DCF Model: Discounted Cash Flow
Project free cash flow, discount at WACC, add terminal value — the intrinsic valuation method every banker pretends is science.
What is a DCF?
A Discounted Cash Flow (DCF) model estimates what a company is “worth” by forecasting Free Cash Flow (FCF) into the future, then discounting those cash flows back to today using a rate that reflects risk. The intuition is blunt: a dollar tomorrow is not worth a dollar today — so future cash gets haircut.
Here is the uncomfortable truth finance professors gloss over: a DCF is precise if your assumptions are right, which they probably are not. The model is still useful because it forces you to spell out what you believe about growth, risk, and terminal economics — and makes those beliefs easy to stress-test.
You should treat a DCF less like a calculator that outputs “the answer” and more like a structured argument. The output is only as honest as your inputs — and every input is a bet about the future. That is why banks put sensitivity tables next to DCFs: the point is not to nail the fourth decimal; it is to show what has to be true for the story to hold.
If you are new to this, keep one skeptical line in your head: DCF is finance dogma dressed up as physics. The spreadsheet will always produce a number. Your job is to know which cells are doing the persuasion.
Step 1 — Project Free Cash Flow
You start from the most recent annual Free Cash Flow: operating cash flow minus capex (sometimes adjusted for one-offs). Then you apply a growth path — a single rate, a step-down schedule, or a two-stage story (faster early, slower late). Most public-company DCFs use 5–10 explicit years.
Why not net income? Because net income is an accounting score; FCF is closer to “cash the owners could actually walk away with after reinvesting in the business.” If you want the philosophy, read our FCF guide — the short version is: cash pays dividends; accruals pay for creativity.
Why FCF forecasts go wrong: Management guidance is not a forecast — it is a story. Cyclical businesses look cheap at peak FCF and expensive at trough FCF if you extrapolate the wrong year. Capex can spike for a new fab, a fleet refresh, or AI infrastructure — your “steady state” may be years away. Working capital surprises (inventory, receivables, payables) can swallow cash even when revenue looks fine. If your model assumes smooth 8% growth forever while reality is lumpy, you have not modeled a business — you have modeled a fantasy with decimal places.
Visual — FCF timeline & discount factors
Place a visual here: five projected FCF bars stepping up (or down), with a second row showing discount factors 1/(1+WACC)^n shrinking each year — readers grok “time value” faster than from formulas alone.
Step 2 — Choose a Discount Rate (WACC)
The Weighted Average Cost of Capital blends the cost of equity (often from CAPM: risk-free rate + beta × equity risk premium) with the after-tax cost of debt, weighted by how much of the firm is funded by each. In practice, WACC is the single input people fight about most after terminal growth.
Rule of thumb: large, mature businesses often land in high single digits to low double digits for WACC — but “typical” is not “correct.” A higher WACC punishes distant cash flows more aggressively, which is why two reasonable analysts can produce wildly different “intrinsic values” without anyone committing fraud.
When you move WACC by 50–100 basis points, you are not “fine-tuning” — you are often shifting the whole fair-value story. That is not an argument against WACC; it is an argument for ranges, not heroics.
Step 3 — Add Terminal Value (Gordon Growth)
You cannot forecast line-by-line FCF to infinity — so you cap the explicit period and lump everything after year *n* into terminal value (TV). A common approach is the Gordon Growth Model (perpetuity growth): TV = FCFₙ × (1 + g) / (WACC − g), where *g* is long-run growth (often anchored near nominal GDP: roughly 2–3% in developed markets).
TV is not a footnote: for many names, 60–80% of enterprise value comes from the terminal chunk. That means your “small” tweak to *g* or exit assumptions is often doing more work than your entire five-year forecast.
Terminal value sensitivity (why your “2.5%” matters): Gordon Growth explodes when (WACC − g) gets small. Example: if WACC is 9% and you move *g* from 2.5% to 3.5%, the denominator shrinks from 6.5% to 5.5% — terminal value jumps ~18% before discounting, and total enterprise value can easily swing 15–25%+ when TV dominates. That is not a bug; it is the math punishing optimism dressed up as precision.
