In the recent McKinsey article discussing designing next-generation credit-decisioning models they outlined four best practices for automated credit-decisioning models for banks as they continue their digital transformations. Digital lending based on high-performance credit-decisioning models, says McKinsey, lead to:
- Increased revenue
- Lower credit losses
- More efficient lending
Sounds good, right?
We work with lenders in a variety of markets and have found that decision modeling, building a visual blueprint of your credit risk decisioning, is a critical success factor in moving to next generation credit decisioning. Decision models graphically show how a business decision is broken down into successively more granular sub-decisions. They show what data each sub-decision needs, precisely. They are linked to the policies, regulations, best practices, and analytic insights that constrain and drive credit decision-making. And they show how you can orchestrate a wide range of machine learning, artificial intelligence and rules-based technology to address your decision-making needs.
Basing your next generation credit decisioning approach on a robust decision model, ideally one built using the industry standard Decision Model and Notation (DMN), will help you achieve all the best practices, McKinsey identified.
First and foremost, decision modeling helps you implement a modular architecture. As McKinsey recommends, a decision model focuses you on developing multiple sub-decisions and aggregating them into an overall decision. This lets you use different data types to create risk scores for specific segments or geographies and so leverage new data sources that might be available only for some borrowers. The decision model shows how to combine these models, allowing you to implement a modular credit decisioning architecture. Here at Decision Management Solutions, our DecisionsFirst approach results in modular, analytic models that are focused, quicker, and easier to build and update, as well as more transparent. And a visual decision model is a vital tool for coordination, allowing business, risk and IT teams to all see how the sub-decisions are combined into an overall credit decision.
Decision modeling focuses a bank’s investment on new data sources. McKinsey strongly recommends that lenders work to expand their data sources, both internal and external, to supplement the data traditionally used for credit risk decisioning. Many such efforts are data-first, spending money to collect and manage data that turns out not to be useful. A DecisionsFirst approach focuses on what kind of data – what kind of prediction – might cause the decision to be made differently. Then new data sources can be explored to see if they can support such a prediction in the certain knowledge that it would change decision-making if it were found.
As this new data is pulled in, McKinsey points out that its critical to mine this data for credit signals and to use it to segment customers and prospects in new ways. It’s also critical to design and run experiments, champion/challenger as well as A/B testing, to see how robust these new signals are. Designing this so that the business owners can still understand what’s going on – preventing this from being implemented in a big pile of code – requires a model-based approach and a decision model is perfect, allowing experiments and different approaches to decisioning for different segments to all be shown visually.
Finally, and most importantly, McKinsey wants banks to leverage their business expertise not just their data. Our DecisionsFirst approach builds the initial decision model by listening to business experts. This puts all the data and analytic work into a strong business context and the visual model ensures that these experts can stay involved through implementation, simulation, experimentation and continuous improvement.
Decision modeling is the best way we have found to think about credit decisioning, indeed about any operational decisioning, and you should be adopting it now.
Reach out today to discuss how we can help your business.
James is the founder and CEO of Decision Management Solutions. He is the leading expert in how to use decision modeling, business rules, and analytic technology to deliver Digital Decisioning. James is passionate about helping companies improve decision-making and effectively adopt advanced analytic technology. He provides strategic consulting to companies of all sizes, working with clients in all sectors to adopt decision making technology. James has spent the last 20 years working in decisioning and has led Decision Management efforts for leading companies in insurance, banking, health management, and telecommunications.