Microsoft
Microsoft — Q1 FY2027 scorecard
Earnings scorecard for Microsoft (NASDAQ: MSFT). Q1 FY2027 reported after the close on 28 Oct 2026, the first print under the Agents and Infra / Devices and Consumer recast. Pre-print snapshot 23 Oct: $524.51, up 34% since the July print and 5% below the $553.72 52-week high (Yahoo Finance).
What happened
A clean beat with a raised guide. Revenue $92.28B (+19%) against a $90.7B Street and the $89.85–90.95B guide; non-GAAP EPS $4.89 against $4.73 (Microsoft IR). Azure grew 48% in constant currency, three points above guide, about $33.17B. Agents and Infra $76.80B (+25%); Microsoft Cloud $63.9B (+30%).
Capex was the surprise: $39.4B including leases, over $10B below the ">$50B" guide. Hood guided Q2 to $96–97.1B and Azure 49–50%, held the FY27 capex frame at $170B, reiterated free cash flow positive. The stock closed flat at $531.01 and traded $544.50 (+2.5%) after hours, short of the $555–575 playbook zone. Grades: four bull, EPS inline one cent under $4.90. Scenario: clean beat.
What the market is pricing
The Street was at $90.7B of revenue (+17%) and $4.71–4.73 of EPS against $4.13 a year ago, in the upper half of Microsoft's $89.85–90.95B guide (MarketBeat). Azure was guided to 44–45% constant currency on the narrower September definition, BNP Paribas modelling 46%; capex "over $50 billion".
Polymarket put 46% on Azure at 48–50% and 41% on 46–48%, so traders expected a two-to-four-point beat; its EPS market gave 79% to a beat. Options implied ±5.5%; the last print rose 15.5% (Earnings Watcher). Consensus was Buy with an average target near $606, Bernstein at $662 and resistance flagged at $551. The debate was whether AI revenue scales fast enough to justify $41B of quarterly capex, with Dell's 11% fall and a Gartner survey (13% of CIOs see significant AI value) setting the mood.
Also listen for
- The recast. Azure now excludes GitHub, developer services and Security Copilot; Microsoft 365 commercial cloud guided to about 17% cc (8-K exhibit 99.1).
- Q2 Azure guide and capacity. Hood said first-half FY27 growth would accelerate; July's wording was "capacity constrained".
- Free cash flow. Q4 FCF $19.6B, down 23%; "remain free cash flow positive" is the line to watch.
- RPO ex-OpenAI. $678B backlog, up 84%, only 25% excluding OpenAI; Copilot seats (30M+) are the monetisation tell.
Reaction playbook
- Clean beat — breaks $553.72 and tests $555–575. If: Azure ≥ 48% cc, revenue ≥ $91.5B, Q2 Azure guided at or above Q1, capex frame held near $175B with free cash flow positive. Narrative: the acceleration is real and already funded.
- In line — $505–535. If: Azure 45–47% cc, EPS about $4.75, capex over $50B as flagged. Narrative: a fine quarter the stock already paid for after a 34% run.
- Miss or guide cut — probes $486 support, then the $480s. If: Azure < 45% cc, capex well above $55B with the frame raised, or the free-cash-flow line dropped. Narrative: spend still rising; AI capex reprices across the Mag 7.
The medium-term story
Fiscal Q4 2026 turned the capex argument. Revenue rose 18% to $90.0B, Azure grew 43% past $100B a year, RPO hit $678B (+84%), and Hood trimmed the capex frame to about $175B (Microsoft). The stock rose 15.5% next day. FY26 capex was $115.9B versus $64.6B.
On 2 Sep Microsoft recast into Agents and Infra ($268.1B in FY26, +23%) and Devices and Consumer ($63.7B, +1%), Azure restated to $101.9B (+40%). The bull case: the most visible AI revenue of any hyperscaler, 30M Copilot seats, Maia 300 silicon. The bear case: a 34% rerating prices the beat, RPO ex-OpenAI grows 25%, free cash flow fell 23% as capex runs $50B a quarter, OpenAI concentration is harder to see.
Track record
- Q1 FY27 (28 Oct 2026) - Clean beat. Revenue $92.28B (+19%, bull), Azure 48% cc (bull), Agents and Infra $76.80B (+25%, bull), non-GAAP EPS $4.89 (inline, one cent short of the $4.90 bull line), capex $39.4B incl. leases vs >$50B guide (bull). Q2 FY27 guided $96–97.1B revenue, Azure 49–50% cc; FY27 capex frame $170B, FCF positive reiterated. Stock +2.54% in extended hours on 28 Oct (nearly 4% at peak), 52-week high at $553.72 still unbroken.
Research log
- 2026-06-11 · ingest — Build: seven in-house MAI models including MAI Thinking 1 (1T-parameter MoE); Frontier Tuning pitch claims a model tuned for McKinsey's tasks beat GPT-5.5 on quality at 10x lower cost. Source: The Next Wave of Enterprise AI
- 2026-06-12 · ingest — Microsoft restricted employees from using Fable 5 in Copilot over Anthropic's 30-day retention policy, the highest-profile enterprise pushback of that launch. Source: Why Fable 5 is the Most Controversial AI Release Ever
- 2026-06-27 · economist — Five hyperscalers plan ~$750bn of 2026 capex, ~$3trn globally 2026–2030; Microsoft repriced its coding assistant so steeply that some programmers returned to writing software themselves. Source: Americas Data-Centre Backlash Puts the AI Boom at Risk (Economist)
- 2026-07-04 · economist — Microsoft reported to be considering DeepSeek's model for Copilot; leading Chinese models sit 6–10 months behind US equivalents and are far cheaper to run. Source: Donald Trumps AI Regime Is Opaque Unpredictable and Unsustainable (Economist)
- 2026-07-30 · ingest — FY26 Q4: revenue $90bn (+18%), Azure +43% past $100bn annualised, RPO +84% to $678bn, M365 Copilot 30M+ seats, capex $41bn (+70%), FCF $19.6bn (-23%); stock +15.5%. Figures Reddit-sourced on the page. Source: Microsoft
- 2026-07-30 · ingest — About $60bn of neocloud compute agreements sit as multi-year opex against a ~$190bn CY2026 capex guide on ~$200bn of operating cash flow; the $678bn RPO is the off-balance-sheet counterweight. Source: Hyperscaler Financial Web
- 2026-08-02 · ingest — Nadella on the FY26 Q4 call: a unified consumer-and-enterprise Copilot app ships this year; Microsoft repositions as a model-agnostic platform with over 11,000 models, MAI as the cost and privacy play. Source: 6 Questions Shaping Enterprise AI (AI Daily Brief)
- 2026-08-22 · economist — The NHS bought 500,000 Copilot licences and depends on Microsoft for email, cloud and AI; a former UK defence procurement adviser names Microsoft and Oracle, not Palantir, as the real kill-switch risk. Source: Why Everybody Hates Palantir (Economist)
- 2026-10-24 · last30days — Pre-print sweep (94 items, three Polymarket markets, fourteen web sources): the Street had bet on 46–50% Azure growth against a 44–45% guide; Azure constant-currency growth set as the swing factor. Source: Yahoo Finance MSFT
- 2026-10-29 · grade — Q1 FY2027 graded: revenue $92.28B (+19%), Azure 48% cc, capex $39.4B against a >$50B guide; four bull, EPS inline. Scenario: clean beat. Social sweep skipped for index lag. Source: Microsoft FY27 Q1 release