PDF Estimating Auto Lending Fraud Losses

WHITEPAPER Auto Lending Fraud Losses in 2017

Visit us at

2

ABOUT POINTPREDICTIVE

PointPredictive Inc. is a leading provider of fraud solutions to banks, lenders and finance companies. It solves the billion-dollar fraud problems of auto lending, mortgage lending and on-line retail fraud with the latest technology platforms, smarter science and business experience by leveraging big data with analytic models. The PointPredictive Auto Fraud Manager solution helps banks, lenders and finance companies solve the problem of fraud, early payment default and dealer risk for automotive lending. Leveraging a collaborative fraud consortium, the pattern recognition models can precisely identify risk and reduce both Fraud and Early Payment Default Losses by 50% or more with low false positive rates. Located in San Diego, Calif., more information about PointPredictive can be found at

2

Estimating Auto Lending Fraud Losses for 2017

3

INTRODUCTION ? AUTO LENDING FRAUD RISK IS RISING

Auto lenders in the US are experiencing increased losses due to fraud and early/first payment default as they struggle to maintain loan quality amid rising auto loan origination volumes. In the last quarter of 2016, uncollectible auto loans soared to over $1.1 Billion dollars and delinquencies on subprime loans hit their highest level since the worst recession in US history.

With auto lending levels soaring to historical highs and delinquency losses rising dramatically, many lenders have a heightened sensitivity to fraud and its impact on their bottom line.

Traditional fraud and risk management processes are also challenged by the changing portfolio risk profile as almost 20% of all new auto loans are issued to non-prime borrowers, according to recent Experian data.

Similarities to the Mortgage Industry

It is not unlike what the mortgage industry experienced between 2003 and 2008. In 2003, mortgage lending volumes were hitting record levels due to an increase in indirect lending through brokers as well as an increase in non-prime lending. By 2008, the mortgage industry was melting down ? in large part due to rampant fraud that was hidden in these broker and non-prime applications and then misclassified as credit loss.

Fast-forward 8 years and the US auto industry is experiencing the same types of rapid growth, which is fueling the same sparks of fraud risk that erupted in the mortgage industry.

To understand the impact of rising fraud, PointPredictive leveraged predictive algorithms and application analysis by fraud experts to estimate auto lending fraud risk for 2017.

Based on that analysis, PointPredictive now estimates the level of annual fraud risk to be between $4 Billion and $6 Billion.

This is double the annual amount that we estimated just one to two years ago. This whitepaper explores the components of that fraud risk and the drivers that are contributing to the rising level of auto lending losses in the industry.

3

Estimating Auto Lending Fraud Losses for 2017

You Can't Fight Auto Fraud With Credit Risk Tools

4

AUTO LENDING FRAUD MAY BE KNOWN OR HIDDEN

There is no central reporting agency for collecting and reporting on auto lending fraud losses in the US. This makes assessing the absolute levels of fraud losses in the US very challenging. Even if there was a centralized reporting structure for confirmed frauds, it would tend to understate the impact of fraud since our research indicates that a substantial portion of application fraud is misclassified as a credit loss.

Because of this, the best approach to this problem is to review past applications, leveraging statistical models and industry experts to form a complete picture of fraud. By applying those findings to current lending estimates, we can arrive at the annual levels of loss on based on new applications, originated loans and expected default rates on those loans. This is the approach used by PointPredictive to estimate annual levels of fraud originations in the US each year.

Known Fraud and Unknown Hidden Fraud

The cost of auto lending fraud does hit the bottom line of banks and lenders but it may not always be categorized or recognized as a fraud loss. In the case of auto lending fraud, there is both Known Fraud and Unknown Fraud.

Known Fraud - Some fraud is identified either before an auto loan is originated or shortly afterwards when a borrower notifies the lender of identity theft or a lender discovers the fraud in their collections process

Known fraud is more common than you might expect and is present on approximately 0.30% (30 basis points) of originated application volume.

