MSFT | Earnings Review

NASDAQ: MSFT | Azure re-accelerates to +43% (guide ~45% cc) on a $90B double beat; CapEx optics ease to ~$175B CY26 while demand still exceeds supply

Revenue beat / miss %

+2.7%

$90.0B vs ~$87.6B LSEG · +17.7% YoY · largest rev surprise in L10 · +3.2% vs company guide mid

EPS beat / miss %

+11.8%

$4.74 adj vs ~$4.24 · ~+5.4% ex +$0.27 discretes (Anthropic / VRP / Xbox) · GAAP $4.81 (+31.8% YoY)

Guidance vs consensus

Azure ~45%

Q1 FY27 rev mid $90.4B vs ~$89.7B · Azure ~45% cc vs ~41% Street · CapEx >$50B vs ~$56B VA

Growth trajectory

Azure accel

Azure 39→43% actual → ~45% guide · consolidated high-teens plateau (+17.7%) · Cloud GM −139/−300 bps

Executive summary / What is new

Double beat with the largest revenue overshoot in 10+ quarters, driven by Azure capacity/efficiency monetization. Microsoft printed $90.0B revenue (+17.7% YoY) vs ~$87.6B Street (+2.7%) and vs April guide mid ~$87.3B (+3.2%). Adj EPS $4.74 beat ~$4.24 by +11.8%; ex +$0.27 discretes (Anthropic $3.2B gain, VRP benefit, Xbox charges) the operational beat is still ~+5.4%. Azure + other cloud hit 43% reported / 43% cc vs 39–40% company guide (+300–400 bps). Intelligent Cloud $39.3B (+31.6%) was the primary dollar beat driver; Microsoft Cloud $59.3B (+27%).

What is new vs last quarter: Azure re-accelerated (FQ3 40% → FQ4 43%) and Q1 FY27 is guided ~45% cc — cleanly above Street ~41%. CapEx $41.0B landed in line with “over $40B”; Q1 guided >$50B after a useful-life change (DC/office 15→25 years) that restates reported CY2026 CapEx to ~$175B from prior ~$190B with investment intent unchanged. Copilot paid seats >30M (net adds more than doubled QoQ); E7 early traction (EY 400k); commercial RPO $678B (+84% incl. OpenAI / +25% ex-OpenAI). FY2026 full year: rev $331.8B (+17.8%), OI $155.2B (+20.8%), OPM 46.8% (+~120 bps), Azure annual >$100B (+41%).

Tone: More confident / aggressive on near-term Azure than the April call; transparent on GM mix and PC drag; CapEx ROI vs year-ago was the soft Q&A answer. Contradictions (9, 3 high): CapEx “moderate” promise reverse; GPU pre-sold vs slow-down optionality flip; 15→25yr DC life restating CapEx optics. Near-term catalysts: Azure 45% delivery + H1 acceleration (next print 2026-10-28); CapEx/Cloud GM stability at 65%; Copilot/E7/consumption ARPU; RPO quality ex-OpenAI; Ignite Nov 17–20.

Total revenue	$90,007M (+17.7% YoY)	Operating income	$40,603M (+18.3% YoY)
Op. margin	45.1% (+~20 bps YoY)	Gross margin	67.2% (−139 bps YoY)
Azure + other cloud YoY	43% reported / 43% cc	Intelligent Cloud	$39,306M (+31.6%)
Productivity & BP	$37,847M (+14.3%)	More Personal Computing	$12,854M (−4.4%)
Microsoft Cloud	$59.3B (+27%); Cloud GM 65%	GAAP diluted EPS	$4.81 (+31.8% YoY)
Non-GAAP / adj EPS	$4.74 (+23% OpenAI-adj per mgmt)	CapEx (company total)	$41.0B; Q1 guide >$50B
Commercial RPO (call)	$678B (+84% / +25% ex-OpenAI)	Q1 FY27 rev guide mid	$90.4B (+16–17%); Azure ~45% cc

FY2026Q4 reported 2026-07-29 (quarter ended 2026-06-30). Fundamentals: Daloopa company_id 135 with inline daloopa.com/src/{id} citations. Street: LSEG / Visible Alpha / FactSet via public secondary sources (Bloomberg/VA MCP unavailable this run). Management tone and guidance from MSFT FY2026Q4 earnings call transcript in the run workspace. Internal SharePoint/OneNote/Outlook unavailable.

