Valuing a bank: price-to-book and the ROE-multiple link

Earnings are volatile and manipulable. A bank's profit in any year swings with loan loss provisions (money set aside for loans that may go bad). One bad quarter of provisions can make P/E look absurd. Book value is steadier.

Regulators think in book terms. Capital rules (how much equity a bank must hold) are all expressed against book equity and assets. The metric that binds the bank is the metric investors watch.

You will often see a refinement: price-to-tangible-book (P/TBV), which strips out goodwill and other intangibles from equity. Tangible book is the harder, more conservative number, and it is what most bank analysts actually quote.

The core link: P/B and ROE

Here is the idea that makes P/B powerful. The multiple a bank deserves is driven almost entirely by one comparison:

• ROE (return on equity): net profit divided by book equity. What the bank earns on each dollar of shareholder money.
• COE (cost of equity): the return shareholders demand for the risk they take.

The simplified relationship, from the Gordon growth model applied to banks:

Justified P/B = (ROE − g) / (COE − g)

where g is the long-term growth rate of earnings.

Strip out growth for intuition (set g to zero) and it collapses to something you can say in one breath:

Justified P/B ≈ ROE / COE

If a bank earns exactly its cost of equity, it deserves to trade at 1x book. Earn more, and it deserves a premium. Earn less, and it deserves a discount.

Worked example

Take a bank with:

• ROE = 12%
• COE = 10%
• g = 2%

Justified P/B = (0.12 − 0.02) / (0.10 − 0.02) = 0.10 / 0.08 = 1.25x

The market should pay 1.25 times book for this bank.

Now flip it. Same COE and g, but the bank only earns an 8% ROE:

Justified P/B = (0.08 − 0.02) / (0.10 − 0.02) = 0.06 / 0.08 = 0.75x

The market should pay 0.75 times book. Below 1.

Why below 1x book means "priced to destroy value"

Read that last result carefully. The bank has, say, 100 of book equity per share. The market pays 75. Investors are refusing to pay full price for equity that already exists.

The reason is mechanical. If the bank earns 8% on equity but shareholders demand 10%, then every dollar of profit the bank retains and reinvests earns less than it costs. Retaining and reinvesting a dollar creates less than a dollar of value. The bank is a value destroyer as long as ROE sits below COE.

A sub-1x P/B is the market's verdict: "We do not trust this management to earn its cost of capital. We would rather they shrank, returned capital, or got taken over than kept reinvesting."

This is why activist investors pounce on chronically cheap banks and push for buybacks, asset sales, or breakups. Returning capital to shareholders (who can redeploy it at COE elsewhere) beats reinvesting it below COE.

Real-world benchmarks (estimates, early 2026)

These figures move daily and should be treated as approximate, order-of-magnitude reference points, not precise quotes. Check a live source such as the FT markets data pages before using any number.

United States. Large US banks have generally traded at healthier multiples than European peers. JPMorgan Chase, the standout performer, has for a stretch traded meaningfully above 1x tangible book (a P/TBV commonly cited in the region of 2x, reflecting a sustained ROE well into the mid-to-high teens). Less profitable US names have often hovered around or just above 1x.

Europe. Many large European banks (Deutsche Bank, Barclays, Societe Generale among the historically cheap) have spent years below 1x tangible book, with sector multiples frequently estimated in the 0.6x to 1.0x range, though 2023 to 2025 rate rises lifted profitability and multiples for several. Some strong performers (for example certain Nordic and southern European banks) have pushed above 1x.

The pattern fits the theory: US banks have generally posted higher ROEs than European banks, and higher ROE relative to COE means higher P/B. The gap is not a mystery. It is the ROE-COE link doing its job.

Estimating cost of equity, quickly

You need a COE to judge whether a P/B is fair. The standard tool is the CAPM (Capital Asset Pricing Model):

COE = risk-free rate + beta × equity risk premium

• Risk-free rate: the yield on a safe government bond (10-year US Treasury or German Bund).
• Beta: how much the stock moves relative to the market. Big banks often have betas above 1 (say 1.1 to 1.4), reflecting their leverage and cyclicality.
• Equity risk premium: the extra return investors demand for stocks over the risk-free rate, commonly estimated around 4.5% to 5.5%.

Quick illustration with round, illustrative numbers: risk-free rate 4%, beta 1.2, equity risk premium 5%.

COE = 4% + 1.2 × 5% = 4% + 6% = 10%

That is the hurdle. A bank earning 10% ROE is treading water. Earning 14% clears it comfortably.

Putting it together: reading a real situation

Suppose you are handed these figures for a European bank (illustrative):

• Tangible book value per share: 30
• Share price: 24
• Consensus ROE next year: 9%
• Estimated COE: 11%
• Long-run growth: 1%

Current P/TBV = 24 / 30 = 0.80x

Justified P/B = (0.09 − 0.01) / (0.11 − 0.01) = 0.08 / 0.10 = 0.80x

The market has this one right. The 0.80x price is exactly what a 9% ROE against an 11% COE deserves. The stock is not "cheap." It is fairly pricing a bank that earns below its cost of equity.

Now the interesting question flips from valuation to strategy. What closes the gap to 1x? Only a higher ROE. The bank could:

• Cut costs (lower the cost-to-income ratio, a key European bank weakness).
• Reduce excess capital via buybacks (fewer shares, higher ROE on remaining equity).
• Exit low-return business lines.

Every one of these lifts ROE toward and past COE, and the P/B follows. This is why bank management teams obsess over ROE targets: the multiple, and thus the share price, is chained to it.

Key Takeaways

• Use P/B (ideally price-to-tangible-book) for banks, because the balance sheet is the business, earnings are volatile, and regulators think in book terms.
• Justified P/B ≈ ROE / COE. A bank earning exactly its cost of equity deserves 1x book. Above COE earns a premium, below COE earns a discount.
• Sub-1x book means value destruction: the market is saying management cannot earn its cost of capital, so reinvesting shareholder money makes each dollar worth less.
• Estimate COE with CAPM (risk-free rate + beta × equity risk premium); big-bank betas above 1 push the hurdle toward 10% or higher.
• US banks have generally traded above European peers on P/B (early 2026 estimates), and the reason is simply higher ROE relative to COE, not sentiment.