How Identity Resolution Protects Financial Institutions against First Party Fraud

December 09, 2019 | Name Matching, Risk Management

Identity Resolution for First Party Fraud

Identity Resolution Defends against a Distressingly Common Source of Fraud

Last year five people were arrested in New Jersey for defrauding banks of over $2.5 million. They were charged with conducting a fraudulent scheme to use stolen and altered identities to open lines of credit at financial institutions. They then used the credit cards they received to purchase items for which they had no intention to pay. The banks were stuck with the losses.

In the course of the crime, the criminals would make reasonable purchases and pay for them for a while using the credit cards they received in order to establish and enhance their creditworthiness.

Taking advantage of the enhanced creditworthiness, the conspirators then did what is termed “busting out”: they used the cards to make large purchases of all kinds with the intention of never repaying.

This is a real-life example of first party fraud (“First party” means it’s the financial institution that’s hurt. For other kinds of fraud, see here.) It’s a major threat to financial institutions. (For more detail on how sophisticated gangs can get in executing this kind of fraud, see here.)

First-Party Fraud is Difficult to Detect

Financial institutions find it difficult to uncover or protect against this kind of fraud. It’s typically carried out by well-organized networks and frequently bank employees are involved, particularly those who approve loans. It’s estimated that 10% to 20% of unsecured bad debt at U.S. and European banks is actually first-party fraud.

The basic problem for the financial institution is that activity looks normal in all the accounts participating in the fraud right up until the bust-out. Even then the financial institution won’t realize what’s going on when the payments stop coming. During the life of the accounts the criminal network will work hard not to violate any rules or thresholds that would raise red flags.

Typically, the only line of defense a bank has is retrospective. They establish cross-functional teams to look at debts that have been written off across the enterprise. Given that the process is totally manual, the teams cover only a small sample and they are years late.

Conventional fraud analytics also falls short as it’s focused on the individual account holder. What’s needed is technology that can track a criminal network of multiple individuals operating multiple accounts in tandem.  There is one such: Identity Resolution.

How Identity Resolution Can Help Detect First-Party Fraud

The key to breaking up a first party fraud is to look at similarities in the data provided for all new applications to the data for current accounts: name, address, phone numbers, etc. Criminals work hard to make sure that there are no obvious connections, but sometimes their memories or records aren’t good enough.

Here are samples of the kind of connections Identity Resolution can find:

  • A recent client is filing a new loan application, and it turns out that a previously failed application by another client contains a very similar mailing address (it’s a good idea for a financial institution to always check all new applications against past failed ones);
  • A name occurring on a new credit card application is very similar to a name on a list of past customers who had defaulted;
  • Several recently opened accounts share the same home phone number even though they do not share the same address or do not seem to have any family relationship.

In these hypothetical scenarios, the software is able to identify the suspicious links among the participants using similarities in names, phone numbers, and mailing addresses. Using these similarities, fraud investigators can build a network of individuals involved in the crime.

Identity Resolution looks for clues by matching in a fuzzy way all kinds of variations. More advanced Identity Resolution products use matching models generated by Machine Learning algorithms that are trained on a very large amount of real-world labeled data.

In sum, Identity Resolution is a high-accuracy, high-volume technology that delivers trustworthy matching results that ensure that financial institutions can discover criminal networks infesting their accounts. It will also prevent these scams from happening if the financial institutions use Identity Resolution to regularly scan their data.