Want the Secret to Approving More Loans for Small Business Customers?
It involves a new type of credit risk modeling called NeuroDecision™ Technology
Is your organization working to strengthen its current small business lending strategy? Are you searching for expanded financial insight to help approve more loans for small business customers? Is small business growth a priority for your organization in 2017? If you answered “yes” to any of these questions, improving your decision making processes is key to growing this segment of clients.
To confidently extend capital to more small businesses, you have to understand the potentials of opportunity and risk. In a prior article, we explained this can mean drawing on multiple data sources—beyond traditional business credit—to create a big-picture view of these businesses.
Yet, for insights that drive optimal predictability, you also need a special type of model—one that can be applied to a risk environment. It’s called NeuroDecision™Technology (NDT), and here’s why and how it can provide improved transparency and predictive power.
Linear versus non-linear modeling
When it comes to analytic models, particularly credit risk models, businesses have long relied on the tried-and-true logistical regression model. The main reason? It’s explainable.
Industry best practices and financial laws and regulations affect how credit risk model outputs are interpreted. In other words, you must be able to track the analytic process and prove how you got from point A to point B. Logistic regression provides you interpretable regression coefficients for each attribute included in the model so you know what increases or decreases the likelihood of a credit risk default.
As simple and effective as it sounds, these models aren’t perfect and can have disappointing consequences for financial service providers. Specifically, credit-worthy applicants who may be hovering below the line are denied, while applicants who are above the line, but trending toward default, are approved.
Then non-linear neural networks came along. Neural Networks allow for an exhaustive combination of model inputs across multiple models so that you can more accurately predict an outcome. The net result of this is that you get predictive improvements, but transparency can suffer as the machine is creating the weights.
The problem is, it’s almost impossible to explain how neural networks generate their outcomes or predictions. And, as mentioned earlier, models used in many credit risk decision environments must be explainable.
Until now, there simply hasn’t been a better solution.
Enter: NeuroDecision Technology
As a patent-pending, advanced hybrid of neural networks, NeuroDecision Technology was specifically built by Equifax data and analytic researchers to resolve: 1) the opportunity and risk gaps associated with regression models; and 2) the lack of interpretability associated with neural networks. The result is an industry exclusive model that’s highly configurable, granular and predictive, yet it also offers actionable reason codes that are easily interpreted to industry standards for analysis.
What’s more, internal Equifax comparisons with legacy-only solutions show that when fueled by NeuroDecision, you can experience up to:
- 30% projected lift in ability to classify good accounts and defaults
- Double the scoreable population
Here’s the first takeaway for financial service providers. NDT is a new, game-changing model that can help you extend capital to more small businesses by enabling you to more confidently serve a greater number of creditworthy businesses, without taking on added risk.
The second takeaway is this: NDT is only available through the Commercial Financial Network (CFN). In addition to our advanced NeuroDecision model, the CFN also gives you direct access to the industry’s fastest-growing commercial database which includes:
- Over 28 million business records, intelligently linked and growing daily
- Five major data categories, expanding well beyond the traditional credit bureau model
- Over 1,300 commercial payment data contributors
Equifax’s new tools can help small businesses grow
As you contemplate ways to expand your small business insight, NeuroDecision gives you the lens to see that businesses that may have historically registered as a good credit risk are actually demonstrating poor performance which could result in delinquency. NeuroDecision analytics when coupled with a network of diverse data assets, not just one source of payment performance, provide insights that drive better commercial credit decisions.
For more information about NeuroDecision Technology, click here to watch a quick video.
At Equifax, our goal is to help grow your business and optimize the value of your customer relationships by giving you a wider view of your customers across the account lifecycle. For more information about the many ways we can help your business, please visit: http://www.equifax.com/business/business-services.
Recommended For You
Why Constrained Models are Optimized for Credit Risk Credit scores are used as a big part of a lender’s credit […]
Why Proxy Model Explanations of Neural Networks Inevitably Fail For financial service providers, credit risk modeling is all about identifying […]
This question sounds absurd. But this is one of the most important questions lenders must ask themselves when making consumer […]
Banks and lenders want to deliver broader financial inclusion that enables them to score and approve more customers. But they […]