Price-to-Book Ratio: What It Says and When to Listen

What the ratio actually measures

Price-to-book divides the stock price by the book value per share. Book value — also called shareholders' equity — is what remains after you subtract total liabilities from total assets. It's the accounting residual: if you sold every asset at its carrying value and paid off every obligation, book value is what shareholders would theoretically take home.

The arithmetic is simple. If a company carries $4 billion in equity on its balance sheet and has 400 million shares outstanding, book value per share is $10. A stock price of $12 gives you a P/B of 1.2x. You are paying $1.20 for every dollar of net assets. A P/B above 1 means investors believe the business generates returns that justify a premium to the balance sheet. A P/B below 1 means the opposite — the market believes the assets are worth less in use than the accountants say they are. Whether that's the market seeing something the balance sheet misses, or the market getting it wrong, is the entire question.

Consider two hypothetical regional banks, both reporting $20 book value per share. One trades at $24 (P/B 1.2x), the other at $16 (P/B 0.8x). The first is earning 13% return on equity in a normal rate environment. The second is earning 6%. The discount on the second bank isn't random. It's the market pricing a business that earns 6% on equity when investors can earn 5% on short-term Treasuries — and finding it worth less than book. That logic is usually correct.

The industry problem nobody mentions

Most explainers present P/B as a universal tool with different normal ranges by sector. That understates the actual problem. For a large class of businesses, book value isn't just a rough guide — it's an actively misleading number. Using P/B on those companies produces conclusions that are structurally incorrect, not just imprecise.

The issue is intangible assets. Most of what makes a software company valuable — its codebase, customer contracts, brand, the institutional knowledge embedded in its engineering team — doesn't appear on the balance sheet at all, or appears at a fraction of its economic value. GAAP accounting requires companies to expense most internal development costs as they occur. A software company that spent $2 billion building a platform over a decade carries $0 on the balance sheet for that platform. Its book value might be $600 million in cash and servers. If the software generates $1.2 billion in recurring annual revenue, the market might value the business at $15 billion — a P/B of 25x. That ratio isn't signaling that the stock is expensive. It's signaling that book value is the wrong denominator.

Compare that to a steel mill. The mill's assets — furnaces, land, rolling equipment, ore inventory — appear on the balance sheet at something close to their actual economic value. Depreciation schedules roughly track real wear. Replacement cost is estimable. The gap between book value and economic value exists, but it's measured in percentage points, not multiples. P/B has real analytical traction here because the denominator is actually informative.

The distinction comes down to one question: can you see the most valuable assets on the balance sheet? For banks and insurers, the answer is mostly yes — loans, bonds, and securities can be valued. For mining companies, steel producers, and capital-heavy manufacturers, mostly yes — the assets are physical and quantifiable. For pharmaceutical companies with drug pipelines, software companies with installed bases, and consulting firms where the assets go home each evening, the answer is mostly no. Applying P/B to that second group isn't analytically sloppy — it's measuring the wrong thing entirely. A software company at 12x book is not expensive the way a bank at 3x book might be. The same arithmetic is producing a number that means something completely different.

Low P/B isn't a buy signal

The intuition is seductive: if a stock trades below the accounting value of its assets, you're getting a dollar for less than a dollar. Benjamin Graham formalized this logic in the 1930s, and it worked well in an era when most large companies were industrial and their assets were tangible and liquidatable. The logic has worn badly in the decades since.

Three things explain sub-1 P/B. Only one of them is an opportunity.

The first is asset impairment the market has already priced. Book value reflects historical cost minus accounting depreciation, not current market value. If a retailer built out 600 stores at costs that made sense in 2012, those leasehold improvements and fixtures sit on the balance sheet at some reduced carrying value. If e-commerce has permanently reduced the profitability of that footprint, the real economic value of those assets may be far lower than the balance sheet shows. The market can see this before the write-down appears. Sears Holdings traded at sub-1 P/B for years before its 2018 bankruptcy filing. The P/B looked like value. The market was correctly pricing the coming impairment.

The second explanation is that the business earns below its cost of capital. A company earning 4% return on equity when investors require 10% is destroying value with every operating quarter. Its assets are worth more in liquidation than in continued operation — which is exactly what a P/B below 1 expresses. This is not a transient signal. A business that persistently earns 4% on equity should trade below book. The question is whether the 4% ROE is permanent or temporary.

