IB Interview Prep · 2026

    DCF Interview Questions for Investment Banking

    Walk me through a DCF, WACC, terminal value, sensitivity, common traps. The DCF questions you will actually be asked — answered with the structure bankers expect.

    DCF is the question that separates candidates who memorized formulas from candidates who understand valuation. The walkthrough is predictable — project unlevered FCF, terminal value, discount at WACC, bridge to equity — but the follow-ups are not. Bankers will push on WACC components, terminal value sensitivity, and the cardinal rule that cash flow type and discount rate must match.

    This guide covers the DCF questions you will actually be asked, with the structured answers we drill candidates on inside Banking Prep AI.

    What a DCF actually values, and why bankers love asking about it

    A discounted cash flow analysis values a business as the present value of the cash it is expected to generate over its remaining life. Unlike comps and precedents, which read pricing off the market, a DCF is the intrinsic-value method — you build the answer from operating assumptions, capital intensity, and a discount rate, with no reference to what other people are paying right now.

    Bankers ask about DCF in interviews because it tests three things at once: technical mechanics (formulas), economic intuition (why each line item matters), and judgment under pressure (which assumption to flex when the interviewer pushes). A clean DCF answer signals that you can model, that you can explain a model, and that you know which numbers to defend versus which numbers to admit are educated guesses.

    The DCF in five mechanical steps

    Memorize the steps, but understand the why behind each one. Bankers will interrupt you on any step and ask you to defend an assumption.

    1. Project unlevered free cash flow

    Unlevered free cash flow is the cash the operating business produces before any financing decisions. The standard formula is FCF = EBIT × (1 − t) + D&A − CapEx − ΔNWC. Start from EBIT (not net income — net income already deducts interest), tax-effect it at the marginal rate, add back non-cash D&A, subtract reinvestment in fixed assets (CapEx), and subtract incremental working capital tied up in growth.

    Forecast horizon is typically 5–10 years. Five for stable mature businesses, ten for high-growth ones where margins and capex intensity are still normalizing. The forecast must end at a steady-state year — terminal-year growth, margins, and reinvestment should look like a mature business, otherwise terminal value is meaningless.

    2. Compute the discount rate (WACC)

    WACC blends the after-tax cost of every capital provider, weighted by their share of enterprise value. Formula: WACC = (E/V) × Re + (D/V) × Rd × (1 − t). E and D are market values of equity and debt, V = E + D. Use a target capital structure rather than today's snapshot if the company is mid-restructuring or recently changed its mix.

    Cost of equity (Re) comes from CAPM: Re = Rf + β × ERP. Risk-free rate is the 10-year or 20-year Treasury yield (match the duration to the business). Beta is unlevered from comparable companies, then re-levered to the target's capital structure. Equity risk premium is typically 5–6% in mature markets. Cost of debt (Rd) is the company's marginal borrowing rate — what new debt would cost today, not the historical coupon — tax-adjusted because interest is deductible.

    3. Estimate terminal value

    Terminal value captures everything beyond the explicit forecast — the present value of all cash flows from year N+1 to infinity. Two methods: Gordon Growth and exit multiple. Gordon Growth: TV = FCF_(N+1) / (WACC − g), where g is the long-term growth rate (typically GDP-like, 2–3% in mature markets). Exit multiple: apply a forward EV/EBITDA to terminal-year EBITDA.

    Both methods should land in the same neighborhood. If they don't, you have an inconsistent assumption — usually an exit multiple that bakes in growth higher than your Gordon Growth g. Always cross-check by computing the implied long-term growth rate from your exit multiple, and the implied exit multiple from your Gordon Growth: if either is off-market, fix the input.

    4. Discount everything to present at WACC

    Bring each forecast-period FCF and the terminal value back to today by dividing by (1 + WACC)^t. Sum the present values to get enterprise value. This is the only step that is mechanically trivial — but it is also where the mid-year convention applies. By default we assume cash flows arrive at year-end (period 1 discounted by 1.0, period 2 by 2.0). The mid-year convention assumes they arrive mid-year (period 1 by 0.5, period 2 by 1.5, etc.) and bumps EV by roughly 5%.

