P/B Ratio: When Price-to-Book Matters & When It Lies

In March 2025, Spandana Sphoorty Financial — a microfinance lender — traded at about ₹235 a share against a book value of roughly ₹369 (Inve data, Q4 FY25). That is a price-to-book ratio of 0.64: the market was paying 64 paise for every rupee of stated net worth. To a screener, that reads as cheap. It was not cheap. Over the four quarters that followed, the company bled another roughly ₹620 crore in losses, its reserves shrank, and the book value the market had been doubting fell from about ₹369 to about ₹274 a share (Inve data, Q4 FY26). The discount wasn't a bargain. It was a forecast — and the forecast came true. (Illustration, not a view on the stock.)

That is the whole subject of this article in one example. The price-to-book ratio is the most context-dependent of the common valuation multiples: it tells the truth loudly in some sectors and lies quietly in others, and it has a nasty habit of looking cheapest exactly when it is most dangerous. The difference comes down to a single question — does the book value on the balance sheet bear any relationship to how this business actually makes money, and is that book value even real?

This is a field guide to that question. For asset-heavy and financial businesses, P/B is a primary lens; for asset-light ones, it's a distraction. And a low P/B, more often than not, is the market pricing in trouble you haven't found yet. Knowing the difference is most of the skill.

What is the price-to-book ratio?

Price-to-book compares a company's market value to its net worth on the balance sheet.

A P/B of 1.0 means the market values the company at exactly its accounting net worth. Below 1.0, the market is paying less than book value — which can mean a bargain, or a warning that the book is overstated. Above 1.0, the market believes the assets are worth more than their carrying value, usually because the company earns a high return on them.

That last point is the key the textbooks bury: P/B is inseparable from return on equity (ROE). A company that earns 25% on its equity should trade at a high multiple of book, because each rupee of book value is a money-machine producing 25 paise a year. A company earning 6% on equity deserves to trade near or below book, because its assets barely beat a fixed deposit. P/B in isolation is half a sentence; P/B read alongside ROE is the whole thought. A "cheap" low P/B paired with a low, falling ROE is not cheap — it's correctly priced.

When does price-to-book actually matter?

P/B earns its keep where the balance sheet is the business — where assets are real, marked roughly to reality, and central to earning power.

Banks and NBFCs. This is P/B's home turf. A bank's "product" is its balance sheet — loans funded by deposits. Book value is a meaningful measure of the equity cushion absorbing loan losses, and the market prices banks primarily on P/B (alongside ROE and asset quality). A bank trading at 3x book is one the market trusts to keep compounding equity at high returns with clean loans; one trading at 0.6x book is usually one the market suspects of hiding bad loans that will eat into that book value. Which brings us to the trap below.

Capital-intensive and asset-heavy sectors. Manufacturing, infrastructure, real estate, shipping, utilities — businesses whose value sits in plant, property, and inventory. Here book value roughly tracks real, saleable assets, so P/B is a useful comparison tool across peers. But treat it as a rough floor, not a precise measure: because assets are carried at historical cost, the stated book often understates fair value, which is why the curriculum's preferred valuation tool for asset-heavy non-financials is the net-asset-value (NAV) approach — the market value of assets minus liabilities — rather than book-based P/B. The "P/B as a primary lens" treatment really belongs to financials, where monetary assets make book value approximate fair value.

Cyclicals at the bottom. When a cyclical's earnings vanish in a downturn, its P/E becomes useless (tiny or negative E). P/B keeps working because steel plants and shipping fleets don't disappear with the cycle. A commodity producer trading below book in a brutal down-cycle, with physical assets that will produce again when demand returns, is exactly the situation P/B was built for — where P/E fails, P/B holds. The catch is the one Spandana illustrates: P/B only "holds" if the assets actually persist. A factory does. A loan book of stressed borrowers may not.

When does price-to-book lie?

The ratio breaks wherever the accounting book and the economic value of the business drift apart — and in modern markets, that's a large and growing set of companies.

