How 2 Dominant Lenders Distort the CRA "Borrower Characteristics" Benchmark
In my previous article I wrote about how two lenders, American Express and JPMorgan Chase captured a market share of 53.3% of all small business loans reported by FDIC and OCC-regulated CRA lenders during 2024. This market dominance has important implications for the other 3,575 lenders examined by the two bank regulators. With this concentrated loan market, the activity of the two top lenders can easily distort the benchmarks used by examiners to evaluate a bank's lending to mom and pop businesses.
One of the important lending tests administered in a CRA exam is the "Borrower Characteristics" test which measures how much lending banks are extending to smaller businesses as measured by "Gross Annual Revenue" of $1 million or less. Examiners compare the percentage of a bank's small business loans to smaller mom and pop businesses to the average "penetration rate" of other CRA-reporters within a bank's local market (assessment area). When a market is dominated by a couple of lenders their activity can easily distort the performance "benchmarks".
The 2024 CRA national data shows how dramatically two dominant lenders can affect the Borrower Characteristics Benchmark. The graphic below summarizes the situation.
The table shows that AMEX extended 19.10% of its small business lending to smaller business borrowers. For JPMorgan Chase, the results were almost diametrically the opposite, with 75.77%, its lending extended to small businesses.
How did the activity of those two lenders affect the market smaller business borrower penetration rate? The graphic below provides the answer.
The smaller business borrower penetration rate for the entire United States was 54.3%. If American Express's relatively (19.10%) low activity is removed from the calculation for the entire country the benchmark would increase from 54.3% to 67.4%. Imagine that, a single lender effectively lowered the national benchmark by 13.1%!
The activity of JPMorgan Chase had an almost equally dramatic impact in the opposite direction. If JPMorgan's activity were excluded from the national data, the benchmark would decrease from 54.3% to 46.7%. So, the activity of the second most active small business lender raised the bar 7.6% for the entire country.
If the activity of both lenders were trimmed from the national average the result would show a penetration rate of 62.7% extended to smaller businesses.
This has big implications for the other 3,575 lenders subject to CRA examination by the OCC or the FDIC.
The 2026 CRA NPR, by reducing the number of mandatory CRA reporters to about 86 will increase the concentration of lending into only a few banks thereby distorting CRA performance benchmarks all the more.
There is a big irony in the NPR.
Although the proposal reduces the number of mandatory reporters it does not reduce by a single bank the number of lenders accountable for performing under CRA.
The "relief" from reporting under CRA is a trap to reduce possible bank opposition to the NPR.
But what sensible and prudent banker would not collect and monitor data that is used by examiners to determine if they pass or fail their CRA exam?
As I have said in other articles, the cost of collecting and reporting CRA data for most community banks should be very small. I estimate $3,000 to $5,000 annually. The FFIEC offers free software and the data entry time and costs for a bank that originates fewer than 50 loans annually has got to be minimal.
The regulators preach that banks should be monitoring their compliance risk. But the NPR discourages almost all banks from capturing the data they need to identify and manage that risk.
