Microsoft moved $15B of capex off the reported line — in the same quarter free cash flow fell for the first time

Microsoft had a genuinely strong quarter: Azure grew 43% and crossed $100 billion a year. But three separate accounting decisions in the same period all moved the reported numbers in the same, flattering direction — a capex figure that fell without the spending falling, a change to how long data centres are assumed to last, and $329 billion of signed leases that sit outside the capex line entirely. None of it is hidden and none of it is improper. It just means the headline number and the filing are telling slightly different stories, and only one of them is signed.

Valuation — a reasonable P/E sitting on a 1.85% cash yield

The under-discussed number is P/FCF of 54× against a P/E of 27×. The equity yields about 1.85% in free cash flow because capex is taking 63% of operating cash. On earnings the multiple asks for roughly what management guided; the stretch is entirely on cash.

Analyst reaction was constructive and unusually dispersed: 56 analysts, mean target $562.73, high $870, low $400, with 40 Strong Buy / 0 Sell and no post-print cut I could find — though BofA ($500) and Phillip ($515) raised to barely above the current price, which is where the real disagreement lives.

The compute position

Microsoft now reports capacity in gigawatts, not GPUs, and the unit change is the tell. It added ~1GW in the quarter, opened 31 datacenters, and expects to roughly double capacity in two years. Demand still exceeds supply, and management declined to say when that ends — or to name whether the binding constraint is power, shells, GPUs or land.

That reframes the silicon story. Maia 200 is an inference part, not a training part, and critically draws ~750W against Nvidia designs at 1,200W+. When your limit is a gigawatt rather than a purchase order, performance-per-watt is the capacity strategy.

Caveat the skeptic pass insisted on: every custom-silicon claim this quarter — Microsoft's "30% better performance per dollar", Amazon's Trainium comparisons, Google's TPU numbers — is vendor-published and unrefereed. MLPerf Inference v6.0 included no Maia and no current-generation TPU results. There is no neutral referee; treat all of it as directional.

The OpenAI relationship is now two-sided. Microsoft holds ~27% as-converted at $135B carrying value and OpenAI committed $250B of Azure purchases — but Microsoft gave up right of first refusal on OpenAI's compute, and OpenAI has since expanded AWS by ~$100B. Products still ship first on Azure unless Microsoft cannot supply. The exclusivity is gone.

Risk — each isolated, do not blur

1. The reclassification (central). Extending assumed asset life from 15 to 25 years makes a given lease term a smaller fraction of that life, which is the mechanism that moved finance leases to operating leases and the guide from ~$190B to ~$175B. All disclosed on the call — but the number that reached most readers was "Microsoft cut capex."
2. The off-statement obligation. $329.1B of leases signed but not commenced, roughly 3× Azure's annual revenue, commencing FY27–FY33, appearing in no capex figure. Sources disagree on the prior-period base and I could not reconcile them.
3. Earnings quality. The FY26 OpenAI line swung $8.6B, from a $3.62B loss to a $4.96B gain — a non-cash dilution gain from the recapitalisation, reversible if OpenAI's valuation compresses. A $3.2B Anthropic gain sits inside the "clean" non-GAAP figure.
4. Depreciation, deferred rather than absorbed. FY26 depreciation was $34.3B against $22.0B. The life extension pushes future depreciation out; management guided FY27 margin "down less than a point." Worth noting it covers buildings and shells, not servers or GPUs, so the usual "they stretched server lives" attack does not apply.
5. Cloud margin compression. Microsoft Cloud gross margin fell 3 points to 65% — while Amazon printed AWS margin expansion the following day. Same capex cycle, opposite direction, and the most decision-relevant comparison here.
6. Related-party concentration. OpenAI is somewhere in a 30–45% band of commercial RPO depending which disclosure you read, and Microsoft does not disclose how much Azure revenue comes from it.

Horizon and sizing (kept separate)

Horizon. Not a next-quarter setup — the stock moved ~25% in three sessions and the guide is in the price. The thesis resolves over 3–5 years: do the FY27–FY33 lease commencements convert to billed Azure revenue at a return above the depreciation just pushed out?

Sizing considerations (not a recommendation). MSFT carries the same AI-capex-cycle beta as NVDA, AMZN and GOOGL, so it adds correlation rather than diversification. Reported earnings now swing on marks on two private companies, a new and lumpy source of volatility. And a ~1.85% free-cash-flow yield leaves less cushion than a 27× P/E implies.