Delinquency and default rates remain relatively stable, according to a recent Experian report. While that may be comforting news, it’s also good to know that advances in fintech capabilities let lenders reduce delinquency and default rates. How? There’s a correlation between vehicle finance fraud and default rates. As many as 70% of early payment defaults (within the first six months) may contain some element of fraud or misrepresentation perpetrated by dealers, fraud rings, or borrowers.
There’s ample evidence that vehicle finance fraud is increasing. Vehicle acquisition by fraud is cleaner and easier than hotwiring or carjacking. The National Insurance Crime Bureau is holding the 2019 Vehicle Finance Fraud Conference, covering topics such as The New Era of Auto Theft, Credit Washing, and Leveraging Technologies.
Fraud Recognition of Four Popular Tactics
The most popular fraud schemes associated with loan applications involve misrepresentation of income, employment, and collateral, and identify theft. The proliferation of personal and financial data associated with billions of electronic transactions, numerous data breaches, and unscrupulous websites that (for a fee) generate false documents have made it easy to provide seemingly credible, yet misleading information on auto loan applications.
False identity can instantly boost an applicant’s credit score to get better loan terms, a higher value car, but more likely, it’s the fast track to vehicle acquisition at minimum cost.
Income misrepresentation won’t boost a subprime applicant’s credit score, but it can give a lender false confidence that the borrower can afford a newer vehicle or get a lower rate and shorter terms.
Employment misrepresentation, particularly for the unemployed, improves the possibility of matching a lender’s credit policies and obtaining a loan, but it’s almost assuredly a sign of pending default.
Collateral misrepresentation is usually a dealer scheme, where vehicle value, add-ons, taxes, and other fees are inflated to increase dealer profit. Although not as frequently as other schemes, they can represent huge losses for lenders. A Missouri dealer is being sued by a lender for over $600,000 in damages. In another instance, Ford Motor Credit is demanding $112M as a result of a dealership chain fraud scheme.
Fraud Recognition Technologies
To combat these troubling trends, auto lenders can use the latest improvements in fraud recognition technologies and techniques. By identifying fraudulent auto loan applications at the “point of entry”, lenders eliminate the risk of loans increasing their delinquency and default rates. Here’s how improved fraud recognition technologies achieve that goal.
Analysis by machine learning of millions of auto loan applications (legitimate, as well as fraudulent) and performance of the associated loans identifies attributes or characteristics, often subtle or hidden, that can signal misrepresentation. When these machine learning algorithms are applied to new loan applications they detect potential fraud quickly, indicating the type of suspected fraud and providing a confidence score. Based on this information, a lender can determine the next steps in the loan origination process.
Modern loan origination solutions that use decision rules and workflow can automate the process and avoid spending time on highly-suspect applications. A few examples illustrate how fraud recognition and automation work together to support well-informed decisions that reduce delinquency and default rates.
|Reason: employment misrepresentation
Confidence score: high
|Automatically initiate an auto-decline, giving the reason as incomplete or unreliable information.|
|Reason: income misrepresentation
Confidence score: low
|Issue a conditional approval with the stipulation that additional income verification is required.|
|Reason: identity misrepresentation
Confidence score: medium
|Have application reviewed by an underwriter to verify applicant identity|
|Reason: collateral misrepresentation
Confidence score: high
|Automatically initiate an auto-decline, and add dealer name to a “watch list.” Analyze portfolio for other loans sourced from the dealer to proactively identify potential risk.|
With machine learning, fraud recognition continues to improve with every application it analyzes, fine-tuning the recognition algorithms to the unique market segments, geographies, and dealer relationships of every lender. Some fraud detection services can monitor applications submitted to all lenders who use a centralized service, and by doing so provide early warning of a dealer who is systematically perpetrating collateral schemes across the US.
Fraud Recognition: The First Step Toward Declining Delinquency and Default Rates
Lenders who use these improved fraud recognition capabilities will catch subtle fraud indicators that would otherwise be missed. The clear advantages are efficiency (rapid recognition of potential fraud) and risk avoidance (no time wasted on high-risk applications).
defi SOLUTIONS, recognized as one of the Top 50 Most Promising Fintech Providers, offers cloud-based loan origination solutions (LOS) that integrate with fraud recognition services to prevent the risk of delinquencies and defaults. Ensure you’re prepared to eliminate fraud by contacting our team today or registering for a demo of defi LOS.
Get in touch with us today and get a demo!