# Benchmarks that tell good from bad: what a healthy bank looks like this year
Two banks report the same profit of 2 billion. One is celebrated by analysts; the other gets its CEO fired within a year. The difference is not the headline number. It is the ratios underneath. This lesson gives you the four yardsticks that let you judge any bank in under sixty seconds.
A bank with a 2 trillion balance sheet will always earn more in absolute terms than one with 200 billion. Absolute profit tells you almost nothing. Ratios normalize for size, so you can compare JPMorgan to a regional lender to Deutsche Bank on the same scale.
Four numbers do most of the work. Learn these and you can read a bank's health the way a doctor reads blood pressure.
ROE = net profit divided by shareholder equity. It tells you the return the bank generates on the capital its owners put in.
Healthy range (as of early 2026, industry estimate): roughly 10% to 12%.
Below 8%, a bank is often earning less than its cost of capital, meaning it destroys value even while reporting a profit. Above 15% sustained is excellent, and sometimes a warning sign of excessive risk-taking.
Context matters. US banks have generally run higher ROEs than European ones for a decade. Large US banks have frequently posted ROEs in the low-to-mid teens; many large European banks spent years stuck below 8% and only recently, helped by higher interest rates, pushed into the 10% to 13% zone. These are broad estimates, not precise figures.
Worked example. A bank reports net profit of 3 billion and shareholder equity of 25 billion.
ROE = 3 / 25 = 0.12 = 12%Twelve percent. Right in the healthy band. This is the single most common calculation you will do in this sector.
CIR = operating costs divided by operating income. It measures efficiency: how many cents the bank spends to earn one dollar of revenue.
Healthy range (2026 estimate): roughly 55% to 60%. Below 50% is very strong.
A CIR of 55% means the bank spends 55 cents to generate 1 dollar of income. A bank at 70% is bloated: too many branches, too much headcount, weak digital investment, or falling revenue. The best-run banks (some Nordic and digitally native players) have pushed well below 50%.
Worked example. Operating costs of 6.6 billion, operating income of 12 billion.
CIR = 6.6 / 12 = 0.55 = 55%Efficient. If costs stayed flat but income fell to 10 billion, CIR would jump to 66%, and the market would notice instantly.
CET1 stands for Common Equity Tier 1. It is the bank's highest-quality capital (mostly common shares and retained earnings) divided by its risk-weighted assets (RWA), which are the bank's assets adjusted for how risky each one is. A mortgage carries less risk weight than an unsecured corporate loan.
This is the core solvency buffer, the cushion that absorbs losses before depositors and the system are threatened. It exists because of the Basel III framework, the global capital rules written by the Basel Committee on Banking Supervision and enforced in the US by the Federal Reserve and in Europe by the European Central Bank (ECB) and the European Banking Authority (EBA).
Healthy range (2026 estimate): comfortably above 13%. Regulatory minimums plus buffers typically sit lower, often around 10% to 11% depending on the bank, so a level above 13% signals a solid margin.
Worked example. CET1 capital of 60 billion, risk-weighted assets of 400 billion.
CET1 ratio = 60 / 400 = 0.15 = 15%Fifteen percent. Strong. A bank drifting toward its regulatory minimum is a red flag: it may be forced to cut dividends, halt buybacks, or raise fresh capital.
For the official rules straight from the source, the Basel Committee's Basel III overview is free and authoritative.
NPL stands for Non-Performing Loans: loans where the borrower has stopped paying, typically 90 days or more past due. The NPL ratio is non-performing loans divided by total loans.
Healthy range (2026 estimate): under 3%. Under 2% is strong.
This is the asset-quality gauge. A rising NPL ratio means borrowers are defaulting, often the first sign of trouble before it hits profit. After the 2010s European debt crisis, some southern European banks carried NPL ratios above 15%. A decade of cleanup brought the European average down dramatically, and it now sits low by historical standards (a low single-digit percentage across the EU, per EBA estimates). US banks have generally maintained low NPL ratios outside of recession spikes.
Worked example. Non-performing loans of 9 billion, total loans of 450 billion.
NPL ratio = 9 / 450 = 0.02 = 2%Two percent. Healthy asset quality.
Take a hypothetical mid-size European bank reporting:
Verdict in one glance: profitable, efficient, well-capitalized, clean loan book. All four in the healthy band. This is a solid performer.
Now flip two numbers: ROE at 6% and CIR at 72%. Same capital and asset quality, but this bank is unprofitable relative to its cost of capital and structurally inefficient. Analysts would push for cost cuts, a strategy overhaul, or a merger.
The power of the scan is that no single number decides. A high CET1 with a terrible ROE means a bank that is safe but not earning. A great ROE with a rising NPL ratio means profits that may be about to evaporate.
🎬 [VIDEO: "How to Read a Bank's Financial Statements" - youtube.com - a clear walkthrough of the income statement and balance sheet items that feed these ratios]
To read benchmarks with judgment, know the landscape (all figures below are broad estimates, not exact):
The structural takeaway: a "healthy" 10% ROE in Europe may be viewed as strong locally, while the same number at a top US bank might read as merely average. Always benchmark a bank against its regional peers, not a global absolute.
Knowledge check
1. Two banks report identical absolute profits, yet analysts praise one and criticize the other. What is the primary reason ratios are preferred over absolute profit figures when judging bank health?
2. A bank posts an ROE of 6%. Why might this be concerning even though the bank is still reporting a profit?
3. A bank reports a sustained ROE above 15%. Why should an analyst not automatically interpret this as unambiguously good news?
4. Select ALL correct answers about interpreting ROE across different banks and regions.
Select all the correct answers.
5. Select ALL correct answers about calculating and applying ROE.
Select all the correct answers.
When you sit down with a bank's results, run this checklist:
1. Compare to peers, not to zero. A 9% ROE is weak in the US and decent in parts of Europe. Pull two or three comparable banks.
2. Watch the trend, not just the level. A CET1 falling from 15% to 13% over two years matters more than the level alone. Direction reveals strategy and stress.
3. Cross-check the ratios against each other. Suspiciously high ROE plus rising NPLs can mean the bank juiced returns by lending to riskier borrowers.
4. Read the footnotes on NPLs. Definitions vary, and banks sometimes restructure loans to keep them off the NPL line. Look for terms like "forbearance" or "Stage 2 loans."
5. Check for one-off items. A great quarterly ROE inflated by an asset sale is not repeatable. Strip out one-offs to see the underlying number.
For live European sector data, the EBA Risk Dashboard publishes aggregate CET1, NPL, ROE, and CIR figures for EU banks each quarter, free to download.