Financials

Financials - What the Numbers Say

Microsoft entered FY2026 as a $282B-revenue software platform earning a 46% operating margin, generating $72B of free cash flow, and sitting on a roughly $51B net-cash balance sheet — financial output most companies in any industry will never match. But the story has shifted: free cash flow is no longer growing because AI/data-center capital spending has more than doubled in two years (from $28B in FY2023 to $65B in FY2025), pushing return on invested capital down from a 37% peak to roughly 24%. The market is paying a TTM P/E near 25x and EV/EBITDA near 15x, betting that Azure and Copilot will absorb the new asset base profitably. The single financial metric that matters most right now is FCF growth versus capex — the moment FCF re-accelerates, the AI investment cycle is validated; if it does not by FY2027, the multiple is at risk.

Revenue, Margins, and Earnings Power

Definitions used below. Revenue is the top line — what customers paid. Gross profit is revenue minus the direct cost of delivering it (data-center costs, hardware, content rights). Operating income is what is left after R&D, sales/marketing, and G&A. Net income is the bottom line after interest and tax. Margins are each of those divided by revenue.

Revenue is up 4.5x since FY2010 with a clear acceleration after FY2018 — the Nadella-era pivot to cloud. The two visible dips, FY2012 and FY2015, were both restructuring/impairment events (Surface RT writedown and Nokia phone writedown) — they did not reflect demand weakness. Operating income has grown faster than revenue, evidence of operating leverage as the cloud and recurring-software mix expanded.

Operating margin has expanded from 38.6% to 45.6% over 15 years. The structural reason is simple: cloud subscription revenue (M365, Azure, Dynamics) carries higher incremental margin than legacy boxed software, and scale has spread the SG&A base. The FY2018 net-margin drop was a one-time U.S. tax-reform charge, not operating weakness. The most recent reading, 45.6%, is the highest in the company's modern history — and it has been achieved while the company simultaneously absorbed the Activision acquisition and AI infrastructure ramp.

Quarterly revenue is up roughly 28% year-over-year on the latest print, and operating margin has actually expanded into the high-40s despite the capex ramp. That is a sign that the revenue contribution from new AI-related services is currently outpacing the depreciation and operating cost of the new data centers — the most important short-term proof point for the investment case.

Cash Flow and Earnings Quality

Definition: free cash flow (FCF) = cash generated from operations minus capital expenditures. It is the cash a company can actually return to shareholders after keeping the business running and investing for growth. When FCF is consistently close to net income, earnings are "real"; when FCF is much lower than net income, accounting earnings are being absorbed by working capital, capex, or other cash needs.

Operating cash flow has consistently exceeded net income — the OCF/NI ratio sits at roughly 1.34 in FY2025 — which is the textbook signature of a high-quality, deferred-revenue-heavy software business: customers pay upfront for subscriptions, so cash arrives before revenue is recognized. The disconnect that matters is the FCF vs net income gap. FCF was $71.6B against $101.8B of net income in FY2025, an FCF/NI ratio of just 0.70 — the lowest since FY2018 (which itself was a tax-distorted year). The cause is not earnings deterioration; it is the surge in capex.

The FCF margin compressed from a 33% peak in FY2021 to 25% in FY2025 — almost entirely driven by capex intensity rising from roughly 14% of revenue to 23% of revenue. The earnings are real, but they are being routed into the asset base. That is fine if the new assets earn an attractive return; it is a value-destroyer if they do not. This is the central test of the bull case and is why the watch metric at the bottom of the page is FCF growth.

Balance Sheet and Financial Resilience

Microsoft is one of only two non-financial U.S. companies carrying a AAA credit rating from S&P (Moody's Aaa). The balance sheet ended FY2025 with about $95B of cash and $43B of total debt, a net-cash position of roughly $51B. The temporary dip to $24B of net cash in FY2024 reflected funding the $69B Activision close — and the company has already rebuilt the cushion in twelve months from internal cash generation. Equity has nearly doubled in three years to $344B as retained earnings compound.

The balance sheet is not a constraint on this business. It is an enabling asset: it lets Microsoft sign multi-year compute commitments with OpenAI, fund $80B+ annual capex without straining liquidity, and outbid almost any rival in an acquisition without bond-market gating.

Returns, Reinvestment, and Capital Allocation

Definition: ROIC (return on invested capital) measures the after-tax operating profit generated per dollar of capital deployed in the business (equity plus debt minus cash). It is the single best measure of whether a company is creating economic value — anything above the cost of capital (typically 8–10%) creates value, anything below destroys it.

