A DCF answers one question: what is this business worth today, based on the cash it will produce? First forecast the free cash flow, which is the cash left after running costs, taxes and investment. Then shrink each future year back to today using a discount rate. Add a terminal value to cover the years beyond the forecast, and subtract debt to reach the value that belongs to shareholders. For the example company, Harbor Tools, this gives $15.24 a share with one terminal method and $13.78 with another.
Why it matters: The model is only as good as its forecast and its discount rate, so test both.
Summary: A DCF forecasts unlevered free cash flow, EBIT × (1 − t) + D&A − capex − ΔNWC, discounts it at the WACC, adds a terminal value for the years beyond the forecast and bridges from enterprise value to equity value. For the illustrative Harbor Tools, it gives $15.24 a share with Gordon growth and $13.78 with an exit multiple.
- Terminal value supplied 78% of Harbor’s enterprise value, so its assumptions deserve the most scrutiny.
- The mid-year convention raised enterprise value by 4.2%; Gordon terminal value is discounted N − 0.5 years, an exit-multiple value N years.
- A 3×3 sensitivity table showed $11.26 to $22.20 a share; it varies one DCF’s inputs and is not the football field of Section 1.10: Triangulating Valuation — Putting It All Together.
- Growth must be paid for with reinvestment: the same 4% terminal growth is worth $11.27 or $18.96 a share depending on the return new capital earns.
A DCF values a business as the present value of all the cash it will generate for its owners in the future. Building one, in practice, moves through five stages:
| Stage | What Happens |
|---|---|
| 1. Project free cash flow | Forecast revenue, margins, taxes, capital expenditure, and working capital changes for 5–10 years to arrive at unlevered free cash flow (FCF) each year: EBIT × (1 − tax rate) + D&A (depreciation and amortization) − capex − increase in net working capital |
| 2. Choose a discount rate | Almost always the Weighted Average Cost of Capital (WACC) — covered in 1.4 — representing the blended return demanded by both lenders and shareholders |
| 3. Discount each year’s FCF | Apply the PV formula from 1.2 to every projected year of cash flow |
| 4. Calculate terminal value | Since a business does not stop existing after year 10, estimate the value of all cash flows beyond the forecast period in a single lump figure |
| 5. Sum and adjust | Add the discounted FCFs and the discounted terminal value to get Enterprise Value; subtract debt and other non-equity claims and add cash to arrive at Equity Value (the EV-to-equity bridge) |
Terminal value deserves special attention, because in most real DCFs it represents the majority of the total valuation (78% of enterprise value in the worked example below). The most common method is the Gordon Growth Model: Terminal Value = FCF in the first year after the forecast ÷ (WACC − long-term growth rate), the growing perpetuity of Section 1.2: The Time Value of Money — The Idea Underneath Everything. The long-term growth rate is almost always kept modest — at or below long-run nominal GDP growth (about 4% for the US) — because no company can outgrow the entire economy forever without eventually becoming the entire economy.
Because terminal value dominates the total, a DCF’s output is extremely sensitive to two inputs: the discount rate and the long-term growth rate. In the worked example below, moving growth from 2% to 3% raises value per share by 19%, and cutting the discount rate from 9% to 8% raises it by 27%. This is precisely why professionals never present a single DCF number — they present a sensitivity table showing valuation across a range of discount rates and growth assumptions, because the honest answer to “what is this company worth” is a range, not a point. (The football field in Section 1.10: Triangulating Valuation — Putting It All Together is a different chart: it compares the ranges produced by different methods.)