Asset-light businesses. Software, IT services, consumer brands, asset-light platforms. Their value lives in things the balance sheet barely records: code, brand, customer relationships, distribution, people. A premier IT firm can carry a P/B of 8 or 12, not because it's "expensive" but because its book value captures almost none of its real worth. Calling such a stock expensive on P/B is like judging a chef by the price of the kitchen. For these businesses, P/B is noise — use ROE, cash generation, and earnings multiples instead.

Companies fat with goodwill and intangibles. After acquisitions, balance sheets fill with goodwill — an accounting plug, not a saleable asset. It inflates book value, dragging P/B down and making the stock look cheaper than it is. When the acquired business underperforms, that goodwill gets written off, book value drops overnight, and the "cheap" P/B was an illusion all along.

Lenders whose book hasn't caught up to its losses. The most expensive lie in Indian markets. A lender's book value is only honest if its loans are honestly provisioned. When a borrower stops paying, the loan is still sitting on the balance sheet at full value until the lender takes the provision against it. Until that moment, book value is a number waiting to be revised downward — a quality-of-earnings problem dressed up as a cheap valuation. A sub-1 P/B on such a lender is the market pre-writing a haircut to book that the accounts haven't taken yet. The cheapness is a forecast of trouble, not a discount. Spandana, again, is the textbook of this — which is the next section.

Why is a low P/B often a warning, not a bargain?

This is the second-order point that separates a field guide from a glossary, so let it run on the case from the very top of this piece. A market that prices an asset-real business below book value is making a statement: it does not believe the stated book will survive contact with reality. With a lender, that statement is almost always about the loan book — the market expects provisions to climb and eat equity. Watch how cleanly that played out at Spandana, and notice that three separate forces all pushed the same way at once.

Force one — the book was a number waiting to fall. Spandana earned ₹51 crore in the June 2024 quarter (Inve data, Q1 FY25). Then the microfinance cycle turned. The next four quarters were losses of roughly ₹204 crore, ₹394 crore, ₹410 crore and ₹329 crore (Inve data, Q2 FY25 through Q1 FY26). Management's own framing of the full year was blunt: "The company reported a net loss of Rs. 1,035 crores for the year on account of higher provisions" (Spandana Q4 FY25 concall, May 2025) — that ₹1,035 crore is the audited full-year figure, which differs slightly from the sum of the four quarterly prints above. Provisions are how a lender admits the loans on its book are worth less than carried. Each rupee of provision is a rupee taken straight out of net worth — which is why Spandana's net worth fell from about ₹3,556 crore in March 2024 to about ₹2,622 crore a year later and then to about ₹2,194 crore by March 2026 (Inve data, balance sheet). The book the screener treated as solid was, in real time, being marked down to what the loans were actually worth.

Force two — the market saw it before the accounts did. Here is the part worth sitting with. At ₹235 against a ₹369 book in March 2025, the stock was already at 0.64x book (Inve data, Q4 FY25) — before the ₹329-crore June quarter loss and the further bleed that followed. The market was not mispricing a sound business. It was pricing in losses the audited book had not yet absorbed. Put the two numbers far apart and feel the gap: the market said the book was worth 64 paise on the rupee; over the next year the company's own provisions cut total net worth by roughly ₹430 crore and the per-share book value fell from about ₹369 to about ₹274 (Inve data, Q4 FY26). (Part of that per-share drop is dilution, not write-downs: the share count rose from about 7.1 crore to 8.0 crore over the period as a rights issue went through, so the cleaner read on the provisioning is the absolute net-worth line, not the per-share number alone.) The "cheap" 0.64x was not a discount to value. It was an accurate down-payment on a write-down. (Illustration, not a view on the stock — the point is the mechanism, not the ticker.)

Force three — the guidance only ever moved one way. On that same May 2025 call, with the worst year behind it, management guided to growth: "I think we can look at about 20% kind of the growth from the current base. That's the number that will work out" (Spandana Q4 FY25 concall). They reiterated roughly 20% the following quarter. What actually happened is recorded in Inve's Promise Tracker: the loan book did not grow 20% — it shrank. The post-write-off AUM had fallen to about ₹3,948 crore by the September 2025 quarter (Inve data, Q2 FY26) — a narrower, written-down figure, not headline AUM — and the FY26 AUM-growth guidance was marked missed (Inve data, Promise Tracker). A management guiding to 20% growth while the book is being marked down quarter after quarter is telling you, if you read the said against the did, that the stated picture and the real one have come apart. That gap is the same gap the 0.64x P/B was pricing — the ratio and the guidance record were two readings of one instrument.