ROIC peaked at 37% in FY2021 and has compressed to 24% — still elite by any standard, but a clear narrative shift. The denominator has expanded faster than the numerator: invested capital is up roughly 75% in three years as Microsoft built out AI data-center capacity and closed Activision, while operating profit has grown 54%. Whether this is a temporary investment-cycle drag or a structural reset is the heart of the underwriting decision. Historically, Microsoft's incremental ROICs on cloud reinvestment have been very high; if Azure-AI follows that pattern, ROIC should recover toward 30%+ by FY2027–28.

The capital-allocation story rotated decisively in FY2024: buybacks were cut roughly in half, dividends grew modestly, and the freed cash went into Activision and the AI capex line. In FY2025, capex alone ($65B) exceeded combined dividends and buybacks ($43B) — a first in modern Microsoft history. Importantly, the dividend was still raised (the company has raised it every year for two decades), signaling management's confidence that base cash generation can fund both growth investment and shareholder returns.

Share count is down roughly 16% over 15 years — a steady, persistent buyback that has consistently offset stock-based compensation dilution and then some. With $12B of annual SBC vs $18B of buybacks in FY2025, the net retirement pace has slowed but is still positive. Management is reinvesting heavily, returning what is left, and not diluting shareholders — exactly the capital-allocation pattern a long-duration owner wants to see.

Segment and Unit Economics

Microsoft reports three operating segments: Productivity and Business Processes (M365, LinkedIn, Dynamics), Intelligent Cloud (Azure, server products, enterprise services), and More Personal Computing (Windows OEM, Xbox, Search/Bing, Devices, Surface). Segment-level financial detail at the line-item granularity required for charts is not included in the dataset for this run, so this section uses the disclosed FY2025 segment mix from the company's reporting.

Where the economics live. Productivity and Intelligent Cloud together account for roughly 81% of revenue and an even larger share of operating profit. Azure (inside Intelligent Cloud) is the highest-growth piece, growing roughly 30%+ year-over-year on a much larger base; M365 Commercial Cloud (inside Productivity) is the most profitable. More Personal Computing carries the lowest margin and is the most cyclical, but it is also the smallest. The implication: an investor underwriting MSFT is fundamentally underwriting (1) Azure's growth durability and (2) M365 ARPU expansion from Copilot. Hardware, gaming, and search remain optional kickers, not core drivers.

Valuation and Market Expectations

What the multiples imply. At ~25x TTM and ~21.5x forward earnings, Microsoft is trading below its 5-year average P/E (which has sat in the high 20s to mid-30s during the AI-rally years) but at a clear premium to the broader market. EV/EBITDA at 15x is roughly in line with mature cloud peers and below high-growth software multiples. The price is not pricing exuberance — it is pricing a continuation of mid-teens revenue growth, gradual margin recovery, and FCF re-acceleration. That is a defensible base case, but it leaves very little margin of safety if AI revenue underwhelms or capex must continue rising into FY2027.

Consensus analyst price target is roughly $560, implying ~33% upside from $422 spot — closer to the bull side of the range. Forty-two of the 54 covering analysts are at Buy or Strong Buy; only three rate Hold and zero rate Sell. The Street's central case is that AI capex pays off; the contrarian case is that 25x forward earnings is the wrong multiple to pay for a company whose ROIC is still falling.

Peer Financial Comparison

Reading the peer gap. Microsoft trades at a meaningful discount to Apple and Oracle on P/E, in line with Alphabet, and at a premium to Meta. It carries one of the highest operating margins in the group (only Meta higher), the second-highest FCF margin (after Apple), and a competitive ROIC. The case for parity with Alphabet is strong: similar growth, similar margins, AAA balance sheet, more durable enterprise mix. The case for a premium over Meta is weak on growth (Meta is growing faster) but strong on diversification and balance-sheet quality. The most defensible reading is that MSFT is fairly priced versus its hyperscaler cohort, with neither a glaring discount nor a stretch premium.

What to Watch in the Financials

What the financials confirm: Microsoft is an elite-quality compounder — AAA balance sheet, 46% operating margin, dominant cash generation, two-decade dividend-raise streak, and steady net buybacks. The cloud-and-software transformation worked, and the income statement reflects it.

What the financials contradict (or qualify): The "always-rising returns" narrative. ROIC has fallen for four straight years; FCF margin is at a multi-year low; capex now exceeds shareholder returns. The bull case has not been disproven, but the financial scorecard has visibly weakened versus the FY2020-22 peak. A premium multiple on falling returns is the single most uncomfortable feature of the current setup.

The first financial metric to watch is FCF growth in FY2026. If free cash flow re-accelerates back to a teens-percent growth rate against a 14–17% revenue gain, the capex cycle is validated, ROIC recovery is on the table, and the multiple is defensible. If FCF stays flat or declines again, the market is paying a 25x multiple for a business whose owner-economics are no longer expanding — the path most likely to drive a meaningful de-rating.