Getting Out from Behind The Curve
Chapter 1 of this series considered the importance of establishing a specific goal to solve using data analytics and proving the ROI in order to justify automation and decisioning using business intelligence in a credit union. Chapter 1 also highlighted two real use cases of success credit unions have had using data analytics to solve real-world problems. According to a recent study conducted by Best Innovation Group (BIG) and Trellance, 45 percent of credit unions don’t currently have a strategy in place, and those that do have a strategy still say it will take three to five years to implement. Credit unions that aren’t making the most of data analytics today could be in even bigger trouble if an economic downturn occurs, as some economists are forecasting. “As we go forward there will be a significant performance difference between those that have invested and those that have not,” says Kirk Kordeleski, senior managing partner at BIG. “We think any downturn in the economy will highlight the advantage that data-oriented FIs will have over their competitors.”
The Cost of Big Data Analytics Implementation
The survey revealed that more than half of the 85 credit unions surveyed have budgets in place for data analytics. Of those, one-third plan to spend more than $200,000, the other two-thirds plan to spend between $50k to $200k. In addition, credit unions need to consider on-going costs. A rough rule of thumb is that a CU with $500 million in assets should budget between $150,000 and $300,000 per year for three years to cover software/hardware, analytic applications, and strategy. Smaller credit unions can find some savings by relying on a CUSO to provide the analytics and associated services.
The following paragraphs are real use cases that credit unions have shown to prove out their investments.
Use Case #3: Predicting Loan Delinquencies in Credit Unions
Joel Hartzler of the Filene Institute conducted a research project that used data analytics to predict delinquencies. Loans are typically decided based on FICO score. The goal is to give as many loans as you can, but only to those who you are pretty sure will pay you back. Filene wanted to show that using multiple factors can make better decisions. They studied loan originations over a year across several credit unions and used DTI, marital status, age gender, income, about 20 factors including credit score. The result was that using multiple factors could result in 1% more loans, but more importantly, reduce delinquencies by 15%. When asked if using age and gender would invite scrutiny and possibly discrimination lawsuits, Joel did stress that this was a research project to demonstrate the potential of predictive analytics. The combination of predictive factors would not necessarily be the ones used in the real world.
Use Case #4: CLIP
First Tech set out with a number of goals: it wanted to improve the profitability of its credit card portfolio by increasing the line of credit for some members, it wanted to avoid additional unnecessary risk, it wanted First Tech-branded cards to be top-of-wallet, and equally important, it was imperative to meet all regulatory requirements. (For example, the Card Act requirements include assessing a borrower’s current income and debt to show ability to pay). The credit union incorporated data from several different sources, including data from its core together with data from external reporting systems such as from a credit reporting bureau. According to Sandi Papenfuhs, Senior Vice President of Consumer Lending, this data analytics usage resulted in two positive results, one intangible and one very tangible. The member experience was greatly improved for those receiving a credit line increase, which hopefully encouraged top-of-wallet usage. More tangible were the financial results. First Tech issued a total of $400 million in credit line increases, resulting in a 16% spend increase. As a result, the credit union expects a non-interest margin improvement of over 17% in the first two years of the program.
This is Part 2 of the Data Use Cases for Credit Unions series. If you missed Chapter 1, read it here.
If your credit union is seeking to make this discipline a reality, contact the data analytics team at Trellance. Contact us and get the power to use rich data to guide your business decisions.