Based on information provided by the CFPB, the quality of the data must improve. Now when data is submitted, it is run through a series of quality and validity checks. Despite this, a review of what is submitted typically finds numerous errors. The CFPB has announced that it is developing a new system for submitting the data, which will allow lenders to send data at any time and have it checked before final submission. Then, if there are problems, they can be corrected. While this is a benefit to lenders, it requires that there be a constant focus on the data and the submission process.

When conducting Fair Lending reviews, which include an examination of the HMDA data against the file, the CFPB will no longer be satisfied with just finding matching information in the file. It will expect the lender to provide it with the specific document from which the data was gathered. A lack of consistency around the use of a specific document for a specific data element can indicate that the credit union has problems controlling the data. Remember that there will also be a lot more data to control.

What Can Credit Unions Do?

There are steps that credit unions can take starting today to be prepared for this new emphasis on HMDA and Fair Lending:

• One of the most important steps to take is to make sure that you have clearly defined your lending area. If your membership is open to the general public, the risk of the CFPB finding Fair Lending issues is greater. You must be able to show that you have reached out to underserved communities, particularly, those in your lending area that include minorities.

• On the other hand, if you have membership limited by profession or some other unique attribute, gather data to determine what the mix of minorities and non-minorities are within this group. From there you want to make sure that you are reaching out to all members equally.

• In addition, you need to make sure that the data in your HMDA Loan Application Register is accurate. In order to do this, it is best to prepare a directive or protocol identifying what documents are to be used to collect the data and provide it to the staff. It is best to have two or three options listed in preferential order. For example, “income will first be pulled from the underwriting worksheet and if not available there, from the final application.” It is also a good idea to have a second underwriter review any denials to ensure the validity of these decisions.

• Next, you should conduct regular reviews of the data. One easy way is to include it is as part of your random quality control process. Another option is to have it tested by a third party. This ensures the validity of results using a statistically valid sample.

• Other reviews that should be done include conducting a matched pair analysis. This involves selecting two loans, one of which is a white male borrower who was approved and a minority with similar (if not identical) loan attributes who was denied. The reviews should focus on why a different decision was made for each loan. Credit unions should also test for disparate lending, which involves the pricing of the loan. In this test, the pricing of loans that are approved for white males should be compared to the pricing for similar (if not identical) approved loans for minorities. Finally, the Equal Credit Opportunity disclosures and timing should be tested.

• In addition to this testing, you should also have the LAR data analyzed. Today there are independent companies that will conduct a series of analyses and identify any potential problems garnered from your data. This can be based on any of the protected classes such as race, sex, national origin, familial status or any other protected class.

The bottom line? While HMDA is becoming more complex, proper preparation can prevent your organization from facing the slings and arrows of an outraged CFPB.

Becky Walzak is president of rjbWalzak Consulting. She can be reached at 561-459-7070 or [email protected].