Mega-Cap Q2 2026 Earnings — AI Capex Readthrough
Scenario Comparison
Mixed: ad strength but capex caution
Hyperscalers raise FY2026 AI capex guidance
Capex held flat with digestion language
All Scenarios
3Positioning Suggestions
→ Treat as the key AI-capex readthrough for NVDA/MU into August
→ Size semis exposure ahead of the cluster deliberately
Predictions
1At least one mega-cap raises FY2026 capex guidance
If capex raised → hold/add semis
Key Questions
- Do MSFT/META/GOOGL raise AI capex?
- Any efficiency/ROI-on-AI framing?
- Cloud growth re-acceleration?
Monitoring Checklist
Post-Event Results
Actual Outcomes
- • Meta reported Q2 2026 results after the July 29 close and narrowed 2026 capital-expenditure guidance to $130B-$145B from $125B-$145B, raising the lower end by $5B. That directly resolves the 70% prediction true.
- • Meta's $60.80B of revenue grew 28% year over year, but $31.08B of Q2 capital expenditures and cost and legal items compressed operating margin to 31% and left $0.78B of free cash flow. The result confirmed durable AI-infrastructure spending while increasing the burden on Meta to demonstrate operating conversion.
- • META, NVDA, and MU closed down 1.31%, 3.55%, and 9.94%, respectively, on July 29. In the first full regular session after Meta's release, META fell 7.95%, while NVDA rose 2.65% and MU rose 18.36%. The divergence is consistent with a positive semiconductor capex read-through but does not prove that Meta's report alone caused the moves.
- • META research.doc already marked Q2 monitoring passed and the catalyst realized. No NVDA or MU monitoring-calendar catalyst matched the July 28-31 window, so no research.doc writeback was needed.
Market Reaction
Trade Recommendations
→ The conditional premise was met because Meta raised the lower end of 2026 capital-expenditure guidance. Hold existing semiconductor exposure; keep additions inside the existing risk budget rather than chasing the first-session move.
→ For NVDA and MU, the result confirms the existing demand mechanism and leaves the two-hyperscaler capital-expenditure-cut trigger untriggered. No thesis re-rating is warranted from this event alone.
→ For META, higher capital expenditure is not independently bullish because the Q2 margin and free-cash-flow burden was material. Require evidence of second-half operating-income and margin conversion before increasing exposure on the spending signal.