Credit Unions powered by AI can help their members weather the storm

Author: Scienaptic Research


Many economists and researchers were warning us about the next recession to hit us by the mid-2020s. And right when we started experiencing an economic slowdown, we are trapped by the novel coronavirus (COVID-19). The businesses were not prepared for the ripples caused by the effects of coronavirus and measures to contain it. Many companies are taking a huge hit and are trying to cut costs by laying off people and reducing overheads. In fact, on April 3, 2020, the U.S. Bureau of Labor Statistics released a news release detailing the unemployment rates for March 2020.



Till date, over 22 million people have filed for unemployment as more and more States enforce lockdowns. The unemployment rate is now the highest for over 90 years. Like it or not, the world we are living in today has been changed by the effects of COVID-19, and it will take months before we see some economic adjustment. It's bringing the recession at our doorstep. Bank of America states the recession is already here and could be the deepest recession on record. According to the bank, they see the first-quarter GDP sinking 7% and continuing that decline with a 30% drop in the second quarter. Besides, they expect a loss of 20 million jobs due to the coronavirus crisis sending the unemployment rate to a high of 15.6%.

Like every recession, along with job losses, we might expect expenditure slowdowns and difficulty in repayments. But all is not gloom and doom - economists such as Steve Rick, Chief Economist for CUNA, believe that the recession might be an opportunity for Credit Unions (CU) to increase their market share, not just survive but thrive in the downturn. We saw how credit unions responsibly increased the lending opportunities for local businesses and families in the great recession of 2007-2009. Credit unions are said to be countercyclical financial institutions that help the local economies by extending more and careful lending opportunities in the economic downturn. Their emphasis on value creation rather than profit maximization helped them gain many members in this financial crisis. In many regions, the fear of liquidation of banks led to an appeal for credit unions amongst Americans.

The role of Credit Unions ahead is enormous - roughly one-third of all Americans belong to credit unions. In 2008, there were 89.9 million credit union members that grew to 117.8 million in 2018, i.e., a 31% increase in 10 years. The increase in credit union membership suggests that many Americans prefer credit unions over other lending institutions. And rightly so, as the credit unions have always been not-for-profit institutions that have excellent member relationships, competitive rates, and their focus is perpetually on giving back to its members. According to CUNA's official website, credit unions provide the same products and services as banks but with unmatched service, convenience, and cost. And they even focus on returning earnings to their members instead of investors.

However, lending during economic slowdowns tends to be tricky. There will be a constant tension between the mission of serving members and prudent management of financial resources. It will lead to careful lending decisions and might mean a loss of credit. In times like these, financial institutions tend to lose the income generated from loans because consumers usually can't afford or are unwilling to borrow as much. They will be more cautious about their lending decisions, and that will mean giving out lesser credit than expected. It will lead to significant pressure on members, causing dissatisfaction amongst them. The risk managers will try to avoid any undue risk, and that could result in an immediate drop in approval rates in the hope that when the recession hits, then losses will not be too bad. Some risk managers might eventually stop lending. While that may save credit unions and other financial institutions some risk, it will lead to jeopardizing member relationships for the future. But is there a better way for credit unions to come on top of the recession?

We believe there are better frameworks available today to make lending businesses recession-resistant and member friendly. Artificial intelligence and machine learning-based techniques are at the heart of these frameworks, and they have been tried and tested for the last five years. They will help you prepare better so that you can make the right decisions for your members and utilize the funds optimally. Their ability to create sharper differentiation between credit-worthy members and high-default risk applicants will help you immensely in serving deserving members in their hour of need.


Yet we see only 5% of credit unions incorporating newer technologies such as AI/ML in their lending business. Many credit union executives still worry about AI/ML replacing traditional underwriters. In contrast, others think it's not very easy to implement these solutions in their ecosystem, mainly because their data sets are not robust enough. Credit Unions often don't have vast and integrated data repositories such as larger financial services firms. And they are sceptical whether their efforts and investments will ever pay off. But AI/ML-based models can manoeuvre through these challenges very swiftly and efficiently.

We all know how standard logistic regression models helped in producing the most resilient loans before 2008. However, AI models are promising a better path for optimizing portfolios for good times and bad. An accurately monitored machine learning model, trained on a broader data set, has a better ability to detect shifts in credit quality. CUs can now use complex AI models that rely on hundreds of more features to assess risk. These new and better technology tools can analyze your overall portfolio risk with a finer lens and use those insights to, for example, adjust the price on incoming borrowers to optimize yield heading into a downturn.

There is every evidence that the algorithms to predict customer behaviors will continue to evolve. But this transition requires collaboration between Credit Unions and decisioning technology providers. Implementation technology is now available to allow this transition; it just needs to be brought into practice. This will help to make sure that better decisioning instruments have a much faster time to market. Member expectations and behaviors are changing rather rapidly. It is only natural that predictive models be revised more frequently.

The minimum expectation from incorporating AI into credit decisioning are as follows:

  • Grow portfolios & lower risk

  • Invest in best customers (CLI) protect your business as needed (CLD)

  • Reduce G2M by 75-90% for new credit instruments / products / strategies

There are genuine concerns on AI/ML usage, but the consensus is quickly emerging that most concerns can be empirically addressed. These are times when we must understand and learn from the experiences of successful AI for credit implementation so that we can take this evolution to fruition and prepare for adverse times.

78 views

Scienaptic Systems Inc.

50 West 47th Street, 16th Floor, Ste. 1614, New York, NY, 10036

Ph: +1 212 244 4030

  • LinkedIn - Black Circle
  • Facebook - Black Circle

© 2019 Scienaptic Systems Inc. All rights reserved. | Grievance officer | Privacy policy Terms of use | Mobile App Terms & Conditions