Prediction: Meta's AI Spending Will Pay Off Bigger Than Wall Street Expects

Meta's first-quarter revenue rose 33% year over year, supported by a 19% increase in ad impressions and a 12% rise in average ad prices, while free cash flow remained strong at $12.4 billion despite $19.8 billion of capex. The company lifted its 2026 capital expenditures forecast to $125 billion-$145 billion, reflecting higher AI infrastructure and memory costs, but management says AI-driven engagement and conversion gains are already improving monetization across Facebook and Instagram. Shares have fallen about 28% from their 52-week high, leaving the stock at roughly 21x earnings.

Analysis

The market is treating the capex step-up as a margin event, but the more interesting second-order effect is competitive entrenchment: Meta is effectively converting excess cash flow into a widening data and inference moat while smaller ad platforms face a rising compute and model-quality bar. If the ad tools continue improving conversion efficiency, the spending should be viewed less like discretionary growth capex and more like a performance-based customer acquisition engine for advertisers — a dynamic that can re-rate the business despite near-term depreciation drag.

The key risk is not that the investment is large; it is that returns arrive unevenly. AI-driven engagement gains can show up quickly, but the depreciation hit is mechanical and front-loaded over years, so reported earnings may lag operating traction for multiple quarters even if the underlying unit economics improve. That creates a timing mismatch that can keep a lid on the multiple and fuel periodic drawdowns whenever capex is revised higher or memory costs move against them.

Consensus appears to be underestimating how much of this is already being validated in the core ads franchise. The combination of higher impressions, higher pricing, and better conversion quality means Meta is monetizing the same attention more efficiently, which is the right way to fund infrastructure-heavy AI bets. The contrarian view is that the stock may still be cheap if this is an infrastructure build for a structurally higher ad yield curve, not merely an arms race on compute.

The cleaner trade is to separate execution from narrative: Meta could remain volatile over the next 1-2 quarters as capex and depreciation dominate headlines, but the underlying business momentum argues against fighting the stock too aggressively. The bigger loser may be peers that lack Meta’s scale to absorb memory inflation and still self-fund AI, because their cost of matching product quality rises just as advertisers become more sensitive to conversion efficiency.

Key Decisions for Investors

• Maintain a tactical long META on 3-6 month horizon; use any post-guidance drawdowns to add, targeting a re-rating once investors see another quarter of ad-price and conversion outperformance. Risk: further capex raises keep compressing sentiment.
• Buy META Jan-2027 call spreads rather than outright stock to express upside from AI monetization while limiting exposure to near-term depreciation-driven multiple compression.
• Short a basket of smaller digital ad and social names against META as a relative-value trade; thesis is that rising model and compute costs widen the gap between scaled and sub-scale platforms over the next 6-12 months.
• If META approaches prior highs on no new earnings proof, trim 25-30% of the position; upside is operational, but the stock is still likely to be headline-sensitive to any incremental capex surprise.