Comparable cohorts
Inspect institution counts, balance-sheet totals and median performance using versioned, disclosed membership rules.
Choose how institutions are grouped into peer cohorts — by asset size, geography, charter, or regulator — and each cohort's institution count, aggregate balance-sheet totals, and median performance metrics are computed for the selected reporting period. Every cohort carries a membership fingerprint, so the exact set of institutions behind a given row can be reproduced later even as the underlying population changes.
Rule: $1B–$10B, $10B–$50B, $50B–$250B, and $250B+ using reported quarter-end total assets.
Null behavior: Institutions without a verified total-assets value are excluded.
Sensitivity: Crossing a threshold changes membership; size alone does not imply a comparable business model.
Source field: ASSET
| Cohort | Institutions | Total assets | Total deposits | Total loans | Median Return on assets | Median Return on equity | Median Net interest margin | Median Equity to assets | Median Noncurrent loans | Median Liquid assets |
|---|---|---|---|---|---|---|---|---|---|---|
| $250B+20260630 · membership d7a67d473c | 18 | 16,853,018,769 | 12,838,029,431 | 7,370,915,778 | 1.18 | 12.48 | 2.95 | 0.09 | 0.01 | 0.34 |
| $50B–$250B20260630 · membership dadf419776 | 35 | 3,866,947,128 | 3,110,572,225 | 2,406,858,363 | 1.24 | 12.44 | 3.31 | 0.11 | 0.01 | 0.25 |
| $1B–$10B20260630 · membership 74f6e0f312 | 902 | 2,527,986,979 | 2,080,669,504 | 1,765,287,060 | 1.26 | 12.18 | 3.70 | 0.10 | 0.01 | 0.21 |
| $10B–$50B20260630 · membership 1b476fe1a1 | 112 | 2,440,915,108 | 1,953,405,049 | 1,607,421,100 | 1.23 | 10.63 | 3.65 | 0.12 | 0.01 | 0.22 |
| Under $1B20260630 · membership da0329ef19 | 3246 | 1,064,180,572 | 893,677,534 | 704,386,234 | 1.22 | 11.46 | 3.92 | — | — | — |
Medians exclude unavailable metric values. 0 institutions were excluded because the selected membership field was unavailable. Totals use the published full-market comparison universe for 20260630. Membership fingerprints make the selected period and constituent set reproducible; they are not ratings.