When a DCF Is Misleading: Interview Guide to Broken Assumptions

"Walk me through a DCF" is table stakes. "When would you not trust a DCF?" is how interviewers see judgment. You need named failure modes and a triangulation plan, not cynicism about all models. The goal is to sound like someone who builds DCFs and knows when they stop being decision-useful.

The short answer

"A DCF is only as good as the cash flow path, WACC, and terminal value. It misleads when cash flows are highly cyclical or binary, when TV is most of the value with soft assumptions, when WACC is tuned to hit a price, or when capital structure will change so much that a single WACC is dishonest. In those cases I lean harder on comps, precedents, scenario analysis, or an LBO / ability-to-pay frame."

Failure modes that land in interviews

Terminal value traps

You already know exit multiple vs perpetuity. The interview add-ons:

• Perpetuity growth above long-run GDP / inflation without a durable monopoly story
• Exit multiple richer than current comps without a quality or margin change
• Mid-year versus year-end convention inconsistency that quietly inflates PV
• Using a peak EBITDA exit multiple on a cyclical name

If TV is 80% of value, you do not have a five-year model with a small tail. You have a terminal-value opinion with a forecast attached. Say that out loud.

Cyclicals and commodity names

For shippers, miners, homebuilders, and other cycle names, a single base-case DCF often lies. Prefer:

• Cycle-normalized FCF or mid-cycle EBITDA
• Multiple scenarios (trough, mid, peak)
• Trading comps on mid-cycle earnings
• Clear statement that you will not capitalize the current spot year forever

Early-stage and binary outcomes

Pre-profit tech, single-asset biotech, and turnarounds can make explicit forecast DCFs look precise while being fiction. Use scenario probabilities, market comps on revenue or users when that is the market language, and milestones. A 10-year hockey-stick discounted at a neat WACC is not rigor.

WACC and capital structure honesty

A DCF assumes a discount rate matched to the risk of the cash flows. Problems start when:

1. Target leverage in WACC never matches the forecast debt path
2. Beta is peer-average nonsense for a unique asset
3. Country or execution risk is ignored
4. Someone reverse-solves WACC to hit the bid

For changing leverage, mention APV or a changing WACC as conceptual fixes, then say in practice you often triangulate rather than over-engineer in a screen.

When other methods win

Precedents win when control premiums and synergies matter and the deal market is active. Synergy DCFs for strategics are useful only if synergy timing and cost-to-achieve are diligence-backed.

How to answer without dumping DCF

"I still build a DCF as a sanity check and to force an explicit view on growth, margins, and reinvestment. I will not hang an offer on it alone if TV is most of value or the FCF path is not underwritable. I triangulate with comps and a returns or synergy frame depending on buyer type."

That tone beats "DCFs are useless," which sounds lazy. A football field with a wide DCF band and a tight comps band is often the honest picture when TV assumptions dominate.

Mini drill

Take any DCF you have built. Compute TV / total EV. Shock g by 50 bps and WACC by 50 bps. Narrate which input moved EV more. That sensitivity is the interview follow-up, and it trains you to know where the model is fragile.

30-second verbal

"A DCF misleads when the forecast is not underwritable, when terminal value dominates with soft assumptions, or when WACC is tuned to a price. I still build one, then I check TV share, stress the key inputs, and triangulate with comps or a sponsor returns frame."

Related

Walk Through a DCF, WACC, Football Field.

Practice on SheetRank

On Project Horizon, compute how much of your EV sits in TV and shock exit assumptions before you defend a single intrinsic number.