Marvell shares are down after earnings. Many analysts are still bullish

Marvell reported Q2 revenue of $2.74B (vs. $2.72B expected) and EPS of $0.94 (vs. $0.93 consensus), but the stock fell ~8% premarket on an expectations mismatch. Despite the selloff, major banks (BofA, UBS, Barclays, Wells Fargo, Citi) reiterated buy/buy-equivalent ratings with price targets unchanged—e.g., BofA $365 (+51% upside) and UBS $310 (+28%). Analysts attribute the dip to near-term expectations (including a prior disclosure that Google can buy up to $12.2B of Marvell shares) while pointing to upside catalysts like an early-October analyst day.

Analysis

The selloff looks more like a positioning reset than a fundamental reset: MRVL has been repriced as an AI infrastructure compounder, so a modest beat cannot clear the bar once the market has already capitalized a large chunk of the Google-related upside. That creates a second-order benefit for the broader AI capex complex: the counterparty signal from GOOGL reinforces custom silicon demand, but the market is increasingly separating “validated strategic demand” from immediate earnings upside, which is why the stock can fall even as the narrative strengthens.

Near term, the risk is de-grossing in crowded semis names rather than a deterioration in end demand. Over the next 2-6 weeks, the stock is vulnerable if analyst-day expectations keep inflating into early October; over 1-3 months, the key catalyst is whether management can raise the implied mid-term growth path, not just defend it. If commentary suggests the Google agreement was already embedded in forward guidance, the name can stay rangebound despite buy ratings.

The contrarian view is that the market may be over-penalizing a high-quality mix shift because it is focused on one quarter instead of the multi-quarter ramp in cloud networking and custom compute attach. That said, the easy money is probably gone unless the company can show margin leverage from mix, not just top-line acceleration. Falsifier: if October investor day fails to lift the implied 2026-2028 growth trajectory or gross margin framework, the rerating thesis stalls.

Key Decisions for Investors

• Tactically buy MRVL on further post-open weakness, but only after the stock stabilizes for 1-2 sessions; risk/reward is better into the early-October investor day than chasing the first print.
• Pair trade: long MRVL / short SOXX for 4-8 weeks to isolate idiosyncratic upside if the selloff is just crowded-position digestion; cover if MRVL fails to reclaim its post-earnings gap by mid-September.
• Use a defined-risk call spread on MRVL into investor day rather than outright stock if volatility remains elevated; the trade is for a catalyst pop, not a smooth grind higher.
• Watch for any guidance language implying the Google partnership is already fully embedded. If management does not raise the long-term growth/margin framework by investor day, cut exposure and treat the move as overdone.