Key metrics trends (12 quarters)

Trajectory: accelerating at the Azure/Intelligent Cloud core, consolidated revenue holding a high-teens plateau, gross margin compressing on AI infrastructure mix. Segment YoY for Productivity / Intelligent Cloud is clean FY26 vs FY25; FY25 vs FY24 is distorted by the FY25Q1 reportable-segment reclassification.

Revenue drivers → consolidated

Metric	FY24Q1	FY24Q2	FY24Q3	FY24Q4	FY25Q1	FY25Q2	FY25Q3	FY25Q4	FY26Q1	FY26Q2	FY26Q3	FY26Q4
Productivity & BP ($B)	18.6	19.2	19.6	20.3	28.3	29.4	29.9	33.1	33.0	34.1	35.0	37.8
PBP YoY	+13%	+13%	+12%	+11%	+52%*	+53%*	+53%*	+63%*	+17%	+16%	+17%	+14%
Intelligent Cloud ($B)	24.3	25.9	26.7	28.5	24.1	25.5	26.8	29.9	30.9	32.9	34.7	39.3
IC YoY	+19%	+20%	+21%	+19%	−1%*	−1%*	0%*	+5%*	+28%	+29%	+30%	+32%
Azure + other cloud YoY	29%	30%	31%	29%	33%	31%	33%	39%	40%	39%	40%	43%
Microsoft Cloud ($B)	31.8	33.7	35.1	36.8	38.9	40.9	42.4	46.7	49.1	51.5	54.5	59.3
Cloud GM %	73%	72%	72%	69%	71%	70%	69%	68%	68%	67%	66%	65%
MPC ($B)	13.7	16.9	15.6	15.9	13.2	14.7	13.4	13.5	13.8	14.3	13.2	12.9
Total revenue ($B)	56.5	62.0	61.9	64.7	65.6	69.6	70.1	76.4	77.7	81.3	82.9	90.0
Rev YoY	+12.8%	+17.6%	+17.0%	+15.2%	+16.0%	+12.3%	+13.3%	+18.1%	+18.4%	+16.7%	+18.3%	+17.7%
Gross margin %	71.2%	68.4%	70.1%	69.6%	69.4%	68.7%	68.7%	68.6%	69.0%	68.0%	67.6%	67.2%
Op. income ($B)	26.9	27.0	27.6	27.9	30.6	31.7	32.0	34.3	38.0	38.3	38.4	40.6
GAAP dil. EPS	$2.99	$2.93	$2.94	$2.95	$3.30	$3.23	$3.46	$3.65	$3.72	$5.16	$4.27	$4.81

* FY25 segment YoY distorted by FY25Q1 reclassification. Azure/Cloud GM/consolidated are clean. Data sourced from Daloopa.

YoY growth bars — revenue, Azure, Intelligent Cloud

Total revenue YoY % (FY25Q1–FY26Q4 + Q1 FY27 guide)

16.0

12.3

13.3

18.1

18.4

16.7

18.3

17.7

16–17g

Q1'25

Q1'26

Q4★

Q1'27g

Azure reported YoY % — re-acceleration path

~45g

Q1'25

Q1'26

Q4★

Q1'27g

Why the trends: (1) Azure is the second-derivative story — multi-year high at 43% with guide to ~45% cc. (2) Consolidated high-teens is Azure + commercial cloud mix, not broad-based (MPC is a multi-quarter drag). (3) Gross margin compression is the price of the AI land-grab (Cloud GM 72–73% → 65%). (4) FY2026 annual shape is a re-acceleration year after the FY2023 trough (+6.9%): rev +17.8%, Microsoft Cloud +26.9% to $214.4B.

Beat / miss analysis

Classification: Consistent Beater — L4/L8/L10 = 100% rev + EPS. Revenue beat magnitude improving (+0.9% → +1.2% → +1.8% → +2.7%). EPS headline inflated by discretes; clean EPS still solid mid-single-digit beat.

