Microsoft (MSFT)
Software / Cloud / IA
Highest-quality franchise (Office, Azure and artificial intelligence): 5y base $614 (+6%/year with dividend) → Fairly valued. The point to watch is artificial intelligence capex, which climbed to $115.9bn and drove the incremental return on capital down to 16.9% — far below the 30.8% earned by the capital already installed.
- Price
- $489.44
- Intrinsic value (5y, base)
- $614
- Total annual return (5y)
- 5.5%
- Status (nominal)
- Fairly valued
- Margin of safety
- +5%
The essentials
- Two quality engines coexist: subscription (Microsoft 365, LinkedIn, Dynamics) and consumption (Azure, +41% in the fiscal year and passing US$100bn annually); fiscal year 2026 revenue of US$331.8bn (+17.8%, or +16% in constant currency) with an operating margin of 46.8%.
- Artificial intelligence capital expenditure climbed to US$115.9bn (against US$64.6bn the prior fiscal year) and the incremental return on capital fell to 16.9%, down from the 30.8% earned by installed capital — still above the 10% bar, but with far less headroom. It is the central disconfirmer of the thesis.
- Contracted commercial remaining performance obligation rose 84% to US$678bn: this is revenue already signed, not an extrapolated trend, and it anchors the growth projection.
- At $489 it trades at 34× maintenance free cash flow, a multiple that compresses toward 15× over five years if the business executes; the base-case total return is around +6% annually.
Intrinsic value — two valuation methods
Total return at 5 years: 5.5%/year = 4.6% appreciation + 0.9% dividend. The target price ($614) is ex-dividend; the $24 in dividends collected over 5 years are added separately.
By both methods, the value today (DCF $521 · Multiples $514) exceeds the market price ($489).
Pillars of the analysis
The verdict — today vs 5 years
Today — fairly valued: at $489 trades ~4.8% below its value discounted to today (~$514); the discount is positive but does not reach the margin of safety we require (≥38%).
At 5 years — En valor: the target price ($614) plus dividends yield between the 4% floor and the 10% average return — a reasonable return, though without the margin of a great investment.
The bridge: the return at 5 years exceeds the risk-free rate (4.5%) — but the discount does not reach the required margin of safety (≥38%). To require a 15% annual return, it would need to be bought at ~$321.
Thesis
The business
Microsoft is one of the highest-quality franchises in the market: three segments combining recurring subscription (Microsoft 365, LinkedIn, Dynamics) and cloud consumption (Azure, the world's second platform, which passed US$100bn annually), with a 46.8% operating margin and a return on capital of 30.8%, far above the 10% bar. The OpenAI partnership and Copilot distribution across the installed base give it an artificial intelligence distribution advantage no competitor matches. This is fortress quality.
The valuation
It is valued on EV/maintenance free cash flow and owner earnings —which expense capital expenditure and stock-based compensation—, not on EV/EBITDA or reported free cash flow. With the base year set at the just-closed fiscal year 2026 (revenue $331.8bn to Jun-2026), at US$489 Microsoft trades at 34× maintenance free cash flow, a multiple that compresses toward 15× over five years as the metric grows.
The base case projects revenue growing +16% → +10% from the closed fiscal year —the starting point is constant-currency growth, not the headline—, the operating cash margin normalized at ~49.3% and terminal-year maintenance free cash flow at ~US$220bn on an exit multiple of 20×. That gives a value of ~US$614 per share over five years → a total return of ~+6% annually (+5% from price and +1% from dividend).
The margin of safety
It trades close to intrinsic value, far from the required margin of safety. The base total return is ~+6% per year (+5% from price and +1% from dividend), read against the method's reference scale: 4% as the floor, 10% the average return of equities and 15% the threshold of a great investment. The resulting verdict is Fairly valued. The quality of the business is not in question; what defines the tier is how much of that quality is already embedded in the entry price.
The element that limits the margin is the investment in artificial intelligence. It is not that it depresses cash flow —that was already known—: it is that the incremental return on capital fell to 16.9% from levels on the order of 25-30%. It remains above the hurdle, but the headroom between what new capital earns and what it costs has narrowed visibly.
What to watch
The central disconfirmer is the return on artificial intelligence capital expenditure. Fiscal year 2026 delivered the first clean measurement of a full investment cycle: US$115.9bn of capital expenditure, invested capital growing US$113.8bn and an incremental return of 16.9%. The thesis holds as long as that number stays above the 10% bar; if it keeps falling, the investment becomes malinvestment and the market re-rates the business to a cash multiple. The leading indicator argues in favour: contracted commercial remaining performance obligation rose 84% to US$678bn, with ~30% recognizable over the next twelve months. In the background: the contraction of More Personal Computing, regulatory scrutiny of the OpenAI partnership, and the fact that the fiscal year's net income includes US$5.0bn of gains from that same stake —the company itself publishes the figure excluding them, US$128.8bn—.
Educational / informational. Does not constitute investment advice.