Choosing the Explicit Forecast Horizon
A DCF needs a cutoff where detailed, year-by-year forecasting stops and a simplified terminal value takes over, and picking that cutoff too short or too long distorts the valuation either way.
Prerequisites: Driver-Based Forecasting of Financials
No analyst can credibly forecast a company's exact revenue, margins and capital spending forty years into the future, year by year — the uncertainty compounds too fast to be meaningful. Every discounted cash flow model instead splits the future into two zones: a handful of years forecast in real detail, followed by a single simplified "terminal value" standing in for everything after. The question of exactly where that split happens — the explicit forecast horizon — is one of the most consequential and most casually chosen decisions in the entire model.
The explicit horizon should end roughly when the company reaches a "steady state" — stable margins, a stable growth rate, and reinvestment roughly proportional to growth — because the terminal value formula that follows assumes exactly that kind of steady state. Ending the horizon before the company gets there smuggles bad, temporary numbers into a permanent formula.
Why the length matters
Terminal value formulas — Gordon growth chief among them — assume the company has settled into a mature, stable pattern that will persist indefinitely. If the explicit horizon is cut too short, the terminal value formula gets applied to a company that is still in a high-growth or margin-expanding phase, and that temporary condition gets frozen into a perpetuity, wildly overstating value. If the horizon is stretched needlessly long, the analyst is just forecasting years 8 through 15 with the same blunt assumptions used for the terminal value anyway — extra length without extra precision, and extra false confidence in numbers nobody can really know.
Worked example
A five-year-old software company is growing revenue 40% a year and margins are expanding fast as it scales past fixed costs. If an analyst uses a common default 5-year explicit horizon and then applies Gordon growth terminal value at year 5, the model locks in whatever growth and margin the company happens to have in year 5 — say, still 25% growth and 30% margins — as if that continues forever. Since no company can grow 25% a year in perpetuity without eventually becoming larger than the entire economy, that terminal value is badly inflated.
Extending the explicit horizon to 10 years, with growth ramping down from 40% toward a plausible long-run 4-5% and margins stabilizing around a mature 20%, means the terminal value formula is only being asked to handle a company that has actually reached a steady state. The same underlying business can produce a dramatically lower, more defensible valuation once the horizon is long enough to let growth and margins actually settle before the perpetuity math takes over.
What this means in practice
There's no fixed rule for the right number of years — a mature utility might need only 3-5 years of explicit detail because it's already near steady state, while a fast-growing tech company might need 10-15. The test is conceptual, not calendar-based: has growth decelerated to something sustainable forever, and has margin stabilized, by the last explicit year?
Defaulting to a round number like 5 or 10 years purely out of habit, without checking whether the company has actually reached a steady state by then, is the single most common way analysts unintentionally inflate a DCF.
Further reading
- Damodaran, Investment Valuation (ch. 12, length of the growth period)