Discounted Cash Flow Model: How to Build a DCF From FCF to Terminal Value
A discounted cash flow model is the cleanest way to value a business on its own merits, because it forces you to write down what the company will earn in cash, decide what those cash flows are worth today given the risk, and stop hiding behind multiples. Every banker, private-equity associate, and sell-side analyst learns the discounted cash flow model on day one, and the same five-step structure runs from a paper LBO scribbled on a steakhouse napkin to a Goldman Sachs fairness opinion (SEC filings routinely include sponsor and target DCFs in DEFM14A proxies, see for example the Activision-Microsoft proxy filed February 2023). This guide walks the entire build in the order you would actually do it on a deal, with named-firm conventions, worked numbers, and the mistakes that get analyst models thrown back across the table.
What a Discounted Cash Flow Model Actually Is
A DCF, short for discounted cash flow, is a valuation method that estimates what a business is worth today by adding up the cash it is forecast to throw off in the future, after adjusting each future dollar downward for the risk and time value of money. The output is an intrinsic value: a number that does not depend on what comparable companies trade for or what was paid in the last similar transaction. McKinsey’s Valuation: Measuring and Managing the Value of Companies (Koller, Goedhart, Wessels, 7th edition, Wiley 2020) frames DCF as the “primary” valuation tool precisely because it isolates the operating drivers, while CFA Institute’s Level II Equity curriculum calls discounted cash flow analysis the foundational technique against which every other approach is benchmarked.
In M&A practice, a discounted cash flow model rarely walks alone. Bankers present it inside a “football field” alongside trading comps, precedent transactions, premiums paid, and 52-week trading range, but it is usually the widest bar on the chart because it is the most assumption-rich. The Delaware Court of Chancery has leaned heavily on DCF in appraisal litigation (see In re Appraisal of Dell Inc., C.A. No. 9322-VCL, 2016, and DFC Global v. Muirfield, 172 A.3d 346 Del. 2017), which is why every public deal proxy you read has a defensible DCF behind the fairness opinion.
Quick-Reference: The DCF Formula in One Line
Every discounted cash flow analysis collapses to one equation. Memorize it and the rest is plumbing.
| Component | Formula | What it means |
|---|---|---|
| Present value of explicit period | Sum from t=1 to N of [FCF_t / (1 + WACC)^t] | Discount each forecast year’s free cash flow back to today |
| Terminal value (Gordon Growth) | TV = [FCF_N+1 / (WACC – g)] | Captures all cash flows beyond the forecast window |
| PV of terminal value | TV / (1 + WACC)^N | Discount the terminal value back to today |
| Enterprise value | EV = PV(explicit FCF) + PV(TV) | Total operating value of the business |
| Equity value | Equity = EV – Net Debt + Cash + non-op assets | What the equity holders get |
| Per-share value | Equity / Diluted shares (treasury method) | Compare to current stock price |
FCF here means unlevered free cash flow (sometimes written FCFF for “free cash flow to the firm”): EBIT times (1 minus tax rate), plus depreciation and amortization, minus capital expenditures, minus the increase in net working capital. This is the cash available to all capital providers (debt plus equity), which is why you discount it at WACC and not at cost of equity. A separate variant called levered DCF discounts FCFE (free cash flow to equity, which is after interest and net debt paydown) at the cost of equity to get straight to equity value, but unlevered DCF is the M&A and corporate-finance default. Aswath Damodaran’s NYU Stern teaching site maintains a free, continuously updated dataset of WACC inputs and DCF templates that practically every analyst eventually downloads.
The Five-Step DCF Build, Start to Finish
Every discounted cash flow model, from a one-tab interview screen to a 30-tab bank model, is the same five steps in the same order. Skip a step and the model is wrong.
| Step | What you do | Output | Typical time on a deal |
|---|---|---|---|
| 1. Forecast FCF | Project revenue, margins, D&A, capex, working capital for 5-10 years | Annual unlevered FCF for the explicit period | 3-10 days (most of the work) |
| 2. Calculate WACC | Cost of equity (CAPM), after-tax cost of debt, weighted by target capital structure | Single discount rate, usually 7-12% | 2-4 hours |
| 3. Discount the FCF | Apply 1/(1+WACC)^t to each year | Present value of each year’s FCF | 30 minutes |
| 4. Terminal value | Gordon Growth or exit multiple, then discount back | PV of TV, typically 60-80% of total EV | 1 hour plus sensitivity |
| 5. Sum + bridge to equity | Add PVs, subtract net debt, divide by diluted shares | Implied equity value per share | 30 minutes |
If you are interviewing for a banking or buyside seat, this five-step structure is the answer to “walk me through a DCF.” Mergers & Inquisitions and Wall Street Oasis both publish hundreds of interview transcripts and the structure does not change between Goldman, Morgan Stanley, Lazard, or a middle-market shop like Houlihan Lokey or Lincoln International.
