Background
Our client, the second-largest non-bank mortgage lender in the United States, employs over 10,000 professionals nationwide and has recently made significant strides in expanding its personal loan division. However, the company faced a critical challenge: as loan volumes scaled up rapidly, so did the risk of fraudulent activities. With each undetected fraud incident costing between $5,000 to $35,000 in reimbursements to investors, the company urgently needed a solution to mitigate these risks without exponentially increasing its team of fraud analysts.
Challenges
➢ Rapid Scaling and Fraud Risk: As the company expanded its personal loan division, the volume of loan applications surged, exposing them to higher potential fraud risks.
➢ Resource Constraints: Hiring additional fraud analysts to keep pace with the escalating application volumes was not a viable option due to cost considerations.
➢ Inconsistent Fraud Detection: Despite efforts, some personal loan applications were slipping through the cracks, resulting in significant financial losses from fraudulent activities.
Implementation
➢ Automated Decision Making: Applications identified as low fraud risk were automatically approved, streamlining the process and reducing the workload for analysts. Conversely, those flagged as high fraud risk were automatically declined, minimizing exposure to potential losses.
➢ Development of Fraud Model: Our team initiated a comprehensive data-driven approach, analyzing historical loan application data to develop a sophisticated fraud detection model. This model categorized loan applications into three risk segments: low, moderate, and high fraud risk.
➢ Focused Analyst Review: Fraud analysts now focused primarily on applications categorized as moderate fraud risk. This targeted approach allowed them to allocate their efforts efficiently, concentrating on specific elements of these applications rather than every detail.
Results
➢ Scalability without Increased Costs: The implementation of the fraud model enabled the company to scale its loan application volume by tenfold without the need for additional hires in the fraud detection team. This resulted in substantial savings in payroll expenditures.
➢ Significant Reduction in Fraud: The overall fraud rate plummeted from 2% to nearly 0%, demonstrating the efficacy of the new model in mitigating risks effectively.
➢ Enhanced Investor Confidence: With improved fraud detection capabilities, investor satisfaction soared as the incidence of fraudulent activities and subsequent financial liabilities decreased substantially.
➢ Cost Savings and Efficiency Gains: The company saved millions of dollars annually by preventing fraud losses and optimizing operational efficiency within its personal loan division.
“Our collaboration with the consulting team was transformative. Their data-driven approach not only improved our fraud detection capabilities but also streamlined our operations, resulting in substantial cost savings and a near elimination of fraud incidents.” – Head of Risk Management, Second-largest Non-Bank Mortgage Lender
