Reduce the chances of being accused of discriminatory lending practices

Making the news recently are the charges against some to the largest auto lenders alleging discriminatory pricing of loans and other discriminatory lending practices.

American Honda Finance Corp and Toyota Motor Credit Corp are currently embroiled with the U.S. Department of Justice and the Consumer Financial Protection Bureau over accusations of discriminatory loan pricing practices through auto dealerships.

In 2013, Ally Financial Inc. paid a total of $98 million in fines and borrower refunds to resolve similar accusations by the Justice Department and the CFPB of discriminatory lending practices.

The practices that got Honda, Toyota, and Ally cross-wise with the DOJ and CFPB had to do with allowing dealers the ability and incentives to mark up interest rates. The regulatory agencies had evidence that loans to minorities were being marked up by dealers differently than to non-minorities. CFPB has issued a bulletin that it would hold indirect auto lenders accountable for unlawful discriminatory pricing at partnering dealers.

Accusations of discriminatory practices that hit the news are one of the worst nightmares any lender can experience. Resolving allegations of discriminatory pricing of products is very costly in terms of fines, bad public relations, legal fees, and staffing.

Reducing the chances of discriminatory loan pricing allegations are well worth the investment. Pricing loans using stochastically derived and statistically validated models will go a long way towards assuring all parties that loan pricing has been done in an objective, non-discriminatory manner.  This article describes how using stochastic methods for pricing loans according to risks and costs unique to each borrower benefits financial institutions and consumers. Using stochastic methods also helps assure regulators that a financial institution is adhering to unfair-lending laws.

Most financial institutions engage in some form of discretionary consumer loan pricing using credit grades or some other non-discriminatory criteria that in their minds, reflects risks unique to individual borrowers. Unfortunately, many financial institutions price loans by watching the competition. By now, it should be obvious that this method is fraught with dangers. For lenders, pricing according to the competition or some other subjective method carries the following risks to name a few:

1. No two lenders have the exact same cost structure especially for borrowers of different credit risk levels. There is no assurance that using the competition’s pricing will result in profitability and/or avoid the unfair practice of one class of borrower subsidizing another.
2. Pricing loans differently according to a borrower’s “profile” without assuring such pricing is arrived at by using statistically validated methods based on a lender’s unique costs and experience leaves a lender open to accusations of discrimination. Discrimination accusations can be difficult to fend off unless a lender can show it used stochastic methods based on its data, lending history and experience to create its pricing model.
3. Using subjectivity or some other non-statistically validated process to set prices opens the gates for a myriad of entities including regulators to accuse a financial institution of discrimination.

A financial institution wants to engage in activities, including loan pricing, that are defendable if challenged by regulators or other groups. A financial institution can feel confident its loan-pricing model meets regulations, is profitable, and minimizes unfair subsidies between classes of borrowers if its loan program follows these criteria:

1. Uses stochastic methods and its borrower database to determine risks inherent in different credit scores and then establishes its credit grade ranges
2. Identifies costs (including operational, collections, charge-offs, cost of funds, etc.), both direct and indirect that are incurred in its lending process
3. Uses Activity Based Costing (or some equivalent method) to assign lending costs to each risk-grade of borrower, loan type, etc.
4. Converts these costs to interest rates according to risk grade, loan type, etc.
5. Determines an acceptable profit margin and assigns rates accordingly to each risk-grade of borrower, loan type, etc.
6. Statistically validates its pricing structure on a regular basis to assure its pricing model accurately reflects costs unique to each credit-risk grade, loan type, etc.

In summary, there are eight primary reasons financial institutions should use stochastic methods for risk-based loan pricing:

1. It allows them to reach “deeper” into the loan pool and profitably serve a wider base of borrowers
2. So they can identify proper rates for each borrower and avoid subjective pricing
3. It assists them in determining correct FICO ranges for establishing rates by credit grade
4. It helps assure they diversify their loan portfolios
5. It assures they are pricing loans such that they can be confident that they are applying the optimum mark-up on loans above their lending costs
6. It helps maximize yields on overall loan portfolios
7. It helps assure that every borrower pays according to their relative risk and that one class of borrower is not subsidizing another
8. It helps assure objectivity in the loan pricing process and therefore helps assure a financial institution engages in loan practices that are considered ethical and legal

The Department of Justice or the Consumer Financial Protection Board are typically not bearers of glad tidings. Use stochastic management tools to reduce the chances of having to cope with them.