Marvell Stock Slides After Q3 Earnings, Acquires Celestial AI

Marvell reported Q3 adjusted EPS of $0.76 beating the $0.73 consensus and revenue of $2.075 billion versus $2.07 billion expected, up from $1.52 billion year-over-year. Management announced the acquisition of Celestial AI and raised expectations, forecasting full-year revenue growth to exceed 40% and stronger-than-anticipated data-center demand into next year. Despite the beats and bullish guidance, the stock fell about 6.33% in extended trading to $87.01, indicating a volatile market reaction that investors should monitor for implications of the acquisition and near-term execution risks.

Analysis

Market structure: Marvell’s beat and bullish data‑center guide reinforce that end demand for networking/storage silicon and custom AI inference is accelerating; direct beneficiaries include MRVL, cloud hyperscalers (AMZN, GOOGL) and any TSMC‑dependent foundry partners, while incumbent inference/FPGA vendors (selected AMD/Xilinx use cases, niche accelerators) face pricing and share pressure. Higher data‑center ASPs implied by Marvell’s guide suggest tighter supply vs demand for advanced nodes over next 2–4 quarters and potential sustained margin expansion if capacity holds.

Risk assessment: Tail risks include M&A integration failure, customer concentration shocks (a 15–25% cloud capex cut in 1–2 quarters would invalidate guidance), or regulatory scrutiny of AI M&A; these are low probability but high impact. Near term (days–weeks) expect volatility around analyst revisions and options IV spikes; medium term (3–12 months) execution on Celestial AI integration and TSMC allocation are the key value inflection points.

Trade implications: Tactical entry on weakness—buy on a pullback to $78–82 with a 12–18 month horizon and 25–40% upside objective; implement funded LEAPs (Jan‑2026 calls) paired with short 3‑month calls to monetize IV. Relative trade: long MRVL vs short AVGO (equal dollar) to express share gains in networking silicon over 6–12 months. Size positions 1–3% of portfolio and use 12–15% stop losses.

Contrarian angle: The ~6% post‑market drop likely overstates strategic risk—historical parallels (NVDA+Mellanox) show accretive M&A can be a multi‑quarter catalyst if execution is clean. Market is focused on headline M&A uncertainty rather than the >40% FY revenue trajectory; if acquisition cost proves immaterial (<500–800 bps of market cap) this retracement is a buying window, but watch for any disclosure of purchase price or dilution within 30 days which would flip the thesis.

Key Decisions for Investors

• Establish a 2% long position in MRVL on a pullback to $78–82 (limit orders), target price $110–120 over 12–18 months, stop-loss at 15% below entry; thesis: data‑center outgrowth and Celestial AI optionality.
• Buy Jan‑2026 MRVL LEAP calls (~12–18 month tenor) sized to 1% notional and fund by selling 1–3 month calls (covered/diagonal), aiming to capture structural AI upside while monetizing near‑term IV.
• Execute a relative‑value pair: long MRVL vs short AVGO (equal dollar exposure) sized to 1–2% portfolio, horizon 6–12 months; exit if MRVL underperforms AVGO by 10% or MRVL < $70.
• Hedge existing MRVL exposure with a 3‑month protective put spread (buy 10% OTM put, sell 20% OTM put) to limit downside to ~mid‑teens while reducing hedge cost; re-evaluate after next quarterly guide.