Why do analysts reach for EV/EBITDA so often?

If P/E is the multiple everyone quotes, EV/EBITDA is the multiple professionals actually compare with. It pairs the "whole business" price (EV) with a measure of operating profit that strips out the noise P/E can't.

What EBITDA is

EBITDA = Earnings Before Interest, Taxes, Depreciation and Amortisation. The name lists exactly what it leaves out — four charges that vary hugely between companies for reasons unrelated to how good the underlying business is:

• Interest — depends on how much debt a company chose to take on.
• Taxes — depend on jurisdiction and accounting.
• Depreciation & amortisation — non-cash charges that depend on past accounting choices.

There are two ways to build it, and which add-backs you apply depends entirely on which line you start from:

• From net income — add back interest, taxes, and depreciation & amortisation. Net income sits after all four have been deducted, so all four come back.
• From operating profit (EBIT) — add back depreciation and amortisation only. Operating profit is already stated before interest and before tax — that is exactly what "earnings before interest and taxes" means — so adding them a second time would count them twice. Fewer moving parts is why this is the cleaner starting line.

Both routes land on the same number. Take a statement with revenue 1,000, cash operating costs 700, D&A 100, interest 40 and tax 40. Operating profit is 200 and net income is 120, so EBITDA is 120 + 40 + 40 + 100 = 300 from net income, and 200 + 100 = 300 from operating profit. Add interest and tax back to operating profit by mistake and you get 380 — 27% too high, and at an enterprise value of 3,000 that turns a true 10.0x into a reported 7.9x. The mistake is invisible in the output, which is why the starting line matters.

Strip the four out and you're left with a rough measure of the cash the core operations throw off, before the financing and accounting overlay. That makes it a cleaner apples-to-apples number across companies.

Why EV and EBITDA belong together

This is the elegant part. EBITDA is profit before interest — it belongs to everyone who funded the business, both shareholders and lenders. EV is the value of the business to everyone who funded it, both equity and debt. So the numerator and denominator are consistent: whole-business value over whole-business profit. Pairing price (equity only) with EBITDA (everyone) would be mixing levels; EV/EBITDA keeps them matched.

Where it beats P/E

Take two companies with identical operations and identical operating profit, but one funded with debt and one without. Their EBITDA is the same, and — crucially — their EV/EBITDA can be the same too, correctly showing them as equally valued businesses. Their P/E, though, would differ, because interest expense drags down the indebted company's net earnings. EV/EBITDA sees past the financing choice; P/E is fooled by it. That's why it dominates when comparing companies with different debt loads, and in industries where heavy depreciation (factories, cables, aircraft) would otherwise distort earnings.

The honest caveats

EBITDA is not cash flow. It ignores the real cost of maintaining those assets — a company that adds back huge depreciation still has to spend real money replacing worn-out equipment. It also isn't standardised, so companies sometimes present flattering "adjusted" versions. And EV/EBITDA is nearly useless for banks, whose "debt" and interest are the business. A tool with a sharp edge — powerful, but not universal.

In the data

The multiple comes published:

Build it yourself and the denominator is the trap. There are two EBITDA figures for one company: a trailing-twelve-month one on the company profile, and one per fiscal year on the income statement. For Apple they currently differ by about a sixth, $168.0 billion trailing against $144.4 billion for fiscal 2025 (read 29 September 2026). Same name, two windows, two multiples. Neither comes with its add-backs shown, so whichever you take, state the window beside the number.

Try it now

1. Read the EV/EBITDA in the table above. Then try to rebuild it and watch the denominator become the trap. The trailing EBITDA is here:

and the fiscal-year one is here:

Divide enterprise value by each in turn. Same name, two windows, two multiples — so whichever you take, state the window beside the number. 2. While you have the income statement open, build EBITDA the two ways this lesson describes. Neither add-back is on this table: depreciation lives on the cash flow statement, and interest expense here is an em dash. Depreciation is a row of the same fiscal year's cash flow statement:

Note that the missing interest line blocks the net-income route for this filer entirely. Then check the operating route against the reported EBITDA row, and reconcile it against EBIT rather than operating income, because this filer's EBITDA is built on the pre-tax line. Adding interest and tax to operating profit is the double-count that quietly turns a true multiple into a flattering one. 3. Now the case where the multiple breaks entirely:

Read its last row. A multiple that means nothing for a bank was not left out; it was printed as a zero. EV/EBITDA is nearly useless where debt and interest are the business. 4. Compare capital-intensive companies against peers in the same industry, three US telecoms on both multiples:

Rank the three on trailing P/E, then on EV/EBITDA, and note where the ranking or the gaps change. Where the two multiples disagree, the disagreement is usually about debt.