Microsoft (MSFT)

Investment Review After Q3 FY2026 Earnings (stock price as of 2026.04.29)

TL;DR

Microsoft is the world’s largest software platform — Azure, M365 + Copilot, Windows, GitHub, LinkedIn. Q3 FY2026: revenue $82.9B (+18% YoY), AI ARR $37B+ (+123%), Microsoft Cloud $54.5B (+29%), Azure +40% cc, Copilot 20M+ paid seats. Stock $424 at 25x P/E(ttm). Main risk: ~$190B CY2026 CapEx pressures FCF/OP to 49% with an open AI ROIC question.

Summary

Finding	Assessment
5-Year Stock Return	Stock currently $424, vs. $210 (Q1 2021, 5 years ago). +102% over 5 years. TTM OP +163% ($57B→$149B). P/E(ttm) compressed 31x→25x. FCF/OP 49% as CapEx ramps to $190B/yr.
Business Model	STRONG. World’s largest software company. Three segments: Productivity (43%), Intelligent Cloud (38%), More Personal Computing (19%). $282B FY2025 revenue. Microsoft Cloud $54.5B in Q3 (record). 20M+ M365 Copilot paid seats, 400M+ M365 base, 1.6B Windows devices.
Competitive Moat	WIDE. Uniquely controls both distribution (Office 400M+, Windows 1.6B devices) AND AI layer (Copilot 20M, Foundry multi-model, GitHub, Agent 365). OpenAI restructured with royalty-free IP through 2032 + rev share through 2030 (rev-share to OpenAI eliminated). Multi-model Foundry (OpenAI, Anthropic 2x QoQ, open source 5,000+ customers) widens platform leverage.
Past 5-YR Drivers	Azure cloud acceleration (+40% cc Q3 FY2026), M365 Copilot ARPU expansion (20M+ paid seats, +250% YoY adds), operating leverage (OM 41%→46%), Activision integration, GitHub Copilot enterprise subs nearly tripled, AI business $37B+ ARR (+123%).
Recent Stock Move	Stock currently $424, vs. $370 (Mar 2026, last quarter end). Round trip: $375 (Mar 2025) → $518 peak (Sep 2025) → $370 trough (Mar 2026) → $424 on Q3 beat. P/E(ttm) re-rated 23x→25x on Azure +40% cc and Copilot 20M seats.
Revenue Trend	DECELERATING Rev YoY: 18% (Q3 FY2026) → 16% (Q4 2026E).
Margin Trend	FLAT OM%: 46% (Q3 FY2026) → 46% (Q4 2026E).
TTM OP Trajectory	FLAT TTM OP QoQ: 4% (Q3 FY2026) → 4% (Q4 2026E). TTM OP $149B.
Key Opportunities	Azure +40% cc with demand > supply + M365 Copilot 20M seats (+250% YoY adds) + Agent 365 cross-cloud control plane. Foundry 300+ customers on track for 1T+ tokens; Cosmos DB +50%; Fabric 35k customers (+60%). Maia 200 (+30% tokens/$) and Cobalt CPU lowering AI cost. User+usage monetization expanding TAM.
Key Risks	~$190B CY2026 CapEx (incl ~$25B from component price) raises ROIC question; ~2/3 short-lived GPU/CPU. FCF/OP fell 67%→49%, projects 34% in Q4. AI commoditizes Office/Windows surface layer just as it monetizes Copilot. Windows OEM -high teens guided in Q4. Capacity constrained through 2026.
Catalysts	Q4 FY2026 guided $86.7-87.8B revenue (+13-15%); Azure +39-40% cc; FY2026 OM up ~1pp YoY; FY2027 guided to double-digit revenue + OI growth. Modest Azure acceleration expected H2 CY2026. Q4 CapEx >$40B; CY2026 ~$190B. M365 Copilot seat trajectory beyond 20M is the ARPU swing variable.
P/E(ttm) Valuation	FAIR at 25x (5-year range: 22x-35x). Below mid-range despite 18% revenue growth and 46% OM. Market still discounting CapEx and FCF risk. If Azure stays >40% cc and Cloud GM stabilizes, P/E re-rates to 28-30x. If AI investment ROI disappoints or capacity constraints persist, 22x.

