Book value and the P/B ratio
The answer
Book value is what the accounts say is left for shareholders after every debt is paid: total assets minus total liabilities. Divide it by the number of shares and you get book value per share. The P/B ratio is the share price divided by that number.
Book value is an accounting opinion, not a sale price. A P/B below 1 does not mean you are buying assets at a discount.
Why this costs you money
There are 2 ways to lose money with this ratio, and they point in opposite directions.
The first is buying below book. A stock trades at 0.5 times book value. It looks like you are paying ₹50 for ₹100 of net assets. If that were true it would be free money, and buyers would already have pushed the price up.
What the accounts call ₹100 is usually not ₹100. For a lender, a large part of "assets" is loans, and a loan is only worth what will actually be repaid. That judgement is made by the company. When a regulator or an auditor forces a re-count, book value falls, sometimes by half, and the stock that was at 0.5 times book was at 1.0 times book all along. For a manufacturer, "assets" may be machinery bought 15 years ago for a product nobody buys now. It sits on the balance sheet at original cost minus depreciation. Its resale value can be close to scrap.
The second is refusing to buy above book. A software company, a rating agency or a consumer brand may own almost nothing physical. Its people, its brand and its customer habits are the business, and none of those appear on a balance sheet. Accounting rules require most research, advertising and training spend to be expensed in the year it is spent, not recorded as an asset. So the balance sheet of an excellent asset-light company is nearly empty by design.
Such a company can trade at 15 times book value and be reasonably priced. Somebody applying a rule of "never pay more than 3 times book" will never own one, and will spend 20 years owning capital-heavy businesses instead.
Both mistakes come from the same false belief: that book value measures what a company is worth. It measures what a company paid.
How it works
The balance sheet has 3 parts and they always balance.
Assets = Liabilities + Shareholders' equity
Rearrange it and you have the definition.
Shareholders' equity = Assets − Liabilities = Book value
Book value is also called net worth, shareholders' funds, or equity. Those are the same thing. In this article it is called book value every time.
Book value per share is book value divided by the number of shares. Use the diluted share count if you can find it, because share options are coming.
P/B is the share price divided by book value per share. A P/B of 1 means the market is paying exactly what the accounts say the equity is. Above 1 means the market thinks the assets will earn more than they cost. Below 1 means it thinks they will earn less, or that the assets are worth less than stated.
Three things inside book value are worth knowing.
Assets are carried at historic cost, less depreciation. Land bought in 1985 sits at the 1985 price. So book value can be far below reality for an old company with land, and far above it for a company whose machines are obsolete. Both errors exist in the same number.
Goodwill. When a company buys another for more than the value of its identifiable net assets, the difference is recorded as goodwill, an asset. It is not cash and it cannot be sold on its own. If the acquisition disappoints, the company writes goodwill down and book value falls without any money leaving the business. Goodwill is the most fragile line in the equity number.
What is missing. A brand built over 40 years of advertising is not on the balance sheet, because the advertising was expensed each year. Research is usually expensed too. So the most valuable asset of many companies is worth zero in book value.
Subtract goodwill and other intangible assets from book value and you get tangible book value. For most purposes it is the more conservative number, and it is the one banking analysts use.
What it tells you, and what it does not
P/B works best where assets are the business. Banks, non-banking financial companies, insurers and investment holding companies hold financial assets that are valued at something close to a real price and turned over regularly. There, book value means something and P/B is a serious tool.
P/B works worst where the assets are people, brands or software. There, the denominator is close to arbitrary and the ratio carries no information at all.
A P/B below 1 tells you the market disagrees with the accounts. It does not tell you who is right. Sometimes the market is being pessimistic. Very often the market is anticipating a writedown that has not been announced.
P/B tells you nothing about profitability. This is the important gap. Two companies can both trade at 2 times book. One earns 25% on its equity every year; the other earns 5%. They are not similarly priced. They are wildly differently priced, and the ratio alone cannot see it.
The decision rule
A P/B is meaningless without the return earned on that book. Read the 2 numbers together, always.
The link between them is simple enough to remember. Return on equity is profit divided by book value. So a company earning a high ROE turns each rupee of book value into a lot of profit, and buyers will pay several rupees for it. A company earning a low ROE turns each rupee of book value into very little, and buyers will pay less than a rupee for it.
That gives you the working rule.
If a company's ROE is roughly equal to what investors need to earn on equity, it should trade near 1 times book. If ROE is well above that, a P/B well above 1 is normal and correct. If ROE is well below it, a P/B below 1 is not a bargain — it is the market pricing a poor return accurately.
The exception that matters: unless the book value itself is about to change. A lender facing a large loan writedown, or a company about to impair goodwill, has a denominator that is going to move. Then the current P/B is describing a company that no longer exists.
Try this now
Five minutes. This is the single most useful table a new investor can build.
- Open your holdings. For each one, write down 2 numbers from the app's fundamentals or financials tab: P/B and ROE. Nothing else.
- Sort the list by ROE, highest at the top.
- Look down the P/B column.
- Circle any holding where the 2 numbers disagree — a high ROE with a low P/B, or a low ROE with a high P/B.
- For each circled holding, ask 1 question. High ROE and low P/B: is the ROE inflated by a one-off profit, or by heavy borrowing, or is a writedown coming? Low ROE and high P/B: is the market expecting the return to improve, or is the book value nearly empty because the real assets are not on it?
What you should see. The 2 columns will move together for most of your holdings. Higher ROE, higher P/B. That relationship is not a coincidence and it is not a rule somebody invented. It is arithmetic, and seeing it in your own portfolio is the point of this exercise.
