This guide is corporate finance math inside a case: invest / do not invest, buy vs lease, fund with cash vs debt. It is not how a bank makes money, and it is not a startup take-rate case.
Last full pass: 24 August 2026.
A finance case interview at McKinsey, BCG, Bain, or a PE-heavy boutique is a capital-allocation conversation. The client is usually a corporation or a fund asking whether a project earns more than its cost of capital after tax and after the working-capital it swallows. Candidates who open a 3C framework here waste four minutes. Candidates who quote a payback period and sit down fail the so-what. If the prompt is “our retail bank’s NIM compressed,” go to the financial services case interview. If it is a payments app’s GMV, go to fintech. If it is a DCF you need to build, also keep the DCF walkthrough nearby. This page is the live-case version: fast, MECE, and allergic to fake precision.
What you are actually deciding
- Invest or kill a project (line, warehouse, software, acquisition bolt-on).
- Price a deal (what is the most we can pay at our hurdle rate?).
- Choose a funding mix (does cheap debt make a bad project look good? usually yes — that is a warning).
- Working-capital vs accounting profit (EBITDA up, cash down).
Ask immediately: cash vs accounting; pre-tax vs after-tax; real vs nominal; and whether the hurdle is WACC, a PE IRR, or a simple payback the CFO likes to see in addition to NPV.
Exhibits you should expect
| Exhibit | Use |
|---|---|
| Capex schedule and useful life | Timing of cash out. Residual / salvage is often hidden. |
| Incremental EBITDA or contribution | Must be incremental and with-without, not the whole plant. |
| NWC days (inventory, AR, AP) | The silent cash drain at t=0 or during ramp. |
| Tax rate and depreciation method | Tax shield is real; mixing EBIT and EBITDA is how people double-count. |
| WACC or hurdle / target IRR | State whether you are discounting unlevered FCF (WACC) or levered (equity IRR). |
| Sensitivities | Volume, price, delay. A base-case NPV of $4m that dies at −10% volume is a no. |
Formulas interviewers actually want spoken
Unlevered FCF (project): EBIT(1 − t) + D&A − capex − ΔNWC.
NPV: −initial cash + Σ FCF_t / (1 + r)^t. If NPV < 0 at WACC, you are destroying value even if IRR looks “high” on a small equity check.
IRR vs NPV. IRR is a rate; NPV is cash. Mutually exclusive projects: NPV wins. A 40% IRR on $2m is not better than a 18% IRR on $80m if the alternative is idle cash at 10% WACC — check NPV.
Payback is a liquidity and risk lens, not a decision rule. Mention it; do not lead with it.
Practice corporate-finance case math
Run a scored case where working capital and after-tax cash have to show up in the NPV, not just EBITDA / capex.
Worked mini-case: the filling line that “pays back in six years”
Prompt. ChemCo will spend $14.0m on a filling line. Operations promises +$2.4m incremental EBITDA per year for 8 years, then scrap $1.0m. Tax 25%. WACC 10%. To run the line they must hold $3.1m extra resin inventory from day one, recovered at the end. The plant manager likes the 14 / 2.4 ≈ 5.8-year payback. Do we build?
Structure. T0 cash, annual after-tax operating cash, terminal release, NPV at 10%. Ignore heroics on mid-year conventions unless asked.
T0 cash. −$14.0m capex − $3.1m NWC = −$17.1m. Already the payback slide is lying: it ignored inventory.
Annual cash (interview approximation). Without building a full depreciation schedule, a conservative operating FCF is EBITDA × (1 − t) = 2.4 × 0.75 = $1.80m, which understates cash if depreciation is large (you would add back D&A and only tax EBIT). Flag that to the interviewer. Using $1.80m is conservative vs the “EBITDA is cash” error.
PV of 8-year $1.80m annuity at 10%. Annuity factor = (1 − 1.10^(−8)) / 0.10 ≈ 5.335. PV ≈ 1.80 × 5.335 = $9.60m.
PV of terminal $1.0m scrap + $3.1m NWC release = $4.1m in year 8. PV ≈ 4.1 / 1.10^8 ≈ 4.1 / 2.144 = $1.91m.
NPV ≈ −17.1 + 9.60 + 1.91 = −$5.6m. Kill the project.
If you had used the plant manager’s payback and skipped NWC, you would have discussed “implementation risk” on a value-destroying line. If you add a full tax shield on straight-line D&A of 14/8 = $1.75m, extra annual cash ≈ 1.75 × 0.25 = $0.44m, PV ≈ $2.3m — NPV still about −$3.3m. Same call.
Recommendation. Do not build. Revisit only if volume is contracted (take-or-pay offtake) or if NWC can be consignment. Risk: the EBITDA case double-counts a mix shift from an existing line. Next exhibit: incremental volume source and inventory days.
What a generic profitability tree misses here
- Profit is not cash. Inventory and AR can eat the entire EBITDA win.
- Time value. A dollar in year 8 is not a dollar today; payback hides that.
- Hurdle rate is a cost. Beating 0% is not a strategy.
- Sunk cost. The feasibility study is gone; only incremental cash matters.
- Leverage theater. Debt can raise equity IRR while NPV at WACC is negative. Say that out loud.
See where you stand on a finance case
Practice NPV, NWC, and the payback trap under time pressure.
Related guides
- Financial services case interview (banks and insurers)
- Fintech case interview (GMV, take rate, MAU)
- NPV case interview
- Valuation interview questions
CoachNed is independent and unaffiliated with McKinsey, BCG, Bain, or other firms named for interview-style practice. Cases on CoachNed are AI-simulated.