Visual — WACC vs. terminal growth grid
Place a heatmap or grid: rows = WACC (e.g. 7.5%–10.5%), columns = terminal growth (e.g. 2.0%–3.5%), cells = implied EV or share price. This is the chart that makes grown analysts cry.
Step 4 — Net Debt Adjustment
The enterprise DCF spits out Enterprise Value (EV) — the value of the whole operating business to all capital providers. To get equity value, you bridge from EV using debt and cash: Equity Value = EV − Total Debt + Cash (same thing as EV minus net debt, where net debt = debt minus cash). Divide by diluted shares outstanding for an implied price per share.
If you forget cash, you punish companies for being liquid; if you forget debt, you reward them for borrowing. Neither mistake ages well in a diligence meeting.
Common DCF Mistakes
- Double-counting growth: You forecast supernormal growth for five years *and* bake an aggressive terminal *g* — you have smuggled two “hockey sticks” into one model.
- Mismatching cash flows and discount rates: FCFF must pair with WACC; FCFE pairs with cost of equity. Mixing them is how you get a number that looks authoritative and is nonsense.
- Ignoring reinvestment: Revenue growth is not free. If margins expand forever with no capex or working capital drag, you are valuing a spreadsheet, not a company.
- Treating synergies as base case in a sell-side pitch: Synergies are optional, timing-uncertain, and politically motivated — good for negotiation, dangerous as a default assumption.
- Using the DCF output as a price target without ranges: A point estimate implies false confidence. Sensitivity tables exist because the honest answer is usually a band.
Red Flags in Your Assumptions
Overly optimistic FCF: FCF growing faster than revenue forever, margins expanding every year with no competitive response, capex flat while revenue doubles — these are not “bull cases,” they are plot holes.
Overly optimistic terminal value: Terminal growth above nominal GDP for a mature business, or an exit multiple at the top of the peer range “because quality.” If the terminal chunk is most of value, your “quality” story is doing almost all the work.
Overly optimistic WACC: Cherry-picking a low beta, stale debt yields, or a compressed ERP to make the math behave. If your WACC is always 0.5% lower than the next analyst’s, ask whether you are modeling Microsoft or modeling wishful thinking.
How to Challenge a DCF in a Pitch
In a live pitch, you rarely “win” by debating Excel settings — you win by exposing which assumptions are doing the heavy lifting. Ask: What FCF base year are you normalizing, and what one-offs were stripped? Why this five-year growth path versus management guidance versus history? What is the implied exit multiple if we back-solve the terminal value? How does the model behave if WACC moves 100 bps?
Follow-ups that separate diligence from theater: What revenue growth is embedded in the out-year FCF, and is it above industry growth? Where does margin expansion come from — mix, price, or cost cuts that competitors can copy? If we haircut terminal growth by 50 bps, what happens to the implied multiple?
Analysts commonly disagree on risk premia and beta (cost of equity), long-run margins (especially in cyclicals), capex intensity, working capital, and terminal methodology (growth vs. exit multiple). If those debates sound boring, good — boring is where money is made or lost.
DCF vs. Comps vs. LBO: When to Use Each
Trading comps tell you what the market is paying *today* for similar businesses — fast, market-anchored, and hard to defend if your peer set is cherry-picked. They are best when comparables are real and the stock is liquid enough that prices mean something.
DCF is an intrinsic, fundamentals-first view — great when cash flows are stable enough to forecast and when you care about drivers rather than “what the market thinks.” It is worst when the business is early-stage, hyper-cyclical, or when value lives mostly off-balance-sheet (some brands, platforms, biotech optionality).
LBO / feasibility matters when a deal must work for a financial buyer with leverage — it is less “fair value” and more “can this cash flow service the debt stack and still hit hurdle returns?” Use it alongside DCF and comps, not instead of them. On Briefed, the Football Field view exists precisely because adults present ranges, not prophecies.
In practice, you triangulate: comps anchor you to market reality; DCF forces a cash-flow story; LBO tells you whether a deal can survive leverage. None of them is “true” — together they are a valuation toolkit, not a religion.
Worked Example — Microsoft (Illustrative, Rounded)
Below is a teaching walkthrough with rounded numbers so you can audit the logic on paper. It is not investment advice, and it will not match “the market price” — real markets embed expectations, optionality, and discount-rate debates a toy model will not capture.
We use Microsoft because filings are clean and FCF is large enough that you can focus on mechanics instead of weird adjustments. If your implied value disagrees with the stock, that is normal: the market may be pricing faster long-run growth, lower discount rates, balance-sheet optionality, or simply a different narrative. Your job is to know which story your model encodes.
Step A — Find historical FCF in the 10-K
Open Microsoft’s latest Form 10-K, go to the Consolidated Statements of Cash Flows. For fiscal 2024 (year ended June 30, 2024), identify cash from operations and subtract capital expenditures (often “additions to property and equipment” and sometimes additional investing outflows, depending on presentation). In a simplified screen, you might see operating cash flow on the order of ~$118B and capex on the order of ~$44B, implying unlevered FCF on the order of ~$74B — verify from the filing before treating any number as gospel.
Step B — Forecast five years of FCF
Suppose you believe a mature-but-still-growing profile and set 5% annual growth for years 1–5 off a $74B base (year 0). Then: Year 1 ≈ $77.7B, Year 2 ≈ $81.6B, Year 3 ≈ $85.7B, Year 4 ≈ $90.0B, Year 5 ≈ $94.5B. Your job in real life is to defend that path with revenue, margin, capex, and working-capital logic — here we are showing mechanics.
Step C — WACC (worked numbers)
Say you estimate WACC = 8.5% after blending cost of equity and after-tax cost of debt (see WACC for the full formula). Discount factors for years 1–5 are 1/(1.085)ᵗ — roughly 0.922, 0.849, 0.783, 0.722, 0.665.
Step D — Terminal value (Gordon Growth)
Set terminal growth g = 2.5% with the same WACC 8.5%. Terminal value at the end of year 5 is: TV₅ = FCF₅ × (1 + g) / (WACC − g) = 94.5 × 1.025 / (0.085 − 0.025) ≈ $1,614B (rounded).
Step E — Enterprise value and implied share price
Present-value the five explicit FCFs using the factors above — the sum lands near ~$336B (rounded). Present-value the terminal value by discounting TV₅ five years: ~$1,614B × 0.665 ≈ $1,073B. Enterprise value ≈ $336B + $1,073B ≈ $1,409B.
To bridge to equity, take EV, subtract debt, add cash (use the balance sheet — Microsoft has been net cash in many periods). Suppose for illustration the net cash position adds ~$50B to equity versus a naive “debt-only” picture; in practice you must pull the exact components from the filing. Divide by diluted shares (order-of-magnitude ~7.5B) to get a ballpark implied price — then compare to the live quote and ask what the market is embedding that your assumptions do not.
Related Reading
How to use this on Briefed
To save hours building a DCF model from scratch, load any public company in Briefed and adjust the assumptions shown in the Assumptions pane. Watch how changes to growth rate or WACC cascade through to enterprise value — no formula rebuilding required. Pro users can export this to PowerPoint for a pitch book.
Pro highlights:
- Sensitivity analysis export — show your range, not just your base case.
- Multi-scenario comparison — side-by-side cases without duplicating workbooks.
- Formatted Excel / deck-ready outputs — fewer nights lost to layout tweaks.
- Sensitivity analysis export — ranges, not just a point estimate
- Multi-scenario comparison without duplicate workbooks
- Deck-ready exports for IC memos and pitch books