Marvell’s Rally: Valuation Questions as Oppenheimer Raises Target to $300

Oppenheimer this week lifted its Marvell Technology (MRVL) target to $300 from $250, implying roughly 19.5% upside from the prevailing share price of $251.01. The broker's case rests on a long-term artificial intelligence demand story where revenue from custom application-specific integrated circuits, or ASICs, is projected to more than double to over $4 billion next year and to top $10 billion by calendar year 2028.

The Google warrant as a market-moving event

The expanded agreement with Alphabet is the central development that has altered market expectations. Announced on August 17, 2026, Marvell issued Alphabet a warrant allowing the purchase of up to 58.97 million shares at $206.58 per share, exercisable through August 2033. The structure is performance-contingent: additional tranches of shares vest for every $500 million of custom product revenue recognized from Q3 FY2027 through FY2033.

Oppenheimer emphasizes that the scope of the warrant covers a suite of data-center infrastructure products - AI inference accelerators, storage, networking, memory interface controllers, and near-memory computing - but explicitly excludes full TPU cores, according to Oppenheimer's read. That framing positions the deal as a platform-level strategic engagement rather than a single-product purchase.

JPMorgan's interpretation of the warrant mechanics produces an illustrative revenue implication: 240 tranches of $500 million equals $120 billion in cumulative commitments over 6.25 years, which the bank translates to an average potential Google ASIC attach of about $19.2 billion per year through fiscal 2033. That pace would outstrip Marvell's prior analyst-day projection of roughly $11 billion in total AI ASIC revenues in CY28.

Revenue and earnings trajectory

Marvell's financials depict a material inflection. After several years of negative EPS driven by the crypto and networking downturns, the company swung sharply into profitability:

Oppenheimer's projections run ahead of current consensus: the firm models $11.8 billion in ASIC revenue for CY26, a 44% year-over-year increase, and $16.8 billion for CY27. Oppenheimer also outlines a path to $11-12 in EPS for CY28, a sizeable premium to the Street consensus of about $6.43 for FY28 (Jan '28). That divergence - the difference between an $11-12 EPS path and the consensus figure - is the numerical heart of Oppenheimer's investment thesis.

Valuation and competing narratives

Market pricing contains both bullish expectations and reasons for caution. At a trading price of $251.01 and a market capitalization of $220.1 billion, Marvell shows multiples that reflect lofty growth expectations: P/E last twelve months of 87.1x, forward P/E of 60.1x, EV/EBITDA of 81.7x, and a reported revenue growth rate of 34.1%.

The PEG ratio cited in coverage stands at 0.14, but that figure is influenced by mathematically large percentage increases coming off prior negative EPS bases. A fair-value quantitative model cited in coverage flags a potential downside of -33.0% to $168.20, an indication that conventional metrics have yet to fully incorporate the company's shifting AI infrastructure profile.

The bullish case rests on robust hyperscaler capital expenditures. Oppenheimer expects cloud service provider CapEx of about $750 billion this year and roughly $1 trillion next year, and notes Marvell has more than 20 custom ASIC and offload wins. In that scenario, MRVL functions as a critical supplier in a structurally constrained market where management's ability to match supply to demand would be a meaningful operational advantage.

The bear case emphasizes concentration and volatility. Revenue dependence on a small number of hyperscalers, a reported beta of 2.25, and an equity return of +252% over the past year leave the stock exposed to significant multiple compression if AI CapEx moderates even modestly.

Technicals and market action

Technically, the shares register a strong bias across short and long timeframes. All monitored time windows from one-minute to monthly produce a Strong Buy signal. The daily relative strength index sits at 61.4, suggesting bullish momentum without an overbought reading, while a weekly average directional index of 44.3 confirms a pronounced trend. On August 20 the stock rose +5.79% to $251.01, and pre-market activity on August 21 nudged the price to $253.92, an increase of +1.16%.

Key technical levels cited include a daily pivot at $236.95, resistance R1 at $245.80 (already breached) and R2 at $254.34, with weekly supports at S1 $208.50 and S2 $194.99.

Analyst positioning and consensus

The Street has shifted bullish. Representative ratings and targets include:

• Oppenheimer - Outperform, $300
• HSBC - Buy, $300
• Barclays - Overweight, $275
• JPMorgan - Overweight, $240
• UBS - (no rating published), $340 target noted in coverage

Wolfe Research described the Google warrant as "potentially transformational," reiterating the interpretation that the warrant profile implies as much as $120 billion of incremental revenue through fiscal 2033.

Conclusion

Oppenheimer's $300 target is anchored in a coherent structural argument: Marvell is positioning itself as a preferred custom silicon partner for hyperscalers seeking alternatives to merchant GPUs. The warrant terms with Alphabet create a long-dated alignment of incentives over a roughly seven-year window, an uncommon arrangement in semiconductor commercial relationships and one that signals deep strategic commitment from both parties.

That said, the stock already trades at steep multiples - 87x trailing earnings - and has delivered a 252% return over the last year, suggesting much of the AI infrastructure narrative is already reflected in the price. The discrepancy between Oppenheimer's $11-12 EPS trajectory for CY28 and the Street consensus of about $6.43 is the central hinge: it represents either a substantial upside opportunity or a misread of demand durability, depending on whether cloud service provider CapEx reaches the levels Oppenheimer forecasts.