Meta's earnings performance has been exceptional—five consecutive quarters of beats—yet equity price appreciation has stalled entirely. This disconnect reveals a fundamental investor bifurcation: growth optimists betting on AI infrastructure payoff versus value purists questioning whether $145 billion in capital expenditure justifies current valuations or represents overinvestment.
The capex scale is historically aggressive for a software-dominant business model. Such spending typically signals either confidence in proprietary AI advantages or desperation to compete in a winner-take-most AI arms race. Meta frames this as foundational infrastructure for inference and model training; skeptics view it as capital inefficiency that suppresses near-term earnings accretion and free cash flow visibility.
Next quarter's guidance and capex commentary will be determinative. If management signals moderation or demonstrates ROI metrics tied to monetizable AI products, sentiment may reverse. If guidance maintains or escalates the spending plan without concrete revenue catalysts, downward pressure could accelerate. NVDA benefits directionally from sustained data center capex, while GOOGL faces parallel capital allocation scrutiny.
Sector implication: This represents a broader tension in mega-cap technology between AI supremacy narratives and disciplined capital allocation. Investors are rationing conviction until proof of AI-to-revenue conversion emerges; expect volatility anchored to quarterly delivery benchmarks rather than fundamental revaluation.