Short answer: The code on page one of your UBPR encodes charter type first and asset size second. Groups 1 to 8 are insured commercial banks, largest to smallest. 101 to 104 are insured savings banks, again largest to smallest. 201 to 203 are credit card specialty banks, 301 is bankers banks, 401 is non-insured non-deposit trust companies, and a four-digit code is the year a de novo opened. The full table is below. Two groups, 5 and 6, hold 2,353 of the 4,315 group members counted in that table, which is the first clue that the group alone will not answer most questions.
How to read the number
The codes look arbitrary until you notice they are built in two layers. The hundreds digit is the charter or business model. Within that family, the lower the number, the larger the bank: group 1 is the top of the commercial ladder and group 8 the bottom, 101 the top of the savings ladder and 104 the bottom. Nothing else is encoded, and since a February 2026 change that cut the commercial groups from 17 to 8, nothing else is considered for a commercial bank either. Office count and metro location used to split the small commercial groups. They no longer do.
The four-digit codes are the exception to the size rule. A bank that opened in 2024 is in group 2024 no matter what it holds, because a balance sheet built from scratch produces ratios that say more about the calendar than about management. The year group is a holding pen until the FFIEC stops treating the bank as de novo.
You do not pick your group and you cannot appeal it. The three things that move a bank are charter type, opening date, and average total assets, and only the last one moves on its own.
Every peer group code, and how many banks are in it
This is the table that is hard to find anywhere in one piece: every code, the FFIEC's own description of it, the number of group members that filed a call report for the quarter ending June 30, 2026, and the median total assets of those members.
| Code | FFIEC description | Banks | Median assets |
|---|---|---|---|
| 1 | Insured commercial banks having assets greater than $100 billion | 29 | $255.0B |
| 2 | Insured commercial banks having assets between $10 billion and $100 billion | 114 | $22.3B |
| 3 | Insured commercial banks having assets between $3 billion and $10 billion | 227 | $4.80B |
| 4 | Insured commercial banks having assets between $1 billion and $3 billion | 529 | $1.55B |
| 5 | Insured commercial banks having assets between $300 million and $1 billion | 1,273 | $512.1M |
| 6 | Insured commercial banks having assets between $100 million and $300 million | 1,080 | $187.3M |
| 7 | Insured commercial banks having assets between $50 million and $100 million | 320 | $77.4M |
| 8 | Insured commercial banks having assets less than $50 million | 133 | $35.4M |
| 101 | Insured savings banks having assets greater than $1 billion | 128 | $1.97B |
| 102 | Insured savings banks having assets between $300 million and $1 billion | 154 | $540.0M |
| 103 | Insured savings banks having assets between $100 million and $300 million | 120 | $184.4M |
| 104 | Insured savings banks having assets less than $100 million | 73 | $63.2M |
| 201 | Credit card specialty banks having assets greater than $3 billion | 7 | $43.5B |
| 202 | Credit card specialty banks having assets between $1 billion and $3 billion | not published | not published |
| 203 | Credit card specialty banks having assets less than $1 billion | not published | not published |
| 301 | Bankers banks | 12 | $687.8M |
| 401 | All non-insured non-deposit trust companies | 58 | $34.2M |
| 2019 | De novo banks opened in 2019 | 12 | $293.7M |
| 2020 | De novo banks opened in 2020 | 6 | $301.0M |
| 2021 | De novo banks opened in 2021 | 7 | $259.5M |
| 2022 | De novo banks opened in 2022 | 14 | $191.1M |
| 2023 | De novo banks opened in 2023 | 6 | $144.2M |
| 2024 | De novo banks opened in 2024 | 6 | $80.0M |
| 2025 | De novo banks opened in 2025 | not published | not published |
| 2026 | De novo banks opened in 2026 | 7 | $44.5M |
Source: BankingLens, from the FFIEC's UBPR peer group file and call reports for the quarter ending June 30, 2026. Banks are group members that filed; median assets are the median total assets of those members. Not published: our peer group file carries no size figures for these three groups, and our scorecard computes no medians or percentiles for a group with fewer than five filers. The Banks column adds to 4,315, which is the number of group members the FFIEC's peer group file counts and not a filer count: it runs 19 above the 4,296 banks that filed for the quarter, so it is not meant to foot to that total. The quartiles further down this page are computed over 4,238 banks, being those 4,296 filers less the 58 non-insured non-deposit trust companies in group 401. Those 58 keep their row above, because describing the group is the point of this page.
Four things in that table are worth a second look before you use your own row.
The distribution is lopsided. Groups 5 and 6 hold 2,353 banks between them, more than half of everything in the table, while 201, 301 and the de novo years are small enough that a single member moves the median. In group 5 your percentile is a position among 1,273 banks and it is stable. In group 301, with 12 members, one peer's bad quarter moves you several percentile points with nothing happening at your bank.
