In nearly all areas of business, there is typically a positive correlation between efficiency and growth, and credit unions are no exception. For instance, the goal in member business lending (MBL) is for credit unions to push through as many credit requests as possible without increasing staff, jeopardizing credit quality and without delaying the approval process. This is challenging, especially as credit unions begin to grow and diversify their loan portfolios following the new principle-based MBL guidance and recent interest rate hikes.
According to
To keep up with demand, using automated decisioning is an excellent way to gain efficiency during the underwriting process while developing sound policies and controls that reduce risk. Whether a credit union already has a concentration of member business loans or will be starting a new lending segment, auto-decisioning can help credit unions manage the increased loan application volume.
While auto-decisioning is a proven tactic for streamlining the lending process, there are several factors to consider before moving to an “underwriter-free” decisioning model. One consideration should be the credit union’s risk culture. Management should determine whether the credit union is willing to trust the decision-making algorithms and let the system verdict stand. If the team is not comfortable with this, then the credit union is not ready for auto-decisioning.
However, there are steps that can help a credit union progress toward auto-decisioning. The initial step involves standardizing products and pricing. The credit union should offer a limited group of products and dollar thresholds that require no financial statement. From there, the credit union can build comfort with the product set, pricing and scoring methods before implementing a truly automated decision strategy.
Once a credit union becomes familiar with standardized products and pricing, and feels comfortable enough to refrain from touching each credit request, the credit union can move forward with minimizing the human element in the lending process. It is important to understand that auto-decisioning cannot happen on day one, as it is driven by data analysis, and data is required to establish auto-decision parameters. Each credit union also varies in terms of risk tolerance – a good score to one credit union may be too risky for another. This information must be analyzed to set the decisioning parameters. To gather the relevant data, a minimum of 300 loan decisions must be analyzed, and this only includes approvals and declined applications, not renewals. This means that if a credit union is starting a new business lending segment, they must process at least 300 business loans to accumulate data and develop decisioning parameters that mimic the underwriter’s judgement. It is crucial that the automated strategy is managed in a way that reflects a credit union’s manual underwriting processes and is compliant with all regulatory guidelines.
Still, auto-decisioning is only applicable for certain loan types and dollar amounts, as higher dollar amounts will likely require manual review. For a credit union new to an algorithm-led decisioning process, it is advised to have the system start with loans at $50,000 or less within a limited product set, such as credit cards, overdraft lines or other small lines of credit. The dollar threshold should be set to an amount that the credit union feels comfortable with the system making the final decision.
Once the parameters are set and loan applications are being processed, credit unions should regularly review their auto-decisioning portfolio. Auto-decisioning is not a “set it and forget it” endeavor. While it can maximize efficiency, it is still important to consistently monitor the system so it aligns with the credit union’s risk management policies. Any reports that are compiled for a credit union’s portfolio, like delinquencies, should also be compiled separately for any auto-decisioned loans. The objective is to ensure that the auto-decisioned portfolio behaves the same as the manually-underwritten portfolio.
With the new MBL guidance and additional interest rate hikes expected in 2017, credit unions will have more opportunities than ever to grow and diversify their loan portfolios. To fully capitalize on the current market environment, credit unions will need a solid auto-decisioning strategy in place to manage the increased volume of loan applications while mitigating risk and delivering top-notch service to members through faster decision times.