This quarter vs Street / guide

Metric	Consensus / Guide	Actual	Variance	Read
Revenue	~$87.62B LSEG guide mid ~$87.25B	$90.0B	+$2.39B / +2.7% +3.2% vs guide	Beat ★ peak L10
Adj / call EPS	~$4.24	$4.74	+$0.50 / +11.8% ~+5.4% ex discrete	Beat (partly discrete)
GAAP dil. EPS	—	$4.81	+31.8% YoY	Statutory
Azure YoY	39–40% guide / ~40% Street	43% / 43% cc	+300–400 bps	Beat / re-accel
Intelligent Cloud	~$38.1–38.2B	$39.3B	+~$1.1–1.3B	Primary $ driver
Cloud GM	Feared mid-60s pressure	65%	−300 bps YoY	Better than feared, still YoY compress
CapEx	>$40B guide	$41.0B	In line	Demand/execution beat, not CapEx undershoot

Mgmt variance drivers: Azure fleet efficiency + earlier capacity delivery “quickly monetized”; GitHub Copilot usage pricing; M365 commercial / on-prem recognition; MPC inventory build above guide; Cloud GM better than expected at 65%.

Heatmap — last 8 quarters (★ = FY26Q4)

Metric	FQ1'25	FQ2'25	FQ3'25	FQ4'25	FQ1'26	FQ2'26	FQ3'26	FQ4'26 ★
Revenue	BEAT	BEAT	BEAT	BEAT	BEAT	BEAT	BEAT	BEAT ★
EPS	BEAT	BEAT	BEAT	BEAT	BEAT	BEAT	BEAT	BEAT ★
Rev surprise %	+1.7%	+0.9%	+0.4%	+0.5%	+0.9%	+1.2%	+1.8%	+2.7% ★
EPS surprise %	+6.5%	+2.5%	+4.8%	+4.3%	+5.9%	+6–7%	+5.4%	+11.8% ★
Azure vs guide	beat	in-line	beat high	beat	in-line	in-line/−1pp	+1pp	+3–4pp ★

Actuals: Daloopa. Street consensus from LSEG / StreetAccount / MarketBeat public aggregations (not Bloomberg/VA).

Guidance deep dive

Headline: Q1 FY27 guide is a clean raise vs Street — rev mid $90.4B (+16–17%) and Azure ~45% cc (vs ~41%) — while CapEx >$50B Q1 and $175B CY26 (lease reclass) eases spend optics. FY27 framed as double-digit rev and OI with OPM down <1 pt.

Q1 FY27 company guide

Metric	Low	High	Mid	vs Street
Total revenue	$89.85B	$90.95B	$90.40B (+16–17%)	+$0.74B vs ~$89.66B
Azure cc	—		~45%	~+410 bps vs ~40.9%
Intelligent Cloud	$40.95B	$41.25B	$41.10B (+33–34%)	Above Street tone
Productivity & BP	$36.7B	$37.0B	$36.85B (+11–12%)	—
MPC	$12.2B	$12.7B	$12.45B (down YoY)	Wide PC range
COGS	$29.6B	$29.8B	$29.7B → GM ~67.1%	—
OpEx	$16.8B	$16.9B	$16.85B (+7–8%)	Implied OI ~$43.85B / OPM ~48.5% flat YoY
CapEx	—		>$50B	Below VA ~$56B (~−$6B optics)
ETR	—		~20%	—

Guidance waterfall

Q1 FY27 revenue guide vs Street ($B)

Street (LSEG)

89.66

Δ vs Street

+0.74

Guide mid

90.40

Guide high

90.95

CY2026 CapEx reported ($B)

Prior guide

190

Lease reclass

−15

New reported

175 (intent unchanged)

FY27 full-year color

| Metric | Frame | |---|---| | Revenue / OI growth | Double-digit both | | OpEx | Mid- to high-single digits | | Operating margin | Down <1 pt YoY | | FCF | Remain free-cash-flow positive | | CapEx | Grow YoY; CY26 reported ~$175B | | Windows OEM & Devices | High-teens revenue decline FY | | Xbox | Reset; return to growth in FY27 (Q1 still guided decline) |

Tone vs April call: More confident on Azure acceleration and AI product monetization; same conviction / cleaner optics on CapEx; honest on mix pressure (not defensive). Sequential N+1 growth quality improved (prior Q4 guide 13–15% → Q1 guide 16–17%).

