Information and education only. Every page, figure and calculator on this site exists to explain how financial instruments work. Nothing here is investment, tax or legal advice, a recommendation, or a valuation you can rely on. Full disclaimer

Discounted cash flow

Also known as: DCF, Intrinsic valuation

The only method that values the business itself. Also the one whose answer moves most when nobody is looking.

5 min read · 816 words

1 · SnapshotThe one idea to remember
Key idea: a discounted cash flow looks like the most rigorous method on this desk and is the most sensitive to assumptions nobody can check. Which is why it is always quoted as a range, and why anybody who gives you a single number from one has told you something about themselves.
2 · BeginnerWhat actually happens?

A business is worth what it will produce for its owners, in cash, from now until it stops. That is the idea behind a discounted cash flow, and it is the only valuation method that says anything about the business itself rather than about what somebody paid for a different one.

Money arriving in ten years is worth less than money arriving today, so each future year is shrunk by a rate before it is added up. That rate is the discount rate, and it stands for what investors could earn elsewhere at similar risk.

Nobody can forecast forever, so the model runs for five or ten years and then stops. Whatever the business is worth after that is squeezed into one figure at the end, called the terminal value.

And here is the thing everybody discovers eventually: that final figure is usually most of the answer. Five years of carefully built forecasts often account for less than half of the result, and the rest comes from one calculation about a period nobody modelled.

11–3 wks22–5 days31–3 days41–2 days52–4 daysForecast agreedRange in the book
The only method that values the business itself rather than by reference to others — and the one most sensitive to assumptions nobody can observe.
  1. 1

    The forecast1–3 wks

    Management's plan is taken apart and rebuilt, because every later number depends on it.

  2. Is the plan credible — The banker and the client decides. A forecast nobody believes produces an answer nobody believes, however good the arithmetic underneath it.

  3. 2

    Free cash flow2–5 days

    Earnings are converted into cash: tax, capital expenditure and working capital are all real money.

  4. 3

    The discount rate1–3 days

    A cost of capital is assembled from inputs that are mostly estimates of unobservable things.

  5. 4

    Terminal value1–2 days

    The value beyond the forecast, which is usually most of the answer and is a division by a small number.

  6. Does the range survive review — The valuation committee decides. A range so wide it contains every plausible price has not said anything, and a narrow one is claiming a precision the method does not have.

  7. 5

    Sensitivity and range2–4 days

    The answer is presented as a range across rates and growth, because a single number would be a claim nobody can make.

Who is on the deal

WhoSideWhat they are actually for
The analystNeitherBuilds the model and owns every assumption in it, whether or not they were chosen.
ManagementSell sideSupplies the forecast, which is the largest input and the one nobody outside can check.
The diligence teamsBuy sideProduce the adjustments that change the answer once somebody has actually looked.
The valuation committeeNeitherTests the work before it goes anywhere, because a published range is a statement the firm stands behind.
Desk
Valuation & Deal Analysis
Values
The business, not by reference to others
Largest input
The forecast, supplied by people with an interest in it
Usually most of the answer
The terminal value
Always presented as
A range, and honestly so

What decides whether it completes

Not how hard this is, and not a rating — there is deliberately no total. It says which of five blockers decides whether this transaction happens at all, in the same order on all 70 transaction types so they can be compared. This publication's own reading; see the notice below.

  • Pricedecides it
  • Financingbarely applies
  • Approvalbarely applies
  • Diligencedecides it
  • Executionmatters

What decides it here. Two inputs decide the answer and neither is observable: the forecast, which comes from people with an interest in it, and the discount rate, which can be moved within a defensible range enough to change the result by a quarter. That is the honest reason this method is always quoted as a range.

