StockData

How the score works

How we estimate fair value

Our fair value asks what a share would be worth if the market paid its usual price for this company's earnings, cash flow, assets and sales — usual for the company itself and for its industry — and checks that against what the business earns on its own. It uses reported numbers only. We don't use analyst forecasts.

Five methods, blended

MethodWeightValue per share
P/E40%Fair P/E × diluted EPS over the last twelve months. Only when EPS is positive. Banks and insurers: 50%.
EV/EBITDA35%Fair EV/EBITDA × trailing EBITDA, minus net debt, ÷ shares. Not used for banks and insurers.
Earnings power15%Owner earnings ÷ cost of equity ÷ shares, a no-growth value. Owner earnings are operating cash flow minus maintenance capital spending (the smaller of capex and depreciation), averaged over three years.
P/B10%Fair P/B × book value ÷ shares. Skipped when return on equity is above 25%, where buybacks have shrunk book value. Banks and insurers: 50%.
P/S10%Fair P/S × revenue ÷ shares. Only for companies without positive earnings or EBITDA, because companies with different margins aren't comparable on sales.

The fair multiple

For each multiple, the fair level mixes the company's own five-year median (weekly, using the figures published at each date) with today's median for its industry group. The company's own history gets more weight the longer it is: years of history ÷ 10, kept between 30% and 70%. Fair multiples are capped to sensible ranges (P/E 5–45, EV/EBITDA 3–30, P/B 0.3–12, P/S 0.1–15).

fair multiple = w × own 5-year median + (1 − w) × industry median
w             = clamp(years of history ÷ 10, 0.3, 0.7)
cost of equity = 10-year Treasury yield + beta × 5%   (beta clamped 0.5–2)

Guards against nonsense

Margin of safety

Margin of safety = (fair value − price) ÷ fair value. A margin of 20% or more passes the margin-of-safety check in the score. Negative values mean the price is above fair value.

When to distrust it

Any estimate built on past numbers is wrong when the future looks different: a company whose profits are about to collapse looks cheap, and a fast grower whose profits are about to jump looks expensive. Cyclical companies look cheapest at the top of the cycle. Use the estimate as a starting point for reading the filings, not as a price target.

Not investment advice.