Looking for any insight from those who have been asked to help determine lender goals for CRA small business and home mortgage loans.

I'm struggling with building a method since the numbers will be measured off of a sample taken by examiners. That being said, 2019 sample sizes are going to be different than 2023 sample sizes. My denominator is constantly changing based on production and the economy.

Thoughts? ISB bank setting goals for the lending test in particular.

Why do the examiners base their CRA PE on a sample size? Is it because you don't collect ad report the data? If you are going to be held accountable for CRA performance (and every insured bank is accountable) why would you go into an exam blind? Why not use the free FFIEC CRA software to collect and edit check your small business, small farm and CD loans? You can't beat the price

If the examiners rely on a sampling of your loans you won't know if the sample is accurate or not and what it may show. You may look good You may look bad. I would be having an anxiety attack in such a situation.

Beginning in 2026 you aren't going to have a choice about collecting small business loan data under Section 1071 (if you do 100 small business loan originations annually - in fact, even if you don't do 100 small bus loans/year you're going to need to collect the data on all your commercial loans because you will need to determine if you become a covered reporter every year!).

Roughly how many commercial loans of $1,000,000 or less does your institution extend annually? If it's anything close to 100 you're going to need to set up a system very soon, so why not get some practice with collecting CRA loans?

Okay, fill me in on the discussion of examiners using a method other than sampling.

I am collecting small business/farm/hmda data. Won't they take all small business loans originated in a given year and pull a sampling of that denominator? ISB bank.

You would need to demonstrate that the integrity of the data you've compiled is reliable. If you've collected the data that's what the sampling is intended to verify. If you have not collected, compiled, and edited your data (which an ISB is not required to do). then examines will do a sampling of your loan files and base the CRA PE on that sampling (in other words, in the first instance the sampling attests to the reliability of your data and in the second instance it becomes the data upon which your exam is based).

It sounds like you are collecting the data so the sampling will be to substantiate how reliable it is. If it is found to be unreliable, then the exam will be done based on a "scientific" sampling of files not an unreliable database. You should have a system in place to demonstrate that you authenticate the data you've been collecting and compiling. Do periodic sampling on your own data and make it SOP.