Harbor Tools Inc. is an illustrative US maker of industrial hand tools (USD millions except per share).
| Input | Value |
|---|---|
| Revenue, Year 0 | $1,000.0 |
| Revenue growth, Years 1–5 | 8%, 7%, 6%, 5%, 4% |
| EBIT margin / tax rate (t) | 14% / 25% (21% federal + 4% assumed state) |
| D&A / capex / net working capital (NWC) | 4% / 5% / 12% of revenue |
| WACC (Sections 1.4 and 1.5) | 8.5% |
| Terminal growth (g) / exit multiple | 3.0% / 9.0x EBITDA (assumed peer level) |
| Debt / cash / diluted shares | $600.0 / $100.0 / 100.0 million |
Step 1, free cash flow, starts from operating profit, because a DCF at WACC values the business for all investors. Year 1: NOPAT (net operating profit after tax) = 151.2 × 0.75 = 113.4; ΔNWC = 0.12 × (1,080.0 − 1,000.0) = 9.6; FCF = 113.4 + 43.2 − 54.0 − 9.6 = 93.0. Each year is discounted at mid-year, because cash arrives through the year (the mid-year convention).
| Year 1 | Year 2 | Year 3 | Year 4 | Year 5 | |
|---|---|---|---|---|---|
| Revenue | 1,080.0 | 1,155.6 | 1,224.9 | 1,286.2 | 1,337.6 |
| EBIT | 151.2 | 161.8 | 171.5 | 180.1 | 187.3 |
| NOPAT = EBIT × (1 − t) | 113.4 | 121.3 | 128.6 | 135.0 | 140.5 |
| + D&A | 43.2 | 46.2 | 49.0 | 51.4 | 53.5 |
| − Capex | 54.0 | 57.8 | 61.2 | 64.3 | 66.9 |
| − ΔNWC | 9.6 | 9.1 | 8.3 | 7.3 | 6.2 |
| = FCF | 93.0 | 100.7 | 108.0 | 114.8 | 120.9 |
| × 1 ÷ 1.085^(t − 0.5) | 0.9600 | 0.8848 | 0.8155 | 0.7516 | 0.6927 |
| PV of FCF | 89.3 | 89.1 | 88.1 | 86.3 | 83.8 |
Step 2, terminal value, two ways. Gordon growth: Year 6 is rebuilt at 3% growth (NOPAT 144.7 less reinvestment, capex − D&A + ΔNWC, of 18.6), so FCF₆ = 126.1 and TV = 126.1 ÷ (0.085 − 0.030) = 2,292.2, discounted 4.5 years because its cash also arrives mid-year. Exit multiple: TV = 9.0 × Year-5 EBITDA of 240.8 = 2,167.0, discounted 5 years because the sale is assumed at year-end.
| Gordon (g = 3.0%) | Exit (9.0x) | |
|---|---|---|
| Sum of PV of FCF | 436.6 | 436.6 |
| Terminal value (TV), end of Year 5 | 2,292.2 | 2,167.0 |
| PV of TV (years discounted) | 1,587.9 (4.5) | 1,441.1 (5.0) |
| Enterprise value | 2,024.5 | 1,877.7 |
| − Debt + cash | −500.0 | −500.0 |
| Equity value / per share | 1,524.5 / $15.24 | 1,377.7 / $13.78 |
Step 3, cross-checks. Compared undiscounted, the Gordon value implies 9.5x Year-5 EBITDA and the 9.0x exit implies perpetual growth of 2.7%; compared at present value, which credits the Gordon value’s half-year-earlier discounting, the figures are 9.9x and 2.4%. Without the mid-year convention the Gordon EV would be 1,943.5, 4.0% lower.
Step 4, sensitivity table: value per share (Gordon) by WACC and terminal growth.
| WACC \ g | 2.5% | 3.0% | 3.5% |
|---|---|---|---|
| 7.5% | $17.70 | $19.70 | $22.20 |
| 8.5% | $13.94 | $15.24 | $16.81 |
| 9.5% | $11.26 | $12.16 | $13.22 |
Try it yourself: DCF value, sensitivity grid and the EV-to-equity bridge
Pre-filled with the illustrative Harbor Tools inputs from the worked example above (USD millions except per share), so the first results match it: $15.24 a share by the Gordon method and $13.78 at a 9.0x exit multiple. Every default is illustrative. Change any input and every table recalculates.
