Three steps to help CUs identify potentially problematic business loans earlier in the lending cycle.
By Erick Smith and James Hartzog
Sponsored by D+H, based on its three-part ebook series
Making the right decision on business loan approvals is only the first step in mitigating risks in your MBL program. You also need the right early warning system to monitor your business loan portfolio and reduce risk once loans are on the books. Traditionally, financial institutions would work to reduce risk after one of two events occurred: 1) loans or borrowers were placed on the watch list after an undeniable trouble indicator (such as being 90 days past due), or 2) a random review resulted in a credit-rating downgrade on a loan; this was great for the loan in question, but left many others unconsidered. The good news is that a rules-based early warning system can help identify more borrowers at risk of default earlier in their loan’s life. Being rules based, this model (or set of models) can be automated, so a meaningful portion of the portfolio can be reviewed much more frequently, increasing a credit union’s ability to identify risk and respond accordingly. Even more good news: The technology now exists to make developing such a system possible for credit unions of all sizes. As recently as five years ago, implementing a rules-based early warning system would have been possible only for large institutions with considerable resources. But today, most loan origination and risk management system vendors have improved their offerings so the consideration has become: “How do we use our tools to get this job done?” Here are three key steps to follow when developing a rules-based early warning system for business loans:
- Determine the loans to include in the analysis. The sample size needs to be large enough to be statistically reliable, but small enough to be manageable, and should include both performing and nonperforming loans. It should also include all business segments, as each will likely have different warning characteristics and therefore different risk models.
- Identify the data that illuminates the warning characteristics. This activity is straightforward for such hard data as overdrafts and late payments stored in the core system. More problematic is “judgmental” data. For example, many credit policy manuals contain the statement “Borrower is required to maintain strong cash flow,” but don’t define mathematically what this means, making it difficult to measure. The business loan relationship manager will be in the best position to develop “soft” data (such as knowledge of the business owner’s personal problems) that doesn’t appear in a financial statement or recent payment history. Additionally, the data set for analysis will need to incorporate external information such as negative news. Ideally, the data will extend back 12 to 36 months.
- Collect and store the data as a routine course of business. The credit union will need to collect and store the necessary data, or be prepared to do so.
A rules-based early warning system can help reduce losses by automatically reviewing your business loan portfolio and identifying the short list of default candidates. Reviewing your portfolio can also help identify areas of decreased risk, such as specific borrowers, industry segments or product types, to provide leads on additional lending opportunities for your credit union. Erick Smith is lending solutions marketing manager and James Hartzog is senior commercial lending expert and product manager at D+H. Get more information about the D+H portfolio of lending solutions online or by calling 800.815.5592. Learn more about CUES’ business lending schools.