What the five mean, and which one decides where →

3 · IntermediateHow it runs in practice

The steps

  1. Forecast free cash flow. Not profit — cash. Start from operating profit, deduct tax, add back non-cash charges, then subtract capital expenditure and the increase in working capital. Those last two are real money and they are where optimistic models go wrong.
  2. Pick a discount rate — see the cost of capital.
  3. Discount each year back to today.
  4. Add a terminal value, either by growing the last year forever at a modest rate or by applying an exit multiple.
  5. Add it all up to get enterprise value, then bridge to equity value.

The bridge, which prevents more errors than anything else

Enterprise value is what the whole business is worth to everybody who funds it. Equity value is what is left for the shareholders. To get from one to the other, subtract debt, add cash, and adjust for pension deficits, leases and minority interests.

Most valuation errors made by newcomers are not modelling errors. They are comparing an enterprise value with an equity value, and the fix is to write the bridge out once and never forget it.

The terminal value, two ways

$$ TV = \frac{FCF_{n}\,(1+g)}{r-g} $$
What the symbols mean
  • Tmaturity, in years
  • Va value
  • Fthe forward or futures price
  • Cthe price of a call option
  • nhow many periods, or how many things
  • ga growth rate, per year

Or an exit multiple applied to the final year's earnings. The two are worth computing side by side: if they disagree wildly, one of the assumptions is wrong, and the disagreement is more informative than either answer.

4 · AdvancedThe numbers & the documents

Why the growth rate is dangerous

The perpetuity formula divides by the difference between the discount rate and the growth rate. If the rate is 8% and growth is 2%, the divisor is 6. Move growth to 3% and the divisor is 5 — the terminal value rises by a fifth from a change most people would not argue about.

And a growth rate above long-run economic growth is a claim that this business eventually becomes the entire economy. That is the discipline: whatever number is used has to survive being said out loud.

Where models are usually wrong in one direction

  • Margins that expand every year for no stated reason.
  • Capital expenditure below depreciation in the terminal year, which means a business shrinking while it is assumed to grow forever.
  • Working capital that does not grow with revenue, which quietly manufactures cash.
  • A discount rate that has not been checked against what the market actually charges this company to borrow.

The mid-year convention, and other small honest fixes

Cash arrives through the year rather than on the last day, so discounting each year from its midpoint is slightly more accurate and raises the answer a little. It is not a trick; it is the kind of detail that separates a model somebody has thought about from one that has been filled in.

What the method is genuinely for

Not to produce a number — the range is too wide for that. It is to make assumptions explicit. A discounted cash flow forces somebody to say what they think margins, growth and investment will be, and that is a conversation that cannot be had with a multiple.

Which is why it sits beside comparables and precedents rather than replacing them. Three methods that are wrong in different directions are more useful than one that is wrong in a way nobody can see.

The formulas above are standard textbook formulations, simplified for teaching. They explain the mechanism — they are not a valuation tool, and they will not reproduce a dealer’s price.

5 · Desk notesHow people on the deal think about it
Desk note: compute what percentage of your answer is the terminal value before showing it to anybody. Above about three quarters, you have not valued the forecast period at all — you have written a note about the discount rate, and it should be presented as one.

Now say it back

Close the page and give Discounted cash flow in four sentences. It takes a minute and it is the only way to find out whether reading it was enough.

  1. Who wants what — name both sides and what each one is actually trying to get.
  2. What has to happen, in order — the three or four stages, not the whole timetable.
  3. Where the money comes from — cash, new shares, or borrowed; somebody has to fund it.
  4. What kills it — the ordinary way, not the dramatic one.

Why these four

Where this transaction shows up elsewhere

  • EasyFairness opinionDealA narrow statement, on a stated date, about one specific offer
  • EasySynergiesDealThe savings a combination is supposed to produce — announced with confidence, paid for at announcement, and checked…
  • MediumMaDeskHow a takeover actually works: the auction, the offer, the vote, the regulator and the long stop date — plus what…
  • MediumValuationDeskThe arithmetic underneath every transaction: discounted cash flow, comparables, precedent transactions, the buyout…