Prices as of 2026-09-25 close · Financials through FY26 Q4 (2026-06-30; 10-K filed 2026-07-29)
Microsoft is excellent, but at 26x earnings the good news is already in the price
The business is strong: FY26 revenue of $331.8B (+17.8%), operating income of $155.2B (+20.8%), Azure above $100B for a full year for the first time, and commercial RPO (remaining performance obligations: contracted revenue not yet recognized) at $678B (+84%). Checking the filings turned up three things to watch. GAAP profit includes a $6.5B (pre-tax) OpenAI dilution gain. The company's adjusted figures strip out only OpenAI, so the $3.2B Anthropic investment gain in Q4 stays inside adjusted EPS. And capex doubled, pulling free cash flow down from $71.6B to $67.0B, while $329.1B of data center leases that have not yet commenced sit off the balance sheet. On the market side, the Fed hiked in September and the 10-year Treasury yield has climbed to 5.18%. The stock is up 7.4% year to date, behind the S&P 500 and the tech ETF.
The market
- Indexes: As of 2026-09-25, the S&P 500 is up 13.1% year to date and 17.2% over 1 year; the Nasdaq is up 16.5% year to date. The past month shows a sharp split: Nasdaq +3.6%, Dow -3.1%, Russell 2000 -5.6%. Money is crowding into big tech.
- Sector: The tech ETF (XLK) is up 36.8% year to date. This stock is up only 7.4%, a laggard among big tech names. Over the past 3 months, though, it has gained 38.7% and beaten the S&P by 33.4 percentage points: it is catching up.
- Rates: The 10-year Treasury yield is 5.18%, up about 52bp in 1 month; the Fed raised rates 25bp on Sept 16. The VIX sits at a low 14.9, so the market is not panicking.
- What it means for this stock (our inference): The stock trades at 26.1x current-fiscal-year earnings, above the 19.4x median of the chosen peers, which makes it more sensitive to a rising discount rate. On the other side, the company holds $76.8B in cash and short-term investments and earned $3.3B of interest income in FY26 (10-K), so higher rates help that line a little. The stock's 1-year beta to the S&P is 0.99, but the correlation is only 0.39: it moves mostly on its own drivers (AI capex, OpenAI).
Indices and sector
| Last | 1W | 1M | 3M | YTD | 1Y | |
|---|---|---|---|---|---|---|
| S&P 500 | 7,743.41 | +1.2% | +0.9% | +5.3% | +13.1% | +17.2% |
| Nasdaq Composite | 27,068.72 | +2.1% | +3.6% | +7.0% | +16.5% | +20.9% |
| Dow Jones Industrial | 51,828.62 | +0.3% | -3.1% | -0.1% | +7.8% | +12.8% |
| Russell 2000 | 2,837.55 | -0.8% | -5.6% | -5.7% | +14.3% | +17.7% |
| Technology sector ETF | 196.27 | +3.6% | +7.5% | +8.5% | +36.8% | +41.9% |
| MSFT | 516.17 | +4.5% | +4.0% | +38.7% | +7.4% | +2.6% |
Rates, volatility and relative performance
Macro and market events
| Date | Event | What it means for this stock | Source |
|---|---|---|---|
| 2026-09-16 | Fed raises rates 25bp to 3.75%–4.00%, its first hike since 2023; the median dot points to one more hike in 2026 | Higher rates → higher discount rates for richly valued growth stocks (our inference). This stock trades above the peer median, so the effect leans negative; interest income on its cash gets a small lift. | CNBC |
| 2026-09 中旬 | 10-year Treasury yield rises to 5.135%, the highest since July 2007; the Dow posts its worst week since March | market.json shows the yield at 5.18% as of Sept 25. Rising long-term rates weigh on the valuations of long-duration tech stocks (our inference). | CNBC |
| 2026-09-09 | US-Iran conflict escalates, oil breaks above $100 a barrel, US stocks fall | Energy prices push up inflation → stronger expectations of further hikes (our inference). Data center power costs may also be hit, but the company's filings do not quantify this exposure, so it could not be verified. | AP / The Columbian |
Events
Press reports: Big Tech's AI build-out reaches the Middle East, and the Iran conflict hits it
Only the yfinance news headline is available; the article itself was not read. Microsoft's 10-K does not break out data center assets by region, so the size of this exposure could not be verified.
