The attention lenders are getting about possible discrimination in their mortgage lending decisions may not be all bad.
If the scrutiny occasioned by discrimination investigations causes banks to find out what constitutes a good loan. it may help them to focus better on loans with higher values.
The disclosure requirements of the Home Mortgage Disclosure Act have fueled a storm of attacks on mortgage lenders from regulators, the conventional media, and advocacy groups. What started out to be a means for the government to look at redlining has become a device for raising questions about discrimination against individuals.
Taking Heat
Bank officials may feel unfairly attacked. They may feel that they've put in place a pretty good system of training and rules of conduct that minimize the room for discrimination.
Unfortunately, the statistical study leased in October by the Federal Reserve Bank of Boston concluded that. while underwriting factors explain many of the minority declines, they do not explain all of them.
In preparing to, defend themselves against charges of discrimination, lenders could do their own retrospective search of files. They could also begin to routinely compile data on the host of factors now considered in a loan application.
Once enough cases have been built up, the bank also could do its own examination of factors affecting application decisions and see if underwriting factors explain the denials.
On the Upside
Even setting up such a system could be a positive defense to regulators in showing good faith. Obviously, if the bank's analysis of its own data showed some differences in minority treatment not explained by underwriting factors, internal housecleaning would be in order.
At this point. the costs of building such a data file might seem excessive, if justified only by their value for regulatory compliance.
But here comes the upside: What if banks used the data not only to seek out defenses against charges of discrimination, but also to improve the underwriting process itself?
The industry knows that the rules of thumb are largely based on "gut" feel and plausible observations about relationships between application characteristics and loan performance.
In fact, the most recent long period of generally appreciating house prices meant that even when poor assessments of borrowers' willingness and ability to pay led to a payment default, the property value was able to bail out the mortgage holder.
Make Use of the Data
In these times of flat or even declining house prices, this bailout is no longer available.
Why not use the data needed to identify discrimination, or defend against such claims, in combination with information on loan performance, to help improve the underwriting decision itself?
A bank could use such a growing file of automated data to link with information on actual loan performance to relate performance to underwriting factors.
From an analysis of these links. the bank would be in a much better position to say what weights should be placed on various underwriting factors.
Much of the guesswork in how to weigh compensating factors could be eliminated. The payoff would be improved performance in new loans made.
]
