P/B ratio: what you pay for what a company owns.
Why some banks command three times book — and others shouldn't even get one
One bought SBI at book. One paid three times that for HDFC Bank.
Rajan's office colleague bought HDFC Bank shares in 2019 at nearly three times book value. Rajan thought she was paying too much. He bought SBI instead — trading right at book. Looked like the obvious deal. A few years passed. Rajan learned something important about what 'cheap' actually means when you're buying a bank.
Book value is what shareholders actually own. Start with everything the company has — buildings, cash, loan accounts, investments, machines. That's total assets. Subtract everything it owes — bonds, borrowed money, customer deposits. What remains belongs to shareholders. Divide by the number of shares and you have book value per share.
P/B ratio asks how much the market charges for each rupee of that book. If a bank's book value per share is ₹100 and the stock trades at ₹300, the P/B is 3. You're paying ₹3 for every ₹1 sitting on the balance sheet. The extra ₹2 is what the market believes the franchise is worth — its reputation, its relationships, its ability to keep earning strong returns.
Banks are the natural home of P/B analysis. A bank's entire business is its balance sheet. Loans are its product. Deposits are its raw material. Earnings can swing badly in a single year if a large borrower defaults. But the loan book itself does not vanish overnight. Analysts anchor bank valuations to book value precisely because assets are more stable than one year's profit.
A clean formula connects fair P/B to fundamentals: Fair P/B ≈ ROE ÷ Cost of Equity. A bank earning 15% return on equity, when investors require 12%, deserves about 1.25x book. Push ROE to 20% and fair P/B rises to about 1.67x. Higher quality always commands a higher multiple — the math insists on it.
Gold rate and making charges
When you buy a gold necklace, you pay two things. First, the value of the gold itself — the actual metal, priced at that day's rate. Second, the making charges — what the jeweller adds for the design, the craftsmanship, the name on the box. Book value is the gold. The P/B premium above 1x is the making charges — what you pay for the franchise, the brand, the trust the market places in this particular business. A jeweller with three generations of reputation charges more. A bank with a clean loan book and consistent returns commands a higher P/B. The question is always the same: are the making charges worth it?
Why this matters
When you screen bank stocks, P/B is the first number to check — not P/E. A low P/B sounds like a bargain. But a bank trading below 1x P/B is telling you the market doubts those loans will be repaid in full. Before you buy 'below book', ask whether that book is real. Warren Buffett set 1.2x P/B as Berkshire Hathaway's own buyback floor until 2018 — he only repurchased his company's shares when price fell that close to book. When someone that careful treats book value as a hard anchor, it is worth understanding why.
Slide the ROE. Watch fair P/B move in real time.
Try the widget below. Drag a bank's ROE up or down and watch its fair P/B shift in real time. The same formula that explains why HDFC Bank and SBI trade at such different multiples will become obvious in seconds.
Fair P/B for banks: drag ROE, watch the multiple move
With 15%% ROE vs 12%% cost of equity, this bank deserves 1.3xx book. A weak bank (8% ROE) gets only 0.7xx. The gap is the price of trust.
Better returns on the book. Higher price for the book.
Where people go wrong
- Applying P/B to IT or consumer brand companiesSoftware engineers, brand loyalty, and distribution networks do not appear on a balance sheet. For Infosys or Nestle, P/B tells you almost nothing useful — their real value lives in intangibles that accounting cannot capture.
- Treating P/B below 1 as automatic cheapnessA bank below 1x P/B usually carries bad loans hiding in the book. The market is not wrong — it is discounting assets it does not trust. Check gross NPA levels before assuming the price is unjustifiably low.
- Comparing P/B across different sectorsA steel company at 2x P/B and a bank at 2x P/B are completely different situations. Asset intensity, required returns, and business risk make cross-sector P/B comparisons close to meaningless.
- Ignoring goodwill inflated by acquisitionsWhen companies acquire others at a premium, that premium lands on the balance sheet as goodwill. This inflates book value artificially — making P/B look deceptively low if you have not adjusted for it.
Book value is what shareholders own after every liability is subtracted from every asset.
Fair P/B = ROE ÷ Cost of Equity — better-run banks always justify a higher multiple.
P/B below 1x is a warning sign — ask whether those balance sheet assets are truly worth face value.
People anchor on 'I'm buying below what the company owns' the moment P/B drops under 1. They forget to ask whether what the company claims to own is actually there.