How to Analyze Bank Stocks: The Metrics That Actually Matter

Analyzing a bank means working through five things in order: net interest margin and its direction, deposit mix and funding cost, loan book composition and concentration, credit quality through nonperforming loans and reserve coverage, and capital adequacy through the CET1 ratio. Only then does valuation make sense, and banks are valued on price to tangible book value read alongside return on tangible common equity, not on the multiples used for ordinary companies. This guide explains why bank financial statements work differently, what each metric measures, and where bank problems usually originate. Educational only, not financial advice.

Why banks break normal analysis

A bank's balance sheet is inverted relative to intuition. Customer deposits, which feel like the bank's money, are liabilities: the bank owes them back on demand. Loans, which feel like money going out the door, are assets that generate income. There is no gross margin because there is no cost of goods sold. Revenue splits into net interest income, the spread business, and noninterest income from fees, and the two behave completely differently through a cycle.

Then there is the allowance for credit losses, the single largest exercise of judgment in the accounts. It is management's estimate of losses on loans that are currently performing, and it flows straight through the income statement. Two banks with identical loan books can report meaningfully different earnings based purely on how conservative their loss estimates are.

The practical upshot: enterprise value is meaningless for a bank because debt is raw material rather than financing, free cash flow is not calculated the usual way, and most standard screens output nonsense. Banks need their own toolkit.

1. Net interest margin

Net interest margin is net interest income divided by average earning assets. It measures the spread between what the bank earns on loans and securities and what it pays for deposits and borrowings, and it is the core profit engine.

NIM is expressed in percentage points and moves in hundredths, which understates how much it matters. A bank with $50 billion of earning assets gains roughly $50 million of annual net interest income for every 10 basis points of margin. That is why a quarterly report leading with a 12 basis point NIM compression can move the stock several percent.

What to look for: the direction over four to eight quarters rather than the absolute level, and the bank's rate positioning. Asset-sensitive banks, whose loans reprice faster than their deposits, earn more when rates rise. Liability-sensitive banks are hurt. Banks disclose this sensitivity in the 10-K, usually as a table showing the modeled effect of a 100 or 200 basis point rate shift.

2. Deposits are the actual franchise

Two banks with identical loan books can be worth very different amounts depending on what funds them, and deposit quality is where durable competitive advantage lives in banking.

Track the noninterest-bearing share of total deposits over time. When it shrinks, customers are moving money into higher-yielding accounts, funding costs rise, and NIM compresses. Track the deposit beta too, meaning how much of a rate increase the bank has had to pass to depositors. A low beta indicates genuine pricing power over its funding base.

Uninsured deposits, those above the FDIC limit, deserve a separate look. A high concentration of large uninsured balances from a narrow customer base is a structural fragility, because that money moves quickly at the first sign of trouble.

3. The loan book

Read the composition, not just the total. Commercial and industrial lending, commercial real estate, residential mortgages, construction and consumer credit carry different loss profiles and behave differently through a downturn.

Concentration is the risk that matters. A regional bank with an outsized share of its loan book in one property type in one metropolitan area is exposed to a single local economy. Construction and land development loans have historically produced the highest loss rates in downturns. Commercial real estate concentration is a standing supervisory concern precisely because it has repeatedly been the source of regional bank failures.

Also check loan growth against the peer group. A bank growing loans substantially faster than its market is either taking share on service or winning it on price and credit standards. The second is common and rarely disclosed. Loan growth well above peers, with a stable or falling reserve ratio, is worth understanding before you accept the earnings.

4. Credit quality

Credit losses are what actually destroys bank equity, and they are nearly invisible in good years. These are the indicators, roughly in order of how early they appear:

• Criticized and classified loans. Internally downgraded credits that are still performing. The earliest visible signal.
• Nonperforming loans. Loans 90 days past due or on nonaccrual, expressed as a percentage of total loans.
• Allowance for credit losses coverage. Reserves as a percentage of loans, and as a multiple of nonperforming loans. Coverage falling while NPLs rise is a clear warning.
• Provision expense. The current-period charge to build reserves. It flows straight through earnings.
• Net charge-offs. Loans actually written off. The lagging confirmation that the earlier signals were real.

The sequence matters because it lets you position yourself ahead of the reported numbers. By the time charge-offs appear in the income statement, the deterioration has been visible in criticized loans for several quarters.