KEY FINDINGS

Known auto lending fraud can run approximately 0.30% (30 basis points) of origination volume at lenders.

Unknown auto lending fraud typically manifest as early payment default, which can run as high as 3% of originations.

Unknown Fraud - Some fraud is never identified ? not during the application process and not even after a loan has been funded and defaults. This fraud, in most cases, results in early or first payment default where the borrower never makes a payment on their loan after they walk out of the dealership.

Data and investigative analysis of early payment default loans indicate that between 40% and 70% of those loans have significant misrepresentation on the original loan application which led to the financial loss.

40% to 70% of early payment defaults have been linked to fraud when the original application is examined.

4

Estimating Auto Lending Fraud Losses for 2017

5

AUTO LENDING FRAUD TYPES CATEGORIES

Based on analysis of application data and loan servicing data by industry fraud experts, PointPredictive breaks down fraud losses in three distinct categories:

Early Payment Default Fraud ? Loans that default within the first 6 months have a much higher probability of containing material misrepresentations in the original loan application than loans that default later. PointPredictive analysis suggests that 40% to 70% of early payment defaults have some element of misrepresentation in the initial application that led to the loss. The range depends on the lender's pre-application fraud controls as well as the underlying risk of the loan program itself.

Dealer Fraud ? Auto lenders, particularly those concentrating in non-prime lending, experience high levels of losses that can be tied back to specific dealers. Some dealers have extremely high levels of early payment default, known fraud and bad loan quality that leads to losses. These losses may not always be categorized as fraud. However, through careful retrospective analysis, lenders often determine that many of the losses they take are due to intentional misrepresentation are clearly the result of a systemic, organized attacks originating from within the dealers' finance organizations.

PointPredictive analysis has determined that, for some lenders, virtually 100% of their fraud losses are associated with fewer than 3% of their dealers and nearly 100% of their early payment default losses come from just 10% of their dealers.

Known Fraud and Misrepresentation ? Most auto lenders do have some tracking systems in place for fraud losses; however, that rarely provides a complete picture of their losses since so much fraud is hidden. Identity Theft, Straw Borrower, Collateral and Dealer Fraud are generally the most common categories of fraud losses that lenders experience.

Known fraud and misrepresentation levels can vary from lender to lender based on their preand post-funding fraud controls as well as the level of reporting that they do concerning fraud. On average, PointPredictive finds that a lender's known fraud is approximately 0.30% (30 basis points) of originated volume.

5

Estimating Auto Lending Fraud Losses for 2017

6

AUTO LENDING FRAUD WILL REACH $4 TO $6 BILLION IN 2017

After analyzing both known and unknown fraud, PointPredictive estimates that there will be $4 to $6 billion in fraud for loans originated in 2017.

This estimate was based on origination volume of new and used car auto loans of approximately $600 billion and the estimated rate fraud for both known and unknown fraud types.

Unknown Fraud (Early Payment Default)

Known Fraud (Misrepresentation)

Total Originated Fraud

Low Estimate 0.40% 0.30%

$4.2 Billion

High Estimate 0.70% 0.30%

$6 Billion

PointPredictive Low Estimate - $4 Billion in Fraud for 2017

PointPredictive's low estimate for 2017 is $4 billion in originations with fraud that will lead to a financial loss. This assumes an average industry early payment default rate of 1% (some lenders are lower, some are higher) and assumes our low range estimate of early payment defaults that are linked to fraud of 40%.

PointPredictive High Estimate - $6 Billion in Fraud for 2017

PointPredictive's high estimate for 2017 is $6 billion in originations with fraud that will lead to a financial loss. This assumes an average industry early payment default rate of 1% (some lenders are lower, some are higher) and assumes our low range estimate of early payment defaults that are linked to fraud of 70%.

6

Estimating Auto Lending Fraud Losses for 2017

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download