Jim Cramer Says Buy 2 AI Stocks up 460% and 1,300% Since 2023 -- Wall Street Agrees

Wall Street targets imply upside for AI leaders: Nvidia median target $300 vs $211 current (+42%) and Meta median target $815 vs $669 (+22%), with investors citing Nvidia’s >80% AI accelerator market share and full-stack platform. Meta also plans to launch a cloud business to monetize its AI infrastructure; capex is cited at $111B (2024-25) rising to $135B in 2026, with earnings expected to grow ~15% annually through 2027. Overall, the piece is constructive on valuation and AI demand, but highlights ongoing uncertainties around competition and returns on Meta’s AI spending.

Analysis

The actionable signal is less “buy AI” and more a shift in how the market should underwrite AI capex. For NVDA, the core setup is still supply scarcity: as long as hyperscalers are capacity-constrained, custom silicon efforts by AMZN, MSFT, and GOOGL are a bargaining chip, not a true substitute. That supports an earnings-duration premium, but it also means the stock is vulnerable to any sign that order growth is normalizing faster than shipment capacity or that ASPs are being traded down to defend share.

META is the more interesting second-order story. If it can externalize excess AI infrastructure into a cloud-like service, the market may start capitalizing its capex at a lower cost of capital than a pure consumer-ad ad platform, which should help multiple support over the next 1-3 quarters. The risk is execution: until there is evidence of third-party demand, this is a narrative bridge, not a new profit pool. If monetization lags, the 2026 spend ramp becomes a balance-sheet and ROIC overhang rather than a catalyst.

Contrarian view: consensus may be overconfident on NVDA’s durability and underconfident on META’s optionality. NVDA likely remains the cleaner fundamental winner over 6-18 months, but the best risk/reward may be a relative-value expression that recognizes NVDA is already the crowded consensus long while META still has room for re-rating if investors believe its infrastructure can earn third-party revenue. Falsifiers are straightforward: NVDA guided data-center growth slowing materially or gross margin compression; META missing any commercialization milestone tied to AI infrastructure within the next 2-3 quarters.

Key Decisions for Investors

• Prefer META over NVDA on a 3-6 month relative-value basis: long META / short a modest basket of AMZN-MSFT-GOOGL if you want to express the idea that Meta’s infrastructure monetization can improve ROIC while cloud incumbents face incremental price pressure. Risk is that the cloud plan remains noncommercial and the pair bleeds on sentiment.
• Use pullbacks to build a tactical long NVDA, but only on weakness after earnings or a capex-related selloff; upside remains intact as long as hyperscaler demand stays supply-constrained. Falsify with any guide-down in data-center growth, margin compression, or evidence of pricing discipline breaking.
• For META, consider a 3-6 month call spread rather than stock if you want to express the monetization optionality with defined downside. The catalyst path is disclosure of customer demand or product launch details; if no third-party traction emerges by the next two earnings cycles, fade the move.
• Watch AMZN/MSFT/GOOGL for secondary pressure if the market starts treating AI infrastructure as a rentable asset outside the traditional cloud oligopoly. Best tell: reacceleration in capex without corresponding cloud revenue growth would argue the market should assign a lower multiple to incremental AI spend.