After reading this, you'll be able to build a working DCF model for any company from scratch — and, more usefully, you'll see exactly which assumptions are carrying all the weight.
TL;DR
- A DCF converts a company's expected future cash flows into what they are worth in today's money.
- Three inputs drive almost everything: (1) how fast cash flows grow, (2) the discount rate, and (3) the "terminal value" — what happens after your forecast ends.
- The terminal value typically represents 60–80% of your estimated total value, so the assumptions that feel furthest away matter most.
- A DCF is not a price-prediction machine. It is a way to make your assumptions explicit and test how sensitive your conclusion is to each one.
- A rough, honest model built on assumptions you understand beats a precise spreadsheet built on numbers you borrowed.
Why discount cash flows?
Start from something you already know: you'd rather receive money today than the same amount a year from now. Not because you're impatient, but because the money today can be invested and earn a return over the next year.
This is the time value of money — and a DCF does nothing more than apply it consistently. It takes the cash a company is expected to earn in the future and converts each future amount into what it is worth right now. Add those present-value figures together and you get an estimate of what the business is worth today.
The word "discounted" means: reduced. We shrink each future cash flow to account for the wait.
Step 1: Estimate free cash flow
Free cash flow (FCF) is the cash a business generates after spending on the assets it needs to run and grow. The simplest version:
FCF = Operating cash flow − Capital expenditure (capex)
Both numbers live in the cash flow statement of any annual report.
Your job is to project FCF for the next 5–10 years. Start with the most recent full-year FCF as your base, then estimate a growth rate you can defend. For a stable, predictable business 5 years is usually enough. For a high-growth company where much of the value lies further out, 10 years is common.
Growth rate tiers — a rough guide
| Stage | Typical FCF growth | Example |
|---|---|---|
| High-growth | 15–30% per year | Young SaaS, post-launch biotech |
| Maturing | 8–15% per year | Established software, consumer brands |
| Stable / large-cap | 3–8% per year | Utilities, large industrials |
| Declining | 0–3% or negative | Commoditising businesses |
Be conservative. No company can grow faster than its addressable market indefinitely. Optimistic FCF projections are the single most common source of overvaluation.
Worked example: Company A
Suppose Company A earned $100 in FCF last year, and you expect FCF to grow at 10% per year for 5 years:
| Year | FCF |
|---|---|
| 1 | $110 |
| 2 | $121 |
| 3 | $133 |
| 4 | $146 |
| 5 | $161 |
These are raw future figures — before any discounting.
Step 2: Choose a discount rate
The discount rate represents the return you require for taking on the risk of owning this business. If risk-free government bonds yield 5%, a riskier business must offer a higher expected return or a rational investor would put the money in bonds instead.
For a first-pass equity model, the most practical approach is to estimate a cost of equity using the Capital Asset Pricing Model:
Cost of equity = Risk-free rate + Beta × Equity risk premium
- Risk-free rate: the yield on a long-term government bond in the currency the company earns in
- Beta: how much the stock moves relative to the broader market (available from any data provider)
- Equity risk premium: the extra return investors historically demand for equities over bonds — commonly estimated at 5–7% for developed markets
Many practitioners simply use a round rate anchored to business risk and adjust from there. The model is rarely sensitive enough to the difference between 10% and 11% to make precision worth the effort.
Discount rate quick reference
| Risk profile | Typical cost of equity |
|---|---|
| Low risk (utilities, large staples) | 7–9% |
| Medium risk (established businesses) | 9–12% |
| High risk (cyclicals, growth) | 12–16% |
| Speculative | 16%+ |
For Company A — a stable, mid-size business — use 10%.
Step 3: Calculate the terminal value
You cannot project cash flows forever. The terminal value captures everything that happens after your explicit forecast period, by assuming the business continues at a slow, sustainable growth rate (the terminal growth rate, or g).
The most common formula is the Gordon Growth Model:
Terminal value = Final year FCF × (1 + g) ÷ (discount rate − g)
Convention: cap g at the long-run nominal GDP growth rate of the economy the company operates in. Using a terminal growth rate close to the discount rate causes the formula to blow up to implausible values.