Stifel upgrades to Buy from Hold, lifts target to $575 from $530
One of the few cautious houses has turned bullish, a sentiment positive. In the same week Oppenheimer ($515 → $570) and Cantor ($522 → $608) also raised their targets (yfinance ratings data). We did not read the underlying note, so the specific reasons could not be verified.
From FY27, two reporting segments: Agents and Infra, Devices and Consumer
Totals do not change, but the Azure definition does: GitHub and Security Copilot move to M365 Cloud, so FY26 Azure growth is restated from 41% to 40% and Q4 from 43% to 42%. Any comparison with the old basis must use the restated figures. FY27 Q1 Azure guidance on the new basis is 44–45% (constant currency).
10-K discloses $329.1B of not-yet-commenced data center leases and $194.1B of purchase commitments
This is capacity built up for AI demand, and it is also future fixed cost: the leases commence in stages over FY27–FY33, and once they do they bring lease liabilities and depreciation. Whether they earn a return depends on whether Azure and AI demand hold up.
FY26 Q4 results: revenue $90.0B (+18%), GAAP EPS $4.81, adjusted $4.74
Revenue, operating income and EPS all beat company guidance; Azure grew 43% and commercial RPO reached $678B (+84%). But the company itself says discrete items lifted EPS $0.27 above guidance, including a $3.2B Anthropic investment gain. Strip that out and it is still a beat, just smaller than the headline 11.8%.
Director Reid Hoffman will not stand for re-election
The 8-K states there is no disagreement with management. A new director was also appointed in May and joined the audit and compensation committees.
Red is positive, green is negative (Chinese market colours), grey is neutral, blue is still unclear.
Analyst rating changes, last 90 days (35; latest 12 shown)
| Date | Firm | Action | Rating | Target |
|---|---|---|---|---|
| 2026-09-23 | Stifel | up | Hold → Buy | 530 → 575 |
| 2026-09-22 | Oppenheimer | main | Outperform → Outperform | 515 → 570 |
| 2026-09-21 | Cantor Fitzgerald | main | Overweight → Overweight | 522 → 608 |
| 2026-09-15 | Citizens | reit | Market Outperform → Market Outperform | 550 → 550 |
| 2026-09-04 | Stifel | main | Hold → Hold | 450 → 530 |
| 2026-09-01 | B of A Securities | main | Buy → Buy | 500 → 600 |
| 2026-08-12 | Wells Fargo | main | Overweight → Overweight | 650 → 700 |
| 2026-08-05 | Tigress Financial | main | Buy → Buy | 680 → 690 |
| 2026-07-30 | Bernstein | main | Outperform → Outperform | 646 → 647 |
| 2026-07-30 | Citigroup | main | Buy → Buy | 570 → 600 |
| 2026-07-30 | RBC Capital | reit | Outperform → Outperform | 640 → 640 |
| 2026-07-30 | Stifel | main | Hold → Hold | 400 → 450 |
yfinance aggregate, not individually verified. up = upgrade, down = downgrade, main = maintain, init = initiate, reit = reiterate.
Raw yfinance headlines (10, unfiltered)
- 2026-09-26 · Yahoo Finance Big Tech's AI build-out ran to the Middle East. Then the Iran war hit.
- 2026-09-27 · TheStreet Anthropic's IPO is a $300 billion test for Amazon
- 2026-09-26 · Barchart Meta’s Muse Could Be the Biggest Threat Google Has Faced in 20 Years
- 2026-09-26 · Investor's Business Daily Dow Jones Futures: Growth Shrugs Off Surging Yields, But Trump Rejects Iran Deal; Micron, SpaceX, Tesla Eye Buy Points
- 2026-09-26 · 24/7 Wall St. JPMorgan Predicts IREN Will Generate $24 Billion in Annual Revenue by 2030
- 2026-09-26 · 24/7 Wall St. 3 Elite Dividend Stocks Hiding in the Nasdaq-100
- 2026-09-26 · Motley Fool The 3 "Magnificent Seven" Stocks That I'm Buying Now
- 2026-09-26 · Barrons.com What Musk’s Tesla and SpaceX Stocks Are Doing After That White House State Dinner
- 2026-09-26 · Motley Fool Most Investors Own All Seven Magnificent Seven Stocks. That's a Mistake.