The circled ones are your reading list. Every genuine opportunity and every genuine trap in this cluster lives in that small group where the 2 numbers disagree. A stock at 0.4 times book earning 3% on equity is not cheap. A stock at 9 times book earning 45% on equity is not obviously expensive.
Three real cases
1. Eastman Kodak, January 2012 (United States) — assets with no future Kodak's revenue peaked near $16 billion in 1996 and profit near $2.5 billion in
- For years afterwards its balance sheet showed large assets: film plants,
chemical facilities, inventory, patents. On paper there was substantial book value. But a film factory is only worth something if somebody will buy film. The company filed for Chapter 11 bankruptcy protection on 19 January 2012 and emerged on 3 September 2013 having sold or shed most of those businesses. Investors who bought Kodak because it looked cheap against book value had valued equipment by what it cost, not by what it could earn.
2. Indian public sector banks, 2015 to 2018 (India) — the book value moved In 2015 the Reserve Bank of India began an Asset Quality Review, requiring banks to classify stressed loans honestly and to provide against them. Reported bad loans across the Indian banking system rose sharply over the following years, and several public sector banks reported large losses. Their book value fell, and some required capital from the government to keep operating. Before the review, many of these banks traded below 1 times their stated book value and looked cheap on that basis. The stated book value was the thing being questioned. This is the clearest lesson available on why a lender's P/B is a judgement about a judgement.
3. Berkshire Hathaway, February 2019 (United States) — the metric retired For decades Warren Buffett opened Berkshire's annual report with the change in book value per share, treating it as the best available proxy for the growth in the company's real worth. In the letter covering 2018, published in February 2019, he stopped. His reason was direct: Berkshire had shifted from owning marketable securities, which are carried at market price, to owning whole operating businesses, which are carried at cost less depreciation. Book value had stopped tracking value. When the investor most associated with book value publicly abandons it as a yardstick, that is worth more than any argument.
The question that resolves it
A novice sees a P/B of 0.6 and asks: why is this trading below its assets?
An expert sees the same 0.6 and asks: what return does this company earn on those assets, and is the asset number still true?
The first reading treats book value as a floor. The second treats it as a claim that has to be checked.
What would make this wrong
If book value measured what a company was worth, then buying companies below book value would produce reliable profits and companies would not trade below book for years at a time. Many trade below book for decades, and the reason is consistent: they earn less on their equity than investors require.
The honest limits are 3.
Book value can be understated as easily as overstated. An old company holding land at a 1970s cost has a book value far below reality, and a P/B of 1.5 on that company may be genuinely cheap. The direction of the error is not always the same.
Accounting rules differ. Two companies in 2 countries with identical operations can report different equity because of different treatment of goodwill, revaluation and leases.
And a low P/B with a high ROE is occasionally exactly what it looks like — a mispricing. The point of this article is not that P/B never works. It is that it never works alone.
In India
Indian accounts follow Ind AS, the Indian version of international standards. Under those rules most internally generated brands and research cannot be recorded as assets, so consumer and technology companies show small book values relative to their size.
Banks and non-banking financial companies are the main users of P/B in India, and the main hazard. For a lender, book value depends on provisions set aside against loans that may not be repaid. The Reserve Bank of India sets rules for classifying a loan as non-performing and for how much must be provided. Two Indian banks with the same stated P/B can carry very different amounts of unrecognised loss.
The check available to any retail investor: in the annual report, find gross non-performing assets, net non-performing assets, and the provision coverage ratio. A bank with low provision coverage has recognised the problem loans but has not absorbed the cost of them. Its book value has further to fall.
Revaluation reserves are a second Indian complication. A company may revalue land or buildings upward, increasing book value without any cash arriving. The reserve is disclosed. A book value that jumped on a revaluation is not the same as one that grew from retained profit.
In the United States
US accounts follow GAAP, and 2 features change the book value number substantially.
Goodwill from acquisitions is large. Decades of takeovers have left many large US companies with goodwill and acquired intangibles making up a big share of stated equity. Under GAAP goodwill is not amortised each year; it is tested for impairment and written down in one large step when the test fails. So book value can be stable for years and then drop sharply on a single announcement. Tangible book value, which removes goodwill and intangibles, is the safer denominator for a US company.
Buybacks reduce book value. A company buying its own shares above book value per share pays out more cash than the book value it retires, so total equity falls. Sustained buybacks can drive book value close to zero and in some cases below it. Several large, profitable US companies report negative shareholders' equity for exactly this reason. Their P/B does not exist. That is a financing choice, not distress, and it makes the ratio unusable.
US banks are the strongest place for P/B in either market. They are regulated closely, and analysts routinely compare price to tangible book value per share.
Where they differ, and what that tells you
The same ratio breaks in 2 different places in the 2 markets.
In the United States, P/B breaks mostly at the top of the balance sheet: goodwill inflates equity, and buybacks deflate or erase it. The number is distorted by financing and acquisition history.
In India, P/B breaks mostly at the lender's loan book: equity depends on provisions, and provisions depend on classification decisions that a regulator may later overturn. The number is distorted by judgement about credit.
What that tells you is which check to run. For a US company, subtract goodwill and intangibles first, and check whether equity has been shrunk by buybacks. For an Indian lender, check gross and net non-performing assets and provision coverage first, and treat the stated book value as provisional until you have.
There is a second difference worth naming. Because Indian promoters typically retain large stakes and Indian companies have historically raised equity rather than retired it, Indian book values grow more steadily and mean more from year to year. American book values are noisier, and the year-on-year change in book value per share tells you far less about an American company than about an Indian one.
Carry this
- Book value is what the company paid, not what the assets are worth.
- Never read P/B without ROE. Low P/B with low ROE is the market being correct.
- For a lender, the book value is a judgement. Check provisions before you believe it.