Magazine article The Journal of Lending & Credit Risk Management

Relationship Information Belongs in Consumer Credit Decision-Making

Magazine article The Journal of Lending & Credit Risk Management

Relationship Information Belongs in Consumer Credit Decision-Making

Article excerpt

The author relates what can happen when a bank acts in a credit quality vacuum. Banks must add relationship information to credit decision-making to gain competitive advantage over ever-increasing new competitors and to maintain and nurture their current consumer clients.

A customer's home equity line had reached its 15-year anniversary. Weeks earlier, in anticipation of the maturity of this line, the bank had performed its risk-focused standard underwriting process, using performance information from the home equity line history and from the customer's credit bureau.

In the vast majority of cases, this process results in the continuation of the revolving line. After assessing this customer's information, however, our underwriting center was convinced that the bank should discontinue the customer's revolving credit access to the line and begin amortization.

The customer receiving this verdict had a healthy deposit relationship with the bank, along with an investment relationship and two credit products, including the home equity line. The underwriting center's conviction triggered an unsurprising response from the customer, who wrote the bank stating that she was moving her total relationship away from the bank.

Where Is the Customer?

The decision-making process and supporting tools used by the bank were risk-focused. That is a good thing. But they were not particularly customer relationship-focused. That is not a good thing. Had we added relationship information into the equation, the bank may have proceeded differently and made a better, profit-based decision. At the very least, we would have enhanced communication with the customer and possibly retained her use of several other bank products.

The goal of banks' underwriting policies is to ascertain a level of risk that is both safe and profitable in the long term. Most of us determine a bank's acceptable risk through the following steps:

* Set a minimum required risk-adjusted return.

* Know the revenue.

* Know the cost of funds.

* Know the cost of operations.

* Solve the equation for allowable cost of credit loss.

Unfortunately, we often run this equation in a product silo or channel silo without knowledge of the customer relationship.

What's Handy versus What's Needed

In the same month that I learned of this customer's problem and reaction, the risk management team that is dedicated to the bank's credit card division approached me with a proposal. They wanted to strengthen our underwriting process used for credit card applications by adding a bankruptcy-focused risk model to our current use of a credit bureau score, a custom credit scorecard, and decision tree logic driven by bureau and application information. Already risk-focused, the process would have an even tighter focus with the addition of a bankruptcy-focused risk model. Meanwhile, the information being used by this process was only that information which was available to every credit card issuer in competition with my bank - and did not include our own relationship data.

My first question was how this approach is in the bank's best interest. Is it possible, I asked, that adding information about the status of the applicant as a customer of the bank could add even greater value to the credit decision process - improving customer-level profit while also outperforming the proposed risk management value of the bankruptcy model?

The answer I was given by our risk management team was, yes, customer information may represent a superior addition to our process. But the value of that addition was unknown and we did not have the ability to access quality customer information to support a real-time underwriting process that uses only seconds to render a decision. The bankruptcy tool was available now, was proven to enhance underwriting performance in other processes, and had already proven valuable in a retro score analysis. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.