Price-to-book compares what buyers pay for a share to the accounting residue behind it. Two equivalent forms, pick whichever matches the data you have: market capitalisation ÷ shareholders’ equity, or price per share ÷ book value per share. A ratio of 1 means investors are paying roughly what the accountants recorded owners as having contributed and kept; below 1 says the market prices those claims at a discount, above 1 says a premium.
That’s the easy half. The hard half – and the part most beginner explanations get backwards – is that neither direction tells you anything until you know what kind of book you’re dividing by.
Book value is history, not an appraisal
Equity comes off the balance sheet as contributed capital plus retained profits, with assets carried mostly at original cost less depreciation. It is not what you’d collect in a wind-down and it isn’t replacement cost either. Buildings frequently appraise above their carried figure; specialised machinery can carry large numbers while being worth scrap. So the denominator reflects decisions made over many years, filtered through accounting rules, rather than today’s value of anything.
Which means the same P/B reading carries opposite meanings depending on what the business actually owns. Compare two firms both trading at 0.8× book. One is a lender: its assets are loans and securities, close enough to monetary that carrying values approximate realisable ones, so 0.8 could genuinely mean paying under the value of the assets. The other is a consultancy: its whole product walks home every evening, and the equipment figure is office furniture. Its “discount” is arithmetic without content – there was never much book value to discount.
High readings deserve the same scepticism in reverse. Software, payments networks, consumer brands and service businesses produce enormous profits from very little recorded capital, because their actual assets – agreements, habits, code, distribution – mostly aren’t allowed on the balance sheet at all. A technology company at 9× book is not automatically expensive; a bank at 1.2× is not automatically conservative. Ratios travel between industries badly.
What the ratio is really asking
There’s a clean way to think about the premium. Paying above book is rational exactly when reinvested capital earns more than it costs. Work the logic rather than memorising thresholds: if a business turns 15% on owners’ equity and keeps most of it, book value per share compounds quickly and a buyer today is purchasing that future accretion – paying somewhat above recorded book can still be cheap. Run the same business at 2% on equity and each retained dollar earns less than it could outside; even a substantial discount to book can be generous. Return on equity does most of the explanatory work, and decomposing it into margin, turnover and leverage tells you whether the premium rests on pricing power or on borrowing.
So P/B becomes genuinely useful in specific territory rather than universally:
- Financial firms, where assets resemble money and regulators police capital, which keeps the denominator comparable across competitors. Analysts of lenders usually reach for tangible book – equity after goodwill and intangibles – precisely because acquired premium isn’t loss-absorbing capital. Deduct those before trusting the number; goodwill impairs late by construction.
- Liquidation and deep-value hunting. Graham built a whole approach around the extreme case: a company whose market price sat below a conservative measure of its net current assets. Those “net nets” are rare today and rarely cheap for no reason, but the reasoning – find situations where the downside has some asset backing independent of earnings forecasts – stays sound, and the screen that encodes it shows how mechanical that discipline gets. His thinking is worth reading directly rather than through summaries.
- Cyclical troughs in heavy industry, where pessimism prices entire franchises like melting ice cubes and book gives you a floor to argue about instead of vibes.
Everywhere else it’s decoration, useful mostly as a prompt to go look.
Four ways people misuse it
As proof of cheapness. Cheap-looking stocks cluster with genuinely deteriorating ones – obsolete products, losing share, fraud quietly discovered late. A low ratio is a question (“why?”), never an answer, and value traps are identified after the fact by definition. Estimating what went wrong belongs in fundamental analysis, not in a division.
Without adjusting for buybacks. Repurchase programmes shrink equity, which lifts return on equity and lowers P/B simultaneously – mechanically flattering both metrics while nothing operational improves. If a company borrowed to buy itself back, an attractive-looking discount can be leverage wearing a costume. Read the cash flow statement’s financing section alongside the discussion of why cash beats profit.
Against a cross-industry average. “The sector trades at 3× and this at 1×” compares business models more often than opportunities. Ask what structurally differs: capital intensity, contract durability, funding costs, regulation.
From a single period. Equity moves on impairments, one-off charges, pension swings and currency translation. Pull several annual columns from the filing itself and see what drove each jump. Financial-site figures are transcriptions and transcriptions drift, sometimes materially.
Reasonable way to use it
Take the number, note the industry, deduct goodwill and intangibles, look up book value per share across several years, then ask the one economic question – does this earn enough on owners’ capital to justify whatever premium or discount it carries? If the honest answer is that you can’t tell, that’s not failure, it’s the correct result for a business whose value lives somewhere the balance sheet can’t record. Move to earnings-based measures and cash generation for those cases, keep an eye on cost – expenses are the reliable enemy of returns – and remember that every shortcut ratio survives precisely because it removes judgement from places where it was doing the most work.
This is educational explanation of a valuation ratio, not investment advice, and nothing here recommends any security. Ratios describe published accounting positions rather than fair value; a low or high reading can be wrong in either direction, markets can stay unreasonably priced far longer than a plan survives, and any holding can lose value including one bought at a large discount to its recorded assets.