EV/EBITDA, P/S and P/B

The P/E ratio prices the equity. But a company is bought and paid for by shareholders and lenders together, and if you want to compare two businesses with different amounts of debt, you have to price the whole thing. That is what enterprise value does. This lesson covers the three multiples that either fix the debt problem or work when earnings do not exist.

Enterprise value

Enterprise value (EV) = market cap + total debt − cash (plus preferred stock and minority interests, if any).

ACME: $4,000M + $600M − $200M = $4,400M

The intuition: if you bought every share of ACME for $4,000M, you would inherit $600M of debt and get $200M of cash in the bank, so the real cost of owning the business is $4,400M. EV is what an acquirer actually pays.

Cash reduces EV because it comes with the purchase. Debt increases EV because the buyer has to service or repay it. A company with a market cap of $1,000M and net cash of $400M has an EV of $600M; the market is valuing the operating business at $600M, not $1,000M.

EV/EBITDA

EBITDA = earnings before interest, taxes, depreciation and amortisation = operating income + D&A.

ACME: $300M + $80M = $380M. EV/EBITDA = $4,400M ÷ $380M = 11.6

EBITDA is a rough proxy for the cash the operations produce before financing, tax and reinvestment. EV/EBITDA is therefore a "whole business" multiple that ignores how the company is financed. It is the multiple that acquirers, private equity and credit analysts use, and it is the one to reach for when comparing companies with different debt levels.

Compare ACME with ACME Heavy from module 2, lesson 2: same business, same EBITDA, but $3,000M of net debt and a $1,400M market cap.

On P/E, Heavy looks cheaper. On EV/EBITDA they are identical, which is correct, because they are the same business. The lower P/E was entirely a product of the leverage, and leverage is risk, not value. This is the clearest example of why EV multiples exist.

Key idea: P/E prices the equity slice; EV/EBITDA prices the whole business. When two companies have different debt, only the EV multiple compares them fairly.

Where EV/EBITDA breaks: EBITDA ignores capex. A business that must spend heavily just to stand still (a telecom, an airline, a miner) has EBITDA that overstates its real cash generation. For those, look at EV ÷ (EBITDA − capex) or at free cash flow yield directly. EBITDA also ignores stock-based compensation when companies present "adjusted EBITDA"; check what was added back.

Price to sales (P/S) and EV/revenue

P/S = market cap ÷ revenue. ACME: $4,000M ÷ $2,000M = 2.0. EV/revenue: $4,400M ÷ $2,000M = 2.2.

The sales multiple exists because revenue is almost never negative and rarely manipulated as heavily as earnings. It is the multiple of last resort for companies with no profits, and the multiple of choice in high-growth sectors where every company is reinvesting all its gross profit.

Its weakness is that a dollar of revenue is worth very different amounts depending on margin. A software company with 80% gross margins and a distributor with 8% gross margins cannot be compared on P/S at all. The rule: P/S is meaningful only within a sector, and even then it needs a margin next to it.

A quick check on whether a sales multiple is sane: what would the P/E be if the company reached a mature margin? A company at 10 times sales that could plausibly earn a 20% net margin at maturity is at 50 times "mature earnings". Whether that is reasonable depends on how fast it gets there.

Price to book (P/B)

P/B = market cap ÷ shareholders' equity, or share price ÷ book value per share. ACME: $4,000M ÷ $1,600M = 2.5, or $40 ÷ $16 = 2.5.

Book value is what accountants say the shareholders own after all liabilities are paid. For most modern businesses it is a poor guide, because the valuable things (brands, code, customer relationships, know-how) are not on the balance sheet, so P/B ratios of 5 or 10 are normal and say little.

P/B matters in three places:

• Banks and insurers. Their assets are mostly financial and carried near market value, so book is a reasonable floor and P/B is the standard multiple. A bank at 0.7 times book is priced for losses; at 1.5 times, for good returns on that book.
• Asset-heavy businesses. Shipping, real estate, some industrials. Book value approximates replacement cost, and the market cycles around it.
• Distress. A stock below book value, with real (not goodwill) assets, has a liquidation argument beneath it. Whether liquidation ever happens is another question.

P/B also links to return on equity (ROE): a business earning 20% on its book value deserves a higher P/B than one earning 5%. ACME's ROE is $210M ÷ $1,600M = 13.1%, and a 2.5 times P/B for a 13% ROE is unremarkable.

Which multiple for which company

Try it: Compute EV for any company with debt: market cap + debt − cash. Then compute both P/E and EV/EBITDA. Find a competitor with a different debt level and compute both for it too. Notice which multiple changes the ranking.

Recap

• EV = market cap + debt − cash; it is what the whole business costs, not just the equity.
• EV/EBITDA compares businesses regardless of leverage; two identical businesses with different debt have different P/Es but the same EV/EBITDA.
• EBITDA ignores capex and often stock-based compensation; check both.
• P/S and EV/revenue work when there are no earnings, but only within a sector and alongside a margin.
• P/B matters for banks, asset-heavy businesses and distress; pair it with ROE.