BofA raises Marvell stock price target to $240 on compute growth

BofA Securities raised its Marvell price target to $240 from $200 while keeping a Buy rating, citing stronger outlook across compute, networking, optics, memory, storage, and security. The firm lifted fiscal 2027/2028/2029 sales estimates to $11.5B/$16.5B and EPS estimates to $4.06/$6.11/$10.02, while Marvell’s preliminary fiscal 2029 custom compute revenue target was set at $10B, up from $4B year-over-year. The note reinforces a constructive AI/data center growth thesis and may support the stock, which already trades near $198.70.

Analysis

The key implication is not just higher estimates for MRVL, but that the market is starting to underwrite a longer-duration AI infrastructure annuity rather than a one-cycle custom silicon win. If custom compute reaches the implied scale, the earnings profile shifts from “project-driven” to a more recurring design-win ladder, which should compress perceived cyclicality and support a structurally higher multiple than legacy semis. That said, the crowded bullish setup means incremental upside now depends less on another guide-up and more on proof that margins and working capital can scale with revenue.

Second-order winners sit in the adjacent bandwidth and interconnect stack, not necessarily in the same direct-custom-silicon bucket. As MRVL gains share in networking/optics/security, the beneficiaries are likely the suppliers of high-speed interconnect, advanced packaging, and test equipment, while smaller custom ASIC competitors face tougher qualification dynamics and more pricing pressure. AMD’s role here is nuanced: it is not the direct loser from this note, but any investor rotation toward a “more diversified AI semis compounder” narrative could create relative multiple pressure if AMD’s own cadence looks less linear.

The biggest risk is timing mismatch: the bull case depends on large program ramps translating into actual shipments over the next 2-6 quarters, while the stock is already discounting a lot of 2028-2029 power. Any slippage in custom program milestones, customer concentration, or gross margin mix could trigger a sharp de-rating because expectations are now anchored to a very high growth path. The contrarian read is that the market may be paying for the destination too early; if execution remains solid but merely in line, upside from here is likely more limited than consensus suggests.

From a portfolio perspective, this favors owning MRVL on dips rather than chasing strength, and expressing the view with defined risk because implied expectations are elevated. The cleaner setup is relative-value: long MRVL versus a basket of lower-quality AI hardware names where revenue visibility is weaker, or versus AMD only if you expect MRVL’s custom-compute narrative to keep compounding while AMD faces greater sentiment volatility around product timing. Near-term, the stock should react more to forward commentary on program conversion and margin trajectory than to headline revenue beats.

Key Decisions for Investors

• Buy MRVL only on pullbacks toward prior breakout support; use a 3-6 month horizon and target a move toward the new street-high valuation band, with a stop if management commentary points to program slippage or margin dilution.
• Express relative value: long MRVL / short a basket of weaker AI hardware exposure over 1-3 months, focusing on names with less revenue visibility and lower design-win durability; the trade benefits if the market continues rewarding earnings quality over narrative.
• Consider a bullish call spread in MRVL for the next two earnings cycles rather than outright stock, to capture upside from further guide-ups while limiting downside if expectations peak earlier than the fundamentals.
• If holding AMD, hedge part of the position with a short MRVL/long AMD pair only if you expect MRVL’s multiple expansion to overshoot fundamentals; otherwise prefer avoiding the pair because MRVL currently has the cleaner estimate revision momentum.
• Watch for an entry in MRVL after any post-earnings consolidation; if the stock holds gains after the next update, that would confirm institutional accumulation and reduce the risk of a crowded-long reversal.