The specialty groups sit outside the size ladder entirely. The seven credit card banks in group 201 had median assets of $43.5 billion, nearly double the median group 2 commercial bank at $22.3 billion, and they are still not in group 2. The 58 trust companies in group 401 had median assets of $34.2 million, less than the median bank in group 8, the smallest commercial group, at $35.4 million. A trust company administers assets that never land on its own balance sheet, so every ratio with assets in the denominator means something different for it. Both facts are arguments for the grouping, not against it.
The de novo years are an age ladder. Median assets fall in order with the opening year: $293.7 million for 2019, $301.0 million for 2020, $259.5 million for 2021, $191.1 million for 2022, $144.2 million for 2023, $80.0 million for 2024 and $44.5 million for the seven banks opened in 2026. Note what that does to group 2020. Its median member now holds $301.0 million, the size of a group 5 commercial bank, and is still measured against five other banks of its vintage rather than the 1,273 banks of its size.
Savings banks are a separate ladder at every size. Group 102 covers $300 million to $1 billion, the same range as commercial group 5, and holds 154 institutions against 1,273. The ladders are separate because the business models are, so never drop a savings bank into a commercial bank's hand-built list for being the right size.
Why the UBPR only compares you to your own group
The Uniform Bank Performance Report puts three numbers next to each other on every line: your ratio, the peer group figure, and your percentile within the group. The comparison is confined to the group for two reasons, one statistical and one political.
The statistical reason is that a ratio only carries information when structure is held constant. Put a credit card bank, a trust company and a $60 million rural commercial bank in one distribution and the percentiles stop describing management and start describing business model. The group is a crude control for structure, and crude beats absent.
The political reason matters more in a board room. The group is assigned by the regulator from the charter and the balance sheet, so nobody at the bank chose it, which is the one thing a hand-built list can never claim. That is why the right sequence is the FFIEC group first, a custom list second, with both on the page.
One mechanical detail is easy to miss. The FFIEC's peer group average report trims banks above the 95th and below the 5th percentile for each ratio before averaging, which means the average for one ratio is computed from a different set of banks than the average for another. They cannot be combined into anything. A median needs no trimming, which is why every figure on this page is one.
Where the group is too coarse to tell you anything
Group 5 runs from $300 million to $1 billion. A $310 million bank and a $990 million bank are both in it, and the median member holds $512.1 million. Those two do not compete for the same loans, cannot carry the same overhead, and are almost certainly not funded the same way, yet both are ranked against the same 1,273-bank distribution. Group 2 is wider still: $10 billion to $100 billion is a tenfold span across 114 banks, with a median of $22.3 billion. Group 1 has no upper bound at all, and its 29 members have median assets of $255.0 billion.
What that costs you shows up in the spread inside a single asset band. Here is the middle half of banks in the $300 million to $1 billion band, the band that matches group 5.
| Ratio, banks with $300M to $1B in assets | Bottom quartile | Median | Top quartile |
|---|---|---|---|
| Return on assets | 0.87% | 1.28% | 1.70% |
| Efficiency ratio | 52.96% | 61.02% | 70.35% |
| Cost of funds | 1.94% | 2.38% | 2.79% |
| Loans to deposits | 69.41% | 81.66% | 91.49% |
| CRE concentration to capital | 85.15% | 153.44% | 227.48% |
Source: BankingLens, computed from FFIEC call reports for the quarter ending June 30, 2026. Quartiles are calculated across the full population of filing banks in each size band, not sampled. Non-insured non-deposit trust companies are left out of every statistic: they take no deposits and make no loans, so a margin or a funding cost computed for them has no meaning. The asset band includes savings banks and other charters, so its population is slightly larger than commercial group 5 alone.
Read the CRE row. The middle half of banks this size run from 85% to 227% of capital in commercial real estate, a gap in strategy so large that the two ends of it are not in the same business. Comparing your concentration, your charge-offs or your reserve coverage to a median drawn across that range answers a question nobody asked. The efficiency row spans more than 17 points and the loan-to-deposit row more than 22.
Some of that spread is size, and the ladder makes it visible. Median CRE concentration runs 80.95% of capital in the $100 million to $300 million band, 153.44% in the $300 million to $1 billion band, and 219.81% in the $1 billion to $3 billion band. A bank moving up through the bands is being compared to a progressively more concentrated crowd, which is a separate reason a percentile can move while your balance sheet stands still.
This matters most within about 10% of a band line. A bank at $980 million is near the top of group 5, where its overhead ratios flatter it against smaller peers. Cross to $1.01 billion and it is near the bottom of group 4, compared to banks up to three times its size, with its own ratios unchanged. Nothing happened except the arithmetic of the group.
What to do about it
The group is the right starting point and the wrong stopping point. Five practical steps, in order.
- Find where you sit inside your own group. Your percentile tells you where your ratio ranks. It does not tell you where your balance sheet ranks. Compare your total assets to the group median in the table above. A $340 million bank in group 5, whose median member holds $512.1 million, is being measured mostly against banks larger than itself, and the scale effects in the efficiency ratio will run against it every quarter.