Step 1: Forecasting Free Cash Flow
The forecast is where 80% of DCF disagreements live. Pick a forecast horizon long enough for the business to reach a steady state (mature growth, normalized margins, normalized reinvestment) but short enough that you are not just making things up. Five years is the corporate-finance default. Ten years is common in sectors with long cash-flow runways (pharma with a patent cliff, infrastructure concessions, oil and gas reserves). Anything beyond 10 years gets compressed into terminal value.
The unlevered free cash flow build, line by line, is:
| Line | How to forecast | Common source |
|---|---|---|
| Revenue | Volume x price, or growth rate off a base year | Management guidance, sell-side research consensus on FactSet or Visible Alpha |
| (-) COGS, (-) Opex | Hold or trend gross margin and EBITDA margin to a normalized level | Historical margins plus industry benchmarks |
| EBIT | Revenue minus COGS minus Opex minus D&A | From the operating model |
| (-) Taxes on EBIT | EBIT x effective tax rate (usually 21% US federal plus state blend, see IRC Section 11) | Effective rate from 10-K tax footnote |
| (+) D&A | Tie to capex schedule, usually 80-110% of capex in steady state | 10-K cash flow statement |
| (-) Capex | As % of revenue, split into maintenance and growth capex if disclosed | 10-K + management guidance |
| (-) Change in NWC | NWC as % of revenue, then take the year-over-year delta | Working capital roll from historical balance sheet |
| = Unlevered FCF | Sum of the above | Plug into DCF |
Two anchor disciplines separate a defensible forecast from a fantasy one. First, terminal-year reinvestment rate (capex plus change in NWC, divided by EBIT after tax) should equal terminal growth divided by return on invested capital, otherwise the model has a hidden inconsistency where the company keeps growing without spending to grow. McKinsey calls this the “value driver” identity: g = ROIC x investment rate. Second, terminal-year margins and capex intensity should look like a mature peer (think Coca-Cola, not a five-year-old SaaS company), not like the projection year before.
Sell-side analysts at Lazard, Centerview, and Houlihan Lokey routinely build a “Base / Upside / Downside” trio of forecasts and run each through the DCF, then present the equity-value range in a football field. Pitchbook’s 2025 M&A Outlook noted that PE buyers were tightening their downside-case revenue assumptions by 200-400 bps versus 2021-22 underwriting after the rate shock, which is the kind of disciplined sensitivity that should live inside the FCF forecast, not as an afterthought.
Step 2: Calculating WACC, the Discount Rate
WACC, the weighted average cost of capital, is the blended rate of return that all capital providers demand. The formula is straightforward; the inputs are where everyone fights.
| WACC component | Formula | Practical source |
|---|---|---|
| Cost of equity (Ke) | Rf + Beta x ERP (CAPM) | 10Y Treasury for Rf, Damodaran or Bloomberg for ERP and beta |
| Risk-free rate (Rf) | 10-year US Treasury yield (or local sovereign for non-US) | US Treasury daily yield curve |
| Equity risk premium (ERP) | Historical (~5-6%) or implied (~4-5%) | Damodaran updated monthly on NYU Stern; Duff & Phelps (now Kroll) annual study |
| Beta | Lever industry unlevered beta to target capital structure: BL = BU x [1 + (1-t)(D/E)] | Bloomberg, Capital IQ, or Damodaran’s industry tables |
| Cost of debt (Kd) | YTM on outstanding bonds, or synthetic credit spread over Rf based on interest coverage ratio | 10-K debt schedule, FINRA TRACE for bond pricing, Moody’s or S&P rating |
| Tax rate | Marginal rate (21% US federal plus state blend) for terminal cost of debt | 10-K tax footnote, state apportionment |
| Capital structure weights | Target (long-run) D/V and E/V at market values, not book | Industry peer median or stated company policy |
WACC equals (E/V) x Ke + (D/V) x Kd x (1 – tax rate). For a typical US large-cap industrial in early 2026, with a 4.2% 10-year Treasury yield, a 5.5% ERP, a relevered beta of 1.10, a 5.5% pre-tax cost of debt, a 25% effective tax rate, and a target 30% debt / 70% equity capital structure, the WACC arithmetic is: cost of equity = 4.2% + 1.10 x 5.5% = 10.25%; after-tax cost of debt = 5.5% x (1 – 0.25) = 4.1%; WACC = 0.7 x 10.25% + 0.3 x 4.1% = 8.4%. Move any one input meaningfully and the WACC swings 50-100 bps, which is why sensitivity tables are mandatory in step 8.
For private companies, add a size premium (Kroll’s annual size premium study tables the historical excess return by size decile) and, for very small or thinly traded businesses, a company-specific risk premium. The CFA Level II equity curriculum walks through both the build-up method and modified CAPM in detail. Kroll’s Cost of Capital Navigator (formerly the Duff & Phelps Valuation Handbook) is the dominant US private-company appraisal reference and gets cited in nearly every fairness opinion produced by middle-market banks.