Q3 FY2026 Key Data

Revenue $82.9B (+18% YoY, +15% cc); adj EPS $4.27 (+21%); OM 46%; net income $31.8B (+23%). Microsoft Cloud record $54.5B (+29%). AI ARR $37B+ (+123%). FCF $15.8B; CFO $46.7B (+26%).
Productivity $35B (+17%); Intelligent Cloud $34.7B (+30%) Azure +40% cc; MPC $13.2B (-1%). Copilot 20M+ paid seats (+250% YoY adds); Dynamics 365 +22%; LinkedIn +12%; Gaming -7%.
CapEx $31.9B (down sequentially); ~2/3 short-lived GPU/CPU; 1 GW added; Maia 200 live in Iowa/Arizona. Q4 CapEx >$40B (incl ~$5B from component prices); CY2026 ~$190B (incl ~$25B prices).

Key Takeaways from Earnings Call

AI monetizing at scale: $37B+ ARR (+123%), Copilot weekly engagement = Outlook, GitHub Copilot enterprise subs nearly tripled. Satya: “daily habit of intense usage”. Per-seat → user+usage shift; GitHub Copilot usage-based June 1.
Azure +40% cc again exceeded guidance; demand still exceeds supply; modest H2 CY2026 acceleration expected. 1 GW added, Wisconsin Fairwater DC online 6 weeks early; dock-to-live time -20% YTD; inference +40%. RPO $627B (+99% YoY).
CapEx ramps to ~$190B CY2026 (incl ~$25B component price); ~2/3 still short-lived — ROIC question dominated Q&A. Amy Hood: TAM expansive, RPO + Copilot acceleration drive confidence. FY2027 guided to double-digit revenue + OI growth.

Key Segments and Revenue Mix (FY2025)

Segment	Revenue	% of Total	Key Details
Productivity & Business Processes	$120,810M	43%	M365 Commercial +14%, Consumer +11% (89M subs), LinkedIn +9%, Dynamics 365 +19%; 60% OM
Intelligent Cloud	$106,265M	38%	Azure +34%, Server products -3%; Foundry, Fabric ($2B+ run rate); 42% OM; cloud GM declining on AI investment
More Personal Computing	$54,649M	19%	Windows OEM +2%, Gaming +9% (Activision), Search +8%; 27% OM
Total FY2025	$281,724M	100%	+15% YoY. OP $128.5B (+17%). Net income $95.8B. Microsoft Cloud 69% GM. 1B Windows 11 users.

Business Model

SUPPLIERS

NVIDIA, AMD (GPUs/CPUs)
OpenAI, Anthropic (AI models, partners)
Data center builders & energy
Content creators (LinkedIn, Gaming)

Infrastructure / Models →

MICROSOFT

Azure cloud + AI Foundry
M365 + Copilot (20M+ Copilot seats)
GitHub Copilot (140k orgs)
Dynamics 365 + LinkedIn
$31.9B Q3 CapEx; $190B CY2026

Cloud / SaaS / Licenses → $ Revenue

CUSTOMERS

Enterprise (Fortune 500)
SMBs (M365, Dynamics)
Developers (GitHub, Azure)
Consumers (Xbox, Windows, M365)
OpenAI, Anthropic (cloud infra)

Moat Assessment: WIDE

MSFT uniquely controls distribution (Office 400M+, Windows 1.6B devices) AND AI layer (Copilot, Foundry, GitHub, Agent 365). No software peer has this combo at scale. Foundry hosts OpenAI + Anthropic + open-source + first-party MAI models.
OpenAI restructured: rev share through 2030 + royalty-free IP through 2032; rev-share to OpenAI eliminated. Satya called it “win-win”. Multi-model: Anthropic 2x QoQ; 5k+ open-source customers; MAI-Transcribe-1 + MAI-Image-2 power Bing/PowerPoint.
Key question: can Copilot justify “user + usage” pricing at scale? Per-seat → seats-plus-consumption: 60% of CS customers on usage credits; GitHub Copilot usage-based June 1; HSBC -30% resolution; LinkedIn Talent $450M+ ARR.

Stock vs. Earnings — 5-Year Performance

Stock currently $424, vs. $210 (Q1 2021, 5 years ago). +102% over 5 years. Peaked $518 (Sep 2025), troughed $370 (Mar 2026, -29%), recovered to $424 on Q3 beat. P/E(ttm) compressed 31x→25x.
TTM OP grew $57B → $149B (+163%) on Azure AI acceleration and margin leverage. Microsoft Cloud hit $54.5B in Q3 (+29%); AI business surpassed $37B ARR (+123%). OM expanded 41%→46%.
FCF/OP fell to 49% as CapEx surged to $31.9B/quarter (Q4 >$40B; CY2026 ~$190B). ~2/3 of CapEx is short-lived GPU/CPU. CFO $46.7B (+26%) in Q3; $10.2B returned via dividends + buybacks; Debt/EBITDA 0.3x.

