Traditionally demand deposit account origination has followed a fairly standard process. Enter the application, check OFAC and "closed-for-cause" databases, open the account. Lather, rinse, repeat. However, with banks looking to shore up deposit account profitability in the wake of the Durbin Amendment by cutting costs and reducing fraud losses, the standard origination process just won’t do it anymore.
Fortunately there are a number of new data sources and analytic models available for banks to incorporate into the origination process to help address these challenges.
The traditional data source used for demand deposit account origination is closed-for-cause data. When a consumer abuses a demand deposit account (usually by overdrafting it and not paying the fees) the account is forcibly closed by the bank and the consumer is reported to the closed-for-cause databases.
Closed-for-cause data is great at identifying historical account abuse. If you have had a checking account closed by a bank in the last five years, you will most likely be listed in one of the main closed-for-cause databases. The logic is that if you have abused a checking account in the past, you are likely to abuse the checking account that you are applying for today.
The logic of using closed-for-cause data during the demand deposit account origination process is perfectly sound, but there are some disadvantages to using this data. First, it is quite expensive (dollars per file), which is a problem for banks looking to reduce their costs (in today’s market this is a priority for every bank).
Second, closed-for-cause data is not predictive for every risk population. It’s great at identifying consumers who have committed account abuse in the past, but that’s not the only group of consumers who pose a risk to banks’ demand deposit account portfolios. What about consumers who have abused credit accounts but never debit accounts? What about fraudsters who steal legitimate consumers’ identities and use them to open deposit accounts and steal money from banks? These risk populations fall outside the purview of closed for cause data.
So what other options do banks have for data to support cost-effective demand deposit account origination?
Well, for predicting future deposit account abuse banks can look at traditional and alternative credit data. If a consumer has abused credit products in the past then there is a strong possibility that they will abuse the demand deposit account they are applying for today. Traditional credit data from the primary credit bureaus can easily support this type of analysis.
Credit data can also be used to catch a portion of the consumer population that has committed demand deposit account abuse in the past. These consumers abuse debit and credit products equally. Thus, these consumers can be identified as risky based on their negative credit histories.
Alternative credit data can also be used to provide an additional perspective on the consumer’s financial responsibility. For example, if the consumer has responsibly used debit and credit products in the past but they took out a payday loan one week ago, that might be an indication that they have recently become financially distressed and are a risk to abuse the checking account they are applying for.
For catching identity fraud during the account opening process, financial institutions have a host of alternative data sources and analytic models to choose from. Banks could look at additional information that is not included in traditional credit files like professional licenses, asset information, education history, utility records, rent payments, address changes, and collections data. This deeper level consumer data can be used to detect synthetic identities, which usually lack anything more than a name, address and social security number. This data can also be used to catch fraudsters who have stolen legitimate consumers’ identities.
Analytic models and scoring algorithms can be used to detect discrepancies in this data that are indicative of fraud. For example, an application that is completely normal except for a phone number that is associated with 13 different addresses should raise a red flag. Discrepancies can also be detected for applications that are submitted online. If a consumer who lives in Cleveland applies for a demand deposit account from a laptop that is accessing the internet in Tripoli, it’s a good chance that consumer’s identity has been stolen.
Credit and fraud data is not very expensive (cents per file) and it can be used to catch significant portions of the various risk populations applying for deposit accounts.
Banks need to ask themselves: Which data source or combination of data sources will be the most cost-effective in managing risk during the account origination process? Closed-for-cause data used to be the only answer. Today, banks have more options.
Paul Thielemann is the vice president of sales and marketing for Zoot Enterprises.