One subtlety that trips up new analysts: cost of debt should reflect the company’s current marginal cost of borrowing, not the average coupon on outstanding debt. If a company issued bonds in 2020 at a 3.5% coupon but new-issue yields for the same credit are now 6.5% (mid-2025 high-yield index per FRED’s BAML HY index), use 6.5%. Same for beta: use a forward-looking adjusted beta (Bloomberg’s adjusted beta blends 67% raw historical beta with 33% market beta of 1.0, following the Blume adjustment) rather than the raw five-year regression beta, since betas mean-revert.
For non-US companies, swap the risk-free rate to the local sovereign yield curve (German bund for Eurozone, JGB for Japan, gilts for UK) and use a country-risk-premium adjustment for emerging markets per Damodaran’s country risk premium tables, which are based on sovereign CDS spreads scaled by equity-to-bond relative volatility. Latham & Watkins’s annual Global M&A Insights report and Houlihan Lokey’s cross-border valuation handbook both walk through the country-premium debate; the simple answer is that an Indonesian or Argentine company needs a WACC several hundred basis points above what its US peer would carry, full stop.
Step 3: Discount Rate Selection (Why It Has to Match the Cash Flow)
The single fastest way to torpedo a DCF is to mismatch the numerator and denominator. The rule, taught in every Wharton, Booth, and Stern corporate-finance class:
| Cash flow you are discounting | Correct discount rate | Output is |
|---|---|---|
| Unlevered FCF (FCFF), pre-interest | WACC | Enterprise value |
| Levered FCF (FCFE), post-interest, post-debt paydown | Cost of equity (Ke) | Equity value directly |
| Dividends | Cost of equity (Ke) | Equity value (Gordon Growth dividend model) |
| After-tax operating cash flow in nominal terms | Nominal WACC | Nominal EV |
| Real (inflation-adjusted) FCF | Real WACC (use Fisher equation) | Real EV (convert to nominal at the end) |
If you are using a nominal forecast (the default), do not discount at a real rate. If your forecast is in US dollars, do not use a Brazilian real cost of capital. If you include the tax shield inside FCF (by levering up), you should not also subtract it again via the after-tax cost of debt; the standard convention is to keep FCF unlevered and capture the tax shield inside the WACC, full stop. Tim Koller’s McKinsey valuation text and Damodaran’s Investment Valuation (3rd ed., Wiley) both hammer this consistency rule.
Step 4: Terminal Value, Gordon Growth vs Exit Multiple
Terminal value typically represents 60-80% of total enterprise value in a 5-year DCF, and 50-70% in a 10-year DCF. Get it wrong and the rest of the model does not matter. There are two accepted methods.
| Method | Formula | When to use | Risks |
|---|---|---|---|
| Gordon Growth (perpetuity) | TV = FCF_N+1 / (WACC – g) | Default for stable, mature businesses; required for academic and appraisal contexts | Tiny changes in g move TV massively; g must be below long-run nominal GDP growth (~3-4% in US) |
| Exit multiple | TV = Terminal-year EBITDA x exit multiple | Sponsor models, M&A pitch books, sectors with deep transaction comps | Implies a terminal growth rate; cross-check by backing out implied g and confirming it is reasonable |
The professional convention, used by every bulge-bracket bank and most middle-market shops including Houlihan Lokey, Lincoln, Piper Sandler, and William Blair, is to run BOTH methods and present them side by side. Goldman Sachs’s fairness opinions in SEC-filed proxies (for example the 2024 EQT / Equitrans Midstream proxy and the 2022 Twitter buyout proxy) typically disclose both perpetuity-growth and exit-multiple terminal values, with the implied “other” assumption shown as a sanity check.
Gordon Growth terminal value is highly sensitive to small changes in g. For a company with year-6 FCF of $100M, a WACC of 8.4%, and a g of 2.5%, TV = 100 / (0.084 – 0.025) = $1,695M. Move g to 3.0% and TV jumps to $1,852M, a 9% swing. Move it to 2.0% and TV drops to $1,563M, an 8% drop. The cap: terminal g should NEVER exceed long-run nominal GDP growth of the economy the business operates in, because no company can outgrow its host economy in perpetuity. For US-domiciled businesses, that is roughly 3.5-4.0% nominal (about 2% real plus 2% inflation), per the FOMC Summary of Economic Projections.
Exit-multiple terminal value uses EV / EBITDA, EV / EBIT, or occasionally P/E. The multiple should equal where you think the business will trade at the end of the forecast period given its then-current growth, margin, and risk profile; that is usually close to today’s trading comp range for mature peers. Always back-solve the implied perpetuity g and check it is reasonable: g_implied = WACC – [FCF_N+1 / TV_exit]. If the implied g is 6% you have set the exit multiple too high.