$Chart 1: Stock vs TTM OP & FCF$

Past 5-Year Key Drivers

Azure accelerated to +40% cc in Q3 FY2026 with demand exceeding supply. Intelligent Cloud $34.7B (+30%), 40% OM. Foundry: 300+ customers on track for 1T+ tokens (+30% QoQ); Fabric 35k paid customers (+60% YoY); Cosmos DB +50%.
M365 Copilot reached 20M+ paid seats; seat adds +250% YoY (fastest since launch). AI ARR $37B+ (+123%). Customers with 50k+ seats quadrupled YoY (Accenture 740k). Copilot weekly engagement equals Outlook. M365 commercial cloud +19%.
OM expanded 41%→46% despite massive AI infrastructure spend. Cloud efficiency, OpEx discipline, headcount declining YoY. FY2026 OM up ~1pp YoY despite ~$900M Q4 retirement-program one-time. FY2027 guided to double-digit revenue + OI growth.

Quarterly Key Metrics

USD	Q3'24	Q4'24	Q1'25	Q2'25	Q3'25	Q4'25	Q1'26	Q2'26	Q3'26	Q4'26
Quarter End	Mar 2024	Jun 2024	Sep 2024	Dec 2024	Mar 2025	Jun 2025	Sep 2025	Dec 2025	Mar 2026	Jun 2026
Stock Price	$425	$457	$430	$422	$375	$497	$518	$484	$370	$424
Stock QoQ	+14%	+8%	-6%	-2%	-11%	+33%	+4%	-7%	-23%	+15%
Rev YoY	+17%	+15%	+16%	+12%	+13%	+18%	+18%	+17%	+18%	+16%
OM%	45%	43%	47%	45%	46%	45%	49%	47%	46%	46%
R40 (Rev YoY + OM%)	+62%	+58%	+63%	+58%	+59%	+63%	+67%	+64%	+65%	+62%
TTM OP ($M)	105,762	109,433	113,090	117,711	122,130	128,528	135,937	142,559	148,957	155,440
TTM OP QoQ	+5%	+3%	+3%	+4%	+4%	+5%	+6%	+5%	+4%	+4%
FCF/OP ttm	67%	68%	64%	59%	57%	56%	57%	54%	49%	34%
Debt/EBITDA	0.6x	0.5x	0.4x	0.4x	0.4x	0.3x	0.3x	0.3x	0.3x	0.3x
P/E(ttm)	33x	34x	31x	30x	26x	34x	34x	31x	23x	25x

Recent Stock Performance & Drivers

Stock currently $424, vs. $370 (Mar 2026, last quarter end). Wild round trip: $375 (Mar ’25) → $518 peak (Sep ’25) → $370 trough (Mar ’26) → $424 on Q3 beat. P/E(ttm) re-rated 23x→25x.
Cash flow risk: FCF/OP at 49% in Q3 FY2026, well below the 60% threshold. CapEx $31.9B in Q3 (down seq); Q4 guides >$40B; CY2026 ~$190B. Two-thirds short-lived. FCF $15.8B vs CFO $46.7B.
TTM OP grew $106B → $149B (+41%) over 8 quarters. OM 46% in Q3; FY2026 OM guided up ~1pp YoY despite ~$900M Q4 retirement-program one-time costs. Cloud efficiency gains continue offsetting AI infra costs.

Opportunities

Microsoft Cloud record $54.5B (+29%); Azure +40% cc with demand > supply. Modest H2 CY2026 acceleration expected. 1 GW added; Maia 200 +30% tokens/$. 300+ Foundry customers on track for 1T+ tokens (+30% QoQ); Cosmos DB +50%.
M365 Copilot inflected: 20M+ paid seats, seat adds +250% YoY (fastest since launch). 50k+ seat customers quadrupled YoY (Accenture 740k; Bayer/J&J/Mercedes/Roche 90k+). Weekly engagement = Outlook. GitHub Copilot enterprise subs nearly tripled.
User-plus-usage monetization opens consumption upside that doesn’t flow through bookings. 60% of CS customers on usage credits; GitHub Copilot usage-based June 1; tens of millions of agents in Agent 365. HSBC -30%; LinkedIn Talent $450M+ ARR.

Challenges

~$190B CY2026 CapEx (incl ~$25B component); ~2/3 short-lived — largest infra build in corporate history. Q4 alone >$40B. ROIC question dominated Q&A; Hood defense: RPO $627B (+99%) + Copilot seat acceleration.
FCF/OP fell 67%→49% in 8 quarters; Q4 2026E projects 34% as CapEx tops $40B. FCF only $15.8B in Q3 vs CFO $46.7B as cash CapEx absorbs most operating cash. Capacity constrained at least through 2026.
AI is double-edged for core franchises. Office UI complexity, Windows lock-in, and licensing switching costs face the same AI commoditization Microsoft wields against rivals. Windows OEM guided -high teens in Q4 (memory price + Windows 10 EOS lap).

