In a live case, NPV is a sixty-second yes-or-no call, not a model. I stop listening the moment a candidate starts writing year-by-year discount factors for a twenty-year stream. The sheet rewards three things: a defensible rate in one sentence, a mental-math present value, and a crisp invest-or-pass recommendation. Everything else belongs in a spreadsheet you will never open in the room.
After this piece you should be able to run the two shortcuts that cover almost every case NPV (perpetuity and finite annuity), name a discount rate without inviting a CAPM debate, and know when payback is the smarter tool than NPV for the decision on the table. Sibling guides cover full DCF setup and break-even; this one stays on the investment decision under the interviewer's clock.
What the interviewer is actually scoring
NPV questions usually arrive as a fork: "Should we build the plant / buy the fleet / digitize the warehouse?" The interviewer already knows the arithmetic is easy if you pick the right shortcut. What they grade is judgment.
| What you do | What the sheet hears |
|---|---|
| Ask for cash flow shape and horizon | You know which formula applies |
| State a rate and one reason | You will not stall on WACC theater |
| Compute PV in under a minute | Math speed without a calculator |
| Recommend invest / pass / conditional | You close the decision |
The formal idea is still the textbook one: future cash is worth less than cash today, so you shrink future flows and subtract the outlay. The CFA Institute frames the same present-value logic for any asset: value equals discounted expected cash flows. In a case you are not valuing a stock. You are answering whether this project clears the hurdle.
The perpetuity shortcut: CF divided by r
When cash flows are roughly level and the horizon is long (roughly ten years or more, or "ongoing"), present value collapses to one division:
PV = CF / r
Work it once out loud. A plant throws off $8 million a year indefinitely. You use 10 percent. PV = 8 / 0.10 = $80 million. If the plant costs $60 million, NPV is about +$20 million. Invest.
Growing streams use the Gordon form the CFA reading also states: PV = CF1 / (r − g), with g strictly below r. In the room, growth is usually a trap. If the case says "grows at 3 percent forever" and you already picked 8 percent, the denominator shrinks and the value balloons. Say the growth assumption and keep moving; do not let the interviewer watch you invent a terminal multiple.
Mental habit: treat "many years of flat cash" as a perpetuity first. If the answer is strongly positive or strongly negative, you are done. Only refine if the call is close.
The annuity shortcut: finite years without a spreadsheet
Finite, level cash flows need an annuity factor, not twenty discount lines. Present value of a level payment for n years is:
PV = CF × [1 − (1 + r)^(−n)] / r
You will not compute that on paper under pressure. Memorize three 10 percent anchors and adjust.
| Horizon at 10% | Annuity factor (approx) | How to remember |
|---|---|---|
| 5 years | 3.8 | Five years of "almost 0.75 average" |
| 8 years | 5.3 | Between 5 and 10, closer to midway |
| 10 years | 6.1 | Ten years of "about 0.6 average" |
Example. A contract pays $5 million a year for eight years. At 10 percent, PV ≈ 5 × 5.3 = $26.5 million. Capex of $30 million means NPV ≈ −$3.5 million. Pass, unless the case adds a salvage or strategic option the numbers ignore.
If the rate is 8 percent instead of 10 percent, nudge the factor up a notch (eight-year factor climbs toward about 5.7). If the rate is 12 percent, nudge down (toward about 5.0). Directional adjustment beats fake precision.
Ned's rule. If the cash stream is flat and longer than about a decade, use CF / r. If it is flat and shorter, use CF times a memorized annuity factor. If it is lumpy, discount the two or three big years only and treat the rest as a short annuity. Never write a full discount table in a first round.
How to state a discount rate without a debate
Candidates lose the room arguing betas. You need a number and a public anchor, not CAPM theater.
As of early 2025, Aswath Damodaran's published global cost-of-capital work put the median firm near about 9 percent in US-dollar terms, with technology toward the high end of the distribution and financials toward the low end; he also uses the upper deciles (around the low teens) for young, high-risk stories. His current data page (last full update January 9, 2026) publishes sector costs of capital you can cite by name. His Data Update 6 for 2025 explains why industry risk, not corporate folklore, should set the hurdle.
The risk-free floor moves with markets. The US 10-year Treasury yield, via the Federal Reserve's H.15 series on FRED (DGS10), sat near 5 percent in mid-to-late September 2026. That is a floor, not a project rate. A stable industrial project sits well above the Treasury; a speculative digital bet sits higher still.
Script that ends the debate:
- "I'll use 10 percent as a round market hurdle, in line with Damodaran's published median cost of capital band for a typical firm."
- "I'll bump to 12–15 percent because this is an early-stage / high-obsolescence bet."
- "I'll drop toward 8 percent if the cash is contracted and regulated."
Then do the math. McKinsey's Valuation (Koller, Goedhart, Wessels) spends chapters on cost of capital for a reason: in real work the rate matters. In a case, a named public band plus a one-line risk nudge is enough. You are not building a firm WACC.
