The multiples that matter

A multiple is a shortcut. Instead of building a full model, you take a company's value and divide it by one number that stands in for its earning power. The whole game is matching the right value to the right number. Get that pairing wrong and the multiple is nonsense, even if the arithmetic is clean.

Here's the one rule that runs the entire topic: value on top must be claimed by the same investors as the metric on the bottom. That's it. Everything below is just working out who has a claim on what.

The two families of multiples

There are two kinds of multiples, and they answer to two different owners of the business.

Enterprise value (EV) multiples put enterprise value on top. Enterprise value is the price of the whole operating business, the cost to buy it free of its financing choices. It belongs to everyone who funds the company: debt holders and equity holders together. So the metric on the bottom has to be a number those same people all share, which means it must sit above interest expense on the income statement. Interest is what gets paid to lenders. If you haven't paid it yet, the money still belongs to the whole capital pool.

Equity value multiples put equity value on top, the piece owned only by shareholders. So the bottom metric has to be a number that belongs only to shareholders, which means it sits below interest, after the lenders have been paid.

That single distinction, above the interest line or below it, decides which multiple you're allowed to build.

| Multiple | Numerator | Denominator | Sits where vs. interest | |---|---|---|---| | EV / Revenue | Enterprise value | Revenue | Above (top of the P&L) | | EV / EBITDA | Enterprise value | EBITDA | Above | | EV / EBIT | Enterprise value | EBIT | Above | | P / E | Equity value (share price) | Net income / EPS | Below |

Revenue, EBITDA, and EBIT are all struck before interest. Net income is struck after it. That's why the first three pair with EV and the last one pairs with equity value.

Why EV/EBITDA is capital-structure neutral (and P/E isn't)

This is the part interviewers actually press on, so reason it through instead of memorizing it.

Take two companies with identical operations. Same revenue, same margins, same everything except one is funded with lots of debt and the other with almost none. Should they be worth the same as businesses? Yes. The operating engine is identical. Only the financing wrapper differs.

Their EV/EBITDA multiples should come out roughly the same, and they do. EBITDA is computed before interest, so the debt load doesn't touch it. Enterprise value ignores the financing mix by construction. Both sides of the ratio are blind to leverage, so the multiple is too. That's what "capital-structure neutral" means: you can compare a heavily levered company to a debt-free one on the same axis.

Now look at P/E. The levered company pays real interest expense, which drags its net income down. Lower net income, and if the market prices both businesses' equity sensibly, a different P/E. Change nothing about the operations, just add debt, and the P/E moves. So P/E is not capital-structure neutral. It bakes in financing decisions, share count, and even one-time tax items.

A quick numerical check

Two companies, identical operations. Both have EBITDA of $100 and, say, an enterprise value of $800.

$\frac{\text{EV}}{\text{EBITDA}} = \frac{800}{100} = 8.0\times \text{ for both}$

Company A is all-equity. Company B carries $300 of net debt. Since $\text{EV} = \text{equity value} + \text{net debt}$, their equity values differ:

• Company A equity value: 800 - 0 = \800$
• Company B equity value: 800 - 300 = \500$

Same business value, same EV/EBITDA, different equity values and different P/E. The EV multiple saw through the leverage. The equity multiple didn't. For the mechanics of that bridge, see Enterprise value vs. equity value.

When to use each

EV/EBITDA is the workhorse. Use it to compare companies with different debt loads, different tax situations, or different depreciation policies, because EBITDA strips all three out. It's the default for most industrials, consumer, and healthcare comps.

EV/EBIT keeps depreciation in. Prefer it when capital intensity genuinely differs between companies, because EBITDA flatters a business that has to spend heavily on equipment. EBIT charges them for using up their assets; EBITDA pretends that's free. A CFO deciding between a capex-heavy and capex-light target cares about that difference, so you should too.

EV/Revenue is the fallback when profits are negative or meaningless: early-stage software, pre-profit growth names, a turnaround. It says little about quality, so treat it as a last resort, not a headline.

P/E is what public-market equity investors quote, and it's unavoidable for banks and insurers, where the capital structure is the business and EBITDA barely makes sense. Outside financials, lean on it less than EV multiples because of the leverage contamination above.

Reading a multiple like a banker

A multiple isn't a fact, it's a claim about growth, risk, and returns squeezed into one number. A high EV/EBITDA usually means the market expects fast growth or low risk. A low one can mean a cheap stock or a broken business. The multiple only becomes useful next to a peer set. On its own it's a number without a sentence.

That's why you never quote one multiple in isolation. You build a set of trading comparables, clean each one for non-recurring items, and read the range. Two things drive whether the number means anything: picking genuinely comparable companies, and using a consistent metric (same LTM or forward basis) across all of them.

And remember the ceiling on this whole approach. Relative valuation tells you what the market pays for similar companies today. It says nothing about whether the market is right. That's the job of a DCF, which values the business off its own cash flows and its own WACC rather than the crowd's mood. You want both. Comps for the market's verdict, DCF for the intrinsic one, and you flag it when they disagree.