Historical performance (inflection points)

Revenue re-accelerated from a +12.3% trough (FY25Q2) into a durable high-teens band (~17–18.5%) held for five consecutive quarters (FY25Q4–FY26Q4). OpenAI-adj EPS holds mid-20s% growth; GAAP EPS is noisy.

8-quarter trajectory

| Metric | FY25Q1 | FY25Q2 | FY25Q3 | FY25Q4 | FY26Q1 | FY26Q2 | FY26Q3 | FY26Q4 | |---|---:|---:|---:|---:|---:|---:|---:|---:| | Revenue ($M) | $65,585 | $69,632 | $70,066 | $76,441 | $77,673 | $81,273 | $82,886 | $90,007 | | Rev YoY | +16.0% | +12.3% | +13.3% | +18.1% | +18.4% | +16.7% | +18.3% | +17.7% | | Rev accel (bps QoQ) | +85 | −377 | +100 | +483 | +33 | −171 | +158 | −55 | | GAAP EPS | $3.30 | $3.23 | $3.46 | $3.65 | $3.72 | $5.16 | $4.27 | $4.81 | | Non-GAAP EPS | — | — | — | — | $4.13 | $4.14 | $4.27 | $4.74 |

Inflection points

FY25Q2 trough: rev YoY +12.3% (−377 bps) — softest print in the window.
FY25Q3→Q4 primary accel: +13.3% → +18.1% (+483 bps) as Azure stepped to high-30s.
FY26 plateau: five quarters in the high-teens band; Q4 −55 bps rate wiggle with absolute record $90B.
Azure step-function: low/mid-30s → 43% (guide ~45%); Intelligent Cloud clean YoY +28% → +32%.
EPS: ignore FY26Q1 air pocket and FY26Q2 OpenAI spike; OpenAI-adj mid-20s% re-accelerating modestly into Q4 (+21% → +23% mgmt frame).

Near-term rate path: Q1 FY27 rev guide +16–17% is a mild ~100–150 bps headline deceleration on MPC/PC comps, while Azure still accelerates — mix quality up, consolidated % slightly down.

Key catalysts

| # | Catalyst | Timing | Watch | Read | |---|---|---|---|---| | 1 | Azure ~45% delivery + H1 acceleration | FY27Q1 print 2026-10-28 | Prove 45% durable, not one-quarter efficiency spike | Highest multiple sensitivity | | 2 | CapEx intensity vs ROI / FCF | Each print; useful-life change FY27 | Cash PP&E + op leases, not just CapEx line; Q1 >$50B | Street fears eased by $175B optics | | 3 | Copilot / E7 / seats+consumption | Ongoing FY27 | Seat adds, E7 mix, Cowork/GitHub usage $ | Bridge from CapEx to ARPU/TAM | | 4 | RPO quality ex-OpenAI | Quarterly | +25% ex-OpenAI; bookings volatility flagged | Demand breadth under multi-cloud OpenAI | | 5 | Multi-model platform + OpenAI economics | Structural | Foundry mix; mark-to-market noise | Reduces exclusivity narrative risk | | 6 | Capacity / silicon (Maia, Cobalt, Vera Rubin, Helios) | Multi-quarter | Dock-to-live, GW adds | Supply still binding constraint | | 7 | Agent 365 / Foundry / Perception | Ignite Nov 17–20, 2026 | Product attach narrative | Narrative upside, not one-Q earnings | | 8 | Xbox / PC optics | FY27 | Xbox return-to-growth claim; OEM high-teens decline | Guided drag, not thesis core |

Priority stack (1–2Q): Azure 45% + H1 accel → CapEx / Cloud GM stability at 65% → Copilot/E7 ARPU → RPO ex-OpenAI → Ignite/PC secondary.

