NVIDIA: Risk-Reward Update, Bullish on Data Center Demand

Overview

Morgan Stanley published a risk-reward update report on NVIDIA (NVDA.US) on May 21, 2026, maintaining its Overweight rating and $288 target price. The core thesis rests on the large language model (LLM) boom driving robust cloud capital expenditures, Blackwell remaining the leading solution for AI workloads, and the Rubin family further solidifying NVIDIA’s performance advantage. The report also updates EPS forecasts for fiscal years 2027–2029 and presents bullish, base-case, and bearish valuation scenarios.

Core views

**Sustained Strength in Data Center Demand**: The report argues that enthusiasm generated by large language models continues to translate into strong cloud capex. Blackwell remains the preferred solution for generative AI workloads, with compute demand persistently outstripping supply. The firm expects ongoing upward pressure on earnings estimates, and the Rubin family is poised to sustain NVIDIA’s performance leadership. **Three-Scenario Valuation**: In the base case, the firm applies a ~22x multiple to MW CY27 EPS ($13.08), aligning with the broader market but reflecting a discount versus peer semiconductor companies (AMD/AVGO/INTC). This discount stems from NVIDIA’s current high market share and gross margins, which constrain near-term multiple expansion. The bull case (target price ~$330, implying +47.67%) assumes a ~23x multiple on bullish CY27 EPS ($14.00), underpinned by continued data center revenue growth and contributions from networking, Vera Rubin systems, AI PCs, autonomous vehicles, and software. The bear case (target price ~$160, implying -28.40%) applies a ~16x multiple to CY27 EPS ($10.00), assuming both key uncertainties evolve negatively—e.g., a sharp slowdown in data center growth, intensified competition, or increased customer adoption of in-house hardware. **Key Financial Metrics**: Forecasts for FY2027–2029 include: FY2027 GAAP revenue of ~$393.0 billion, MW gross margin of 74.4%, and MW EPS of $8.96; FY2028 revenue of ~$598.8 billion, gross margin of 72.5%, and EPS of $13.08; FY2029 revenue of ~$783.9 billion, gross margin of 72.0%, and EPS of $17.63. Days of inventory are projected to rise from 123 days in FY2026 to 141 days in FY2027, then gradually decline.

Analysis framework

The firm employs a risk-reward framework, integrating fundamental analysis with option-implied probabilities to assess potential returns across different scenarios. Core analytical steps include: first evaluating demand strength based on AI capex trends, product cycles (Blackwell → Rubin), and competitive dynamics; second, assigning distinct valuation multiples to three EPS scenarios (bull/base/bear) to derive scenario-specific target prices; and third, leveraging options market data to estimate the risk-neutral probability of each scenario, yielding a comprehensive risk-reward distribution.

Methodology notes

• Valuation MethodologyP/E and PEG Valuation

  P/E (Price-to-Earnings) valuation method: Target price derived by multiplying forecasted earnings (EPS) by an appropriate valuation multiple.

  In this report, the firm uses a ~22x multiple on MW CY27 EPS ($13.08) as the base-case valuation, benchmarked against both the broad market and peers (AMD/AVGO/INTC). It notes a relative discount versus peers due to NVIDIA’s high market share and gross margins limiting near-term multiple expansion. Bullish and bearish scenarios employ alternative P/E multiples and EPS assumptions to define the target price range.

• Cyclical and Industry Trend FrameworkInflection Point Analysis

  Assessing industry cyclical positioning by analyzing trends in AI capex and the relative balance between demand and supply.

  The firm judges the AI industry to remain in a phase where demand consistently exceeds supply. Blackwell’s status as the leading solution validates sustained high industry cyclicality, and the upcoming Rubin launch is expected to reinforce NVIDIA’s performance leadership—supporting valuation.

• Corporate Fundamentals and Financial FrameworkFree cash flow analysis

  Assessing profitability and operational efficiency via forecasts of GAAP revenue, gross margin, EPS, and inventory metrics.

  The report provides key financial forecasts for FY2027–2029, including GAAP revenue, MW gross margin, MW EPS, and days of inventory. The projected rise in inventory days followed by a gradual decline reflects the firm’s view on inventory turnover dynamics.

• Industry/Industrial Analysis FrameworkSupply-demand framework

  Analyzing how supply-demand dynamics in the AI chip market affect NVIDIA’s financial performance.

  The report repeatedly cites 'compute demand persistently exceeding supply', noting that Blackwell—the preferred solution—remains in short supply. Future shifts in supply-demand balance constitute a central assumption in the bear case.

• Quantitative/Factor/Portfolio TheoryBeta/alpha analysis

  Estimating risk-neutral probabilities of scenarios using options-market implied volatility to assess excess return relative to the market.

  The firm uses options market data as of May 20, 2026, to estimate the risk-neutral probability that NVIDIA’s stock price will reach or exceed each scenario’s target price within the next 12 months.

Asset mapping & comparison

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

• NVIDIA CORP (NVDA.US)

  Core research subject, benefiting from AI capex growth and product leadership

  Strengths

  Blackwell is the preferred solution for generative AI workloads; Rubin family expected to sustain performance leadership; new growth vectors including AI PCs, autonomous vehicles, robotics, and software

  Weaknesses

  High market share and gross margins limit near-term multiple expansion; valuation discount versus peers AMD/AVGO/INTC

  Comparison

  Trades at a valuation discount versus computing semiconductor peers (AMD/AVGO/INTC), as its growth drivers are already largely priced in

  Risks

  Order-related risks: Tariff headwinds and export controls impact worse than expected; competitive risks: AMD regains GPU competitiveness; cloud customers (beyond Google) develop competitive in-house chips; demand risks: End-market AI adoption lags expectations, prompting major GPU procurement cuts

Key data

• Target Price$288.00Morgan Stanley target price
• Current Stock Price$223.47Closing price as of May 20, 2026
• Implied Upside+28.88%Based on difference between target price and current stock price
• MW CY27 EPS Estimate$13.08EPS used in base-case valuation
• FY2028 GAAP Revenue Forecast$598.809 billionFirm’s forecast
• FY2028 MW EPS Forecast$13.08Firm’s forecast

Impact & implications

The report maintains NVIDIA’s Overweight rating, affirming its position as a core beneficiary of the AI capex cycle. Supported by Blackwell’s leadership and the iterative advancement of the Rubin family, NVIDIA is well positioned for robust growth in AI training and inference. Additionally, emerging drivers—including AI PCs, autonomous vehicles, robotics, and software—may further expand its total addressable market. However, its high market share and gross margins constrain near-term multiple expansion, representing the primary reason for its relative valuation discount versus peers.

Risks

• Order-related risks: Tariff headwinds and export controls impact worse than expected
• Competitive risks: AMD re-emerges as a competitive GPU alternative; cloud customers (excluding Google) develop competitive in-house chips
• Demand risks: AI end markets underperform expectations, prompting customers to sharply reduce GPU procurement
• Valuation risks: High market share and gross margins limit near-term multiple expansion, resulting in a relative valuation discount versus peers

What to watch

• AI Capex Trends: Cloud customers’ (especially large ones’) capital expenditure growth
• Product Rollouts: Shipping progress and market acceptance of Blackwell and Rubin families
• Competitive Landscape: GPU product developments from AMD and other competitors, and progress on cloud customers’ in-house chip initiatives
• New Business Lines: Commercialization progress of AI PCs, autonomous vehicles, robotics, and software