Much ado about delinquencies
It’s no secret that delinquencies are a huge issue in the credit union industry right now. Auto delinquencies 60+ days past due increased to 1.33 percent in Q1 2024, up from 1.19 percent last year. The monthly debt burden for consumers grew 32 percent from 2020 to 2023 — nearly double the rate of inflation (18 percent). And to make matters even worse, delinquencies on credit cards are higher than they’ve been since the Great Recession, setting off alarm bells for credit unions, their members, and the NCUA.
The traditional, “way-we’ve-always-done-it” reaction to this news is to hold back. Lend only to A paper. Restrict, be careful, and reduce risk. But there are a few problems with this strategy:
1. Inaccurate scoring methods
Over-relying on top credit tiers through generic scoring models leads to inaccurate, biased decisioning — which is not very useful for mitigating risk or finding good borrowers in other tiers and giving them access to credit. Also, focusing primarily on A paper hurts members. In tough economic environments, it’s critical that all members are still given a fair shot.
2. Inconsistent manual reviews
Another characteristic of using generic models is that, typically, you’re required to manually review a lot of applications, especially those in the middle tiers. Manual review can only look at so many factors, is often inconsistent, and is not an effective way to reduce delinquencies, particularly when it’s a significant percentage of your application volume.
3. Ineffective credit policies
Lastly, credit policies could be harming, not helping, your assessments and decisioning. They might layer additional inaccuracy and increase unnecessary manual reviews, introducing more opportunities for inconsistent decisions and risk that slips through the cracks.
Fortunately, credit unions have already found another way to combat delinquencies with precision and speed. AI-automated underwriting has been proven to reduce risk by providing deeper insights that lead to far more accurate assessments than generic models, and automate more decisioning that ensures consistency in accuracy.
AI: credit unions’ secret weapon
By now, you might know that lenders can use AI to increase approvals and automate decisions. AI-automated underwriting decisions at least one in six U.S. credit union members*, and this number will only grow with time. While increasing approvals and automating decisions is still incredibly useful for your credit union, it’s only one application of AI. AI-automated underwriting is a flexible tool that allows lenders to toggle their risk and approval rates to accommodate economic fluctuations.
Underwriting with best-in-class machine learning models allows your credit union to be two to four times more accurate than generic scoring models and reduce risk by 20 percent or more, holding approvals constant.* These metrics are a game-changer for the industry, using more data and better math to get ahead of the status quo.
For instance, Commonwealth Credit Union ($2.2B in assets) is a great example of how credit unions can adopt AI to achieve impactful results. Commonwealth uses AI to decision 70-83 percent of all consumer loan applications, leading to significantly lower delinquency and charge-off ratios than their peers. According to Jaynel Christensen, Chief Growth Officer, relying on AI-automated underwriting allows their credit union to stand out amidst the current economic environment.
“With climbing delinquencies and charge-offs, Commonwealth Credit Union sets itself apart with 30-40 percent lower delinquency ratios than our peers.* Zest AI’s technology is helping us manage our risk, strategically continue to underwrite deeper, say yes to more members, and control our delinquencies and charge-offs.” — Jaynel Christensen, Chief Growth Officer
Read Commonwealth Credit Union’s case study and watch their success story video.
Credit unions’ battlefield strategists
Technology can do a lot for your organization, just like it has enhanced our own daily lives. But powerful AI technology without experts is like using a supercomputer for creating a Word document — there’s a lot of potential and razor-sharp intelligence left unused.
Implementing AI-automated underwriting helps your credit union make more accurate decisions, which will inherently decrease delinquencies. But to implement these tools to their fullest potential, you need a team of experts at your side to ensure you know where the potential is — from automation to risk mitigation and how to meet your goals effectively. This is a holistic, strategic endeavor that is not one-size-fits all.
Even if you have the best AI tool, credit policies can still derail attempts to reduce risk, especially if you are manually reviewing more applicants than necessary. An effective technology partner will help you define cut-offs and identify and remove non-performant policies and optimize others to ensure you’re mitigating risk at every step of the way.
Breaking the cycle
Delinquencies are high this quarter. In the next quarter, they might rise again. And who knows when unforeseeable events, such as a pandemic, will disrupt the economy and create changes that are tough to predict. No matter what happens, there will always be periods of economic booms and busts.
Some credit unions might view high delinquencies as an unavoidable consequence of the economy. Others have already gotten ahead of the game, adopting innovative technology that allows them to continually adapt, adjust, and optimize. The future of lending is flexibility, and the credit unions that can adapt and remain agile are the organizations that will thrive through any economic climate.
So, there is much ado about delinquencies… what’s your attack plan?
Footnotes:
