1. Home
  2. Developed a statistical scoring model to identity theft in income tax filings

Finding identity theft in income tax fillings

Client

One of the large prepaid card issuers in the U.S.

EXL Analytics Solution

The EXL team developed a statistical scoring model for estimating the probability of a tax filing being a case of identity fraud.

Background

Client is looking at providing tax filers with an option of availing a Refund Anticipated Loan (RAL) for their tax returns through prepaid debit cards.

A huge challenge for providing RAL is the large number of identity theft cases encountered by the Internal Revenue Service (IRS) in U.S.

If the client provides a loan for a filing tagged by the IRS as a case of identity fraud, it can result in a monetary loss for the client.

The EXL team was tasked with identifying suspicious patterns related to identity theft to mitigate the risk of providing an RAL.

Results

  • 62% Fraud capture in three deciles
  • 40% Exposure loss mitigated