Meta's Earnings Got a Major Tax Boost. Here Are the Adjusted Figures You Need to See.

Meta’s Q1 revenue rose 33% year over year to $56.3 billion, but adjusted EPS grew only about 14% after excluding an $8.03 billion tax benefit, highlighting a widening gap between sales growth and underlying profit growth. Management raised 2026 capital expenditure guidance to $125 billion-$145 billion from $115 billion-$135 billion, underscoring a much heavier spending outlook. The article is constructive on ad growth and AI/infrastructure execution, but cautious on margin pressure and the earnings impact of escalating costs.

Analysis

META is still compounding at a rate most mega-caps can’t match, but the market is increasingly being asked to underwrite a capex supercycle just as operating leverage is fading. The key second-order issue is not this quarter’s margin level; it’s the growing probability that incremental revenue gets increasingly absorbed by depreciation, cloud commitments, and memory/networking inflation over the next 6-18 months. That means the stock can look optically cheap on forward earnings while being expensive on normalized free cash flow if data-center buildout stays this aggressive.

The beneficiaries are less obvious than META’s direct competitors. Semiconductor memory, networking, and power-infrastructure vendors should see demand duration extend, but the incremental winner is likely the infrastructure stack with the tightest supply constraints, because pricing power shifts upstream when hyperscalers become time-sensitive buyers. Conversely, ad-tech and social peers face a tougher read-through: if META is still growing impressions and pricing simultaneously, it suggests ad demand is healthy, but it also raises the bar for rivals to defend share without matching product and AI spend.

The contrarian risk is that consensus is treating this as a clean ‘AI optionality’ story, when the near-term reality is a balance-sheet and FCF story. If ad pricing decelerates even modestly, or if memory costs stay elevated into the second half, the market could re-rate META from ‘cheap growth’ to ‘capital-intensive platform,’ which typically compresses multiples faster than sell-side models adjust. The biggest catalyst to reverse sentiment would be evidence that capex is translating into measurable AI monetization or a sharp decline in component costs that restores operating leverage before 2027.

Key Decisions for Investors

• Stay underweight META into the next 1-2 quarters if already long; use rallies to trim until the market shows capex-to-revenue conversion, because downside is driven by free-cash-flow disappointment rather than headline earnings.
• If expressing a positive view, prefer a call spread over outright stock: buy META 6-12 month call spreads financed with nearer-dated upside sales; this captures multiple expansion if AI monetization appears while limiting exposure to capex compression.
• Long NVDA vs short META on a 3-6 month horizon as a relative-value expression: META’s spending intensity supports the AI supply chain, but META itself carries the burden of monetization risk; use as a hedge if wanting AI exposure without capex overhang.
• Look for a basket long in infrastructure beneficiaries rather than META itself, especially names tied to memory, networking, and datacenter power, with a 6-12 month horizon where backlog conversion can outpace the market’s model refresh.
• Avoid adding to INTC as an AI-infrastructure proxy here; META’s custom-silicon commentary signals hyperscalers are more likely to vertically integrate where possible, limiting the long-term TAM for generic compute suppliers.