Oracle (ORCL)
Software (base de datos) / Nube de infraestructura (OCI)
Mission-critical database franchise (a high-margin annuity) betting big on the AI cloud (OCI): the RPO exploded to $638bn, but the data center capex ($55.7bn) sank free cash flow to −$23.7bn, financed with debt. It has already corrected ~40% from its peak. At $145: Fairly valued — a coin flip on OCI's economics.
- Price
- $145.13
- Intrinsic value (5y, base)
- $176
- Total annual return (5y)
- 5.7%
- Status (nominal)
- Fairly valued
- Margin of safety
- +5%
The essentials
- Hybrid: mission-critical database franchise (~70% historical margin, deep switching cost) + OCI, the AI cloud (capital-intensive, ~16% gross margin). The cloud is already 51% of revenue; OCI grows +76%.
- The RPO (contracted backlog) exploded from $138bn to $638bn on the AI-cloud mega-contracts — but only 12% is recognized within 12 months (long, concentrated). The $55.7bn capex sank free cash flow to −$23.7bn, financed with debt ($129.5bn).
- It has already corrected ~40% from its peak. At $145: Fairly valued, with an enormous dispersion — it is the malinvestment test (§3) in its purest form: does the OCI capex earn its cost of capital, or is it low-return backlog financed with debt?
Intrinsic value — two valuation methods
Total return at 5 years: 5.6%/year = 3.9% appreciation + 1.7% dividend. The target price ($176) is ex-dividend; the $13 in dividends collected over 5 years are added separately.
By both methods, the value today (DCF $216 · Multiples $153) exceeds the market price ($145).
Pillars of the analysis
The verdict — today vs 5 years
Today — fairly valued: at $145 trades ~5.1% below its value discounted to today (~$153); the discount is positive but does not reach the margin of safety we require (≥38%).
At 5 years — En valor: the target price ($176) 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 ~$96.
Thesis
The business
Oracle is a hybrid: a very high-margin mission-critical database franchise (the durable moat) betting big on AI cloud infrastructure (OCI, capital-intensive and thin-margin). The $638bn RPO turns Oracle from "mature low-growth value" into a contracted-growth story — if it executes the capex and the AI customers pay. But the bet is expensive: free cash flow is −$23.7bn (the data center capex, 1.7× operating cash flow), financed with a $129.5bn debt load that is growing. The quality of the backlog (concentrated, ~16% margin) is the question.
The valuation
It is valued by EV/EBIT: free cash flow is deeply negative because of the AI capex, so it does not work as a metric (EV/FCF is N/M); operating income ($20.6bn, +17%) is solid and growing. The cash burn is shown explicitly — it is a heavy investment phase. The base scenario projects ~$176/share over five years.
The central fact: Oracle has already corrected ~40% from its June peak (from ~$213 to $145). The question is no longer whether it was priced for perfection (it was) but whether, at $145, the backlog supports the price. It trades at ~24× operating income; the multiple compresses toward ~13× over five years if OCI scales. The base return at market price is +6%.
The margin of safety
The verdict is Fairly valued: It trades close to intrinsic value, far from the required margin of safety.. The 40% correction took out much of the "perfection" premium; at $145 the expected return is around the market's average return, but with an enormous dispersion that is the heart of the thesis. In the favorable scenario, OCI scales profitably, the backlog converts and value rises sharply; in the adverse one, OCI's economics disappoint (thin margin, a key customer fails to pay, asset impairment), the debt weighs and value ends up well below the price. It is neither a deep-value bargain nor a clear case of overvaluation: it is a binary-outcome bet on the economics of the AI cloud.
What to watch
The disconfirmer is the malinvestment test: does the ~$55.7bn/year of capex, financed with debt and at ~16% gross margin on the AI contracts, earn its cost of capital? The indicators to follow: (1) the conversion of the $638bn RPO into revenue and, above all, into margin; (2) the ability to pay of the AI customers —few and giant— whose contracts sustain the backlog; (3) the trajectory of the debt ($129.5bn and rising; S&P one notch above junk) and of free cash flow (today −$23.7bn), which must turn positive as OCI ramps. As a governance risk, Ellison's control (40.6%, with 346M shares pledged) and the transition to two new co-CEOs.
Educational / informational. Does not constitute investment advice.