Continuing the Company A example:
- Final year FCF: $161
- Terminal growth rate (g): 3%
- Discount rate: 10%
Terminal value = $161 × 1.03 ÷ (0.10 − 0.03) = $165.83 ÷ 0.07 ≈ $2,369
Compare this to the sum of the five projected FCFs ($671 un-discounted). The terminal value is already more than three times larger — and once discounted (step 4) it still represents the majority of total estimated value. This is why the terminal growth rate assumption matters so much: small changes in g move the estimated value significantly.
*Sensitivity of terminal value to g***
| Terminal growth rate | Terminal value |
|---|---|
| 2% | $2,070 |
| 3% | $2,369 |
| 4% | $2,770 |
| 5% | $3,358 |
A 3-percentage-point shift in g changes the terminal value by more than 60%. Know your assumption before you rely on the output.
Step 4: Discount everything back to today
Every future figure — the projected FCFs and the terminal value — must be converted into today's dollars. The formula for a single amount in year n:
Present value = Amount ÷ (1 + discount rate)ⁿ
Company A: full valuation
| Year | Amount | Discount factor | Present value |
|---|---|---|---|
| 1 | $110 | 1.10¹ | $100 |
| 2 | $121 | 1.10² | $100 |
| 3 | $133 | 1.10³ | $100 |
| 4 | $146 | 1.10⁴ | $100 |
| 5 | $161 | 1.10⁵ | $100 |
| Terminal | $2,369 | 1.10⁵ | $1,472 |
| Total | $1,972 |
The yearly FCFs discount to approximately $100 each because FCF growth (10%) equals the discount rate (10%) — a coincidence in this example that makes the table clean. In practice, if a business grows slower than the discount rate, near-term FCFs carry more weight; if faster, later years matter more.
$1,972 is the estimated enterprise value under these assumptions. To get equity value, subtract net debt (or add net cash) and divide by diluted shares outstanding. That per-share number is your estimate of intrinsic value — the price at which you would expect a fair return.
Common mistakes
1. Copying growth rates without decomposing them. If a sell-side report says "15% FCF growth," ask: what revenue growth does that imply? What margin assumption? Borrowing a number without understanding the mechanism embeds someone else's bias into your model.
2. Setting the terminal growth rate too high. A terminal growth rate of 8% in a 10% discount rate model means the denominator shrinks to 0.02 and the terminal value becomes astronomical. Use a rate that reflects long-run nominal GDP growth — usually 2–5% depending on the economy.
3. Projecting earnings instead of free cash flow. For capital-heavy businesses, earnings and FCF diverge significantly. FCF is what actually accrues to owners after maintaining and growing the asset base; earnings can flatter a business that needs constant reinvestment.
4. Treating the output as precise. Build a quick sensitivity table: run the model with your base case, then with growth 2% higher and 2% lower. If intrinsic value swings from $10 to $40 per share, the thesis rests on a narrow assumption — that's worth knowing before you size the position.
5. Not cross-checking with multiples. If your DCF implies a 50× EV/FCF for a slow-growth business, the assumptions are almost certainly too optimistic. Multiples are a sanity check on the model, not a substitute for it.
Build the thesis in your notes, not the spreadsheet
The most useful version of a DCF is not a fifty-tab workbook. It is a clearly written record of three numbers — growth rate, discount rate, terminal growth rate — and the specific reasoning behind each one.
Write it as you would explain it to a sceptical colleague: "I'm using 10% FCF growth for 5 years because the company is taking share in an underpenetrated market and margins have room to expand; 3% terminal growth because it is a domestically focused business unlikely to outgrow nominal GDP long-term."
That sentence is worth more than the precise formulas, because it makes your thesis visible to your future self — the one who will revisit these notes when results diverge from expectations.
See [What Is an Investment Thesis](what-is-an-investment-thesis.md) for how to structure the reasoning before you build the model, and [The Investing Decision Journal](the-investing-decision-journal.md) for how to track what you assumed versus what happened.
Try this: Find the cash flow statement for a company you already follow. Look up its most recent annual operating cash flow and capex, compute FCF, and pick a 5-year growth rate you can defend in one sentence. Plug the numbers into Step 4 above. Don't optimise — just write down why you chose the numbers you chose. That discipline is where the real value of a DCF lives.