Revenue vs. EBITDA vs. Free Cash Flow: What Each Multiple Is Really Pricing

Every multiple is a price divided by something. The something decides how much of the business you're looking at, and how much you're taking on faith.

A valuation multiple is just a price divided by a measure of the business. Enterprise value over revenue, enterprise value over EBITDA, some value over free cash flow. Candidates usually learn these as three interchangeable shortcuts and pick whichever one the comps page happens to show. That works until an interviewer asks why you picked it.

The useful way to think about it: each denominator sits at a different point on the income statement, and everything below that point is something the multiple quietly assumes. A revenue multiple assumes the most. A free cash flow multiple assumes the least. EBITDA sits in the middle, which is a big part of why it's the default.

First, the matching rule

Before getting into what each one prices, there's one rule that catches people in interviews more than any other. The numerator and the denominator have to belong to the same group of investors.

• Enterprise value belongs to everyone who funds the business, debt and equity together. So it pairs with metrics calculated before interest is paid: revenue, EBITDA, EBIT, and unlevered free cash flow.
• Equity value belongs only to shareholders. So it pairs with metrics calculated after interest: net income (which gives you P/E) and levered free cash flow.

EV over net income is the classic mismatch. The numerator includes the lenders' claim, the denominator has already paid them, and the resulting number doesn't mean anything. If you only remember one thing from this post for a technical round, make it this.

EV/Revenue: pricing the margins you hope to get

Revenue is the top line, so a revenue multiple says nothing about costs. When you pay a multiple of sales, you're implicitly betting on what margin the company will eventually earn on those sales. The margin isn't in the number. You're supplying it in your head.

There's a simple identity that makes this concrete. EV/Revenue equals EV/EBITDA times the EBITDA margin. A company at 2.0x revenue with a 20% EBITDA margin is trading at 10.0x EBITDA. Flip it around: a software company at 10.0x revenue that you think can reach a 30% margin at maturity is being priced at roughly 33x those future earnings. Saying that out loud is a much better answer than “software trades on revenue.”

When it's the right tool

• The company has negative or near-zero EBITDA, so an earnings multiple is meaningless. Early-stage and high-growth companies are the usual case.
• Margins across the peer set are distorted by heavy reinvestment, like sales and marketing spend that will shrink as a share of revenue later.
• You need a rough sanity check fast and the earnings data is messy.

The weakness is the flip side of the strength. Two companies with the same revenue multiple can have completely different economics. A grocery chain and a software company might both do $500 million of sales, and nobody would pay the same price for both. Revenue multiples are only comparable across businesses with similar margin potential.

EV/EBITDA: pricing operating profit, before the bills that vary

EBITDA is earnings before interest, taxes, depreciation, and amortization. Stripping out interest makes it neutral to how the company is financed. Stripping out taxes removes differences in tax jurisdiction and structure. Stripping out D&A removes accounting choices about how fast assets get written off. What's left is a rough read on the operating profit of the core business, which is why it compares well across peers and why it's the default multiple on most comps pages and in most LBO models.

The catch is what it leaves out. Depreciation is a non-cash expense, but it exists because the company spent real cash on equipment, buildings, or technology at some point, and it will have to keep spending to stay in business. EBITDA ignores capital expenditure entirely. So does EV/EBITDA.

A quick example

Take two companies, each with $100 of EBITDA and each with an enterprise value of $1,000. Both trade at 10.0x EBITDA. On a comps page they look identical.

Now add capex. Company A needs $10 a year to maintain its assets. Company B is capital heavy and needs $50. Ignore taxes and working capital for a second to keep the math clean. EBITDA minus capex is $90 for A and $50 for B. Divide enterprise value by that and A trades at about 11.1x, while B trades at 20.0x. Same EBITDA multiple, wildly different price for the cash each business actually throws off.

This is why some practitioners look at EV/(EBITDA minus capex) for capital-intensive industries, and why an interviewer who asks “why might two companies with the same EV/EBITDA deserve different multiples” is usually fishing for capex intensity, growth, or margin stability.

Watch the adjustments

Most companies report some version of “adjusted” EBITDA, and the adjustments aren't standardized. Stock-based compensation is the big one. It's non-cash, so companies often add it back, but it's a real cost to shareholders because it dilutes them. If one comp adds it back and another doesn't, the multiples aren't measuring the same thing. Restructuring charges that show up every single year are another flag. A charge that recurs is a cost of doing business.

Free cash flow multiples: pricing what's actually left

Free cash flow is the cash the business generates after paying for what it needs to keep running. Unlevered free cash flow starts from operating profit, takes out taxes, adds back D&A, then subtracts capex and the increase in net working capital. It's calculated before interest, so it pairs with enterprise value. Levered free cash flow also takes out interest and required debt repayment, so it pairs with equity value. Price to levered free cash flow, or its inverse, the free cash flow yield, is how a lot of public equity investors frame it.

This is the most complete of the three. Capex is in it. Working capital is in it. Taxes are in it. It's also the number a DCF is built on, so an FCF multiple is the closest thing to a shortcut version of an intrinsic valuation.

The problem is that it's lumpy. A company that builds a new plant this year will show a terrible free cash flow number, even if the plant pays for itself many times over. A company that stretches its payables or cuts back on maintenance can make one year look great. Working capital swings around quarter to quarter. That noise is why FCF multiples show up more as a check on mature, steady businesses than as the headline number on a banker's comps page.

Side by side

• EV/Revenue prices the top line and leaves margin, capex, taxes, and working capital to your assumptions. Best for unprofitable or early-stage companies. Weakest when margins differ a lot across peers.
• EV/EBITDA prices operating profit before financing, taxes, and D&A. Best for comparing established companies with different capital structures. Weakest when capex intensity varies or adjusted EBITDA gets aggressive.
• EV/Unlevered FCF and Price/Levered FCF price the cash that's actually left over. Best for mature, stable businesses. Weakest when a big investment year or a working capital swing distorts one period.
• P/E prices net income to shareholders. Common in public equity and for financial companies, but it's affected by leverage and tax structure, so it compares less cleanly across companies with different balance sheets.

How this comes up in the interview

The direct version is “what are the most common valuation multiples and when would you use each.” A strong answer names the multiple, the matching rule, and one sentence on what the denominator leaves out. You don't need all four. Two done well beat four recited.

The indirect versions are more common and more revealing. Some you should have ready:

What to actually practice

Be able to say what each multiple ignores in one breath. Be able to run the EV/Revenue equals EV/EBITDA times margin identity in your head. Have the two-company capex example ready, with numbers. And when you see a comps page, check whether everyone's EBITDA is adjusted the same way before you trust the median. See reading a comps set for that side of it, and strategic vs. financial buyers for why the same EBITDA can be worth different prices depending on who's buying.