A third, less common method is the “value-driver” terminal value, which expresses TV as NOPAT_N+1 x (1 – g/ROIC) / (WACC – g). This is the McKinsey-preferred formulation because it embeds the reinvestment-rate identity directly and prevents the most common DCF abuse: assuming a high terminal growth rate without funding the capex to support it. Tim Koller’s Valuation textbook (Wiley, 7th ed.) devotes an entire chapter to this method and Aswath Damodaran’s Stern DCF mechanics lecture walks through the algebraic equivalence.
One field-tested sanity check: take your terminal-year EV and divide by terminal-year revenue. If the implied EV/revenue multiple is wildly out of line with how mature peers trade today (think Procter & Gamble, Colgate, Coca-Cola, Linde, Honeywell, Air Products for industrials), the terminal value is suspect. The Corporate Finance Institute publishes worked DCF examples that consistently apply this revenue-multiple cross-check.
Step 5: Sensitivity Analysis, the Most Read Page in the Book
Investment committees and corporate boards do not look at the central case. They look at the sensitivity grids. Three are mandatory.
| Sensitivity table | X axis | Y axis | Output |
|---|---|---|---|
| WACC x terminal growth | WACC (e.g., 7.5%-9.5%, 50 bps steps) | g (e.g., 1.5%-3.5%, 50 bps steps) | Implied equity value per share |
| WACC x exit multiple | WACC (50 bps steps) | Exit multiple (0.5x steps) | Implied equity value per share |
| Revenue growth x EBITDA margin | Long-run revenue CAGR | Terminal EBITDA margin | Implied equity value per share |
A typical bank sensitivity grid is 5×5 or 7×7, with the base case in the middle cell. Shade cells inside the precedent-transaction range, outside the trading-comp range, etc. Anyone reading the model should be able to glance at the grid and answer “what does this business need to be worth for [insert deal price] to be defensible?” in under 30 seconds. The Corporate Finance Institute and Wall Street Prep training materials, which are the most widely used in banking analyst classes, both standardize on this two-way table format.
Beyond two-way grids, sponsor and corporate-development teams often run a Monte Carlo simulation on the DCF, drawing 10,000 paths through joint distributions of WACC, terminal g, revenue growth, and margin. The output is a probability-weighted equity-value distribution rather than a single point. Palisade’s @RISK Excel add-in is the industry standard; Bain & Company and the major PE diligence groups (Alvarez & Marsal, AlixPartners, FTI Consulting) all build Monte Carlo overlays for large LBO underwrites. The output usually informs the IC discussion in the form of “the P10 equity value is $X, P50 is $Y, P90 is $Z; we are bidding at the P30 to leave returns asymmetric to the upside.”
Scenario analysis (Base / Upside / Downside) is the simpler cousin and still mandatory. The convention used at Goldman, Morgan Stanley, JPMorgan, and Bank of America: Upside assumes peer-leading revenue growth and EBITDA margin expansion, Downside assumes mid-cycle recession (revenue down 10-15% in a single year) and margin contraction. Both Upside and Downside DCFs anchor the football-field bar around the Base. Boards expect to see all three on a single page before they vote on a deal, per the Delaware General Corporation Law business judgment rule defense framework.
Step 6: The Football Field, Where DCF Lives Among Its Peers
A DCF that stands alone on a page in a pitch book is doing it wrong. The standard banker presentation puts DCF inside a “football field” alongside other methods. Each method produces a low/high implied equity value range; the field shows the overlap (or lack of it).
| Valuation method | What it captures | Typical width on the field |
|---|---|---|
| 52-week trading range (public only) | What investors paid in the past year | Narrow, but reflects sentiment not intrinsic value |
| Trading comps (EV/EBITDA, P/E) | What similar public companies trade for | Medium |
| Precedent transactions (control premium baked in) | What strategic and PE buyers paid for similar companies | Medium-wide |
| Premiums paid (public deals only) | One-day / four-week premium over unaffected price | Narrow (usually 25-40% premium) |
| DCF, Gordon Growth | Intrinsic value, perpetuity TV | Wide |
| DCF, exit multiple | Intrinsic value, multiples-anchored TV | Wide |
| LBO analysis | What a PE buyer can pay to hit a target IRR (usually 20-25%) | Tells you the floor a sponsor will bid |
| Sum-of-the-parts (if conglomerate) | Each segment valued separately, then summed | Depends on segment dispersion |
The reason DCF tends to be the widest bar is that small input changes (WACC, terminal g, terminal margin) translate into large value swings. That is a feature, not a bug. It signals the assumption sensitivity to the investment committee. See our deeper walk on discounted cash flow business valuation and the business valuation formula methods and math page for how these methods stitch together in a private-company sale context.