When payback beats NPV in the room
NPV answers "does this create value at our hurdle?" Payback answers "how fast do we get the cash back?" Sometimes the second question is the one the client (and the interviewer) cares about.
Use payback when:
- The client is cash-constrained and cannot wait for distant value.
- Technology or regulation will obsolete the asset in a few years, so distant cash is fantasy.
- The interviewer frames the decision as a screen ("would you even look at this?") before deeper work.
- Liquidity risk dominates: a positive NPV project that returns cash in year twelve can still kill a stressed balance sheet.
Payback is Capex / Annual CF for a flat stream. A $40 million outlay saving $8 million a year pays back in five years. If the CFO's hard rule is "under four years or no," you recommend pass even when a 10 percent NPV looks fine. Say that tradeoff out loud: "NPV at 10 percent is positive, but payback is five years against a four-year cash rule, so I would not proceed."
ROI (gain / cost) is even coarser. Use it only as a sanity check, never as the final call when timing matters.
One fully worked investment decision
HarborLine Logistics runs regional freight consolidation. Management wants a $28 million automated sortation line. Labor and damage savings are estimated at $4.5 million a year, starting next year, for the practical life of the equipment (well over a decade; treat as ongoing). No major mid-life capex is flagged.
Step 1 — Rate. Mature logistics operations with mostly labor savings, not a venture bet. Use 10 percent, citing Damodaran's published median band for a typical firm, rounded for mental math.
Step 2 — Perpetuity PV. PV ≈ 4.5 / 0.10 = $45 million.
Step 3 — NPV. 45 − 28 = +$17 million. Directionally clear: invest.
Step 4 — Stress. At 15 percent (if automation obsolescence is real): PV ≈ 4.5 / 0.15 = $30 million; NPV ≈ +$2 million. Still positive, but thin. Say you would want a hard look at technology life before signing.
Step 5 — Payback cross-check. 28 / 4.5 ≈ 6.2 years. If HarborLine's board insists on recovery inside five years, the project fails the cash screen despite a positive NPV. Your recommendation then becomes: "Value-creating at 10 percent, but fails the five-year cash rule; I would renegotiate scope or phasing, not approve as-is."
That five-step loop — rate, shortcut, NPV, stress, payback — is what a strong first-round answer sounds like. No year table. No CAPM. A decision.
Practice this today
Open the free quick math drill and run ten timed "CF, r, n → invest or pass" prompts. Then take one full case on /start (three typed turns, instant scores, no account) and force yourself to say the rate sentence before you divide. Everything on CoachNed is open for seven days with no card; after that it is $120 for a recruiting season or $49 a month.
Frequently asked questions
What discount rate should I use in a case interview if none is given?
Use 10 percent as the default round hurdle and justify it in one line with Damodaran's published cost-of-capital data for a typical firm. Nudge up for early-stage or high-obsolescence bets and down for contracted, regulated cash. Do not open a CAPM discussion unless the interviewer asks.
How do I calculate NPV in a case without a calculator?
Classify the stream: long and flat → CF / r; finite and flat → CF times a memorized 10 percent annuity factor (about 3.8 for five years, 5.3 for eight, 6.1 for ten); lumpy → discount the big years only. Subtract capex and recommend.
When should I use payback period instead of NPV?
Use payback when cash recovery speed, liquidity, or short technology life dominates the client's decision. State both: "NPV is positive at 10 percent, but payback misses the five-year cash rule, so I recommend pass or restructure."
Is perpetuity accurate enough for case interviews?
Yes, when cash is roughly level and the horizon is long. Interviewers prefer a clean approximation with a stated assumption over a slow "exact" table. Flag that real life is finite, then close the call.
What is the difference between NPV in a case and a full DCF model?
Case NPV is a decision shortcut under time pressure. A full DCF builds an explicit forecast, reinvestment, and terminal value — the territory of longer modeling work, not a first-round fork. Keep the room answer on rate, shortcut, and recommendation.
Sources
- Discounted Dividend Valuation (CFA Institute) — present-value identity, perpetuity V0 = D/r, and Gordon growth form CF1/(r − g). Checked 2026-09-24.
- Damodaran Online: Data for current year — published sector costs of capital and related discount-rate datasets (last full update January 9, 2026). Checked 2026-09-24.
- Data Update 6 for 2025: The Hurdle Rate Question (Aswath Damodaran) — global cost-of-capital distribution, median near ~9%, and use of upper-decile hurdles for high-risk stories. Checked 2026-09-24.
- FRED series DGS10 (Federal Reserve H.15) — US 10-year Treasury constant-maturity yield as the observable risk-free floor (near ~5% in September 2026). Checked 2026-09-24.
- Koller, Goedhart, and Wessels, Valuation: Measuring and Managing the Value of Companies (Wiley / McKinsey & Company) — standard treatment of cost of capital as the project hurdle in corporate valuation. Checked 2026-09-24.
CoachNed is independent and not affiliated with McKinsey, BCG, Bain, Deloitte, or any firm named above.