Why It's Time For Credit Unions To Start Embracing AI Like The Big Banks

Subscribe to Our Blog

The future of lending lies in machine learning. ML can help lenders increase originations, reduce charge-off risk, and more effectively market to their customers. But so far, the vast majority of early adopters have been large financial institutions, while many credit unions and other small lenders believe the complexity, costs, and skills necessary to incorporate ML into their operations makes it a hard to justify the business case.

Consider the findings of a recent Fannie Mae survey of senior executives from nearly 200 lenders. Although executives at more than half of the larger institutions (52%) expected to be regular users of machine learning or artificial intelligence technology in the next two years, only around 15% of credit unions and other smaller institutions had similar adoption plans.

...the results are every bit as powerful for smaller lenders as larger ones, if not more so.

Our experience suggests that ML can benefit institutions of all sizes. In fact, the results are every bit as powerful for smaller lenders as larger ones, if not more so. Nonetheless, we often encounter a degree of healthy skepticism from those worried about the risks. So, let me share with you what I call the “Four Wonders of the ML World” – the four biggest concerns that I hear most often from executives at smaller lenders and why you should feel more at ease.

I’m wondering even where to begin

When we meet with prospective clients, we find that most executives are not so much fearful of machine learning as they are overwhelmed. The Fannie Mae survey, for instance, found that nearly 1 in 5 credit union executives (20%) said among the biggest reasons they had not adopted ML was that they didn’t know where to start.

That’s completely understandable. Not only is ML a new technology, but the marketplace is fraught with hype. The truth is that ML can help you achieve impressive results by allowing you to use far more powerful models to identify the most creditworthy borrowers. But it doesn’t have to disrupt your business. Just think about an ML model as a turbocharged version of the traditional credit decisioning scorecard that you use today.

Another worry that small lenders especially have are concerns about the quality and availability of their data. Nearly 1/3 of all credit unions executives (31%) cited it as a significant hurdle. The reality? ML is about making better use of the information that you already have by capturing the complex interactions between different variables. That’s why more traditional sources, such as credit bureau and borrower application data, are usually far more compelling than most alternative sources, like social media “likes” or Spotify playlists, which capture the imagination. The more data obviously, the better. But if you have at least 100,000 records or 1,000 defaults, you probably have enough data to start.  

I wonder if I need a fancy IT infrastructure to put my ML model into production

For most clients, the complexity of integrating an ML model into their existing loan operating systems and technology infrastructure is also a significant concern. The Fannie Mae survey, for example, found that 43% of credit union executives rated this among their two biggest roadblocks to ML adoption. Traditionally, many institutions invested heavily in developing ML models, only to struggle to get them into production because of the technical challenges and tough compliance requirements. What’s more, it was incredibly disruptive to most organizations’ IT roadmap.

It’s time for credit unions to stop wondering and start incorporating ML into their core business operations.

Today, that is no longer the case. New tools and more powerful automated approaches have accelerated the model development process. Institutions have greater flexibility, too. The latest model development tools are built to work in the cloud and on-prem, offering an array of deployment options. What’s more, it often can be helpful to have an experienced partner by your site.

In fact, we’ve found that when our modelers and programmers collaborate with our clients, we typically can expedite the ML model development so that it takes less time than the process for implementing a traditional scorecard.

I wonder if I am going to need to hire a small army of whiz kids to explain model results

The arms race for the best technology talent is intensifying – not surprising when nearly a quarter of financial executives across all institutions cited a lack of necessary skills as a major impediment to ML/AI adoption. Traditional financial institutions must compete with Silicon Valley stalwarts, like Google and Facebook, as well as a raft of fledgling fintech companies to hire the best data scientists. In fact, several of the biggest banks have announced plans to spend billions of dollars in ML/AI as they build out entire data science and modeling teams. Most smaller institutions simply don’t have the resources to make such an investment.

Here, an outside partner which can fully support the model development and maintenance processes, or collaborate with your existing team, can be hugely beneficial. So, too, can the proper technology. Advanced model building technology means ML models are being built to be fully interpretable, so you won’t need a roster of whiz kids. Interpretable models  mean you should be able to rely on easily digestible readouts from your model that can explain how it arrived at every credit decision for every borrower – an even provide accurate adverse action reasons as well.

I wonder if my ML model will pass muster with my board and regulators

The last major concern that I commonly encounter regards whether this new technology is safe to use. Although this received the lowest number of responses in the survey – only about 9% of all financial executives and 6% of credit union executives rated it a significant hurdle – it perhaps has the most significant consequences. Simply put: failure here is not an option.

The good news is that ML no longer needs to be a black box. Advances in complex math mean that today’s ML models can make credit decisioning fully transparent and explainable. And automation can expedite the documentation and validation of underwriting models. Since you can fully explain, document, and validate the way your model arrived at every lending decision, you’ll be able to satisfy all internal risk reviews as well as regulatory compliance requirements.

Stop Wondering. Start Implementing  

It’s time for credit unions to stop wondering and start incorporating ML into their core business operations. The software has caught up with the concerns that have historically slowed its adoption and is poised to transform how lenders of all sizes market, underwrite, and service their customers. Credit unions have the chance to embrace this game-changing opportunity; they shouldn’t be left behind.

This post originally appeared in the Credit Union Journal.