The Five DCF Mistakes That Get Models Sent Back
Wall Street Oasis, Mergers & Inquisitions, and AlphaSights interview transcripts all surface the same recurring errors. If you make any of these, expect to redo the model.
| Mistake | What goes wrong | How to spot it |
|---|---|---|
| Terminal value too big a % of EV | TV is 90%+ of EV; the explicit-period forecast adds almost nothing | If TV > 80%, extend the forecast period or use a more conservative g |
| WACC mismatch with cash flow type | Discounting FCFE at WACC, or FCFF at cost of equity | Check: if you started from EBIT, you must use WACC |
| Capex / D&A divergence in terminal year | Net investment != growth, so steady-state assumption breaks | Force terminal capex ~= D&A x (1 + g); reinvestment rate = g / ROIC |
| Terminal growth above long-run GDP | Implies the business eventually swallows the economy | Cap terminal g at 3-4% nominal for US, lower for slow-growth peers |
| Working capital ignored or inconsistent | NWC drops to zero in terminal year, inflating FCF | Hold NWC as a stable % of revenue in terminal year |
| Tax rate inconsistent | Used statutory rate in forecast but effective rate in WACC | Use marginal long-run effective tax rate (often 23-26% blended for US) |
| Stock-based compensation ignored | Especially in tech, treating SBC as non-cash overstates FCF | Treat SBC as a real cash cost or fully dilute share count later |
The terminal-value sanity check is the single most important DCF audit. If terminal value is more than 80% of enterprise value in a 5-year forecast, either the forecast horizon is too short (the business has not yet reached steady state) or the explicit-period growth is too low relative to the terminal-period growth, which makes no economic sense.
DCF vs Comparable Company Analysis vs Precedent Transactions
Bankers and investors triangulate value with three distinct methods because each captures something the other two miss.
| Method | What it answers | Strengths | Weaknesses |
|---|---|---|---|
| DCF | What is the business intrinsically worth based on its future cash? | Captures unique growth, margin, and capex profile; ignores market mood | Highly sensitive to WACC and terminal assumptions; garbage in / garbage out |
| Comparable company analysis (trading comps) | What does the public market pay for similar businesses today? | Real, observable multiples; minimal assumptions | Requires truly comparable peers; reflects market sentiment, including bubbles |
| Precedent transaction analysis | What have strategic and PE buyers paid for similar companies? | Embeds control premium and synergy value | Stale (deals are 1-5 years old); deal-specific dynamics (auction vs negotiated, distressed vs healthy) |
The professional default is DCF as the anchor, comps as the market check, and precedents as the M&A premium overlay. We cover the comp side in depth on business valuation formula methods and math. Damodaran calls this the “trinity” of valuation and his Stern teaching site publishes industry-by-industry multiples updated each January that the buyside uses as a free Bloomberg substitute.
DCF in M&A Transactions, From Pitch to Closing
In sell-side M&A, the DCF lives inside the sellside banker’s “valuation deck” delivered to the board before launch. Boutiques and bulge brackets all use it the same way: anchor the asking-price range with an aggressive central case, justify the price expectation in the buyer education materials, and use the DCF (alongside comps and precedents) as the basis for the fairness opinion at signing. The fairness opinion is itself a regulated document under Delaware corporate law; under the duty of care in Smith v Van Gorkom, 488 A.2d 858 (Del. 1985), boards are expected to have a documented valuation analysis (typically including a DCF) before approving a sale.
On the buyside, strategic acquirers run a DCF on the target standalone, then a DCF on the target with synergies (cost synergies, revenue synergies, and net-present-value of those), and the difference funds the offer premium. The Wachtell Lipton M&A memo series (wlrk.com) and Skadden’s M&A Insights both walk through the synergy-valuation discipline boards apply when authorizing the premium; Davis Polk’s M&A client updates regularly summarize synergy-related controversies from recent SEC-filed proxy fights. PE sponsors run an LBO model rather than a standalone DCF, since their return is governed by debt financing and exit multiple, not WACC, but they almost always cross-check with a DCF to make sure they are not paying more than fundamentals justify. Our deep dives on the LBO model from scratch and the LBO model step-by-step guide walk the sponsor approach in detail; the paper LBO example walkthrough covers the napkin version.
If you are an sell-side analyst in a sector group at a bulge bracket, the DCF is your daily bread. If you are a PE associate, you build the LBO and then back-check it with a DCF on every IC memo. If you are an M&A advisor running a private-company sale, the DCF is one input into the asking-price recommendation, alongside comps and the buyer-universe analysis.
DCF for Private Companies, DLOM and 409A
Private-company DCFs require two extra adjustments that public DCFs do not. First, a discount for lack of marketability (DLOM), to reflect that private-company shares are illiquid and cannot be sold on a public exchange. DLOM ranges typically run 10-35%, depending on the company stage and the time-to-liquidity expectation; the Mandelbaum, Stout, and FMV restricted-stock studies are the standard empirical anchors cited by appraisers under IRS Revenue Ruling 59-60 (the foundational guidance on closely-held business valuation, see IRS).
Second, a discount for lack of control (DLOC) for non-controlling minority interests; this is typically 15-30% based on control-premium studies (Mergerstat publishes these annually). The two discounts are NOT additive in a simple sense; appraisers apply them multiplicatively to the marketable, controlling DCF value.
The IRS guidance on DLOM in estate and gift tax contexts (Mandelbaum v. Commissioner, T.C. Memo 1995-255, and subsequent Tax Notes coverage of the FMV restricted-stock studies) is the most-cited body of law on the topic. The Journal of Accountancy and the Tax Adviser publish updates on appraisal court cases each quarter that test DLOM ranges; expect IRS challenges on any DLOM above 30% without a strong empirical defense.
For startups and pre-IPO companies, a Section 409A valuation under IRC Section 409A and Treasury Regulation 1.409A-1(b)(5)(iv)(B) requires the company to determine fair market value of common stock for stock-option strike-price purposes. The Big Three independent 409A providers (Carta, Pulley, Eqvista) all run a DCF (sometimes a hybrid OPM / PWERM model) and apply DLOM. Carta’s published 409A methodology walks the framework. The penalties for an unsupported 409A are severe under IRC Section 409A: immediate income recognition plus a 20% additional federal tax on the employee, which is why nobody skips this.
Excel / Template Mechanics, the Build Itself
The DCF tab in any bank model follows a near-identical layout. From top to bottom:
| Section | Rows | Anchored to |
|---|---|---|
| Forecast period header | Years 1-N + terminal year | Calendar year columns |
| FCF build | EBIT, taxes, D&A, capex, NWC change, FCF | Links to operating model tab |
| Mid-year convention toggle | Discount period = year – 0.5 if mid-year, else year | WACC tab cell |
| PV factor row | = 1 / (1+WACC)^period | WACC + period |
| PV of FCF row | = FCF x PV factor | Product |
| Terminal value, two methods | Gordon Growth and exit multiple, side by side | Terminal FCF / multiple, WACC, g, exit multiple |
| PV of TV row | = TV / (1+WACC)^N | TV + WACC |
| Enterprise value | = Sum of PV(FCF) + PV(TV) | Sum |
| Bridge to equity | EV – net debt + cash – minority interest – preferred | Latest balance sheet |
| Per-share value | = Equity / diluted share count (treasury method) | 10-Q share count, plus dilutive options/RSUs |
| Sensitivity tables | WACC x g, WACC x exit multiple | Data table function |
Mid-year convention is worth flagging. By default, the formula 1/(1+WACC)^t assumes cash flow arrives at year-end, which is conservative because most businesses generate cash throughout the year. Mid-year convention applies a discount period of (t – 0.5), which lifts the present value by roughly half a year of WACC, or about 4-5% on total EV. Most bank models include a toggle. Investment-banking textbooks like Rosenbaum & Pearl’s Investment Banking: Valuation, LBOs, M&A, and IPOs (3rd ed., Wiley 2020) cover the mechanics in detail; the toggle is standard at Goldman, Morgan Stanley, JPMorgan, Lazard, Centerview, Evercore, and Moelis.
Worked Example: A $50M EBITDA Industrial Company
Below is a fully worked DCF on a mid-market US industrial company doing $300M in revenue and $50M in EBITDA, growing 5% per year, with a 70/30 equity/debt capital structure target. All figures in $M.
| Y1 | Y2 | Y3 | Y4 | Y5 | Terminal | |
|---|---|---|---|---|---|---|
| Revenue | 315 | 331 | 347 | 365 | 383 | 395 |
| EBITDA | 53 | 56 | 59 | 62 | 65 | 67 |
| D&A | (11) | (12) | (13) | (13) | (14) | (14) |
| EBIT | 42 | 44 | 46 | 49 | 51 | 53 |
| (-) Taxes @ 25% | (11) | (11) | (12) | (12) | (13) | (13) |
| NOPAT | 31 | 33 | 34 | 37 | 38 | 40 |
| (+) D&A | 11 | 12 | 13 | 13 | 14 | 14 |
| (-) Capex | (13) | (13) | (14) | (15) | (15) | (15) |
| (-) Change in NWC | (2) | (2) | (2) | (2) | (2) | (1) |
| Unlevered FCF | 27 | 30 | 31 | 33 | 35 | 38 |
| PV factor @ 8.4% | 0.923 | 0.851 | 0.785 | 0.724 | 0.668 | |
| PV of FCF | 25 | 26 | 24 | 24 | 23 |
Sum of explicit-period PV of FCF: $122M.
Terminal value (Gordon Growth, g = 2.5%): TV = 38 / (0.084 – 0.025) = $644M. PV of TV = 644 / (1.084^5) = $430M.
Terminal value (Exit multiple, 9.0x terminal EBITDA): TV = 9.0 x 67 = $603M. PV of TV = 603 / (1.084^5) = $403M.
Enterprise value range: $552M (Gordon Growth) to $525M (exit multiple), midpoint roughly $538M. At 10.7x trailing EBITDA, this is in the middle of the mid-market industrial trading range (peer set typically 9-12x in 2025-26 per Pitchbook and Capital IQ data). If net debt is $80M, implied equity value is $445-472M, midpoint $458M.