5. Capital adequacy

Capital is what absorbs losses before depositors and bondholders are touched. The headline measure is the CET1 ratio: common equity tier 1 capital divided by risk-weighted assets. US banks face regulatory minimums plus buffers, and the largest institutions face additional requirements set through stress testing.

Read the headroom rather than the level. A bank operating a few tenths of a percent above its requirement has no capacity to absorb a credit surprise, fund loan growth or buy back stock, and it may be forced into a dilutive equity raise at exactly the worst moment. A bank with several points of headroom has options.

Tangible common equity to tangible assets is the simpler cross-check, since it ignores risk weightings, which are themselves model outputs. Where the two measures diverge sharply, the risk weightings are doing a lot of work and deserve scrutiny.

Valuation: price to tangible book and ROTCE

Banks are valued on price to tangible book value, and the multiple is only interpretable next to return on tangible common equity. The logic is straightforward: if a bank earns more on its equity than the return investors require, that equity is worth more than its accounting value, and the stock should trade above book. If it earns less, it should trade below.

This is why a bank trading at 0.7 times tangible book is not automatically cheap. If it earns 6 percent on tangible equity against a cost of equity near 10 percent, the market is pricing the economics correctly. Buying banks purely on a low P/TBV screen is one of the more reliable ways to end up holding a value trap.

The price to earnings ratio has some use for banks, but it flatters them at the top of a credit cycle, when provisions are low and earnings look their best precisely as risk is highest. Analysts often normalize by applying a through-cycle charge-off rate to the loan book.

Where bank problems actually start

Almost every bank failure traces to one of four things, and usually to several at once: credit losses concentrated in a single sector or geography; funding that reprices or flees faster than assets can be liquidated; interest rate positioning that went the wrong way and left large unrealized losses in the securities portfolio; and thin capital that turned a survivable loss into a solvency event.

These are correlated, which explains why bank trouble tends to arrive suddenly after a long quiet period rather than gradually. The unrealized losses in held-to-maturity securities portfolios are worth checking specifically, because they do not flow through reported capital until the securities are sold, and a bank forced to sell them for liquidity crystallizes the loss at the worst possible time. Analysts who work through loan-level and deposit-level detail at scale increasingly do it by querying the underlying data in plain English rather than rebuilding the same spreadsheet each quarter.

A practical checklist

1. Net interest margin: level, direction over eight quarters, rate sensitivity disclosure.
2. Deposit mix: noninterest-bearing share, deposit beta, uninsured concentration, reliance on brokered funding.
3. Loan book: composition, concentration by sector and geography, growth versus peers.
4. Credit: NPL ratio, reserve coverage, provision trend, criticized loan movement.
5. Capital: CET1 headroom above requirement, tangible common equity ratio, unrealized securities losses.
6. Efficiency ratio and its trend.
7. ROTCE, then price to tangible book value read against it.

That is a genuine afternoon per bank if you do it from the filings. Investables.ai assembles it automatically: enter any US bank ticker and the bank stock analysis card pulls margin, funding, loan composition, credit quality, capital ratios and peer valuation, then lays out the bull case, the bear case and the risk flags. It is informational research, not advice, and it never tells you what to buy.

Common questions

What is a good net interest margin? Most US commercial banks run somewhere between 2.5 and 4 percent, with the level driven heavily by business mix. Consumer lenders and credit card issuers run far higher because they take more credit risk. Compare against direct peers and against the bank's own history, never against the market as a whole.

What is the efficiency ratio? Noninterest expense divided by revenue, so it measures overhead per dollar earned, and lower is better. Large US banks commonly operate in the 50 to 60 percent range. A ratio drifting above roughly 65 percent suggests the cost base is outgrowing the franchise.

Are bank stocks good dividend investments? Banks have historically paid meaningful dividends, but those dividends are subject to regulatory capital requirements and are among the first things cut when credit deteriorates. Dividend sustainability at a bank is a question about capital headroom and credit quality, not about the payout ratio.

How do rising interest rates affect bank stocks? It depends on positioning. Asset-sensitive banks generally benefit as loans reprice upward faster than deposits. But rising rates also increase funding costs as depositors demand more, create unrealized losses on existing fixed-rate securities, and eventually pressure borrowers' ability to repay. The net effect varies by bank and by how far into the cycle you are.

For research and educational purposes only. This article is not financial advice and not a recommendation to buy or sell any security.