How to compare bank stocks: JPMorgan vs Bank of America

Banks are the one kind of company where most of the standard toolkit breaks. Debt is not a risk measure but the raw material of the business. There is no cost of goods sold and no operating-income line. Cash flow statements swing wildly with deposits and loans. So when TickerVerdict compares JPMorgan Chase with Bank of America, half the usual table is deliberately blank, and two measures do most of the work: price-to-book and return on equity. This guide explains why, how those two fit together, and what you still need to look up yourself.

Why the usual ratios fail

An industrial company borrows to fund factories and pays interest as a cost of financing. A bank borrows — mostly in the form of customer deposits — in order to lend the money out at a higher rate. Its borrowings are its inventory. That one difference breaks a whole family of ratios:

• EV/EBITDA would treat customer deposits as debt to be "repaid", and banks report no EBITDA.
• Net debt to EBITDA, interest coverage and the current ratio measure a burden that, for a bank, is the business itself.
• Gross and operating margin need a cost-of-sales and an operating-income line that bank accounts do not have.
• Return on invested capital would be dominated by deposits.

We leave all of these out for banks rather than print numbers that look precise and mean nothing. Even revenue needs care, because banks earn it in two ways: net interest income from lending, and non-interest income from fees and trading. Bank of America reports a total of the two; for JPMorgan, whose latest quarterly filings carry no total that our reader finds, we add them ourselves. Over the past twelve months that gave revenue of $199.41bn for JPMorgan and $119.24bn for Bank of America.

The two numbers that matter: book value and the return on it

A bank's book value — shareholders' equity — is the capital that absorbs losses before depositors and lenders are at risk. Because a bank's assets are mostly loans and securities recorded close to their value, book value is a meaningful anchor in a way it is not for, say, a software company.

Price-to-book compares the share price with that capital. Return on equity measures how much profit the bank earns on it. The two belong together:

The logic is simple. A bank that earns more on its capital than investors require — the cost of equity, usually estimated at around 10% for large banks — creates value with every dollar it retains, and deserves to trade above book value. A bank that earns less than its cost of equity destroys value by retaining capital and tends to trade near or below book.

JPMorgan earns 17.79% on its equity and trades at 2.38 times book. Bank of America earns 11.23% and trades at 1.19 times book. The market is paying roughly double the book multiple for JPMorgan because it earns markedly more on the same kind of capital. Whether that premium is right depends on whether JPMorgan's higher returns persist — which is the real question an investor in either bank is answering.

How our six-factor score treats banks

Because so many inputs are deliberately missing, bank comparisons behave differently under our rules. On the JPMorgan vs Bank of America page:

• Valuation goes to Bank of America, which is cheaper on both multiples that apply — price-to-sales of 3.01 against 4.47, and price-to-book of 1.19 against 2.38.
• Growth goes to JPMorgan: revenue grew 12.34% a year against 6%, and earnings per share 18.31% a year against 6.1%.
• Income goes to Bank of America: a 2.14% dividend yield and 6.92% buyback yield, against 1.79% and 3.55%.
• Quality and balance sheet are not scored, because the margin, return-on-capital and leverage measures they rely on do not exist for banks. Momentum is not scored on any page.

The result is Bank of America 3.5, JPMorgan 2.5. Read the factors rather than the total: the score says Bank of America is cheaper and returns more cash, while JPMorgan is growing faster and — on return on equity, which the quality factor cannot use for banks — earning more. The six-factor score explained shows why missing inputs compress bank scores towards a draw.

Comparing a bank with a non-bank

We also publish Apple vs Bank of America, because people search for it. Treat cross-sector comparisons like this with care. Bank of America's P/E of 11.87 against Apple's 39.33 does not mean the bank is the bargain: banks structurally trade on lower earnings multiples, because their profits depend on credit cycles and interest rates and because regulators require them to retain capital. The useful question is not which share is cheaper, but which kind of business — a lender or a device-and-services company — you want to own. Our P/E guide explains why comparing multiples across industries misleads.

What our pages cannot show you

Our bank figures come from the same SEC filings as everything else, but proper bank analysis needs numbers that filings report in tables and supplements our reader does not yet use. Before forming a view on any bank, look up:

1. Net interest margin — the spread between what the bank earns on loans and pays on deposits. It drives most of a traditional bank's revenue.
2. Credit quality — loan losses written off, and the reserve set aside for future ones. Profits can look excellent right up until a credit cycle turns.
3. Capital ratios — chiefly the Common Equity Tier 1 ratio, which regulators use to judge how much loss a bank can absorb, and which limits how much it can pay out.
4. The efficiency ratio — costs as a share of revenue, the closest bank equivalent of an operating margin.
5. Deposit mix and stability — how much funding is from sticky everyday accounts versus rate-sensitive ones.

All five are in each bank's 10-K and 10-Q, and in the quarterly earnings supplements banks publish alongside them. How to read a 10-K explains how to find your way around the filings, and the numbers we refuse to print lists every ratio we withhold for banks and why.