Here is the number that reframes the whole DCF conversation: terminal value routinely carries 60-80% of the total. Most of your valuation lives beyond the forecast, in the part of the model with the fewest inputs and the most assumption. Interviewers know this, which is why terminal value questions are really fragility questions: does the candidate know where the model is weakest and do they check it? The candidates who volunteered the checks, unprompted, were the ones I remembered.
The two methods, in words
- Perpetuity growth. Take the final forecast year's free cash flow, grow it one year at a long-run rate and divide by the discount rate minus that growth rate. It answers: what is a slowly growing cash stream worth forever? Elegant and violently sensitive to the two inputs in the denominator.
- Exit multiple. Apply a market multiple, usually EV to EBITDA, to the final year's metric, as if selling the business at the horizon. It borrows the market's current pricing, which is both its convenience and its philosophical problem: you have smuggled relative valuation into an intrinsic method. Knowing that tension exists and saying so, is itself a scoring line.
Practitioners run both and cross-check, which is not diplomacy; it is the first sanity check below.
The three checks that score
The three-check routine · run it every time
1 · The growth ceiling: the perpetuity rate cannot exceed long-run nominal growth of the economy, because a company growing faster than the economy forever eventually becomes the economy. A rate around inflation, give or take, is the defensible zone.
2 · The cross-check: compute the exit multiple your perpetuity assumptions imply and the growth rate your chosen exit multiple implies. If your perpetuity maths implies selling at a multiple far above where the sector trades, per your comp set, one of your assumptions is dreaming.
3 · The steady-state test: the terminal year must look like a company at rest. Margins normalised, capex roughly at or a touch above depreciation, working capital growing with revenue. A terminal year mid-investment-surge capitalises a temporary state forever.
The second check is the one that impressed, because it connects the DCF to the comps work in the site's comparable companies article and shows the candidate treats methods as cross-examining witnesses rather than parallel rituals.
The follow-ups
The most instructive one: 'your terminal value is 85% of the total, is the DCF useless?' The scoring answer refuses the bait: no, the model is telling you truthfully that this company's value is long-duration, common in growth businesses; the response is to test the terminal assumptions harder, perhaps extend the explicit forecast until steady state is genuinely reached and present the output as a sensitivity range rather than a point. Which leads to the presentation question: strong candidates speak in two-way tables, discount rate against growth rate, because they know the honest product of a DCF is a range with reasons, the same humility the WACC article in this series argues for on the rate itself.
The fumbles
Discounting the terminal value from the wrong year, the classic mechanical slip the anchor DCF article flags. Growth rates above the discount rate, which the perpetuity formula punishes with nonsense and interviewers punish faster. Terminal years still growing capex at double depreciation. And the meta-fumble: presenting the terminal value without ever mentioning its weight, which tells the examiner you built the model without noticing where it lives. In timed modelling tests, per the modelling tests article, the terminal block is where markers look first for exactly these errors.
Run the three-check routine aloud on any company you have modelled and this topic moves from fragile to fortified in an afternoon. Having the follow-ups pressure-tested live is what the IBD Recruiting Review is for.
FAQ
What growth rate should I use for terminal value?
At or below long-run nominal economic growth, roughly inflation give or take. A company growing faster than the economy forever eventually becomes the economy.
How do I sanity-check a terminal value?
Cross-check methods: compute the exit multiple your perpetuity assumptions imply and the growth rate your exit multiple implies, then test the terminal year for steady state.
Is a DCF useless if terminal value is 80% of the total?
No. It is telling you the value is long-duration. The response is harder testing of terminal assumptions, possibly a longer explicit forecast and presenting a sensitivity range.
Former Executive Director in TMT Investment Banking at Nomura and M&A banker at BNP Paribas. Top-rated mentor with 300+ sessions. About
