Case Study: UPDATING A CREDIT CARD ORIGINATIONS RISK MODEL FOR BETTER DECISIONING
One of the nation’s largest commercial credit card providers needed to revalidate its current underwriting risk model with current commercial financial data and modeling attributes.
Our client — one of the largest commercial credit card issuers in the U.S. — needed to update its business credit card origination risk model, which is crucial to determining whether applicants are extended credit and, if so, on what terms. This card issuer realized their old risk model was outdated and was in need of revalidating. A longstanding client, they turned to us.
Identifying the Right Data
Our client knew that the breadth and quality of Equifax’s commercial data was unparalleled to the competition and chose to leverage our Data and Proxy data for Commercial Risk Model Development.
With access to more data than any other commercial entity and modeling data and analytics capabilities that are second to none, we counseled our client at the onset of the project about what data was available. We went through a detailed exercise to chart relevant data options and then collaborated to determine what was best for their business purposes. This is a crucial step, as clients don’t always know what data is available to them and what data can be most effective. We worked with them throughout the process to introduce data and analytics that would be most powerful for their risk modeling purposes.
Refreshing Data and Analytics
Working closely with our client’s modeling and analytics team, we took a data set of their current commercial card holders, and appended commercial financial data and attributes at two points in time: origination and an older date.
We also appended proxy performance data, or information that might not have been available on the customer file — for example, an existing non-financial trade opened prior to or on a specific date. We appended the proxy data during the same two time periods (origination and a later date). Then, we scored both to provide our client with a comparison of data and attributes at the two different times.
Partnering to Ensure Success
We moved nimbly in terms of delivering the refreshed information. While we have a strong model development capability, our client — which has its own internal analytics department — was able do in-house risk model development based on the combined data and proxy performance. We collaborated closely with our client throughout the entire process. Once our client had the data in house, we established weekly calls to ensure the project was on target and to problem-solve any issues that arose.
Data and Proxy Append for Commercial Risk Model Development is a customer origination score model based on our unparalleled commercial and proxy performance data.
To learn more about how you can update your risk model for better decisioning, reach out to email@example.com