- 2026-09-26 · Stocktwits Microsoft Copilot Is Turning Into An ‘AI Operating System’ — Here’s What It Means For MSFT Stock, According To Oppenheimer’s Brian Schwartz
Rating and scorecard
Our rating
Publisher's call: Neutral, in line with the rule (weighted scorecard 3.20). A great company, but the price already counts the good news.
- One of the strongest businesses in the group. FY26 revenue was $331.8B (+17.8%) and operating income rose 20.8%. Azure growth sped up from 40% to 43%, the 45.1% operating margin is the highest among peers, and the balance sheet holds net cash.
- But the price is not cheap. At 26.1x this fiscal year's (FY27) earnings the stock costs about 35% more than the peer median of 19.4x. It sits only about 4% below its high after beating the S&P 500 over three and six months.
- The profits carry some froth. GAAP profit includes a $6.5B (pre-tax) OpenAI dilution gain, and adjusted EPS still includes a $3.2B Anthropic investment gain in Q4. Capex has doubled, leaving free cash flow at only 0.50x net income, and $329.1B of data-center leases not yet commenced sit off the balance sheet.
What would change our call: a pullback that clearly narrows the valuation premium, or Azure accelerating further while profits stop leaning on investment gains, would mean an upgrade. AI capex rising faster than revenue, or slower growth in the FY27 Q1 results on Oct 29, would mean a downgrade.
Ratings are research opinions based on public information, written for all readers and not tailored to anyone’s finances. Rules and distribution: Rating definitions
A higher risk score means lower risk. The five dimensions are weighted into a suggested rating, which the publisher confirms.
Why these scores
Revenue growth sits in the 10–20% band (4 points). Operating margin fell from 48.9% to 45.1% over the last 4 quarters, but that is mostly seasonal: Q4 a year earlier was 44.9%, so it is roughly flat year on year and no point is deducted. Gross margin fell 1.4 percentage points year on year (68.6% → 67.2%), which the 10-K puts down to AI infrastructure spending. Azure growth is accelerating (40% → 43%), but total revenue growth did not exceed 20%, so no 5.
Revenue TTM +17.8%, Q4 +17.7%; Q4 operating margin 45.1% (prior year 44.9%); ROE 34.0%
Current-fiscal-year (FY27) P/E is 26.1x, about 35% above the 19.4x median of the chosen peers, which puts it in the 2-point band. PEG is about 1.3 (26.1 ÷ 19.8% expected EPS growth), so no adjustment. Among the peers, AAPL is mostly hardware and ORCL is highly leveraged, so comparability is limited and confidence in this score is only moderate.
Current-FY P/E 26.1x vs peer median 19.4x; expected EPS growth FY27→FY28 19.8%
Strong short term, weak long term: the stock beat the S&P over 3 and 6 months but lagged over 1 year and year to date. It trades above its 50-day (+9.0%) and 200-day (+19.5%) moving averages; the 1-year maximum drawdown was -34.5% (trough on 2026-06-25), and it now sits -4.0% from its high. The signals are mixed, so 3 points.
Excess return 3M +33.4%, 6M +23.7%, 1Y -14.6%, YTD -5.7%
Two points off a starting 5: ① FY26 other income was $10.7B, 6.4% of pre-tax income, but 20.7% in Q2 alone, mainly from the OpenAI dilution gain, an accounting gain rather than cash (10-K, 2026-07-29); ② adjusted EPS still includes the $3.2B Anthropic investment gain in Q4, and full-year adjusted other income was $4.2B (8-K and 10-K, 2026-07-29). FCF / net income is 0.50, but the cause is AI capex the company has disclosed, so no deduction. Receivable days are flat year on year (82 vs 83 days, est.), so no deduction.
