SAN DIEGO -

PointPredictive, a provider of machine learning fraud solutions, recently announced the results of a new study that detected high levels of fraud in early payment auto finance contract defaults.

The firm explained its study encompassed data from millions of auto finance applications submitted by dealers all over the U.S. across all vehicle types.  PointPredictive auto fraud models analyzed each application and gave it a fraud score.

While built to detect fraud, PointPredictive said its scientists were surprised to find that it did extraordinarily well in the detection of early payment default, a term finance companies use to indicate when contracts default within the first six months.

The study found that by scoring auto finance applications with models built to detect fraud, finance companies could detect 50 percent or more of their early payment default (EPD) prior to funding than if they relied on traditional credit scores alone. 

“Our analysis and experience suggest that many auto loans that default within the first six months have fraudulent misrepresentation on the loan application,” PointPredictive chief executive officer Tim Grace said.

“When we ran fraud pattern recognition models on the application information provided on EPD loans, we found strong evidence of fraud,” Grace continued. “This is the same type of behavior mortgage lenders discovered prior to the mortgage crisis when it was determined that up to 70 percent of mortgage EPD was fraud related. 

“The study confirms that using credit scores alone cannot detect fraud or risk of default,” he went on to say.

The PointPredictive analysis proved that fraud scoring could:

• Detect 14 times more fraud for finance than current solutions while flagging less than 5 percent of the total applications.

• Prevent 50 percent or more of a finance company’s early payment default losses by identifying those applications that had misrepresentations that would lead to loss.

• Identify risky dealers that submit multiple fraud applications up to three months sooner and reduce losses due by 70 percent due to early detection of bad players.

In 2007, a study by BasePoint Analytics found that between 30 and 70 percent of mortgage loans that defaulted within the first six months contained serious misrepresentations on the original application. These misrepresentations on borrowers’ income, employment, collateral or even intent to occupy had a material impact on the performance of the loan but were often considered “hidden fraud” since they were never detected in the application process.

PointPredictive insisted that auto financing fraud, like mortgage fraud, occurs when information on an auto finance application is intentionally misrepresented either by the borrower, a sophisticated fraud ring, or an unscrupulous dealer. 

When information is manipulated and the finance company does not know about it, PointPredictive noted that institutions may underwrite the application assuming the information is valid. Intentional fraud presents a problem for auto finance companies since loans that have misrepresentation are more likely to result in EPD.

To counter auto finance fraud and better detect EPD, PointPredictive Auto Fraud Manager uses pattern recognition, a technique that scientists have perfected to detect fraud based on historical data mining. The solution works by analyzing historical patterns of fraud, EPD and risky dealer activity and then scores each application as it comes in from a dealer. 

Finance companies are automatically alerted when an individual application has a significant number of anomalies or fraud patterns related to the income, employment, collateral, borrower or dealer. The lender can review the application and take action before it is approved. Over time, if a particular dealer submits many applications with similar fraud patterns, the solution will alert the lender to that as well so they can take the appropriate action.

The full results of the study have been published in the PointPredictive whitepaper titled “You Can’t Fight Fraud with Credit Risk Tools,” which is available at no charge by emailing info@pointpredictive.com.