Next 12 Months

Q4 FY2026 guided: revenue $86.7-87.8B (+13-15%); Azure +39-40% cc; Productivity +12-13%; Intelligent Cloud +27-28%; MPC $11.75-12.25B. COGS $29.4-29.6B (+22-23%); OpEx $19.3-19.4B (+~7%). FY2026 OM up ~1pp YoY. Microsoft Cloud GM ~64%.
Azure capacity remains constrained through 2026; modest H2 CY2026 acceleration expected. CapEx Q4 >$40B; CY2026 ~$190B. A beat with Azure >40% cc and Cloud GM stabilizing could push P/E from 25x toward 28-30x.
FY2027 guided to another year of double-digit revenue + OI growth despite continued AI investment. OpEx mid-to-high single digits; headcount declines YoY. Lapping Windows 10 EOS, OEM inventory, Office/server transactional comps lowers PBP/MPC bar.

Next 12-24 Months

Azure supply-demand to ease late CY2026/CY2027 as ~$190B CapEx delivers capacity. MSFT on track to double footprint in 2 years. Maia 200, Cobalt CPU expanding. Azure could sustain >40% and FCF/OP recover toward 50%+.
OpenAI restructured: rev share through 2030, royalty-free IP through 2032; rev-share to OpenAI eliminated. Locks in frontier access while MSFT diversifies (Anthropic +2x QoQ) and develops first-party MAI models. Reduces a long-term margin headwind.
User-plus-usage monetization is the path to re-rating. If consumption + outcome-based pricing scales (HSBC, LinkedIn Talent, GitHub Copilot usage-based, Agent 365), MSFT gains an Azure-style consumption layer. P/E could re-rate 25x→28-32x.

Core Metrics Scorecard

	Q1'22	Q2'22	Q3'22	Q4'22	Q1'23	Q2'23	Q3'23	Q4'23	Q1'24	Q2'24	Q3'24	Q4'24	Q1'25	Q2'25	Q3'25	Q4'25	Q1'26	Q2'26	Q3'26	Q4'26
Quarter End	Sep21	Dec21	Mar22	Jun22	Sep22	Dec22	Mar23	Jun23	Sep23	Dec23	Mar24	Jun24	Sep24	Dec24	Mar25	Jun25	Sep25	Dec25	Mar26	Jun26
R40 > 40%	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅
R40 accelerate	✅				✅		✅	✅	✅	✅	✅		✅		✅	✅	✅		✅
TTM OP QoQ > 10%
TTM OP QoQ accelerate							✅	✅	✅					✅		✅	✅
TTM FCF/OP > 60%	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅
Debt/EBITDA < 3.5	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅
Total passes	4	3	3	3	4	3	5	5	5	4	4	3	4	4	4	5	5	3	4	3
Stock QoQ	+4%	+19%	-8%	-17%	-9%	+3%	+20%	+18%	-6%	+15%	+14%	+8%	-6%	-2%	-11%	+33%	+4%	-7%	-23%	+15%
P/E(ttm)	31x	35x	31x	25x	22x	23x	27x	31x	28x	30x	33x	34x	31x	30x	26x	34x	34x	31x	23x	25x

R40 >40% has passed every quarter for 20 quarters straight — the structural quality marker. R40 in 58-67% band, peaking 67% in Q1 FY2026. 15-18% revenue + 45-49% OM unmatched in large-cap tech. TTM OP QoQ >10% never passes; growth steady 3-6% QoQ.
Strongest stock rallies coincided with 5-pass readings. Q3-Q4 FY2023 (5 passes): +20% / +18% QoQ. Q4 FY2025 (5 passes): +33%. Q3 FY2026 = 4 passes lined up with the +15% rebound to $424.
R40 acceleration toggling is the swing factor for re-rating from 25x. Q3 FY2026 R40 stepped up to 65%; Q4 2026E projects 62% (deceleration). Sustained Azure +40% cc + 46% OM keeps R40 >60%; further acceleration is the swing variable for 5 passes.

Chart: R40 vs Stock Price (Since 2020)

Chart: Scorecard Total Passes vs Stock Price (Since 2020)

Source: SEC EDGAR annual and quarterly filings, company earnings conference calls and presentations.

Disclaimer: This report is for educational purposes only. NO investment advice.