Non-operating income as a share of pre-tax income: FY 6.4%, Q2 20.7%; FCF $67.0B / net income $133.7B
One point off 5: not-yet-commenced leases of $329.1B equal about 8.6% of market cap and 1.8x annual operating cash flow, above the threshold (10-K Note 13). Nothing else is deducted: 1-year volatility of 32.8% and maximum drawdown of -34.5% are both within thresholds; OpenAI is 7.3% of revenue and 7.4% of receivables, below 10%; debt principal is $46.1B against $76.8B of cash and short-term investments, a net cash position.
Not-yet-commenced leases $329.1B; OpenAI revenue $24.1B (7.3%); volatility 32.8%
What Wall Street thinks
Rating spread (55 firms)
Price-target range
yfinance aggregate of 52 analysts; update time unknown. We show the spread and do not endorse any target.
Inside the filings
Quarterly results
| Quarter end | Revenue | QoQ | YoY | Gross margin | Op. margin | Net income | Diluted EPS | Op. cash flow | Free cash flow |
|---|---|---|---|---|---|---|---|---|---|
| 2025-06-30 | $76.4B | +9.1% | +18.1% | 68.6% | 44.9% | $27.2B | – | $42.6B | $25.6B |
| 2025-09-30 | $77.7B | +1.6% | +18.4% | 69.0% | 48.9% | $27.7B | 3.72 | $45.1B | $25.7B |
| 2025-12-31 | $81.3B | +4.6% | +16.7% | 68.0% | 47.1% | $38.5B | 5.16 | $35.8B | $5.9B |
| 2026-03-31 | $82.9B | +2.0% | +18.3% | 67.6% | 46.3% | $31.8B | 4.27 | $46.7B | $15.8B |
| 2026-06-30 | $90.0B | +8.6% | +17.7% | 67.2% | 45.1% | $35.8B | – | $55.4B | $19.6B |
Source: SEC XBRL. Free cash flow = operating cash flow − capex. Q4 EPS is often missing from XBRL; company figures prevail.
Valuation
| Basis | EPS | P/E |
|---|---|---|
| GAAP, trailing 12 months | 17.96 | 28.7x |
| Adjusted, trailing 12 months (street basis) | 17.28 | 29.9x |
| Current fiscal year estimate (0y) | 19.76 | 26.1x |
| Next fiscal year estimate (+1y) | 23.68 | 21.8x |
FY27 EPS estimates range from $19.21 to $20.70, a narrow spread; FY28 runs from $22.08 to $26.00, a much wider one. Adjusted trailing 12-month EPS of $17.28 matches the company's reported FY26 non-GAAP EPS, but it includes about $0.45/share (est.) of non-OpenAI investment gains.
At the current price. Forward EPS is the yfinance consensus mean, on an adjusted basis.
What the filings say
yfinance's revenue, operating income, net income, GAAP EPS, operating cash flow, receivables and inventory match XBRL quarter by quarter with a 0.0% difference. The problems sit in the derived metrics and in context that only the filings provide:
- OpenAI gains inside profit: yfinance shows Q2 net income of $38.46B, above operating income, and Q4 GAAP EPS up 31.7% year on year. The 10-K says that of $10.7B in full-year other income, $6.5B (pre-tax) was the net OpenAI gain, which it describes as mainly dilution gains from the OpenAI restructuring. Microsoft holds about 25% (as-converted) and accounts for it under HLBV (measured by what it would receive in a hypothetical liquidation) (10-K, 2026-07-29). Excluding it, full-year net income grew 22%, not 31%.
- Anthropic gains inside adjusted EPS: yfinance shows Q4 adjusted EPS of $4.74, an 11.8% beat. The press release states that Q4 includes a $3.2B Anthropic investment gain and that discrete items together added $0.27/share relative to company guidance (8-K, 2026-07-29). Full-year adjusted other income was $4.2B (prior year -$0.1B), roughly $0.45/share at a 19% tax rate (est.), and all of it is inside the $17.28.
