Key Findings
The AI model lowered its Q2 FY2026 EPS estimate from $4.02 to $4.00, with revenue trimmed to $83.1B from $83.4B. While still 3.6% above consensus of $3.86, the adjustment signals caution on margin expansion.
Thesis
GPT's revision centers on 'higher AI-infra D&A/cost-of-revenue pressure and more conservative non-operating/FCF assumptions.' The model explicitly notes it is 'not assuming aggressive operating leverage' and keeping 'D&A elevated to reflect ongoing AI/datacenter buildout.' Non-operating income volatility from FX and mark-to-market swings is flagged as a key risk.
What This Means
This forecast highlights the tension between AI revenue opportunity and infrastructure investment costs. GPT's 54% confidence reflects genuine uncertainty about how quickly Microsoft can monetize its AI investments versus the depreciation drag. The specific callout of capex cycle timing provides insight into what the AI sees as the binding constraint.