Morgan Stanley Raises Marvell's Long-Term Outlook and Target Price to $195, Maintains Neutral Rating

Overview

Morgan Stanley released a research report noting that while Marvell’s latest quarterly results met expectations, management significantly raised its long-term outlook for fiscal years 2027 and 2028. Driven by robust demand in data centers and AI networking, the firm substantially increased earnings forecasts and raised its target price to $195. However, considering the stock’s already elevated valuation, lack of relative value compared to its top pick NVIDIA, and high stock-based compensation expenses, Morgan Stanley maintains an 'Equal-weight' (neutral/market-perform) rating on Marvell.

Core views

The core thesis centers on broad-based strength in AI infrastructure demand. First, the Data Center segment has become the primary growth engine, with FY27 revenue expected to grow approximately 50% year-over-year and accelerate further to 55% in FY28. Within this, network interconnects represent the most significant near-term upside driver. The firm raised its FY27 interconnect growth forecast to over 80%, driven by TIA/drivers annualized run rates approaching $1 billion and Scale-up optics scaling from $150 million to over $300 million. Management positions Scale-up networking as a critical area for AI infrastructure investment over the next few years. Second, visibility into the Custom Silicon/XPU business has improved markedly. While FY27 growth remains modeled at around 25%, confidence in FY28 has strengthened significantly, with custom revenue expected to double year-over-year. This is based on existing XPU programs, attach opportunities, and large-scale production ramping with a new Tier-1 customer next year. Although this segment still carries optionality, enhanced customer engagement and program expansion have improved modeling confidence. On the financial front, Marvell continues to invest heavily in optics, switching, and custom silicon to complete its AI connectivity platform. Despite higher investments, operating leverage remains strong, with FY28 operating expenses projected to grow only in the high-teens percentage range—far below the ~45% revenue growth. Diluted shares outstanding increased from 856 million to 893 million, with another 22 million shares expected next quarter.

Analysis framework

The firm employs a bottom-up sum-of-the-parts and scenario analysis approach. First, it breaks down the Data Center segment into sub-components (interconnect vs. custom compute) to assess individual growth trajectories and market share dynamics. For networking, it closely tracks adoption of high-speed optical modules (Inphi PAM4/400ZR) and Scale-up architectures; for custom silicon, it monitors mass production timelines and attach rates for key customer programs. Second, it compares management guidance against consensus estimates to identify expectation gaps. Finally, for valuation, it uses a relative P/E multiple approach, anchoring on CY27 estimated EPS (including stock-based compensation), benchmarking against high-growth AI semiconductor peers, and cross-comparing with NVIDIA’s EPS and valuation to assess margin of safety and relative attractiveness.

Methodology notes

• Valuation MethodologyPE/PEG valuation

  Assigning a fixed P/E multiple based on forward EPS

  The firm uses approximately 40x CY27 estimated EPS (including stock-based compensation) as its valuation anchor—a common relative valuation method for high-growth semiconductor stocks—to assess whether the current price has priced in two years of future growth.

• Industry/Thematic FrameworkVolume-price decomposition

  Breaking down revenue growth into volume (units) and price/product mix effects

  When analyzing the interconnect business, the firm not only looks at overall growth but also decomposes contributions from high-value components like TIA/drivers and Scale-up optics (e.g., scaling from $150M to $300M) to validate the structural quality of growth.

• Company Fundamentals & Financial FrameworkOperating/Financial Leverage Analysis

  Analyzing the gap between revenue growth and expense growth to assess margin expansion potential

  The report highlights that despite increased R&D and investment spending, FY28 revenue growth (~45%) far outpaces operating expense growth (high-teens %), and this operating leverage is key to supporting margin expansion and EPS upside.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Marvell Technology (MRVL.O)

  Direct beneficiary, dual-driven by AI interconnects and custom chips

  Strengths

  Inphi optics leadership in high-speed interconnects, strong progress in custom XPU programs, robust operating leverage

  Weaknesses

  Premium valuation vs. peers, high stock-based compensation diluting shareholder returns, slow recovery in storage and enterprise markets

  Comparison

  Compared to top pick NVIDIA, Marvell’s FY27 EPS is only ~$6 (vs. NVDA’s ~$13) despite similar share price, resulting in lower valuation attractiveness

  Risks

  Weaker-than-expected AI end-market demand, intense competition for networking share, delays in custom program ramps

Key data

• FY27 Data Center Revenue Growth Expectation~50%Significantly raised from prior expectations; main growth driver
• FY28 Data Center Revenue Growth Expectation~55%Accelerating further from a higher base
• FY27 Network Interconnect Growth Expectation~80%Firm’s model forecast, reflecting strong AI networking demand
• Latest Quarter Revenue$2.42bnUp 9% QoQ, 28% YoY, slightly above expectations
• Latest Quarter EPS$0.80In line with both consensus and firm expectations
• New Target Price$195.00Raised from $172, implying ~40x CY27 EPS
• FY27 Revenue Forecast$11.5bnRaised from $10.87bn
• FY28 Revenue Forecast$16.2bnRaised from $14.29bn

Impact & implications

The report argues that Marvell’s long-term AI narrative is now substantiated by data, particularly the stronger-than-anticipated breakout in interconnects. This signals Marvell’s successful transformation from a legacy storage/enterprise chip vendor into a core AI infrastructure connectivity platform provider. However, for investors, this means higher valuation requires sustained proof of growth durability. The firm cautions that while fundamentals are positive, capital may prefer other AI leaders (like NVIDIA) offering better valuation or higher certainty at current prices.

Risks

• AI end-market demand falls short, leading customers to sharply cut GPU and associated networking purchases
• Persistent weakness in storage and enterprise data center markets drags overall performance
• Inventory adjustments in interconnect business offset benefits from custom ASIC programs
• High stock-based compensation continues to weigh on investor sentiment and net income
• Competitors (e.g., AMD) re-emerge as threats in GPU or custom hardware segments

What to watch

• Whether interconnect market share and content per system continue to expand
• Production ramp progress and customer confidence in large custom XPU programs
• Direction of future earnings estimate revisions
• Early signs of recovery in storage and enterprise networking markets