Will Alphabet's FY2026 CapEx exceed $185B (above the top of guidance)?

The $175-185B guidance range is management's commitment. Historically, Google has generally stayed within or below CapEx guidance. FY2025 CapEx of $91.4B was 'in line with expectations.' The question is whether competitive dynamics force spending above the range. Peer hyperscalers (MSFT $50-65B, AMZN $50-75B, META $35-45B) are also ramping but none approaching Google's scale. The $240B Cloud backlog could pull demand forward, but management typically sizes guidance to include known demand. Exceeding guidance would signal either demand surprise or competitive escalation. Setting at 35% — more likely within range than above.

The guidance range of $175-185B already represents a near-doubling of FY2025. Management set this range knowing the demand pipeline ($240B Cloud backlog, Gemini scaling, AI infrastructure needs). They would not set a range that they expect to exceed — the reputational cost of exceeding guidance on the largest CapEx commitment in history is high. The more likely scenario is that CapEx comes in within the $175-185B range or potentially below if some projects slip. The CFO has incentive to manage this number carefully given market focus on capital allocation discipline. Setting at 30%.

The AI CapEx arms race creates upside pressure. If Microsoft or Amazon significantly increase their commitments, Google may need to respond to maintain infrastructure parity for Cloud customers. The dual-class governance structure means shareholders cannot constrain this — management has full discretion. TSR-based compensation incentivizes growth over capital efficiency, which could bias toward higher spending. Additionally, AI model training costs are escalating with frontier model sizes (Gemini 3, etc.). Setting at 38% — acknowledging the competitive dynamics and governance structure that could push spending above guidance.

Management gave a $175-185B range. They'll probably land at the high end or modestly above. Exceeding $185B by a meaningful amount requires competitive pressure or demand surprise. Most likely outcome: $178-188B. But the question is strict — above $185B. I'd say 33% chance given the tendency for CapEx to run hot in an investment cycle.

The guidance range is wide ($10B spread). Management typically lands within the range. The construction timeline for data centers means much of FY2026 spending was committed before the guidance was given. Hard to significantly exceed a range that was set based on known commitments. Setting at 28%.

The competitive dynamics are real. Microsoft just announced a multi-year $100B+ AI infrastructure commitment. Amazon is scaling aggressively. If a major Gemini model training run requires additional GPU clusters, or if Apple's Foundation Models partnership demands more capacity than anticipated, management won't hesitate to spend above $185B. The governance structure permits it and compensation incentivizes it. Setting at 37%.

Management set $175-185B range. Likely stays within. Competitive pressure could push above. Setting at 32%.

Management has track record of in-line CapEx execution. $10B wide range provides buffer. Low probability of exceeding top end. Setting at 25%.

AI CapEx arms race is intensifying. Governance doesn't constrain. Possible to exceed $185B if competitive dynamics escalate. Setting at 35%.