Three forces, one company, one direction: a book quietly being written down, a market pricing the write-down early, and guidance that kept pointing up while the business fell. None of them alone proves the case. Stacked on the same name, they do. That convergence — not any single number — is why a clean-looking low P/B on a lender should make you reach for the transcript, not the buy button.

The honest counter-statement is that this is a snapshot of a hard cycle, not a verdict on the company; Spandana returned to a small profit of about ₹5 crore in the March 2026 quarter (Inve data, Q4 FY26), the cycle is turning, and a lender that survives a clean-up can compound off a smaller, healthier book. A two-year transcript record cannot tell you how this management behaves across a full credit cycle. All true. It changes the forecast; it does not change the lesson — which is that the 0.64x was never the bargain it looked like.

So the discipline with a low P/B is inversion: don't ask "why is this so cheap?" Ask "what would have to be true for the market to be right that this book is worth less than stated — and does the evidence rule that out?" For Spandana in early 2025, the evidence ruled it in: the provisions, the loss run, and the one-way guidance all said the book would fall. The answer was in the Q&A and the loss line, not the screener. Across Inve's tracked universe there are thousands of recorded dodges — specific questions management deflected or non-answered — and about a third of all tracked companies — 534 of 1,519 — have at least one commitment that quietly went silent (Inve data). When the topic being dodged is asset quality and the P/B is below 1, those two facts are usually the same fact, viewed from two angles.

How to use P/B without being fooled

A practical sequence for any stock where P/B might apply:

1. First, decide if P/B even applies. Asset-light? Set P/B aside and use ROE and cash-flow multiples. Asset-heavy or financial? Proceed.
2. Never read P/B without ROE. High P/B is justified by high, durable ROE; low P/B paired with low ROE is fair value, not a bargain. The two numbers only make sense together. And when you compare against peers, use both the industry mean and median — a single richly-rated outlier can skew the mean and make an ordinary stock look mispriced.
3. Interrogate the book itself. Treat the stated book value like the odometer on a used car: it's the number on the dial, not proof of what's under the bonnet. For acquisitive companies, strip goodwill and re-check P/B on tangible book. For lenders, the book is only as good as the provisioning — and as Spandana showed, the dial can read fine right up until the engine is opened and ₹1,035 crore of provisions come out.
4. For a sub-1 P/B, find the reason. A clean, productive business rarely trades below book without a reason the market has already spotted. Don't ask "why so cheap?" — ask "what would make the market right that this book is overstated, and does the evidence rule it out?" For a lender, the evidence is the loss line, the provisioning, and whether management's guidance keeps pointing up while the book shrinks. It's usually qualitative, and it's usually in the Q&A.
5. For cyclicals, lean on P/B at the trough, not the peak. It's the lens that keeps working when earnings collapse.

The hardest of these — step 4 — is reading management's behaviour across quarters: which questions get answered straight, which get deflected, which commitments quietly disappear. Doing that by hand across a portfolio is the job Inve's Promise Tracker automates, and you can scan the asset-quality and guidance commentary across companies on the KPI Screener to see where a cheap book is matched by deteriorating fundamentals.

Frequently asked questions

Price-to-book is not a universal valuation tool — it's a specialist's instrument that's superb in the right hands and misleading in the wrong context. Its truthfulness depends on one thing: whether the book value on the balance sheet means anything for how the business earns, and whether that book is even real. Where assets are real and durable — infrastructure, cyclicals at the bottom, a lender with an honestly-provisioned book — P/B is one of the sharpest lenses you have. Where value lives off the balance sheet, or where the book hasn't yet absorbed losses already in the post, it's a mirage.

So skip the screener's verdict and ask the owner's question instead. If you were going to hold this lender for five years, what would have to be true for the book value you're paying 0.6x for to still be standing in year five — and after the loss line, the provisions, and what management guided versus what the book actually did, does the evidence let you believe it? If you can't answer that, you don't have a cheap stock. You have an unfinished forecast that the market has already started to fill in.