- Free cash flow is falling: Operating cash flow was $182.9B (+34%), but purchases of property and equipment reached $115.9B (prior year $64.6B), leaving free cash flow of $67.0B, below last year's $71.6B. On top of that, $24.6B of new finance-lease assets (non-cash) and $19.9B of other investing outflows, described as mainly for component purchases, are not counted in capex (10-K). FY27 Q1 capex guidance is over $50B, including the effect of a change in depreciation lives; the details of that change were not found in the filings read and could not be verified.
- The 34.7% sequential jump in receivables is seasonal: Q4 is the fiscal year-end billing peak. Receivable days are 82, against 83 a year earlier (est.). Typical payment terms are 30–60 days; OpenAI receivables are $6.0B (10-K).
- Debt has not grown: No bonds were issued in FY26, and $3.0B was repaid; long-term debt principal is $46.1B, of which $9.25B matures in FY27. No 8-K under Item 8.01 or 2.03 has been filed since the most recent fiscal quarter.
Buried in the filings
- $329.1B of leases not yet commenced, mostly data centers, commencing FY27–FY33 with terms of 1–20 years; the 10-K says some are contingent on specific contractual conditions being met. Recognized finance-lease liabilities rose from $46.2B to $66.6B (10-K Note 13).
- Contractual obligations total $743.8B: including $194.1B of purchase commitments ($169.0B of it within FY27) and $34.6B of construction-in-progress commitments (10-K). For comparison, cash plus short-term investments is $76.8B.
- OpenAI is both a related party and a major customer: FY26 revenue from OpenAI (including revenue share) was $24.1B, 7.3% of the total; investment commitments are $13.0B, of which $11.9B has been funded (10-K). FY27 Q1 commercial bookings guidance goes out of its way to exclude the prior-year OpenAI contract (8-K, 2026-09-02), which shows OpenAI was a large share of last year's bookings, though the figure is not disclosed.
- Segment and metric definitions change from FY27 (8-K, 2026-09-02); see the event timeline.
- Non-GAAP basis: unchanged; it has only ever excluded OpenAI. No wording such as "no longer exclude" was found.
Where the data vendor got it wrong
| Item | yfinance | Filing (with the math) |
|---|---|---|
| freeCashflow | $16.5B | $67.0B = operating cash flow $182.9B − purchases of property and equipment $115.9B (10-K) |
| forwardPE | 21.8x | Uses FY28 (ending 2028-06) EPS of $23.68; on FY27 EPS of $19.76 it is 26.1x |
| totalDebt | $128.8B | Bonds are only $40.3B; add finance leases of $66.6B and operating leases of $21.9B, and the three sum to exactly $128.8B (10-K) |
| Quarterly total_debt | $56.8B | Roughly bonds of $40.3B plus long-term operating leases of $16.5B (est.): two definitions mixed together |
| Normalized Income | Q2 $38.2B, Q4 $33.4B | Company-adjusted: Q2 EPS $4.14 (GAAP $5.16); Q4 net income $35.3B (8-K). yfinance's exclusions differ in both direction and amount |
| earningsGrowth / profitMargins | 31.7% / 40.3% | Inflated by OpenAI gains; adjusted net income grew 22%, net margin 38.8% ($128.8B / $331.8B) |
| Size of the Q4 beat | +11.8% | Includes the $3.2B Anthropic gain; the company says discrete items added +$0.27 relative to its own guidance (8-K, 2026-07-29) |
| dividendYield | 0.76 | It is already a percentage, i.e. 0.76% |
Price action
Peers
| Ticker | Name | Mkt cap | 1M | YTD | 1Y | P/E TTM | P/E CY | Rev growth | Gross margin | Op. margin |
|---|---|---|---|---|---|---|---|---|---|---|
| MSFT | Microsoft Corporation | $3.83T | +4.0% | +7.4% | +2.6% | 28.7x | 26.1x | +17.7% | 67.9% | 45.1% |
| GOOGL | Alphabet Inc. | $4.21T | +0.6% | +10.1% | +40.3% | 17.3x | 16.7x | +24.2% | 60.9% | 34.0% |
| AMZN | Amazon.com, Inc. | $2.69T | -4.1% | +8.2% | +14.4% | 20.1x | 19.4x | +19.6% | 50.8% | 13.7% |
| ORCL | Oracle Corporation | $414.6B | -7.9% | -29.0% | -52.4% | 21.5x | 16.8x | +29.6% | 64.0% | 35.6% |
| AAPL | Apple Inc. | $4.98T | +8.8% | +25.8% | +33.3% | 39.1x | 38.7x | +16.4% | 48.7% | 32.6% |
| META | Meta Platforms, Inc. | $1.91T | +30.6% | +14.2% | +0.7% | 28.3x | 24.1x | +28.0% | 81.7% | 34.8% |
| Peer median | $2.69T | +0.6% | +10.1% | +14.4% | 21.5x | 19.4x | +24.2% | 60.9% | 34.0% |
yfinance aggregate, not checked against filings. Current-year P/E = price ÷ current fiscal-year (0y) EPS, avoiding forwardPE’s next-year basis. Peer source: --peers. Median excludes this stock.
