Exclusive report on the state of credit modeling in 2020

The lending industry is in the calm before the storm says a first-ever benchmarking study on credit modeling from Cornerstone Advisors (commissioned by Zest AI). The global virus crisis and ensuing consumer recession have scrambled the reliability of traditional credit scores. That makes getting the art and science of credit decisioning right more essential than ever.

Lenders are only now ramping up the necessary investments in people and technology to get ahead of what could be a bumpy recovery. Many lenders want to shorten their model review times: About half the lenders surveyed say they plan to change what’s now a semi-annual review to a quarterly review, and some plan to go to monthly reviews. Many want to drive higher levels of automation, as well, when the recovery hits. But are they ready to execute quickly and nimbly with the aging tools and small teams they have?

The results are mixed. Most lenders have only a couple of FTEs dedicated to modeling. Only ten percent of financial institutions with more than $1 billion in assets (and 20 percent of smaller lenders) say they’re “very confident” in their models’ abilities to maintain their effectiveness post-COVID. Only 15 percent of lenders with more than $1 billion in assets said they are “very prepared” to adapt. Most lenders, especially the bigger ones, admit that it’s somewhat or very difficult to make updates to their credit models, a process that can take many months.

There’s also a lot more progress to be made on automation in loan decisioning. Among smaller lenders (assets below $1 billion), 37 percent don’t auto-decision at all, in contrast to nine percent of larger lenders. Even among the big lenders, only a quarter of them is auto-decisioning more than 50 percent of the loans they process. (And regardless of asset size, there is still a hefty amount of manual review).

The consumer credit market is changing too fast to be stuck with a slow and manual modeling cycle and decisioning operation. Are your automation levels where you want them to be? Do you have the right tools and technologies to adapt as the industry comes back from the pandemic?

Download this inaugural survey of industry credit modeling practices to find out where you stack up on all the important metrics:

• Approaches to credit modeling and automation
• Model build and deploy cycle times
• Attitudes on model performance and the use of alternative data
• The impact of COVID-19 on models
• Use of AI and machine learning in underwriting
• PLUS: Conclusions and recommendations from the experts at Cornerstone.

‍

Get the full report here.