Sensitivity (Gordon Growth, EV in $M):
| WACC \ g | 1.5% | 2.0% | 2.5% | 3.0% | 3.5% |
|---|---|---|---|---|---|
| 7.5% | 620 | 650 | 684 | 723 | 768 |
| 8.0% | 578 | 602 | 629 | 660 | 695 |
| 8.4% | 551 | 571 | 594 | 620 | 649 |
| 9.0% | 515 | 532 | 551 | 572 | 596 |
| 9.5% | 489 | 503 | 520 | 538 | 558 |
Notice how a 50-bp move in WACC or g moves EV by 4-6%. That sensitivity is exactly why bankers triangulate with comps and precedents instead of betting the deal on a single DCF cell.
Cross-checking this DCF against trading comps and precedent transactions is the final discipline. A 2024-2025 mid-market industrial peer set might trade at 8-11x trailing EBITDA on EV/EBITDA per Pitchbook’s 2024 PE Breakdown and Lincoln International’s Senior Debt Index. Precedent transactions in the same period (per Mergermarket and Refinitiv) often closed at a 1-2 turn premium, putting strategic buyer offers in the 10-13x range. Our $50M EBITDA target’s 10.7x DCF midpoint sits right in the trading-comp range, suggesting it is a fair standalone valuation; a strategic buyer with $5M of run-rate cost synergies (capitalized at the 9x exit multiple, that is another $45M of value) could justify bidding to 11.6x or roughly $580M EV without leaving money on the table.
If you were running this as a PE sponsor LBO instead of a strategic-buyer DCF, the math runs differently: assume 50% debt at SOFR + 525 bps (per S&P Global Market Intelligence’s LCD mid-market loan tracker, 2025 average), 5-year hold, exit at 9.5x EBITDA, and back-solve to a 22% IRR target. The LBO would support an entry of roughly $480-520M, materially below the strategic DCF range. That gap is exactly why strategics tend to win competitive auctions for high-quality assets; sponsors win on operational improvement plays, distressed situations, or where the strategic universe is thin.
Walk Me Through a DCF: The Interview Answer
If you are interviewing at a bulge bracket, an elite boutique (Centerview, Evercore, Lazard, Moelis, PJT, Guggenheim, Perella Weinberg), a middle-market bank (Houlihan Lokey, Lincoln, Piper Sandler, William Blair, Raymond James), or a megafund private-equity associate role (KKR, Blackstone, Apollo, Carlyle, TPG, Warburg Pincus), expect “walk me through a DCF” inside the first 15 minutes. The right answer is 90 seconds, structured.
“A DCF values a business as the present value of its future unlevered free cash flows. I project FCF for an explicit period, usually five to ten years: start from EBIT, tax it at the marginal rate, add back D&A, subtract capex and the change in net working capital. Then I calculate WACC, weighting cost of equity (CAPM: risk-free plus beta times equity risk premium) and after-tax cost of debt by target capital structure at market values. I discount each year’s FCF at WACC to get present value. For terminal value I run both Gordon Growth, FCF in the year after the forecast divided by WACC minus g, and an exit multiple on terminal EBITDA. I discount the terminal value back at WACC and add it to the sum of PV of the explicit-period FCF, which gives me enterprise value. Subtract net debt and add cash to get equity value, divide by diluted shares for per-share value. Then I sensitize WACC against g and against exit multiple, and present the result in a football field alongside trading comps and precedent transactions.”
Common follow-ups: Why use unlevered FCF and not levered? (Because we want to value the operating business independent of capital structure, so we can layer the actual capital structure on at the end.) What if WACC < g? (Gordon Growth breaks, division by negative; means you have set g too high or WACC too low.) What is wrong with a 90% terminal value contribution? (The explicit forecast adds no information; extend the horizon or revisit terminal assumptions.) For a deeper grind on the M&A interview canon, see Mergers & Inquisitions and Wall Street Oasis, both of which publish hundreds of recent interview reports from the funds and banks listed above.
DCF in Appraisal Litigation and Delaware Case Law
Discounted cash flow has been the dominant valuation method in Delaware appraisal litigation under 8 Del. C. Section 262. In In re Appraisal of Dell Inc., C.A. No. 9322-VCL (Del. Ch. 2016), Vice Chancellor Laster found that the DCF analyses produced a value above the deal price, eventually overturned on appeal in 2017 when the Delaware Supreme Court weighted the deal price more heavily in Dell Inc. v. Magnetar Global Event Driven Master Fund, 177 A.3d 1 (Del. 2017). In DFC Global v. Muirfield Value Partners, 172 A.3d 346 (Del. 2017), the Supreme Court likewise emphasized the deal-price anchor where the sale process was thorough and well-shopped.
