Avoid 'Guesstimates' in Fair Lending
A Deutsche Bank subsidiary, MortgageIT, recently entered into a $12 million conciliation agreement to resolve allegations that it violated fair lending laws. The allegations were that African-American borrowers and Hispanics on average paid eight to 10 points higher in price and were 65% and 72% more likely to receive higher priced loan types. In addition, they purportedly paid on average $707 and $906 more in fees than nonminority borrowers. Further, their denial rates allegedly were 45% and 35% higher, respectively.
Clients often have told me that they know they lack fair lending problems because of their compensation plan or general pricing policies. The idea is that if they are not encouraging overages, fair lending violations cannot occur. Of course, denial rates are not based on overages. Further, pricing exceptions given disproportionately to one group, even by accident, can be just as damaging on a fair lending analysis as overages. Moreover, the selection of product type is also impossible to determine. When one considers that a pricing disparity of only eight to 10 basis points leads to a $12 million settlement, it is hard to have comfort in one’s ability to guesstimate lending patterns adjusted to the characteristics of similarly situated borrowers that may span several years.
The fact is that the cost of doing a regression analysis has come down. Depending on loan volume, many lenders can analyze a year of data for $5,000 and have counsel confidentially review a report. Such reports cover all aspects of fair lending, not just price, and compare companies to competitors. When one compares the cost of a $5,000 analysis to a $12 million settlement, its fairly clear why it pays not to guesstimate.