Following earnings from Microsoft, Meta, and Alphabet, the combined AI infrastructure spending from Big Tech is tracking toward the $1 trillion threshold — a figure that is reshaping how investors frame the risk/reward in mega-cap tech. Microsoft reported FY2025 revenue of $281.7B (+14.9% YoY) with a 68.8% gross margin and $13.64 diluted EPS; Meta posted $201B (+22.2% YoY) with a 30.1% net margin; and Alphabet logged $402.8B (+15.1% YoY) with a 32.8% net margin. These are healthy top-line growth rates, but the central question is whether the capex cycle is already baked into current valuations.

The spending wave touches not just MSFT, META, and GOOGL directly, but also the infrastructure layer — hyperscaler suppliers, networking, and power. For the three operators themselves, the bull thesis rests on whether AI monetization (Copilot, AI Overviews, Meta AI) can grow into the capex base fast enough to sustain or expand margins.

The bear tension is real: at nearly $1T in combined AI spend, any deceleration in enterprise AI adoption or ad market softness could expose the capex commitment as premature. Free cash flow yield compression is the mechanism to watch — elevated capex reduces FCF even when revenue grows, which matters for a cohort that trades on FCF multiples.

The setup is genuinely two-sided. Revenue growth across all three names is robust and margins remain high by any historical standard. But the 'spending bubble' framing in the headline reflects a legitimate concern that the market may be entering a phase where investors demand proof of AI ROI rather than simply rewarding the spend itself. Watch for any guidance revision or capex step-down as the catalyst that resolves the tension.