How the pandemic changed credit modeling
How the pandemic changed credit modeling and why it’s time for a change
By Ibo Dusi, Happy Money Chief Risk & Revenue Officer (Prepared for financial professional use.)
When I arrived in the U.S. I came with a good education, a stable job and savings in my bank account, yet it took me several years to establish my creditworthiness because I was starting from scratch in this country. This experience is not unique. Millions of Americans struggle with creditworthiness for reasons they are unable to control, be it their immigration status, age, or lack of access to lending. Moreover, the signals traditionally used to determine creditworthiness — primarily past responsible usage of credit — are backward-looking, narrow, incomplete, and have little to no emphasis on their actual income or spending patterns. This traditional approach to risk modeling can generate a “cold-start” problem given that FICO, by its design, lacks a forward-looking, broader, more holistic picture of an individual’s credit worthiness. It’s time for that to change.
The unprecedented macroeconomic crisis that began in 2020 left tens of millions of U.S. citizens unable to go to work and earn their regular incomes, which only increased the number of people who may not have access to affordable credit. Lockdowns resulted in the government, financial institutions, and fintechs stepping in to provide liquidity and payment relief. While this approach helped people in financial distress and mitigated damage to the country’s economy, it also created an unintended consequence: The meaning of credit data and scores changed dramatically. The backward-looking approach fell apart during rapidly changing times, making it more difficult to truly assess borrowers at an individualized level and affecting many people in ways similar to my experience.
As an example, prior to 2020, receiving a forbearance signaled financial stress. And yet, in this past year, a number of consumers chose to delay their mortgage payments not because they lost their job or could not afford their mortgage, but rather due to the overall economic uncertainty and a desire to conserve cash. Additionally, while FICO scores reached all-time highs in 2020, consumers also lost sources of income and relied on government assistance, which created some confusing signals for lenders who use only traditional methods of measuring creditworthiness.
Throughout 2020, the Happy Money team relied heavily on real-time cash flow analytics and income verification to truly assess income and employment stability during the pandemic. This kept the pipeline of creditworthy borrowers full, and we were able to help them refinance their debt and free up cash flow. Our analysis indicates that real-time cash flow analytics and income verification is not only helpful during times of economic stress but also in benign credit environments. This is good news as we approach a turning point in the macro environment.
At the beginning of the pandemic, Happy Money developed what we called a “turning point monitor,” designed to support our risk and analytics team by signaling what we would consider the end of the pandemic. Using data points like unemployment, consumer sentiment, and other economic health signals, we stayed in the red for all of 2020. In March 2021, less than a year after the low point of the pandemic, this turning point monitor finally moved (and has stayed) in the “green” for the first time since the onset of the pandemic. As we look across the country, we are seeing more Americans getting vaccinated and mask mandates lifting. And analytically speaking, we expect to see a shift towards a post-COVID “new normal.”
People helping people
While we aren’t out of the woods yet, things are looking up. And as financial institutions, we’re faced with a great opportunity to incorporate new, more holistic methods of evaluating creditworthiness into our credit modeling long-term. We can’t wait for the environment to return to “normal”; instead we must take underwriting to the next level — for our consumers and our businesses. For Credit Unions in particular, there is a significant opportunity to double down on their charter and seek underwriting methods that foster greater inclusion, transparency, and access throughout the lending process.
In addition to the cash flow analytics and income verification Happy Money implemented in 2020, there are several additional data sources that have proven useful in assessing creditworthiness, including savings and investments, rental and utility payments, and education. With the right modeling, lending institutions can automate this process and more efficiently incorporate these data sources into creditworthiness assessments.
At Happy Money, we use proprietary data and machine learning to better predict ability and willingness to pay. We’ve even created our own measurement of creditworthiness called the Happy Money Score, which helps underwrite a superior asset and enable us to say “yes” to more consumers. In the first quarter of 2021, Happy Money’s portfolio had an average FICO Score of 703 yet had a Happy Money Score of 771. This score is matched to FICO in terms of risk, which means that even though the average FICO score was 703, the portfolio had the same risk level as a portfolio with average FICO scores of 771. By looking at our applicants more holistically, we have been able to find members who have been overlooked by other lenders and banks and offer them better terms. This is one example of how augmenting FICO with additional data, coupled with diligent rebuilding and refitting of credit models, can drive better outcomes.
Innovation is the new normal
As we continue to operate in a rapidly-changing and unprecedented environment, here are some strategies any lending institution can consider employing:
- Augment traditional underwriting data with alternative data sources
- Incorporate “verified” cash flow into your underwriting
- Monitor model performance closely
- Rebuild and refit models more frequently
Much like many aspects of our lives before 2020, we cannot go back to the old ways of lending. Now more than ever, we are seeing just how much the methods of determining one’s creditworthiness lacked innovation and left out certain segments of consumers. Given the changing backgrounds and needs of U.S. consumers and the ever-evolving economic environment, now is the time to innovate how we as financial institutions evaluate creditworthiness. Not only will this benefit consumers by opening up credit opportunities for them in increasingly inclusive and transparent ways but it will also open up growth opportunities for financial institutions and further speed up the economic recovery by encouraging consumer confidence and unlocking pent-up demand. As Happy Money’s Founder Scott Saunders would say, doing the right thing pays off.