Meta is still burning money on AR/VR

Meta reported $4 billion in quarterly Reality Labs losses, bringing cumulative losses to $83.5 billion across 21 quarters, or about $4 billion per quarter on average. The company also projected 2026 capital spending of $125 billion to $145 billion, above prior estimates and analyst expectations, as AI infrastructure and compute needs keep rising. Despite Q1 net income of $26.8 billion and revenue up 33% to $56.3 billion, shares fell more than 5% after hours on spending concerns.

Analysis

Meta’s message is no longer “metaverse optionality”; it is that the company has become a perpetual reinvestment machine with no externally visible ceiling on compute intensity. That shifts the equity from a classic ad compounder into a quasi-utility on capex, where incremental earnings power can be overwhelmed by faster-growing infrastructure depreciation and working-capital drag. The market should care less about the headline profit number and more about the widening gap between operating cash generation and free cash flow conversion as capex remains structurally elevated for multiple years.

The second-order winner is the AI infrastructure stack, especially GPU suppliers, networking, optics, and high-density power/cooling vendors. If Meta is still underestimating capacity needs, that implies hyperscaler ordering cycles remain tight and spot pricing for constrained components can stay firm longer than consensus expects. The loser is Meta’s own multiple: investors may tolerate heavy spend when the payoff is clearly monetizable, but repeated guidance resets around capex reduce confidence that management can accurately forecast the slope of returns.

Near term, this is mostly a sentiment and positioning event over days to weeks, but the larger risk is 12-24 months: if AI monetization ramps slower than infrastructure spending, the market will start capitalizing Meta on lower free cash flow durability, not current EPS. The contrarian view is that the stock may actually be too cheap if the market is over-penalizing a company with unmatched distribution and data advantages; however, that bull case requires evidence that AI engagement is translating into durable ad load or new revenue, not just usage anecdotes. Until then, capex uncertainty is the overhang, and any further upward revisions to 2026-2027 spending are likely to compress the multiple again.

Key Decisions for Investors

• Short META on a 2-8 week horizon into any post-earnings bounce; use 1-2x notional via put spreads to express downside if capex guidance keeps drifting higher. Risk/reward favors this because revisions to free cash flow expectations can outrun near-term revenue growth.
• Long a basket of AI infrastructure enablers (e.g., NVDA, AVGO, ANET) versus META as a pair trade for 1-3 months. The thesis is that Meta’s spending intensity transfers margin to suppliers even if Meta’s own equity multiple de-rates.
• Buy META 6-12 month put spreads only if the market rallies back toward pre-earnings levels; target a 2-3:1 payout if investors reprice 2027 capex risk. This is a clean expression of the downside from repeated capex underestimation.
• Avoid chasing the stock on ‘AI monetization’ headlines until management shows free cash flow stabilization for at least two quarters. The setup favors waiting for proof, not paying for optionality that is being financed at rising cost.