Street Q&A

Scorecard: 6 analysts / 8 scored parts — 6/8 Well Answered (75%); 2/8 Deflected (CapEx ROI YoY; open-model materiality). Short 6-question call, hard stop, no second round.

Q1 — Keirstead (UBS): open/custom model materiality + how MSFT benefits vs frontier-lab exposure

Nadella: harness separable from model; multi-provider customers 5×; Hood: Azure infrastructure “pretty fungible.” Strong platform thesis; no quantification of open-model share/timeline.

Part 2 well answered · Part 1 deflected (materiality)

Q2 — Thill (Jefferies): Azure 43% — still capacity constrained or just better execution?

Hood: demand still exceeds supply; beat from fleet efficiency + dock-to-live process — “quickly monetized.” Linked to ~45% Q1 guide and H1 still accelerating.

Well answered

Q3 — Moerdler (Bernstein): overbuild risk + component inflation vs margins/pricing

Hood: ~2/3 CapEx short-lived — “if demand changes, just slow down”; late-bind expensive components; RPO growth outside frontier labs; new contracts can reflect pricing. Strong framework, no numeric sensitivity.

Well answered (framework) / low on numbers

Q4 — Wood (MS): Copilot pilots vs broad deploy; seats vs E7 vs consumption

Nadella: time-to-high-usage months→days; engagement parity with Outlook/Teams; seat + usage model. Hood: >30M paid seats; E7 early traction; M365 cloud acceleration expected through FY27. Best commercial answer of the call.

Well answered

Q5 — Zelnick (DB): Project Perception / cyber in agentic era

Nadella: multi-agent red/blue/green; multi-model cost/resilience (MAI-Cyber ~90% of tasks); consumption offering as Perception exits private preview. Product-strong, no Security $ ARR bridge.

Well answered (product) / low on P&L

Q6 — Borges (GS): CapEx ROI vs a year ago; remaining levers

Hood declined comparative math (“my math has changed”); substituted TAM confidence + levers (1P silicon, model diversification, portfolio mix). Weakest answer on the call — no IRR/payback/rev per CapEx $.

Deflected / avoided

Dogs that didn’t bark: OpenAI RPO concentration, Anthropic $3.2B mark, Xbox recovery path, useful-life 15→25 challenge, Cloud GM floor, FCF under >$50B CapEx, PC cycle depth.

Contradictions

9 found (3 high) — CapEx/risk-narrative conflicts dominate. Pattern: question-dependent framing and CapEx guidance that did not moderate as first promised.

C-1 · High

CapEx “will moderate / decline YoY” → growth “higher than FY25” → CY26 ~$190B

FY25Q4: CapEx growth “will moderate” / “growth rate will decline YoY.” FY26Q1: “FY26 growth rate higher than FY25.” FY26Q3: CY26 ~$190B (incl. ~$25B component price). Not soft clarification — same growth-rate language reversed within one quarter. Guidance-accuracy flag for FCF/ROIC underwriters.

C-2 · High

GPUs “already sold for entire useful life” vs “if demand changes, just slow down”

FY26Q2 (Moerdler): majority of GPUs “already contracted… sold for the entirety of their useful life… not the risk.” FY26Q4 (Moerdler again): short-lived assets, late-bind, “you just slow down.” Risk narrative flips with the street’s fear (duration mismatch → overbuild). Unresolved residual: who bears mid-life demand normalization.

C-3 · High

“Monetization over the next 15 years” → useful life 15→25 years; CapEx restated to ~$175B “unchanged”

Three quarters of 15-year long-lived framing, including the same Q3 call that set $190B CY26 CapEx. Q4: life extended to 25 years; more leases finance→operating; reported CapEx ~$175B while investment “unchanged.” Watch cash PP&E + lease liabilities, not CapEx alone. Q1 >$50B guide already embeds new optics.

C-4 · Med-High

Capacity-constraint end-date keeps sliding while Azure accelerates

June 2025 hope → Dec 2025 → end FY26 → through 2026 → still constrained Jul 2026 with no new terminal date. Supports demand duration bull case; CapEx/FCF normalization timelines sold in 2025 have not been earned. Treat “demand > supply” as structural until they stop saying it.