- Value per share, Gordon
- $15.24equity $1,524.5m ÷ 100.0m shares
- Enterprise value, Gordon
- $2,024.5mPV of FCF 436.6 + PV of TV 1,587.9
- Value per share, exit multiple
- $13.789.0x Year-5 EBITDA; EV $1,877.7m
- Terminal value share of EV
- 78.4%Gordon method
Free cash flow build, USD millions
| Year 1 | Year 2 | Year 3 | Year 4 | Year 5 | |
|---|---|---|---|---|---|
| Revenue | 1,080.0 | 1,155.6 | 1,224.9 | 1,286.2 | 1,337.6 |
| EBIT | 151.2 | 161.8 | 171.5 | 180.1 | 187.3 |
| NOPAT = EBIT × (1 − t) | 113.4 | 121.3 | 128.6 | 135.0 | 140.5 |
| + D&A | 43.2 | 46.2 | 49.0 | 51.4 | 53.5 |
| − Capex | 54.0 | 57.8 | 61.2 | 64.3 | 66.9 |
| − ΔNWC | 9.6 | 9.1 | 8.3 | 7.3 | 6.2 |
| = FCF | 93.0 | 100.7 | 108.0 | 114.8 | 120.9 |
| × 1 ÷ (1 + WACC)^(t − 0.5) | 0.9600 | 0.8848 | 0.8155 | 0.7516 | 0.6927 |
| PV of FCF | 89.3 | 89.1 | 88.1 | 86.3 | 83.8 |
Terminal value and the EV-to-equity bridge, USD millions
| Gordon (g = 3.0%) | Exit (9.0x) | |
|---|---|---|
| Sum of PV of FCF | 436.6 | 436.6 |
| Year-6 FCF rebuilt at g | 126.1 | — |
| Terminal value, end of Year 5 | 2,292.2 | 2,167.0 |
| PV of TV (years discounted) | 1,587.9 (4.5) | 1,441.1 (5.0) |
| Enterprise value | 2,024.5 | 1,877.7 |
| − Debt + cash | −500.0 | −500.0 |
| Equity value | 1,524.5 | 1,377.7 |
| Value per share | $15.24 | $13.78 |
Sensitivity: value per share (Gordon) by WACC (rows) and terminal growth (columns)
| WACC \ g | 2.5% | 3.0% | 3.5% |
|---|---|---|---|
| 7.5% | $17.70 | $19.70 | $22.20 |
| 8.5% | $13.94 | $15.24 | $16.81 |
| 9.5% | $11.26 | $12.16 | $13.22 |
How to read this: Free cash flow is NOPAT plus D&A minus capex and the increase in working capital, discounted at mid-year. The Gordon terminal value rebuilds Year 6 at the terminal growth rate and is discounted 4.5 years; the exit-multiple value is discounted 5 years. The sensitivity grid moves the WACC by one point and growth by half a point around your inputs, with your case highlighted in the center.
Assumptions: unlevered free cash flow at a constant WACC, mid-year convention, ΔNWC = NWC % × change in revenue, net debt = debt − cash, no minority interests, preferred stock or leases in the bridge. All defaults are the illustrative Harbor Tools figures of the worked example; they are not a forecast for any real company. Figures stay unrounded until display, so totals can differ from the rows by 0.1.
Read the output as a range, $13.78 to $15.24 a share. Terminal value supplies 78% of the Gordon value, so its assumptions deserve the most scrutiny.
Set terminal growth at or below long-run nominal GDP growth (about 4% for the US) and no higher than your reinvestment supports: g = reinvestment rate × RONIC, the return on new invested capital. If terminal value exceeds about 85% of enterprise value, lengthen the forecast. Run both terminal methods; if they differ by more than about 15%, reconcile the implied multiple and growth first. Both triggers are heuristics; early-stage companies breach the first.
