Investor Relations · Second quarter 2026 · First edition
The Quarterly Median Profile Report
The Quarterly Banking Profile is a report in which eighteen banks and four thousand others are averaged together, and the result is announced as the condition of "the industry," the way a man with one foot in boiling water and one in ice is on average comfortable.
What this is. Every chart in the FDIC’s 2Q2026 Quarterly Banking Profile, recomputed for the median bank. The QBP adds up every bank’s dollars and divides — the industry as one bank. The QMPR takes the same 4,238 banks, computes each ratio bank by bank, and reports the one in the middle, with the quartiles around it and the share of banks that do the thing at all. Same data, same universe, same definitions. The dollar-weighted column in every table here reproduces the published QBP to its printed decimal (86 of 93 rows exactly; the other seven are off by between $21,000 and $3.4 million on figures in billions).
Eighteen is the number of banks over $250B in assets at 6/30/2026. The column "QBP ex-18" is the dollar-weighted figure for the other 4,220.
Page 1 — What the typical bank is
| Loan type, % of loans | QBP | QBP ex-18 | Median bank | Middle half of banks | Banks with any |
|---|---|---|---|---|---|
| Nonfarm nonresidential RE | 13.9 | 23.6 | 23.7 | 12.2–36.2 | 97.5% |
| Residential mortgages | 19.0 | 19.9 | 23.6 | 13.2–37.7 | 97.5% |
| Commercial & industrial | 18.2 | 16.3 | 9.7 | 5.6–15.4 | 96.3% |
| Construction | 3.3 | 5.5 | 5.4 | 2.2–9.8 | 91.4% |
| Other consumer | 2.5 | 3.6 | 1.0 | 0.3–2.5 | 95.1% |
| Nondepository financial institutions | 10.9 | 5.5 | 0.0 | 0.0–0.0 | 14.9% |
| Credit cards | 8.5 | 4.9 | 0.0 | 0.0–0.0 | 15.8% |
| Loans to purchase/carry securities | 4.1 | 1.8 | 0.0 | 0.0–0.0 | 4.3% |
| Leases | 0.9 | 0.8 | 0.0 | 0.0–0.0 | 15.8% |
The industry’s loan book is 24% cards, NDFI, margin and leases. The median bank’s is 0%. Those four products exist at 4–16% of charters. The median bank is a real-estate lender — half its book is nonfarm nonresidential and residential mortgage; the industry’s is a third. Take the eighteen out and nonfarm nonresidential jumps from 13.9% to 23.6%, landing on the median; the card and NDFI books fall by less than half, because the second tier of card and warehouse lenders sits in the $10B–$250B class.
Page 2 — What it earns
| Net interest margin, quarterly | QBP | Median bank | Middle half of banks | Banks |
|---|---|---|---|---|
| Industry | 3.32 | 3.92 | 3.42–4.40 | 4,237 |
| >$250B | 2.94 | 2.94 | 1.94–3.24 | 18 |
| $10B–$250B | 3.95 | 3.65 | 3.16–3.99 | 140 |
| $1B–$10B | 3.93 | 3.78 | 3.33–4.19 | 890 |
| $100M–$1B | 4.02 | 4.00 | 3.50–4.46 | 2,647 |
| <$100M | 3.99 | 3.94 | 3.43–4.56 | 542 |
QBP ex-18: 3.95%. The entire 60-basis-point gap between "the industry" and the typical bank is eighteen institutions. Inside every size class the median sits within 30 bp of the class figure; across classes the dollar-weighted number is pulled to the one class that is different. The typical bank’s margin is not a point; it is a band from 3.4% to 4.4%, and the published 3.32% sits below the 25th percentile of it.
| Returns, quarterly | QBP | Median bank | Middle half of banks |
|---|---|---|---|
| Return on assets | 1.37 | 1.26 | 0.85–1.71 |
| Return on equity | 13.76 | 12.00 | 7.89–16.22 |
Wider margin, lower return. The median bank gives back its NIM advantage in capital and in the absence of a fee business: ROE is 1.8 points below the industry’s, and inside the $100M–$1B class the median ROE equals the class figure, so the industry-level gap is again the eighteen.
Page 3 — What it loses
| Credit | QBP | Median bank | Middle half of banks | Banks with any |
|---|---|---|---|---|
| Net charge-off rate, quarterly | 0.573 | 0.004 | 0.000–0.077 | 77.5% |
| — >$250B | 0.662 | 0.328 | 0.088–0.491 | |
| — $10B–$250B | 0.603 | 0.152 | 0.057–0.292 | |
| — $1B–$10B | 0.296 | 0.029 | 0.000–0.122 | |
| — $100M–$1B | 0.157 | 0.001 | −0.001–0.052 | |
| — <$100M | 0.065 | 0.000 | 0.000–0.013 | |
| Past-due + nonaccrual rate | 1.44 | 1.01 | 0.40–2.11 | 94.3% |
| Non-owner-occupied NFNR past due + nonaccrual | 1.68 | 0.00 | 0.00–0.91 | 39.6% |
| Reserve coverage (allowance / noncurrent) | 173 | 201 | 93–519 | 87.8% |
The industry charge-off rate is a card-bank number. The median bank in every class charges off less than a third of its class’s dollar-weighted rate, and below $1B the median rounds to zero; 22.5% of banks had no net charge-offs in the quarter, and 25% of banks below $1B had net recoveries. The headline that would be written from the QBP — "charge-offs at 0.57%" — describes a rate that lives in about six percent of charters.
