Investors

• Stay long SNDK but trim size into strength; use the recent pullback as a tactical add only if price stabilizes above the post-selloff support level. Risk/reward favors continued upside over 1-3 months, but near-term drawdown risk remains high because the name is crowded and momentum-driven.
• Pair trade: long SNDK / short a lower-quality NAND or memory laggard with weaker pricing discipline over the next 1-2 quarters. The trade expresses the view that contracting and supply control will widen the quality gap, while reducing outright AI beta.
• Buy SNDK call spreads 2-4 months out instead of common stock for a defined-risk expression of continued rerating. This is attractive if you expect the stock to re-accelerate on further contract wins but want to cap downside if the AI trade keeps de-risking.
• Reduce exposure to AVGO tactically until AI capex enthusiasm re-stabilizes; the negative tape in adjacent AI infrastructure names suggests a near-term multiple reset risk over the next 2-6 weeks.
• Monitor for follow-through in contract coverage disclosures; if SNDK materially expands NBMs over the next quarter, add on confirmation because that would justify a higher terminal multiple and lower earnings volatility.