Menu
Banking Exchange Magazine Logo
Menu

Getting the point(s) of trended data

Predictive analysis brings understanding plot of credit “movie” over the snapshot

  • |
  • Written by  Paul DeSaulniers, Experian
 
 
Understanding a consumer's relationship to their credit can be better with trended data, a "movie" instead of a "snapshot." But there's more to it than just gathering more data points. Understanding a consumer's relationship to their credit can be better with trended data, a "movie" instead of a "snapshot." But there's more to it than just gathering more data points.

Making smart lending decisions is fraught with challenges. Consumer credit behavior is notoriously difficult to interpret. When lenders pull a credit score, they really can’t determine if it’s improving or deteriorating. Essentially, that score is just a snapshot that captures a moment in time.

Two consumers may have identical balances. However, there isn’t a way to differentiate between the two with a traditional score.

On one hand the lender has a person who consistently pays more than the minimum amount; has demonstrated the ability to pay; and doesn’t show signs of payment stress. On the other, the lender has one who is making only minimum payments and experiences increasing payment stress.

Yet, this difference is critically important to lenders that must make decisions that will impact their credit portfolio’s profitability.

Trended data has been promoted as the solution to this problem, offering lenders vital information about consumers. But trended data on its own is nothing more than the proverbial emperor’s clothes. “Naked” data alone offers no real insights.

And the sheer volume can be overwhelming.

Too much of a good thing?

Trended data analysis typically looks at five fields over a 24-month period: balance amount; original loan/limit amount; scheduled payment amount; actual payment amount; and last payment date.

For example, in the case of a single consumer with ten trades on file, the trended data would capture approximately 1,200 data points. Now, multiply that by a file of 100,000 consumers and the lender suddenly is dealing with more than 120 million data points—a veritable Mount Everest of raw information.

Lenders need to analyze this abundance effectively. When analyzed properly, trended data can provide detailed insight that will enable lenders to predict consumer behavior to better manage risk and optimize profitability.

Unfortunately, most organizations simply don’t have the analytical resources to handle such an enormous volume of credit information.

Trended data identifies segments with potential

Historical information combined with the current credit bureau report can be a gold mine for lenders, but only when a thorough data analysis is performed first. Armed with such analysis, lenders can arrive at insights that drive smart lending and profitability decisions.

Knowing how a consumer uses credit or pays back debt over time can help lenders offer the right products and terms to increase response rates; determine upsell and cross-sell opportunities; prevent attrition; identify profitable customers; and avoid consumers with payment stress, to limit loss exposure.

Once analyzed, trended data can yield insights that allow lenders to offer products and terms that keep their best customers and acquire new ones who stick around longer.

By looking at trade-level fields, lenders can identify critical factors, such as consumers who are big spenders and who pay off their purchases every month. And because historically we know these individuals like to be rewarded for their spending with airline miles, rebates, or other incentives, lenders can target programs, promotions, and higher credit lines more effectively.

By contrast, an inability to personalize products and offers that satisfy the demands of these transactors can lead to weak performance manifested by poor response rates, booking rates, and activation rates, as well as early attrition.

When historical credit information is mined effectively, it also can provide a more accurate assessment of future behavior, such as predicting balance transfer activity. Card-to-card consumer balance transfers are estimated to be between $35 billion and $40 billion a year, which presents a sizable growth opportunity for lenders who wish to proactively increase revolving product lines or simply retain current customers.

Predictive analytics can combine past balance transfer history; historical transfer amounts; current trades carried and used; payments and spend to identify consumers who are most likely to transfer a balance in the future. The ability to target in this way looms especially large given the high cost of direct marketing.

Unlocking the power of trended data

By using predictive analytics, lenders can leverage trended data to assess consumer credit behavior over time and obtain a more complete understanding of the bigger picture.

While the sheer volume of trended data is overwhelming on its own, when clarified through effective analytics, it provides lenders with an opportunity to improve performance.

About the author

Paul DeSaulniers is senior director, Risk Scoring and Trended Data Solutions, at Experian.

back to top

Sections

About Us

Connect With Us

Resources

WEBINAR

Mitigating loss: Understanding the fraud triangle

Time/Date: Wednesday, December 11th, 2024, 2:00 ET

Fraud continues to be top of mind for bank executives, with hard dollar losses growing at an all-time high.

In this session, we will discuss the fraud triangle and gain valuable insights into the psychology behind fraud, and the tangible and intangible losses incurred due to fraud schemes.

You will come away with a comprehensive understanding of how the fraud triangle applies to your customers, various types of fraud affecting community banks, and actionable steps to mitigate their impact.

REGISTER NOW!

This webinar is brought to you by:

Abrigo logo

Banking Exchange logo