The non-owner-occupied CRE line the QBP charts by size class is the same shape: 1.68% industry, zero at the median, and six in ten banks report none past due at all. Reserve coverage is the one credit measure where the median bank is more conservative than the industry — 201% against 173%, and above 200% in the $1B–$10B and $100M–$1B classes — with 12% of banks holding an allowance against no noncurrent loans whatsoever (coverage undefined; 514 banks dropped).
Page 4 — What it holds
| Capital and balance sheet | QBP | QBP ex-18 | Median bank | Middle half of banks |
|---|---|---|---|---|
| Core capital (leverage) ratio | 8.98 | 10.51 | 10.96 | 9.70–13.00 |
| Equity / assets | 9.92 | 11.03 | 10.56 | 9.06–12.73 |
| Tier 1 risk-based (n = 2,440 reporting) | 13.75 | 13.92 | 14.62 | 12.53–18.70 |
| Total risk-based (n = 2,440) | 14.98 | 15.21 | 15.73 | 13.66–19.84 |
| Loans & securities repricing 3–5 yr, % assets | 8.9 | 11.3 | 16.0 | 10.2–21.6 |
| Loans & securities repricing 5–15 yr | 11.6 | 15.9 | 15.0 | 8.6–21.9 |
| Loans & securities repricing >15 yr | 12.5 | 11.2 | 6.4 | 2.1–14.0 |
| Unrealized loss on securities, % amortized cost | −5.5 | −5.2 | −4.9 | −7.9 to −2.2 |
Two points more leverage capital at the median than the industry carries. Below $10 billion, the median bank holds less capital than its class’s dollar-weighted figure (<$100M: 15.2% class vs 12.8% median), because the smallest and most over-capitalized charters dominate the sum; above $10 billion it holds more. Only 57.6% of banks report risk-based ratios; the rest are on the community bank leverage ratio, so the RBC medians describe the 2,440 that do.
The median bank’s long end is short: 6.4% of assets repricing beyond 15 years against 12.5% for the industry, and the 3–5 year bucket nearly double. The eighteen hold the long-duration book.
Page 5 — Who "the industry" is
| Loan growth, quarterly | Industry (dollar-weighted), % | Median bank, % | Middle half of banks |
|---|---|---|---|
| Industry | +1.78 | +1.65 | −0.13 to +3.75 |
| >$250B | +5.52 | +1.69 | +0.79 to +3.95 |
| $10B–$250B | −3.26 | +1.67 | +0.67 to +3.05 |
| $1B–$10B | −0.81 | +1.78 | +0.24 to +3.27 |
| $100M–$1B | −0.09 | +1.64 | −0.13 to +3.84 |
| <$100M | +0.57 | +1.53 | −1.41 to +4.81 |
This is the page. Read by the QBP, the largest banks grew loans 5.5% in a quarter while the $10B–$250B class shrank 3.3% and the two classes between $100M and $10B shrank or stood still. Read by the median bank, every class grew about 1.7%, within a quarter of a point of each other, six times. The dollar-weighted class figures are not lending; they are reclassification — one bank’s book moved into its buyer’s class when the charters merged in June, and banks crossing the $10B and $1B lines take their loans with them. The industry total (+1.78%) is right. The class rows are an artifact of the method that nobody reads as one.
The same decomposition for the headline ratios:
| QBP | QBP ex-18 | Median | Where the gap lives | |
|---|---|---|---|---|
| Net interest margin | 3.32 | 3.95 | 3.92 | entirely the 18 |
| Nonfarm nonresidential RE share | 13.9 | 23.6 | 23.7 | entirely the 18 |
| Leverage ratio | 8.98 | 10.51 | 10.96 | mostly the 18 |
| Credit cards share | 8.5 | 4.9 | 0.0 | half the 18, half the next 140 |
| NDFI loans share | 10.9 | 5.5 | 0.0 | half the 18, half the next 140 |
| C&I share | 18.2 | 16.3 | 9.7 | the top 158, not just the 18 |
| Net charge-off rate | 0.57 | — | 0.004 | the card banks, in two classes |
| ROE | 13.8 | 13.0 | 12.0 | the 18 and the fee businesses |
Shown the tables, the editor said the arithmetic was the bank’s and the opinions were the Chief Economist’s, and declined to improve either.
Edited by Mark Twain (posthumous)
Notes to users
- Universe: 4,238 FDIC-insured institutions at 6/30/2026, every size class equal to the published count. Values at the FDIC’s 8/15/2026 vintage. Growth priors are 1Q2026 and 2Q2025 as the 2Q2026 QBP reads them (8/17/2026 cut, pooled mergers folded in).
- A ratio’s median is the median of bank-level ratios; banks with a zero or negative denominator are dropped and counted. Growth is per-bank on the same-bank panel. No mean of ratios appears anywhere.
- Coverage is the share of banks in the group with a nonzero value. Where coverage is under 50%, the median is zero by construction and the interesting number is the coverage itself.
- Not computed: problem institutions (confidential), failed institutions, and the four count-of-banks rows, which have no per-bank value.