Peers are hand-picked large cloud and platform companies (GOOGL, AMZN, ORCL, AAPL, META). yfinance's automatic peers are security and software names such as PLTR, PANW and CRWD, which are not comparable in size or profit stage, so they were not used. This stock trades at 26.1x current-fiscal-year earnings, above the 19.4x peer median; its 45.1% operating margin is the highest in the group; its 17.7% revenue growth is below the peer median. Its 1-year gain of +2.6% beats only ORCL and META. Note: peer data are yfinance aggregates and were not checked against each company's filings.
Smart money
Who holds it, who is trading it: 7 of the investors we track hold MSFT
| Investor | Data | Period | Value / range | Portfolio weight | Change |
|---|---|---|---|---|---|
| Bill Ackman | 13F holdings | 2026-06-30 | $2.3B | 11.9% | Added |
| Chase Coleman | 13F holdings | 2026-06-30 | $846M | 3.5% | Trimmed |
| Duan Yongping | 13F holdings | 2026-06-30 | $281M | 1.5% | Trimmed |
| Ray Dalio | 13F holdings | 2026-06-30 | $266M | 1.1% | Trimmed |
| George Soros | 13F holdings | 2026-06-30 | $71M | 0.9% | Trimmed |
| Cathie Wood | 13F holdings | 2026-06-30 | $2M | 0.0% | Trimmed |
| Renaissance Technologies | 13F holdings | 2026-06-30 | $2M | 0.0% | Trimmed |
| David Tepper | 13F holdings | 2026-06-30 | $0 | 0.0% | Sold out |
From each person’s latest published 13F, insider filing or disclosure. 13F data lags by about 45 days. Smart Money →
What to watch
Next FOMC meeting
The dot plot points to one more hike this year; watch whether the 10-year yield holds above 5%. Check the Fed calendar for the exact date.
FY27 Q1 results (first report under the new segments)
Consensus EPS $4.72, revenue $90.66B. Company guidance: revenue $89.85–$90.95B, operating margin flat year on year, other income about -$0.1B (excluding OpenAI), meaning no investment gain like Q4's. Watch: whether Azure reaches 44–45% (constant currency, new basis); whether capex tops $50B; the details of the depreciation-life change.
Ex-dividend date
Dividend paid Dec 10 (yfinance calendar).
Annual shareholder meeting
Reid Hoffman's term ends here (8-K, 2026-06-05).
OpenAI and not-yet-commenced leases
Under HLBV, OpenAI gains and losses will keep making GAAP profit swing; watch how fast lease liabilities and depreciation rise as the $329.1B of leases commence, and whether free cash flow recovers.
Research notes only; not investment advice.
Sources
Market and analyst data: yfinance + SEC EDGAR, as of 2026-09-25 close, delayed. Financials: SEC EDGAR (XBRL and original filings). Macro and news items link to their sources. Red means up, green means down.