- Source: federalreserve.gov
If you are considering credit models based on alternative data, Treliant knows the benefits and risks of using alternative data and algorithms to market, underwrite, price, or service loans. Model risk management and compliance programs must consider the unique features of alternative data, including the potential for data to be biased, inaccurate, or lacking a logical relationship to credit risk. Treliant has the experience in designing effective model risk management programs, evaluating data and model impacts on fair lending, and assessing compliance risk management systems.
- On December 3, 2019, the Board of Governors of the Federal Reserve System, the Consumer Financial Protection Bureau, the Federal Deposit Insurance Corporation, the National Credit Union Administration and the Office of the Comptroller of the Currency (jointly, “Agencies”) issued an Interagency Statement on the Use of Alternative Data in Credit Underwriting. In encouraging the responsible use of alternative data, the Agencies recognized the potential benefits of alternative data in improving the speed, cost, and accuracy of credit decisions; lowering the cost of credit; and expanding access to credit.
- However, much of the Interagency Statement was dedicated to the risks of using alternative data. The Agencies reminded lenders that the use of alternative data and analytical methods must comply with the requirements of the Equal Credit Opportunity Act and the Fair Credit Reporting Act, including provision of accurate adverse action notices, as well as avoiding unfair, deceptive, or abusive acts or practices. Lenders using alternative data may need more robust compliance management systems, including testing, monitoring, and controls to manage consumer protection risks properly.
- Banks must also implement appropriate data controls, including rigorous assessments of the quality and suitability of the data to support prudent banking operations. The Agencies referred banks to existing interagency guidance on model risk management for more information on managing model risk, including models whose features include alternative data, in a safe and sound manner.