Terminal value: the part of a DCF that does most of the work

A discounted cash flow model cannot forecast a company forever, so it forecasts explicitly for a number of years and then adds a single figure, the terminal value, to represent everything after that. In many DCFs this single number accounts for well over half of the total value. That makes it the part of the model that most deserves skepticism.

What terminal value represents

Suppose you forecast ten years of cash flow. The terminal value is the value, at the end of year 10, of all cash flows from year 11 onward. It is then discounted back to today like any other cash flow. For the surrounding steps, see DCF basics.

There are two standard ways to estimate it.

Method 1: growing perpetuity (Gordon growth)

Assume cash flow grows at a constant rate g forever, and discount at rate r:

Terminal value = FCF in final year x (1 + g) / (r - g)

This requires r to be greater than g. As g approaches r, the value rises explosively, which is a warning sign about the realism of the inputs. The discount rate r is explained in discount rates and the CAPM.

Method 2: exit multiple

Apply a market multiple to a final-year metric:

Terminal value = final-year EBITDA (or EBIT) x exit multiple

This is simple and anchored to how businesses actually trade. But it imports today's market mood into the model, so a DCF built to check whether a stock is expensive can end up assuming the market will still be pricing it generously. An exit multiple also implies a growth rate, and it is worth backing that growth rate out to see whether it is plausible.

Example (illustrative numbers)

Final-year (year 10) free cash flow is $100. Here is the terminal value at year 10 across different discount rates and growth rates:

For example, at 9% and 3%: 100 x 1.03 / 0.06 = 1,717.

The lowest figure is 1,275 and the highest is 2,600, so reasonable-looking choices produce a spread of more than two to one in the terminal value. Discounted back, that carries straight through to the share value. This is why the same company can look cheap or expensive depending on a couple of percentage points that no one can know precisely.

Choosing the growth rate

Practical guardrails:

• Keep it low. A common convention is a rate in the low single digits, at or below the long-run growth of the economy and often below the risk-free rate. A business cannot grow faster than the economy indefinitely, or it would eventually become the economy.
• Link it to inflation and real growth. Because terminal growth is nominal, it includes expected inflation. See inflation and stocks.
• Make it consistent with reinvestment. Growth requires investment. A useful consistency check is that growth equals the reinvestment rate multiplied by the return on new capital. If the model assumes 3% growth with no reinvestment, it is getting growth for free.
• Be cautious about returns on new capital. A conservative assumption is that, in the long run, competition pushes returns on new investment toward the cost of capital, so growth adds little extra value. See return on invested capital.

Reducing the terminal value problem

• Longer or staged forecasts. A fade period in which growth declines gradually toward the terminal rate avoids a jarring jump from high growth to low.
• Normalize the final year. Use a mid-cycle margin and a realistic level of capital spending rather than a peak-year figure.
• Show the share of value. Report what percentage of value comes from the terminal value. If it is very high, the model is mostly an expression of one assumption.
• Run a reverse DCF. Start from the current price and ask what terminal assumptions it implies; see reverse DCF.

Common mistakes

• Using a growth rate at or above the discount rate, which produces nonsense.
• Applying a terminal growth rate to a final year that is abnormally high or low.
• Using a peak-cycle exit multiple.
• Mixing the discount rate across parts of the model, or applying a perpetuity formula to a company still in a high-growth phase.
• Treating the terminal value as a fact because it comes out of a formula. It is an assumption with a formula attached.

Where to see it on StocksXray

StocksXray's fair value includes a 10-year DCF scaled to the market's average, blended in equal weight with peer multiples and a fundamentals-adjusted multiple. Because typical model error is around 40%, the site shows fair value as a range. See the methodology page.

Key takeaways

• Terminal value stands in for all cash flows after the forecast period and often supplies most of a DCF's value.
• The two methods are the growing perpetuity, FCF x (1 + g) / (r - g), and the exit multiple.
• Small changes in g or r move the result a great deal, as the sensitivity table shows.
• Keep terminal growth modest, tie it to reinvestment, and state how much of the value it represents.