Letting terminal growth come for free. Harbor reinvests 18.6 of its 144.7 Year-6 NOPAT (12.85%) to grow 3%, which implies RONIC = 3.0% ÷ 12.85% = 23.3%. Raise g to 4% with no extra reinvestment and value jumps from $15.24 to $18.96 (+24%); paying for that growth at 23.3% gives $18.00. If competition pulls RONIC down to the WACC, growth adds nothing: TV = NOPAT₆ ÷ WACC. At 4% growth NOPAT₆ is 146.1, so TV = 1,718.5, worth $11.27 a share, barely above the $11.16 at 3%. One unstated assumption spans $11.27 to $18.96 at the same 4% growth. Write down the RONIC your terminal value implies and ask whether competitors could earn it.
What is a reasonable terminal growth rate in a DCF?
For a mature US company, usually about 2% to 4% nominal. Federal Reserve projections from September 2026 put longer-run real growth and inflation at 2.0% each, so about 4% is the ceiling for a company that keeps its share of the economy forever, and any rate needs matching reinvestment.
What is the mid-year convention in a DCF?
It discounts each year’s cash flow by t − 0.5 years instead of t, because cash arrives through the year. It raised Harbor’s enterprise value by 4.2%. A Gordon terminal value is then discounted N − 0.5 years, while an exit-multiple terminal value keeps N years because the sale happens at year-end.
What is the difference between unlevered and levered free cash flow?
Unlevered FCF is cash before any payment to lenders, so it belongs to all investors and is discounted at WACC. Levered FCF deducts interest and net borrowing, belongs to shareholders and is discounted at the cost of equity. Volume I’s operating cash flow minus capex is closer to levered, since US GAAP operating cash flow is after interest.
A DCF discounts unlevered free cash flow at the WACC, adds a terminal value by Gordon growth or an exit multiple, and bridges to equity value; Harbor Tools comes out at $13.78 to $15.24 a share, with 78% of value in the terminal value. Mid-year discounting, a sensitivity table and a reinvestment-consistent terminal year are standard practice.
Four questions on this chapter. Decide on your answer first, then click “Reveal Answer.”
1. A company has EBIT of $200 million, a 25% tax rate, D&A of $30 million, capex of $45 million and a $10 million increase in net working capital. What is its unlevered free cash flow?
- $125 million
- $175 million
- $135 million
- $95 million
Reveal Answer
Answer: A. FCF = 200 × 0.75 + 30 − 45 − 10 = $125 million. Skipping tax gives $175 million, skipping ΔNWC $135 million, and skipping the D&A add-back $95 million.
2. Year-6 free cash flow is $126 million, WACC is 8.5% and terminal growth is 3%. What is the Gordon terminal value at the end of Year 5?
- $2,360 million
- $2,291 million
- $1,482 million
- $4,200 million
Reveal Answer
Answer: B. TV = FCF₆ ÷ (WACC − g) = 126 ÷ 0.055 = $2,291 million. Multiplying by 1.03 again double-counts a year of growth; dividing by WACC or g alone is wrong.
3. Under the mid-year convention, how many years is an exit-multiple terminal value discounted in a five-year DCF?
- Five and a half years, to reach the first later year
- None, since a multiple is already a present value
- Five full years, as the sale is assumed at year-end
- Four and a half years, like a Gordon terminal value
Reveal Answer
Answer: C. Mid-year discounting reflects cash arriving through the year; an exit is assumed to happen at the end of Year 5, so its value is discounted 5 years, while a Gordon value uses 4.5.
4. How does a DCF sensitivity table differ from a football field chart?
- It compares methods; a football field varies inputs within one DCF
- It suits private firms; a football field suits only listed companies
- It shows equity values; a football field shows only enterprise values
- It varies inputs within one DCF; a football field compares methods
Reveal Answer
Answer: D. A sensitivity table shows one model’s output across, for example, WACC and growth; a football field (Section 1.10: Triangulating Valuation — Putting It All Together) lines up the value ranges from DCF, comps, precedents and other methods.
- 26 U.S.C. §11, Tax imposed (Cornell LII) — 21% federal corporate rate
- Federal Reserve, FOMC Summary of Economic Projections, September 16, 2026 — Longer-run medians: real GDP 2.0%, PCE inflation 2.0% (terminal growth ceiling, 1.3)