- Keep a same-size cut beside the group figure. Not instead of it. For a bank in the bottom half of group 5, the honest second comparison is banks from roughly $300 million to $600 million, which strips out the size advantage without letting anyone choose the names. Show both columns and the page survives the question of why you added one.
- Split on the one structural thing your group ignores. For commercial banks the group is asset size and nothing else, so whatever drives your bank most is invisible in it: a CRE concentration, a farm book, a large trust department, a Subchapter S election, the community bank leverage ratio framework. Pick the one your board asks about and cut the peer list on it once. Our guide to building a peer group works through each of these splits.
- Watch the band lines, both yours and your peers'. If you are within 10% of a boundary, put both groups on the page for the year before you cross and the year after. Do the same when a peer crosses out of your group: it leaves the distribution, and the median it leaves behind is not the one you reported last quarter.
- Re-verify the group code every quarter. It is one line on the UBPR and it changes without notice, either because you crossed an asset line or because the FFIEC changed the definitions, as it did in February 2026 when 17 commercial groups became 8. A peer figure carried forward from an old deck can rest on a grouping that no longer exists.
On how we use these groups: BankingLens ranks every bank inside its assigned FFIEC peer group, because a percentile computed against a grouping we invented would be worth less than a free UBPR. We are not limited to it. The asset-band quartiles above come from the whole filing population rather than one group, less the group 401 trust companies, and a watchlist puts any set of banks side by side whatever group they were assigned. None of that is needed to use this page: the codes, the counts and the ranges are the same whether you pay anyone or not, and the UBPR is free.
Frequently asked questions
What do the FFIEC peer group numbers mean?
The number encodes charter type first and asset size second. Groups 1 through 8 are insured commercial banks from largest to smallest, 1 being banks over $100 billion and 8 being banks under $50 million. The 100s are insured savings banks, 101 to 104, again largest to smallest. The 200s are credit card specialty banks, 301 is bankers banks, and 401 is non-insured non-deposit trust companies. A four-digit code is a year: a bank in group 2024 is a de novo that opened in 2024, and it stays in the year group rather than the size ladder.
Which FFIEC peer group is my bank in?
The code is printed on page one of your Uniform Bank Performance Report, which the FFIEC publishes free on its Central Data Repository. You do not choose it and you cannot change it except by changing charter, opening date, or asset size. For a commercial bank, the only test after charter is average total assets, so crossing an asset line moves you into a different group and changes every percentile on the report.
How many banks are in an FFIEC peer group?
It ranges from a handful to more than a thousand. For the quarter ending June 30, 2026, group 5 held 1,273 commercial banks with $300 million to $1 billion in assets and group 6 held 1,080 with $100 million to $300 million, so those two groups alone account for 2,353 of the 4,315 members the FFIEC peer group file counts across every group. At the other end, 401 held 58 trust companies, 301 held 12 bankers banks, and 201 held 7 credit card banks.
Why does the UBPR compare my bank only to its own peer group?
Because a ratio only carries information when structure is held constant. A trust company that takes no deposits, a credit card bank whose yields run several times a commercial bank's, and a two-year-old de novo still building a balance sheet would each distort a shared distribution. Confining the comparison to one group also removes the argument that management picked flattering peers, which is exactly what an examiner or a director will test first.
Is the FFIEC peer group too broad to be useful?
It is too broad to be the only comparison. Group 5 runs from $300 million to $1 billion, which puts a $310 million bank and a $990 million bank on the same page, and the median member holds $512.1 million. Inside that asset band the middle half of banks posted efficiency ratios from 52.96% to 70.35% and CRE concentrations from 85.15% to 227.48% of capital. The group is the right starting point and the wrong stopping point.
What is peer group 401?
Non-insured non-deposit trust companies, 58 of which filed for the quarter ending June 30, 2026. Their median total assets were $34.2 million, less than the median bank in group 8, the smallest commercial group, because a trust company administers client assets that never touch its own balance sheet. Every asset-denominated ratio behaves differently for them, which is why they get their own group and why they should be kept out of any hand-built commercial bank list.
Where these numbers come from
Group codes and descriptions come from the FFIEC's UBPR peer group file. Group counts and median assets are computed from the call reports filed for the quarter ending June 30, 2026, and cover every group, including the 58 non-insured non-deposit trust companies in group 401. The asset-band quartiles are computed across the full filing population rather than a sample, over 4,238 banks: those trust companies are left out of every statistic on the site, since they take no deposits and make no loans. You can see a real bank ranked inside its own FFIEC peer group, ratio by ratio, on the sample scorecard for Frost Bank, which is public and needs no sign-up. Plans that cover your state or every bank start at $29 a month (pricing), and a watchlist holds the banks you follow, peer group or not.