Microsoft

Five weeks after Microsoft's April 29, 2026 print, the stock sits near $414 — still below where it traded into earnings, even though Q3 FY2026 was a clean beat: revenue of $82.9 billion (+18% reported, +15% constant currency), diluted EPS of $4.27 (+23% GAAP), and Intelligent Cloud revenue of $34.7 billion (+30%). Microsoft Cloud reached $54.5 billion. The overhang is CFO Amy Hood's calendar-2026 capex disclosure of roughly $190 billion (+61% YoY) — about $35 billion above the $154.6 billion consensus — with around $25 billion of that increase attributed not to extra GPU buys but to higher component pricing, chiefly HBM memory. Q3 PP&E spend alone hit $22 billion (+45% YoY), and Hood guided Q4 capex above $40 billion.

The bull's offset is demand. Azure and other cloud services grew 40% YoY — the fifth straight quarter of acceleration — and Microsoft's AI business surpassed a $37 billion annualized run-rate (+123% YoY). The standout fresh data point: Microsoft 365 Copilot paid seats crossed 20 million, up roughly 33% sequentially from 15 million in January and with seat additions up about 250% YoY — the clearest sign yet that Copilot is monetizing. The restructured OpenAI agreement, finalized late 2025, converted Microsoft's profit-share into a 27% stake in OpenAI Group PBC (valued around $135 billion) with IP access through 2032, while dropping exclusivity so OpenAI can use other clouds and Microsoft can ship rival models.

The biggest smart-money headline this cycle is Bill Ackman's Pershing Square opening a brand-new MSFT position worth roughly $2.1 billion in the Q1 2026 13F window — a notable vote of confidence from an investor who had previously avoided the name. Across the filing season, roughly 3,308 institutions added MSFT versus about 2,679 trimming, leaving the buyers comfortably ahead. The anchor holders remain Vanguard, BlackRock, State Street, FMR and Norges Bank.

Among the concentrated managers BMInsider tracks, Terry Smith of Fundsmith (London-based, ~$25B AUM) has held MSFT continuously since 2014 and treats it as a Buffett-grade compounder where regulatory risk is materially lower than at Alphabet or Apple and where even $190 billion of capex is being absorbed by 40%+ Azure growth. Insider activity has stayed routine: Satya Nadella and Amy Hood have transacted only through scheduled 10b5-1 plans, with no unusual selling around the capex disclosure. The next watch item is the proxy season detail on how Microsoft has structured incentives around Mustafa Suleyman's roughly 500-person superintelligence team — a tell on how aggressively the board is funding model independence.

Microsoft trades near $414 per share. On a trailing basis the P/E sits around 24.7×; on forward FY2026 EPS the multiple is roughly 31×, a premium to Alphabet but supported by Azure's higher-margin profile, the OpenAI equity stake and lower antitrust exposure. The post-earnings drift has compressed the valuation from its early-2026 highs. Sentiment on the sell side stays strongly constructive: across roughly 71 analysts the consensus is a Strong Buy (about 53 buy, 3 hold, 0 sell), with a median 12-month price target near $560 — implying mid-30s percent upside from current levels. The gap between bullish targets and a flat tape reflects a single debate: whether $190B-plus capex earns an acceptable return before depreciation bites.

1. July 28, 2026: Q4 FY2026 earnings (after close) — the first read on whether the capex shock is digested; Amy Hood is expected to confirm Q4 capex above $40 billion and frame fiscal-2027 spend, which she has signalled will grow at a slower rate with a greater mix of short-lived assets
2. Second half of 2026: Microsoft's own MAI frontier models, which Suleyman says will ship during 2026 — the first real test of model independence and a potential de-risking of the OpenAI partnership overhang
3. Late 2026: Copilot seat trajectory and enterprise AI pricing updates — with paid seats already past 20 million, continued momentum is the metric bulls most want to see sustain into FY2027

Microsoft is the most expensive of the four mega-cap AI bets I track, and the fact that the stock still sits near $414 — below where it was into the April print — tells you the market is pricing the limits of capex-funded growth in real time. I stay patient here. The 20-million Copilot seat milestone is the data point I care about most: it shows the AI layer is becoming software revenue, not just a compute bill. The 27% OpenAI stake is genuinely valuable, Azure's 40% is holding, and Suleyman's superintelligence push attacks the one structural risk that has dogged this name since 2023. But at roughly 31× forward you are paying for execution that still has to land. My add-trigger is the July 28 print showing fiscal-2027 capex decelerating as guided — that would mean Microsoft has found the ceiling on its AI commitment. I would buy more aggressively in the $400-410 zone, where the multiple slips toward 30×. The thesis breaks if Azure growth falls below 30% before fiscal 2027.