When conducting internal audits & monitoring, how are you representing & quantifying your HMDA findings? do you give an error rate based on the number of loans with errors to the total number of loans in your sample? do you give an error rate based on the data fields with errors vs total number of data fields in your sample?
Just looking for input back on what other institutions are doing to represent this data for management and show improvement vs. decline.

Why would you choose to do anything different than your regulator? Measuring internally against a different stick really does no one any good.

I gave an error rate based on number of files; and then a table showing which fields had errors. I also identify whether those errors are key errors or not; and causation of the error if readily identified.