The third explanation — genuine undervaluation — does exist. It appears when earnings are temporarily depressed by something fixable: a one-time write-off, a cyclical trough, a management transition, a macro shock that has clear boundaries. Cyclical industrials in a demand trough sometimes hit this condition. Banks in a credit cycle occasionally reach it when the credit losses are real but bounded and the franchise is intact. The way to distinguish genuine undervalue from the first two cases is to trace the ROE trajectory. If ROE has been chronically below the cost of capital for five or more years, sub-1 P/B is a statement about the business, not a temporary discount. If ROE was 13% two years ago and is 3% today because of a specific identifiable event with a credible recovery path, the sub-1 P/B is worth investigating seriously. The rule: never read a P/B below 1 without immediately checking trailing ROE and asking whether it reflects normal operations or a bounded disruption.

The ROE unlock: how to read P/B and not get fooled

There is a clean theoretical relationship between P/B and ROE that most investors understand in vague terms but rarely work through explicitly. The logic comes from DuPont analysis and the Gordon Growth Model. The justified P/B for any company is approximately equal to ROE minus the long-run growth rate, divided by cost of equity minus that same growth rate. The shorthand: a company earning exactly its cost of equity deserves to trade at 1.0x book. A company earning 20% ROE in a world where equity investors require 10% should trade at a meaningful premium to book. A company earning 5% ROE in that same world should trade at a discount.

This is why comparing P/B ratios across sectors without adjusting for ROE differences produces nonsense. Large US money-center banks have historically traded between 1.2x and 2.0x tangible book when they're earning 12–15% ROE in a normal rate environment. Consumer staples companies with high, durable ROE — Procter & Gamble has averaged above 25% ROE over the past decade — trade at significant premiums to book. Not because someone forgot to apply a sector discount. Because their returns warrant it. The premium is the rational output of durable excess returns, not multiple expansion disconnected from fundamentals.

The two-variable logic also makes the value trap diagnostic precise. A stock at 0.8x book needs an explanation. If the trailing ROE is 14% and the business is intact, that combination is unusual and worth investigating — the market is pricing equity below its economic production capacity. If the ROE is 4%, the 0.8x multiple is approximately fair. The trap is reading the first number in isolation and treating it as a signal without checking the second.

The price-to-earnings ratio doesn't have this problem in quite the same way — earnings already reflect return on equity implicitly. P/B strips that return component out, which is why it requires ROE to be reattached manually. The interactive tool below is built around exactly this two-variable framework. A P/B in isolation is an incomplete sentence. P/B plus ROE plus sector is the full statement.

Where P/B still earns its keep

The metric is in active daily use in several specific contexts. These are not legacy habits — they reflect genuine cases where book value is an informative denominator.

• Banks and insurance companies. Bank equity is mostly financial assets — loans, securities, derivatives — that are marked to market or can be valued with reasonable precision. Tangible book value per share (which excludes goodwill and other intangibles) is the standard valuation anchor in bank analysis, not just a secondary reference. When Wells Fargo trades at 1.4x tangible book and a regional bank trades at 0.75x, that spread contains real information about relative ROE, credit quality expectations, and growth outlook. Bank analysts build tangible book value per share projections as primary outputs, not footnotes. See also: bank stock valuation.
• REITs. Real estate investment trusts use a close cousin: price-to-NAV, net asset value. The underlying assets are properties with estimable market values — often appraised quarterly or annually — and the gap between NAV and stock price tells you whether the market is applying a premium or a discount to the portfolio and management team. Same intellectual structure as P/B, more directly applicable because the assets are regularly marked.
• Capital-intensive industrials. Steel producers, mining companies, and heavy equipment manufacturers are legitimate territory, especially in distressed or cyclical analysis. When Nucor or Cleveland-Cliffs hits a demand trough and trades near or below book, the question of whether the physical plant is intact and whether ROE will recover above cost of capital is exactly the right analytical question. P/B structures the investigation.
• Liquidation and sum-of-parts analysis. If a company is being wound down or broken up, the gap between accounting book value and actual asset market values is the whole question. Activist investors and distressed specialists use book value as a starting point, then mark up or down based on appraisals. The analysis is about the quality and realizable value of the balance sheet, which is exactly what P/B is built to interrogate.

The common thread: in all of these cases, the assets are real, identifiable, and valued with some connection to market prices. The moment you step outside that set — into software, pharmaceutical pipelines, media IP, or any business where the value is predominantly intangible — P/B loses its footing. The metric isn't wrong in general. It applies to a specific subset of the market, and using it outside that subset produces conclusions that look rigorous and aren't.