A discounted cash flow is the most literal way of valuing a business: work out how much cash it should produce in each of the coming years, convert each of those future amounts into what it is worth today, and add them up. The conversion is the “discounting” — a dollar arriving in five years is worth less than a dollar now, because the one you have now can be put to work in the meantime and because the future one might not arrive at all.
In practice a DCF has two halves. The forecast period is a handful of years modelled explicitly, usually five or ten. The terminal value stands for everything after that, collapsed into a single figure. On most companies the terminal half is the larger of the two, which is worth knowing before trusting the total.
FCF is free cash flow in year t; k is the discount rate, which for a whole-company model is the weighted average cost of capital; n is the last forecast year. Subtracting net debt is what turns the value of the business into the value of the equity.
Take a three-year forecast of $100m, $110m and $120m of free cash flow, discounted at 10%:
Notice what discounting did: three growing cash flows became three near-identical present values, because a 10% rate is eating the growth almost exactly. Push the horizon out far enough and each additional year contributes almost nothing — which is why a fifty-year sum is a convergence setting rather than a claim about holding a share for fifty years.
Build it, then attack it. The useful output of a DCF is not the headline figure but the answer to “which input is this resting on?” Move the growth rate, the margin and the discount rate one at a time; whichever swings the answer most is the assumption you actually need a view on.
Check how much of the total is terminal. If four fifths of the value sits after the forecast period, you have not valued the next five years — you have valued a guess about the year after them, and the forecast is decoration.
In the app, Advanced DCF and Simple DCF are two of the four methods on the Intrinsic Value tab: the first forecasts revenue, margin, tax, capital spending and working capital separately, the second grows one cash flow per share at one steady rate.
A DCF earns its keep on a business with cash flows that are positive, reasonably steady and traceable to something you can describe — subscriptions, tolls, consumer staples, utilities. It is also the right tool whenever the share price implies expectations you want to make explicit: run the model backwards in your head and ask what growth today’s price is assuming.
It is the most assumption-hungry method there is, and small changes compound: a percentage point on the discount rate and a percentage point on terminal growth can move the answer by half. It describes a bank or an insurer badly, because for them debt is raw material rather than financing, and it says nothing useful about a company whose free cash flow is negative while it is being built. Finally, it produces a number to the cent from inputs that are guesses — the precision is entirely cosmetic.