    5. Bridge to equity value and price per share

    Enterprise value is the value of the operating business. Equity value is what is left for shareholders after senior claims. Bridge: Equity Value = EV − Debt − Preferred Stock − Minority Interest + Cash. Divide by diluted shares outstanding (not basic — include the dilutive impact of options, RSUs, and convertibles using treasury stock or if-converted method). The output is implied price per share, which you compare to the current trading price for public targets, or use as a reference value in M&A.

    Gordon Growth vs exit multiple, side by side

    The two terminal value methods price the same thing differently. Use both, and reconcile.

    Terminal value methodologies
    MethodFormulaBest whenWatch out for
    Gordon GrowthFCF × (1 + g) / (WACC − g)Mature, stable business with credible long-term gg approaching WACC blows up the model
    Exit multipleTerminal-year EBITDA × forward EV/EBITDACyclical or transition businesses where comps are richer than g would implyBakes in the cycle of the chosen multiple — pick across cycles
    Cross-checkImplied g from exit multiple; implied multiple from Gordon GrowthAlways — both methods must reconcileDivergence > 1.0x EV/EBITDA means an assumption is off

    The cardinal rule: cash flow type ↔ discount rate

    This is the single most common DCF mistake in interviews and the fastest way to lose credibility. Unlevered free cash flow belongs to all capital providers — debt and equity together — so it must be discounted at WACC, the blended cost of all capital. The output is enterprise value.

    Levered free cash flow (after interest expense) belongs only to equity holders. It must be discounted at cost of equity (Re), not WACC. The output is equity value directly, with no EV → equity bridge needed. Mixing them — discounting unlevered FCF at cost of equity, or levered FCF at WACC — produces a number that has no economic meaning. Bankers will spot it instantly.

    Pairing rule
    Unlevered FCF  ↔  WACC  →  Enterprise Value
    Levered FCF  ↔  Cost of Equity (Re)  →  Equity Value
    Unlevered is the standard in IB because it isolates operating performance from financing choices and is comparable across capital structures.

    Sensitivity: how to present the answer like a banker

    A point estimate is a junior mistake. Real DCF output is a range, presented as a sensitivity table.

    • Build a 5×5 grid. Rows: WACC ±100 bps in 25 bps steps. Columns: terminal growth ±50 bps in 25 bps steps (or exit multiple ±1.0x in 0.5x steps). Output: implied per-share price in each cell. The diagonal you read off the table is your football-field width.
    • Quote a range, not a number. Instead of 'the company is worth $42 per share', say 'the DCF supports $36–$48 per share, with the base case at $42 driven by a 9.5% WACC and 2.5% terminal growth'. Bankers want the range plus the base case driver.
    • Disclose where you are most exposed. If terminal value is 75% of EV, say so: 'most of the value sits in the terminal, so the answer is sensitive to long-term growth and exit multiple — I would stress-test those before presenting to a client'.

    The sensitivity is not a 'nice-to-have' — it is the answer. A clean DCF walkthrough that lands on a single number signals you have not stress-tested your own model. A walkthrough that lands on a range with explicit drivers signals you think like a banker.

    The two formulas you must own cold

    Two formulas appear in nearly every DCF interview. Memorize the form, and be able to explain every term. Bankers test the why, not just the what — for each variable, expect a follow-up on where you got the number and how sensitive the answer is to it.

    Unlevered free cash flow
    FCF = EBIT × (1 − t) + D&A − CapEx − ΔNWC
    Terminal value (Gordon Growth)
    TV = FCF × (1 + g) / (WACC − g)
    Alternative terminal: exit multiple × terminal-year EBITDA. Use both and reconcile.

    How to walk through a DCF in three minutes

    Use this script for any 'walk me through a DCF' prompt. Three minutes, five beats.

