A multiple is a full valuation with the assumptions hidden. That is the sentence I wanted to hear from candidates and almost never did. When you say a company trades at a given EV/EBITDA, you are compressing growth, margins, risk and capital intensity into one number and every interview question about multiples is really a question about what got compressed. Answer at that level and the whole topic opens up.
The consistency rule
The pairing rule · one box to memorise
Numerator and denominator must belong to the same investors.
Enterprise value pairs with metrics before interest: revenue, EBITDA, EBIT, because those belong to all capital providers.
Equity value pairs with metrics after interest: net income, earnings per share, book equity, because those belong to shareholders alone.
Mixing levels, price to EBITDA, EV to earnings, produces a number that compares nothing.
This rule is downstream of exactly one concept, the EV and equity split covered in its own article on this site and interviewers use it as a fast probe: ask why EBITDA pairs with EV and the candidate either derives it from whose-cash logic in a sentence or reveals the rule was memorised bare.
When EV/EBITDA lies
- Capital intensity is invisible to it. EBITDA sits above depreciation, so it treats a business that must replace heavy assets every few years identically to one that never spends. Comparing an asset-heavy operator to an asset-light one on EV/EBITDA flatters the heavy one systematically; EBIT or EBITDA less capex is the honest cross-check and saying so unprompted scores.
- It is not cash flow. The oldest trap in the book and I still heard 'EBITDA is basically cash flow' from finalists. It ignores capex, working capital, tax and interest; a growing company can post excellent EBITDA while consuming cash relentlessly.
- Accounting choices leak in. Lease treatment and capitalisation policies move EBITDA between otherwise similar companies, which is why a comp set needs cleaning before the multiples mean anything, as the comps article on this site covers.
When P/E lies
- Leverage distorts it. Two identical businesses with different debt loads print different P/Es, because interest runs through earnings. P/E differences across a set can therefore be capital-structure noise rather than valuation signal; EV multiples strip that out, which is precisely why practitioners default to them for operating comparisons.
- The E is fragile. One-off items, tax-rate differences and buyback-shrunk share counts all move EPS without moving the business. A P/E built on unadjusted earnings inherits every distortion below the operating line.
- It fails where there is no E. Loss-making growth companies push analysis up the statement, to revenue or sector-specific metrics, which is standard in early-stage technology and covered in the TMT questions article.
Sector conventions and the exception that proves the rule
Conventions exist because each sector gravitates to the multiple that hides least: cash-generative operating businesses toward EV/EBITDA, mature dividend-payers toward earnings and dividend measures, high-growth software toward revenue multiples. The instructive exception is financial institutions, where EV barely means anything because debt is raw material rather than financing, so banks trade on price to earnings and price to book instead; the FIG article on this site explains why that world runs on different physics. In an interview, naming the convention for your claimed sector and one reason for it, is the difference between fluency and flashcards.
The question forms to expect
Two companies, same EBITDA, different EV/EBITDA: why? Growth, margins, risk, capital intensity, in roughly that order and being able to rank them shows judgement. Which multiple would you use for a given business and why: derive it from what the multiple hides. And the sleeper, what does a multiple assume: this is where the opening sentence of this article becomes your answer. Rehearse those three aloud and the topic is banked; rehearse them against someone who asked them for fifteen years and that is the IBD Recruiting Review.
FAQ
Should I use EV/EBITDA or P/E?
Follow the consistency rule: enterprise value pairs with pre-interest metrics like EBITDA, equity value with post-interest metrics like earnings. Then choose the multiple that hides least for the business.
When does EV/EBITDA mislead?
In capital-intensive businesses: EBITDA ignores capex, so asset-heavy operators look artificially cheap next to asset-light ones. EBIT or EBITDA less capex is the honest cross-check.
Why do two similar companies trade at different multiples?
Growth, margins, risk and capital intensity, roughly in that order. A multiple is a full valuation with the assumptions hidden; the gap is the assumptions.
Former Executive Director in TMT Investment Banking at Nomura and M&A banker at BNP Paribas. Top-rated mentor with 300+ sessions. About