C-5–C-9 · Medium / Low

C-5: Xbox “return to growth in FY27” vs Q1 still guided mid-SD decline + Q4 −10% / impairments.
C-6: DC builds “quite flexible” vs 25-year life and multi-year site pipeline.
C-7: FY26 “expanding operating margins” → FY27 OPM down <1 pt on same call.
C-8: Efficiency narrative vs multi-quarter company GM 67% / Cloud GM 65% compression.
C-9: RPO +84% / ex-frontier sequential growth celebrated while WAD shortened 2.5→2.3 yrs.

Watch next: cash PP&E + operating-lease payments; whether Xbox actually turns; first quarter without “demand exceeds supply”; Cloud GM YoY direction.

Indirect read-throughs

Macro that matters: AI capacity shortage + component inflation + soft PCs/gaming — not a rates/CPI treatise. Demand still exceeds supply “in a relatively extreme moment”; spot asset prices elevated. Enterprise AI demand broadening (~90% cloud rev non-frontier; RPO growth outside frontier labs). CRM longer sales cycles is the main classic IT caution flag.

Named companies / ecosystem

Name	Role	Signal	Read-through
NVIDIA	Silicon supplier	Next-gen Vera Rubin; fleet modernized with latest NVIDIA	POSITIVE — hyperscaler GPU demand intact
AMD	Silicon supplier	Helios rack-scale; “among first” deployers	POSITIVE — diversifies GPU spend
OpenAI	Partner / customer	Still core; RPO +84% incl. / +25% ex; multi-cloud regime	MIXED — concentration optics vs platform breadth
Anthropic / Mistral / xAI	Model partners	$3.2B Anthropic mark; Mistral sovereign; multi-lab catalog	POSITIVE multi-lab distribution
Arm / Adobe / Elastic	Cobalt customers	Named Cobalt VM workloads	POSITIVE Arm server / Azure attach
EY / HSBC / KPMG / NHS	Copilot / E7	EY 400k E7; HSBC 200k; NHS 505k; KPMG 276k	POSITIVE mega-deal breadth
PC OEMs (HPQ/DELL…)	Windows ecosystem	Component inflation → device pricing → soft units; inventory build	NEGATIVE near-term unit growth
CRM peers (e.g. CRM)	Dynamics competitor	CRM “longer sales cycles”; ERP healthy	YELLOW — industry deal velocity
Hyperscaler peers	AWS / GCP / ORCL	Azure +43% / guide ~45%; capacity monetized	Constructive cycle signal if demand industry-wide
Power / DC infra	Supply chain	CapEx still rising; constraint open-ended	POSITIVE multi-year power/cooling demand

Key takeaways: (1) Shortage not recession is binding for Azure. (2) Component inflation dual-edged (CapEx up / PC units down / cloud TCO argument). (3) Multi-polar model layer (OpenAI + Anthropic/Mistral/xAI/MAI). (4) Consumer split: software resilient, hardware/gaming soft. (5) Constructive for NVDA/AMD next-gen; yellow flag for CRM cycle length and PC OEMs.

Bottom line

Azure is re-accelerating into FY27 with a Credible Beater print behind it; CapEx narrative risk is optics + promise-tracking, not a demand fade. Underwrite the commercial trajectory (Azure 43%→45% guide, IC +32%, Cloud $59B, Copilot >30M seats, RPO +25% ex-OpenAI) and separate the capital story (CapEx still rising, Cloud GM still compressing YoY, ROI number not disclosed, three high-severity CapEx narrative contradictions). Next proof: FY27Q1 on 2026-10-28 — Azure 45%, CapEx >$50B, Cloud GM “relatively stable” QoQ.

Data sourced from Daloopa (company_id 135). Management commentary and Q&A from MSFT FY2026Q4 earnings call transcript (2026-07-29) in tickers/MSFT/data/review_workspaces/2026-08-01/. Consensus from LSEG / Visible Alpha / FactSet public secondary reports. Internal SharePoint / OneNote / Outlook / Excel models unavailable this run. Trace tasks 1–8 remain in the review workspace (not copied).