Why I Won't Stop Buying Meta Before September

Meta’s Q1 2026 results show a major upside vs expectations: EPS $10.44 vs $6.66 consensus (+56.8%) and revenue $56.31B (+33.1% YoY), supported by 19% growth in ad impressions and 12% higher average ad price. The investment thesis leans on AI chip “Iris” mass production starting in September and heavy compute buildout (capex guide $125–$145B), while returns remain strong with operating margin at 41.4% and ROE of 30.24%. Despite notable AI/Reality Labs losses ($4.03B in Q1; $19.2B in 2025) and slower free-cash-flow growth to 11.74% YoY, the piece highlights ample funding capacity (quarterly operating cash flow $32.23B, cash $23.43B) alongside $26.25B buybacks in 2025.

Analysis

META’s edge here is not “AI spend” in the abstract; it is the ability to turn AI into lower unit cost of distribution and higher ad monetization while keeping the economics inside the moat. If the custom chip ramp lands on time, the market should look through the near-term capex spike and start capitalizing a structurally higher operating margin profile, because inference cost is one of the few line items large enough to move long-run earnings power at this scale.

The second-order winner is TSM: every successful hyperscaler ASIC program shifts more wafer and advanced packaging demand toward foundry capacity without requiring TSM to win the headline model race. AVGO is a more ambiguous beneficiary than the market narrative suggests — near-term design and IP revenue can rise, but the longer the hyperscalers prove they can internalize their own compute stack, the more the supplier mix becomes a timing bridge rather than a durable growth annuity. That is the real competitive signal: custom silicon reduces merchant GPU pricing power and may eventually slow the cadence of incremental accelerator demand growth.

The main risk is execution lag, not the thesis itself. Capex hits cash flow now, while monetization from better recommendation economics and chip savings is staged over quarters; if rollout slips or ad pricing softens, the stock can de-rate even with strong reported earnings. The contrarian angle is that the market may still be underestimating how much of META’s AI spend is defensive infrastructure rather than incremental growth, which means the upside is real but the path likely remains choppy for 1-3 months before the September catalyst is fully visible.

Key Decisions for Investors

• Buy META on any post-rally consolidation or broad-market weakness; use a 6-12 month horizon and expect the stock to re-rate if Iris stays on schedule and ad pricing remains firm. Falsifier: capex rises again without a corresponding improvement in engagement or ad yield.
• Pair long META / short AVGO in modest size for a 3-6 month relative-value trade if the market keeps pricing Broadcom as the primary AI capex winner. This works only if hyperscaler custom silicon continues to substitute for merchant accelerator spend; exit if AVGO AI revenue guide re-accelerates meaningfully.
• Long TSM as the quieter beneficiary of the ASIC buildout over 6-18 months. This is the cleaner way to own the manufacturing pull-through with less balance-sheet and product-cycle risk than betting on any single chip designer.
• Consider META call spreads rather than outright stock if you want exposure to the September catalyst but want to finance the execution risk embedded in the capex ramp. Best entry is after a pullback, not after a fresh sentiment spike.