Not investment advice. Educational reading. See Disclaimer.
L.9 · ADVANCED · 2 MIN
Sanity-Checking a DCF (Exit Multiples, ROIC Convergence)
A DCF is a stack of assumptions that compounds into a number. Without sanity checks, the model can talk itself into any answer the assumer is hoping for -- and that anchoring risk is exactly what makes DCFs unreliable on the desk. Three sanity checks separate a trustworthy DCF from an over-confident one: implied exit multiple, implied terminal ROIC, and the gap between DCF output and the observed market price. When two or three of these flash red simultaneously, the right inference is almost always that the model assumptions are aggressive -- not that two or three independent cross-checks are coordinating to be wrong. A lifelong investor uses the cross-checks to keep the DCF honest with itself before letting it influence a portfolio decision.
Implied terminal EV / terminal EBITDA -- compare to peer median and the company's own historical multiple range
Implied multiple more than 30-40% above the comp median or historical mean is the model assuming an unprecedented re-rating
ROIC-convergence check
Terminal-year NOPAT divided by terminal invested capital -- compare to mature historical ROIC and the industry's structural ROIC ceiling
Terminal ROIC well above the company's history or the industry's structural ceiling is the model assuming competitive dynamics that have never held
Implied-vs-observed EV check
DCF intrinsic EV compared to the current trading EV -- and the implied upside or downside in percentage terms
Gaps of 40%+ on a mature liquid name almost always reflect a model assumption issue rather than a market mispricing
§ 02
The cross-checks are not optional; they are the discipline that prevents a DCF from manufacturing its own conclusion. A model that survives all three is one a lifelong investor can act on. A model that fails even one needs the failing assumption found and stress-tested before the number leaves the spreadsheet.
§ 03
Take a DCF you have built (or one in your watchlist's analyst-coverage notes). Compute the implied terminal EV/EBITDA from the terminal-period free-cash-flow and terminal multiple. Compare it to the median peer multiple in **Fundamentals**. Now compute implied terminal ROIC (NOPAT divided by invested capital at terminal year). Compare it to the company's mature historical ROIC. If either cross-check fails by more than 30%, find the assumption driving the gap before the DCF output influences any portfolio decision.
§ 04
A DCF returns intrinsic value 18% above the current trading price. The implied exit multiple matches the peer median to within 5%. Implied terminal ROIC equals the company's historical mature ROIC. What is the disciplined reading?
§ 05
DCFs do not lie -- the assumer does. The three sanity checks (exit multiple, terminal ROIC, implied-vs-observed) are the discipline that keeps the model honest with itself. A DCF that survives all three is a tool worth acting on. A DCF that fails any of them needs the failing assumption found and stress-tested before the number influences anything.
Five questions · AI feedback
●○○○○
Sit with the ideas.
Your DCF on a mature industrial returns an intrinsic enterprise value of $14 billion. Terminal-year EBITDA is $1 billion, implying an exit EV/EBITDA of 14x. The trading comp set (eight peers) is averaging 7.5-9x with 8.5x at the median; the company itself has averaged 8x EV/EBITDA across the prior cycle. Implied terminal ROIC is 22% even though the company's mature historical ROIC is 12% and the industry's structural ROIC is closer to 10-13%. The implied EV is roughly 60% above the current trading price. What is the most disciplined sanity-check reading?
