Valuation Multiples and What Each One Assumes

Open any stock screener and the valuation columns sit side by side. P/E, EV/EBITDA, price to sales, price to book. They look interchangeable. They aren’t, because each one divides a price by a different line from the financial statements, and the line it picks is a bet about what drives the business, so a stock that looks cheap on one column can look dear on the next. Knowing the bet is the whole skill.

P/E, the ratio everyone quotes

The price to earnings ratio divides the share price by earnings per share. Take a hypothetical company with 125,000,000 shares at $40. Its market value is $5,000,000,000. It earned $250,000,000 last year. That’s $2.00 a share. The P/E is 20.

Turn it upside down and you get the earnings yield. One divided by 20 is 5%, the profit you buy with each dollar of price.

P/E assumes earnings are what matter. It also assumes last year’s were typical. Both assumptions fail often. A one-off gain from selling a division inflates net income and drags the ratio down, while a company that spends heavily on growth, or simply had a bad year, can show a P/E of 60 or a negative one that tells you almost nothing about what it will earn later. It is also blind to debt. Interest has been paid before net income is counted.

EV/EBITDA, which counts the debt

Enterprise value is the price of the whole business. You pay for the shares, you take on the debt, and you get the cash in the till. EBITDA is earnings before interest, taxes, depreciation and amortization, a rough measure of operating profit before financing decisions touch it.

Now change one thing. A second hypothetical company has the same $5,000,000,000 market value, the same $600,000,000 of EBITDA and the same $500,000,000 of cash, and owes nothing. Its enterprise value is $4,500,000,000. Its EV/EBITDA is 7.5. Same share price, same operating profit, and the debt-free one is clearly cheaper, a gap that a P/E column could miss completely if the interest bill happened to be small.

EBITDA has a blind spot of its own. It leaves out depreciation. For a business that must keep replacing trucks, plants or servers, that is a real cost. A capital-hungry company can look cheap here and still turn little of its profit into cash.

Price to sales and price to book

Price to sales divides market value by revenue. The first hypothetical company, with $2,500,000,000 of sales, trades at 2 times sales. The ratio still works when there are no profits, which is why young companies are so often quoted on it. Margins break it. A grocer keeping two cents of every sales dollar and a software firm keeping thirty cents earn wildly different amounts from the same revenue, and a ratio that treats their sales alike will call the grocer cheap every single time.

Price to book divides market value by shareholders’ equity. It suits banks. Their assets are mostly loans and securities carried near market value. For a company whose worth sits in brands, patents or code, book value says little, because accounting keeps most of that off the balance sheet, and years of buybacks can shrink equity until the ratio looks absurd.

Multiple	Divides price by	Assumes	Misleads when
P/E	Earnings per share	Last year’s profit is typical	Earnings are one-off, negative or cut by debt
EV/EBITDA	Operating profit before D&A	Debt belongs in the price	The business must keep replacing assets
Price to sales	Revenue	Margins will be normal	Margins differ widely
Price to book	Equity	Assets are marked near value	Value is intangible or buybacks shrank equity

Compare inside one industry

Every multiple packs growth, risk and margins into a single figure. Across industries those differ too much. A utility at 15 times earnings and a software company at 40 might both be fair. Screen the whole market for the lowest P/E and you mostly get banks, energy producers and companies in trouble.

Inside one industry the businesses share customers, costs and cycles, so the ratios start to mean something, and an outlier becomes a question worth an hour with the filing. Say a group of hypothetical parts suppliers trades between 8 and 11 times EBITDA and one sits at 6. What does the market know? Sometimes nothing. More often there’s debt coming due, a big customer leaving or a margin about to fall.

Screeners make this sorting quick, and the guide to screener performance columns covers the rest of the table. The broader ideas sit on the investing hub.

Where multiples run out

A multiple tells you how the market prices a company against its neighbors. That’s useful. It’s also circular, since a whole industry can be mispriced at once and every stock in it will still look normal next to the others. A discounted cash flow model, the next lesson, builds a value from the company’s own cash and ignores the neighbors entirely.