The takeaway for practitioners: DCF in a Delaware appraisal must be defensible against expert cross-examination on every assumption. Beta, ERP, terminal g, capex assumptions, and tax rate are all litigated separately. The Delaware courts have rejected DCFs that depart materially from management’s contemporaneous projections, used non-standard ERP figures, or applied size premiums in inconsistent ways. This is why bank DCFs in fairness-opinion exhibits are documented down to the cell formula. See Cooley’s Delaware appraisal case tracker and Sullivan & Cromwell’s M&A litigation memos for ongoing case-law summaries.
Outside of appraisal, DCF analyses also surface in IRS estate and gift-tax valuation disputes (Tax Court frequently sides with appraisers who anchor to a DCF plus comparable-company analysis blend), in bankruptcy-court valuation hearings under Section 1129(b) of the Bankruptcy Code where reorganization plans require a going-concern valuation, and in family-law equitable-distribution proceedings. Each forum has its own evidentiary standard, but the underlying DCF methodology is the same. Kirkland & Ellis and Sidley Austin both publish bankruptcy-valuation client alerts that walk through Chapter 11 DCF litigation precedent year by year.
Bloomberg’s M&A league tables and Reuters M&A coverage are good starting points to find the recent Delaware appraisal docket, which in 2024-2025 included In re Manti Resources Inc. Appraisal, ongoing post-Dell deal-price-floor debates, and a string of public-buyout proxy contests where the DCF backing the fairness opinion was scrutinized line by line. WSJ’s M&A desk and Forbes M&A coverage regularly summarize the most consequential of these cases for non-litigators.
Quick Reference: DCF Conventions by Asset Class
| Asset class | Cash flow type | Discount rate | Forecast horizon | Terminal method |
|---|---|---|---|---|
| Mature public industrial | Unlevered FCF | WACC, 7-9% | 5 years | Gordon Growth + exit multiple |
| Mid-market PE target | Unlevered FCF (DCF cross-check) + LBO | WACC + sponsor target IRR | 5-7 years | Exit multiple |
| High-growth SaaS | Unlevered FCF, SBC expensed | WACC, 9-11% | 10 years (long ramp to FCF) | Exit multiple on revenue or EBITDA |
| Pharma with patent cliff | Unlevered FCF, asset-by-asset | Risk-adjusted WACC per asset | Patent life + 5 years | Steep decline, then Gordon at low g |
| Bank or insurer | Dividend discount or excess returns | Cost of equity (CAPM) | 5-10 years | Gordon Growth on dividends |
| REIT or infrastructure | FCFE or AFFO | Cost of equity | 10 years | Cap rate or Gordon Growth on AFFO |
| Distressed / restructuring | Unlevered FCF, multiple scenarios | Higher WACC reflecting risk | 3-5 years | Liquidation value floor |
| Private startup (409A) | Multi-scenario PWERM or OPM | Cost of equity + DLOM | 3-7 years to exit | Exit multiple x scenario probability |
Final Pre-Send Checklist
Before you send your discounted cash flow model up the chain, run this checklist. Every analyst has one taped to the cube wall.
- FCF starts from EBIT (unlevered), not net income. Taxes computed on EBIT at marginal rate.
- Discount rate is WACC if FCF is unlevered; cost of equity if FCF is levered. Never crossed.
- Capital-structure weights for WACC are at target market values, not book and not current.
- Terminal value < 80% of EV for a 5-year forecast (extend horizon if not).
- Terminal g <= long-run nominal GDP of the operating geography (US: 3.5-4.0% nominal max).
- Terminal capex roughly equals D&A x (1 + g); steady-state reinvestment rate = g / ROIC.
- Working capital remains a stable % of revenue in the terminal year.
- Stock-based compensation expensed in FCF, not added back. Otherwise dilute share count.
- Mid-year convention toggled on or off consistently with the rest of the model.
- Sensitivity tables provided for WACC x g and WACC x exit multiple, minimum 5×5.
- Football field shows DCF range alongside comps, precedents, and (for public) 52-week range.
- For private targets: DLOM and DLOC applied; for startups: 409A methodology documented.
- Every assumption sourced to a footnote: 10-K, Damodaran, Bloomberg, FactSet, Pitchbook, or named report.
One last operational point: every assumption in the model should be color-coded. Bank-standard convention, taught at the Goldman Sachs analyst training program and at every elite-boutique training week, is blue font for hard inputs (numbers you typed), black font for formulas, green font for cross-tab links, red font for “do not change” hardcodes, and underline for any output cell linked into the pitch book. Color discipline is what lets the next analyst (or your VP at 2 AM) audit the model in five minutes instead of five hours. Wall Street Prep, Breaking Into Wall Street, and the in-house training decks at McKinsey, Bain, and BCG all teach the same convention.
A discounted cash flow model done right takes a week the first time and a day after that. It is the most powerful tool you have for separating value from price, and the most-cited valuation method in M&A pitch books, equity research, private-equity investment committee memos, and Delaware Chancery courtrooms. The math is simple. The discipline is everything.