The S&P 500 is up 8.3% in 2026. Microsoft is down 12%. Do the arithmetic on market capitalization and the result is striking: no other company has subtracted more value from the index’s gain than MSFT. Meta and Tesla are next in line, but their combined drag is a fraction of what Microsoft’s slide has cost the benchmark. The team at YourDailyAnalysis walks through what is happening at a company posting genuinely strong operating results and still watching its stock price head in the wrong direction – a paradox that says something uncomfortable about how markets price AI investments.
Start with the raw numbers. For fiscal Q3 2026, ended March 31, Microsoft reported revenue growth of 18% year over year, operating income up 20%, and non-GAAP diluted EPS rising 21%. Azure and other cloud services grew revenue 40%. Microsoft’s AI business crossed an annual run rate of $37 billion. The commercial backlog hit approximately $627 billion. Outstanding results. The complication is capex. Microsoft spent $31.9 billion in the quarter – roughly two-thirds of that on GPUs and CPUs – and CFO Amy Hood guided Q4 capex to exceed $40 billion. For calendar 2026, Microsoft now expects total capex of approximately $190 billion, a 61% increase over fiscal 2025’s $64.5 billion and roughly 23% above what analysts had modeled. Hood noted that approximately $25 billion of that increase reflects higher component prices – primarily memory and GPU pricing pressures turning AI hardware procurement into a bidding war. YourDailyAnalysis spotlights this component-inflation factor as the element that most surprised institutional desks: it means spending growth is partly involuntary, driven by market conditions rather than aggressive investment choice.
There is a third scenario beyond the bull and bear case that YourDailyAnalysis identifies. The bull case: CEO Satya Nadella has called this a once-in-a-generation infrastructure investment, comparable to early cloud buildout. Azure margins at a similar stage were worse than current AI margins. The bear case: $190 billion in capex is roughly triple what Apple spends annually on its entire business, and return timelines remain opaque. The third scenario is structural – supply-chain constraints force Microsoft into GPU bidding wars that simultaneously inflate costs and delay capacity, meaning the company absorbs the pain without the offsetting revenue acceleration. Hood acknowledged Microsoft expects to remain capacity-constrained through 2026.
The market’s P/E response is instructive. Microsoft’s trailing P/E has compressed to approximately 23.4 – well below its five-year average of 32.4. From a purely mechanical valuation standpoint, the stock is the cheapest it has been in years relative to earnings power. At the same time, Amazon plans $200 billion in capex for 2026, Alphabet is aggressively spending, and Meta saw its stock fall more than 6% after its own earnings call before recovering. The entire hyperscaler cohort faces the same bind: markets demanded proof of AI monetization, got reasonable answers, and still couldn’t shake anxiety over the denominator – the spending. YourDailyAnalysis sizes up the cross-sector read as follows: this is not a Microsoft-specific problem but a hyperscaler-wide transition from the easy part of AI investment (announcing it) to the hard part (proving the return on a $600-billion-plus annual industry capex bill). Estimates for combined AI-related capex across the four major hyperscalers in 2026 now exceed $725 billion.
Our reporters also catch the Windows data as a secondary signal worth noting. Personal computing revenue slipped 1% to $13.2 billion, with Windows sales down 2% and Xbox content off 5% year over year. The core enterprise and cloud franchise is healthy; consumer-facing businesses are softening. Enterprise AI products like M365 Copilot carry better unit economics than consumer licenses – that mix shift matters for long-run margin guidance, even as Azure continues to accelerate. Microsoft’s operating margin hit a 20-year high of 45.6% in fiscal 2025, and management expects it to expand by another point year over year despite the capex surge.
The nearest term catalysts: Q4 fiscal guidance at $86.7–$87.8 billion in revenue (13–15% year-over-year growth), the pace of Azure capacity expansion, and whether memory prices ease in H2 2026. Your Daily Analysis leaves readers with a single operational point: the inflection in MSFT’s stock price will trace the inflection in GPU and memory pricing – not the AI revenue line, which is already strong. Watch semiconductor cost curves, not earnings beats.
