Valuation Metrics: How to Tell If a Stock Is Cheap

A low price-to-earnings ratio does not tell you a stock is cheap. It tells you the market expects less from this company than from the ones trading higher. Whether that expectation is wrong is the entire job, and no ratio is going to do it for you.

Every valuation ratio has the same shape

Put the price of the business on top. Put some measure of what the business actually produces underneath: earnings, book value, cash flow, dividends. Divide, and you have a multiple. It answers exactly one question, which is how much you are paying for one unit of that particular thing, and it answers nothing else. That is why a multiple read on its own is not information. It becomes information the moment you set it against something: the same company five years ago, the three or four businesses that genuinely compete with it, or the market as a whole. On its own it is just a number with a decimal point.

Price to earnings: what one unit of profit costs

The price-to-earnings ratio divides the share price by earnings per share. Here is an illustration with round numbers. A share trades at 50 and the company earned 5 per share last year, so the P/E is 10 and you are paying 10 for each 1 of annual profit. A competitor also trading at 50, but earning 2 per share, sits at 25. The market is charging you two and a half times as much for the same 1 of profit, and it is doing that for a reason it has not written down anywhere.

What the ratio hides is everything sitting behind the word earnings. Accounting profit contains non-cash charges and can be shoved around by one-off events: a legal settlement, a writedown, the sale of a building. One bad quarter shrinks the denominator and makes a P/E look absurd; one large gain does the opposite and makes an expensive company look like a bargain. Check, too, whether the figure you are reading uses last year's reported earnings or next year's forecast. Trailing and forward P/E can be a long way apart, and almost nobody labels which one they are quoting.

PEG: the P/E with a growth assumption bolted on

A company growing profits quickly deserves a higher P/E than one standing still. That is the honest objection to using P/E alone, and PEG is the attempt to answer it: take the P/E and divide it by the expected annual earnings growth rate written as a percentage. A P/E of 30 alongside 30 percent expected growth gives a PEG of 1. The idea is associated with Peter Lynch, who wrote in One Up on Wall Street about comparing what you pay for a company with the rate at which it is growing. The weakness shows up as soon as you say it out loud. The denominator is a forecast, usually somebody else's, and forecasts of fast growth are among the least reliable numbers in finance. PEG also flatters any company recovering from a weak base year, and it falls apart completely when growth is negative, because dividing by a minus sign produces a small tidy figure that means nothing at all.

Price to book: built for balance sheets full of real things

Book value is what the balance sheet says would be left for shareholders after every liability is subtracted from every asset. Divide by the share count and you get book value per share, and dividing the price by that gives you P/B. For a bank or an insurer this is genuinely informative, because most of what such a company owns is financial instruments carried at values that get marked and audited. For a software house or a consultancy it is close to useless. The things that make those businesses valuable, mostly research spending and brand, are expensed in the year they happen rather than recorded as assets, so the balance sheet never sees them. Years of buying back shares above book value shrink reported equity as well, which can push P/B upwards or even negative while nothing whatsoever is wrong with the business.

EV/EBITDA: the multiple that refuses to ignore debt

Two companies can trade on an identical P/E while one carries no debt and the other is financed to the hilt. The EV/EBITDA multiple closes that hole. Enterprise value takes the market value of the equity, adds the debt and subtracts the cash, which is roughly what it would cost to buy the whole business and settle what it owes. EBITDA sits high up the income statement, above interest, so the comparison is neutral about capital structure. It asks what the operations cost you, before the separate question of who paid for them.

The catch is the D and the A. Depreciation and amortisation get added back as though they were bookkeeping fiction, and for an airline or a manufacturer they are anything but: the aircraft wears out, and the money walks back out of the door as capital spending a few years later. EBITDA is also not a defined measure under either US GAAP or IFRS. In the United States the SEC's Regulation G requires a company that publishes a non-standard measure like this to reconcile it to the nearest reported accounting figure, and the European regulator ESMA asks issuers for the same discipline in its guidelines on alternative performance measures. So when a company leads with adjusted EBITDA, the reconciliation is right there in the filing. Go and read what was adjusted.

Free cash flow yield: the number hardest to dress up

Free cash flow is the cash generated by operations minus the cash spent on property and equipment, and both figures live on the cash flow statement rather than the income statement. Divide it by the market value of the equity and you get a yield you can hold up against a bond. As an illustration: a company worth 1,000 in the market that produced 80 of free cash flow last year is on an 8 percent free cash flow yield. Cash is harder to massage than profit, though not impossible and far lumpier. A single year can be flattered by paying suppliers more slowly or by postponing the factory upgrade. Look across five years instead of one, and check whether capital spending happened to be unusually low in exactly the year that produced the attractive number.

Dividend yield: when a big number is a warning

Dividend yield is the annual dividend per share divided by the price, and it rises for two completely different reasons. One is good news: the board increased the payment. The other is not: the share price collapsed while the dividend stayed where it was, which is the market saying out loud that it does not expect the payment to survive. Illustration: a dividend of 2 on a share priced at 40 is a 5 percent yield, and if the price halves to 20 the yield doubles to 10 percent without the company handing over a single extra cent. The two-minute check is the payout ratio, meaning the share of earnings or of free cash flow being paid out. A payout swallowing nearly everything the business generates leaves nothing for debt repayment or reinvestment, and the dividend is usually the first thing a board cuts when trading turns.

The same multiple means different things in different industries

A regulated water utility with dull, predictable cash flows and a fast-growing software company will never trade on the same multiple, and neither of them is mispriced because of it. Compare a business with its own history and with the small number of companies genuinely doing the same thing. Cyclicals invert the logic entirely. A steel producer or a homebuilder looks cheapest on P/E precisely when profits are peaking at the top of the cycle, and looks most expensive at the bottom, when earnings have collapsed and the recovery has not arrived yet. Buying the low P/E there is the classic way to get the timing exactly backwards.

Cheap for a reason

Most statistically cheap stocks are cheap for a reason a screener cannot see. The customer base is shrinking. A patent expires next year. A regulator is circling. The accounts are aggressive in a way that only shows up in the notes. That is the value trap, and it is not a rare accident, it is the normal case, because the market is a reasonably good machine and a low multiple usually means somebody has already worked out something you have not. So treat a low ratio as a question rather than a finding. The question is this: what does this price imply about the future of this business, and do I have a specific, checkable reason to believe that implication is wrong?

What to do once the screen has given you a shortlist

Screening on multiples is the cheap part, and turning a screen into a shortlist worth your time is a process in itself, covered separately in how to find undervalued stocks. The expensive part is the reading: the annual report, which for a company listed in the United States is the 10-K and is free to anyone on the SEC's EDGAR database. Whatever value you eventually arrive at, insist on a margin of safety and buy meaningfully below it. Your estimate is a considered guess with error bars around it, and the discount is what pays you for the times the guess is wrong.

Try this on one company you already hold. Work out its P/E, its free cash flow yield and its price to book, then write a single sentence for each saying what that number implies about what the market expects. If you cannot write the sentence, the ratio was never information. It was decoration.

Written by

Federico Romaldi

Co-Founder, Worthmap

Published: August 14, 2026

Federico is a co-founder of Worthmap, a wealth-intelligence platform built for serious investors. With a background in software engineering and a long-standing passion for value investing, he created Worthmap to bridge the gap between net-worth tracking and investment analysis.