UBPR Peer Groups: Percentiles and Why Yours Changed

The most useful column on a UBPR is not your bank’s ratio. It is the one next to it: where that ratio ranks against your peer group. That single percentile turns a number you cannot interpret in isolation (“our efficiency ratio is 61%”) into one you can act on (“that is the 78th percentile of our peers, worse than three-quarters of banks our size”). But the peer group is assigned by rule, it moves as your bank changes, and the percentile column is read backward more often than any other figure on the report. This guide clears that up.

What Is a UBPR Peer Group?

A UBPR peer group is the set of banks the Uniform Bank Performance Report compares your bank against, so your ratios are measured against similar institutions instead of the whole industry. Comparing a $300 million rural bank to JPMorgan Chase tells you nothing. Comparing it to other small banks with a similar footprint tells you whether it is running well or poorly for what it is.

Since February 26, 2026, the FFIEC sorts every commercial bank into one of 17 numbered peer groups by asset size alone, measured as the bank’s average assets for the quarter. Before that date, groups for banks under $300 million were also split by number of banking offices and by metropolitan or non-metropolitan location; the FFIEC dropped both factors in Announcement 2026-02. The largest banks sit in the low-numbered groups; the smallest banks fall into the higher-numbered ones. The FFIEC revises the asset boundaries from time to time, so the cut points are not permanent, which is the source of the “my peer group changed” surprise addressed below.

The PCT Column: What the Percentile Actually Means

On the UBPR, most ratios appear with three things: your bank’s value, the peer group average, and a PCT (percentile) figure from 0 to 99. The PCT is your bank’s rank within its peer group for that ratio. A PCT of 90 means your bank’s value is higher than 90% of your peers. A PCT of 10 means only 10% of peers are below you.

The trap is assuming a high percentile is good. It depends entirely on the ratio:

• For a ratio where higher is better, such as return on assets or net interest margin, a high percentile is a strong result.
• For a ratio where lower is better, such as the efficiency ratio, the net charge-off rate, or a nonperforming-loan ratio, a high percentile is a bad result. Being in the 90th percentile on the efficiency ratio means your costs eat more of your revenue than 90% of your peers.

So the percentile is never good or bad on its own. Read it together with the direction of the metric. A veteran reading a UBPR checks the direction first, then the percentile, every time, because the report does not flip the scale for you.

Why Did My Bank’s Peer Group Change?

This is the most common UBPR question, and the answer is almost always mechanical. Your peer group is assigned from average assets. When that figure crosses a boundary, or the boundaries themselves move, the assignment moves:

• You grew or shrank across an asset threshold. Cross the line between two size bands and you land in the adjacent peer group, now compared against larger or smaller banks.
• The February 2026 change moved you. If your bank had under $300 million in assets, its group before February 26, 2026 also depended on its office count and metro location. Comparisons across that date mix two different peer sets.
• The FFIEC revised the peer group definitions. The boundaries themselves get updated periodically. When they do, banks can shift groups without changing at all.
• A merger changed your size overnight. An acquisition can push a bank across a size boundary in a single quarter.

When your peer group changes, your percentiles can move even though your own ratios did not, because you are now ranked against a different set of banks. A bank that was a strong performer among small institutions can look average when it grows into a group of larger, more efficient ones. That is not deterioration. It is a new comparison set. Always note which peer group a UBPR percentile is drawn from before reading a trend across the change.

Why Is My UBPR Showing N/A?

An N/A on the UBPR means the ratio could not be computed, not that the data is missing or hidden. The usual causes:

• The denominator is zero or does not apply. A ratio needs something to divide by. A bank with no loans of a certain type will show N/A for a ratio built on that category.
• The bank is too new. Ratios that require a prior-period or average balance cannot be computed until the bank has enough history, so a recently chartered bank shows N/A for those until it seasons.
• The item is not collected on the bank’s form. As covered in the guide to the three call report forms, the FFIEC 051 collects some items less often, so a quarter can legitimately have no value to compute from.

N/A is information. It usually tells you the bank does not do the thing the ratio measures, or has not existed long enough to measure it.

Using Peer Comparison Without Being Misled

Three habits keep peer analysis honest. First, confirm the peer group before reading the percentile, especially across a quarter where the bank grew or merged. Second, read the percentile with the metric’s direction, not as a raw score. Third, watch the trend in the percentile, not just the level: a bank sliding from the 40th to the 70th percentile on its efficiency ratio over a year is deteriorating relative to peers even if the raw ratio looks stable, because its peers are improving faster.

The one real limitation of the standard peer group is that it is assigned, not chosen. Sometimes the FFIEC group is not the comparison you actually care about, for instance if you want to benchmark against banks in your state, or against a hand-picked set of competitors rather than every bank in your size band. That is where a custom peer group, built from the same underlying call report data, does what the standard UBPR group cannot.

Building the Peer Comparison You Actually Want

BankRegReports computes every UBPR ratio for every bank, shows each one against its peer group with the percentile, and lets you build custom peer groups so you can benchmark against the banks you choose rather than only the FFIEC’s assigned group. If the standard peer group keeps shifting under you, or never matched your real competitors, a custom group gives you a stable, relevant comparison across the full history.

Explore bank rankings and peer benchmarks →

---

Frequently Asked Questions

What is a UBPR peer group? A UBPR peer group is the set of similar banks the Uniform Bank Performance Report compares your bank against. Since February 26, 2026, the FFIEC sorts commercial banks into 17 numbered groups by asset size alone, so ratios are measured against institutions of similar size.

What does the PCT column mean on a UBPR? The PCT column is your bank’s percentile rank within its peer group for that ratio, from 0 to 99. A PCT of 90 means your bank’s value is higher than 90% of its peers. Whether a high percentile is good depends on the metric: for return on assets higher is better, but for the efficiency ratio or charge-off rate a high percentile is worse.

Why did my bank’s UBPR peer group change? Peer group assignment follows the bank’s average assets for the quarter, so crossing a size boundary moves the assignment. Growth, shrinkage, a merger, or an FFIEC revision of the group definitions (such as the February 26, 2026 change that dropped office count and metropolitan location) can all move a bank to a new peer group, which also changes its percentiles.

Why does my UBPR show N/A for a ratio? N/A means the ratio could not be computed, not that data is missing. Common causes are a denominator of zero or one that does not apply to the bank, a bank too new to have the required prior-period balances, or an item that is not collected on the bank’s call report form that quarter.

Is a high UBPR percentile good or bad? It depends on the ratio’s direction. For ratios where higher is better, such as return on assets or net interest margin, a high percentile is good. For ratios where lower is better, such as the efficiency ratio, net charge-off rate, or nonperforming-loan ratio, a high percentile is bad.

Can I compare a bank against a custom peer group instead of the UBPR group? Yes. The standard UBPR peer group is assigned by the FFIEC, but because the underlying data is public call report data, you can build a custom peer group from banks you choose, such as competitors in your state or market, and benchmark against that stable set instead.