    1. Frame the method in one sentence. 'A DCF values a company as the present value of its future free cash flows discounted at the cost of capital.'
    2. Project unlevered FCF for 5–10 years. Mention the formula (EBIT × (1 − t) + D&A − CapEx − ΔNWC) and the horizon. Don't dwell — interviewers know the formula.
    3. Compute terminal value. State both methods (Gordon Growth and exit multiple), pick one as your primary, and say you would cross-check with the other.
    4. Discount everything to present at WACC. Mention CAPM for cost of equity, marginal pre-tax for cost of debt, and the after-tax adjustment. If you have time, mention mid-year convention.
    5. Bridge to equity and land on a range. EV − net debt = equity value. Divide by diluted shares. Quote a sensitivity range, not a single number, and name the two assumptions that move the answer most.

    Three minutes is the right length at summer-analyst level. Associate-level answers add nuance — mid-year convention, normalized terminal CapEx, treatment of operating leases, stub period — but the structure is identical. Practice the script out loud until it is automatic.

    Common DCF interview mistakes

    • Mixing levered and unlevered. Discounting unlevered FCF at cost of equity (or levered FCF at WACC) is a structural error. The pairing rule is non-negotiable.
    • Starting FCF from net income. Net income is post-interest. Starting from there means you are already mixing financing into the operating cash flow. Always start from EBIT or operating income.
    • Picking a terminal growth rate above long-run GDP. g > GDP forever means the company eventually becomes the entire economy. Cap g at 2–3% in mature markets unless you can defend higher with hard reasoning.
    • Forgetting tax-shield on debt. Cost of debt in WACC is after-tax: Rd × (1 − t). Using pre-tax Rd overstates WACC and understates EV.
    • Using book values for E and D in WACC. WACC weights are market values. Book value of equity is meaningless for this purpose; book value of debt is acceptable as a proxy if there is no traded debt.
    • Letting terminal value dominate without sensitizing it. If TV is 80%+ of EV, your DCF is essentially a multiple. Stress-test g and exit multiple, and disclose the dependency.
    • Quoting a single price point. Real DCF output is a range. A point estimate signals you have not stress-tested your model. Always present a sensitivity.
    • Using basic shares instead of diluted. Diluted shares include the impact of in-the-money options, RSUs, and convertibles. Basic shares overstate price per share.

    Worked example: building the answer end-to-end

    Imagine a mature industrial target with $200M of LTM EBITDA, EBIT of $150M, a 25% tax rate, $30M of D&A, $40M of CapEx, and $10M of working capital build per year. Year-1 unlevered FCF = 150 × (1 − 0.25) + 30 − 40 − 10 = $92.5M. Grow that 4% per year for five years, then estimate terminal value via Gordon Growth at 2.5% long-term growth. Discount everything at a 9.5% WACC.

    The five forecast-year FCFs sum to roughly $510M of present value. Terminal value at year 5 is approximately FCF_6 / (WACC − g) = ~$1,650M, discounted back to ~$1,050M present value. Total enterprise value: ~$1,560M. Subtract $400M of net debt and you get equity value of ~$1,160M, or ~$23 per share on 50M diluted. Sensitivity flexing WACC ±100 bps and g ±50 bps produces a range of roughly $19–$28 per share — that is the football field you would present.

    How to practice DCF under pressure

    Reading a DCF guide is not enough. The walk-through is a verbal performance — you have to deliver the five-step script out loud, repeatedly, until the structure is automatic and you can spend cognitive bandwidth on the interviewer's follow-ups (why this WACC? why this g? why this horizon?) instead of remembering the next step.

    Practice in three layers: solo reps in front of a mirror to lock the script, peer mocks to stress-test under social pressure, and AI mock interviews for high-volume reps with instant structured feedback on accuracy, structure, and clarity.

    Try it free

    Run a DCF mock interview today

    Get pushed on WACC mechanics, terminal value choices, and sensitivity reading with structured feedback after every answer.

    Q&A

    Frequently asked DCF interview questions