From 2008-2023: How Recessionary Periods Put a Spotlight on Credit Scores
No matter the industry, the last few years have without a doubt been challenging for businesses. From a global pandemic, to rising inflation, the business landscape has seen an uptick in volatility. Data from PwC found that 81% of executives are anticipating another recession to hit as soon as the next six months. With the prospect of recession looming, lenders are preparing themselves and their risk assessment practices by looking back at the recession of 2008.
But the comparison to 2008 doesn’t quite align with the circumstances we’re seeing unfold today. There’s something uniquely different about the underlying conditions for this impending recession. In a recent conversation, I had with Discover’s former Chief Risk Officer, Brian Hughes, he summarized it well, saying the previous recession began as a “bank crisis that became a consumer crisis, that became a problem for the government that they had to react to.”
What we’re faced with today has flipped that notion entirely on its head. And as a result, it actually points to a few positive signs for us, suggesting this recession may be shorter and shallower than what we all experienced in 2008. While uncertainty looms, lenders have arrived at a tipping point in how they approach today’s marketplace. The focus has gravitated toward credit building products and secured cards, igniting discussions around credit scores and whether they are the best way to determine a consumer’s ability to pay. Considering these new variables, let’s take a closer look at where lender concerns lie and the data sources that enable them to maintain growth without adding risk.
Setting the Stage
Credit scoring algorithms fully absorb and embody credit building and secured products. This means when you get a 650 credit score, it's the same 650 that anyone else would get, regardless of the products involved. With that fact, it should logically follow that a lender can trust a credit score. But in practice, in a volatile economic environment, these lenders want to gain an understanding of the underlying information behind that number.
With lenders wanting more insight into products, we start to see more scrutiny, particularly with regard to consumers that use secured products. This set of consumers may be higher risk and possess smaller income profiles. In the event that lenders decide to reconsider credit risk assessment measures, these consumers may be at the front of the line. There are several other signs that lenders watch for including credit card utilization and credit run-ups. On its face, utilizing one of these products shouldn’t end up being a detriment to the consumer.
In Europe, we’ve seen some means of relief – encouraging consumers to use their bank account information as a means of gaining access to credit. However, in the U.S., conversations have started heating up around alternative data sources. According to Nova Credit's recently unveiled State of Alternative Data in Lending Report, 75% of lenders believe that traditional credit data and scores don't deliver a complete picture of a consumer's creditworthiness. And as a result, most (59%) are turning to a variety of alternative data during the underwriting process.
As an example, let’s look at cash flow and bank transaction data. Credit gives a view on how likely a consumer is to make a payment and bank data shows us their income levels, balance, checking account stability and high value expenses. All of these together give lenders a wide range of insights to improve financial inclusion and bring in people that are locked out of the traditional credit field.
Taking a Deeper Look at Data
Recently, I attended a briefing on how credit data was used to manage risk during the pandemic. Over the course of discussions with policy makers, advocates, regulators and others in the lending space at the event, a number of solutions were discussed for the challenges observed in credit data. One of which was to incorporate the use of cash flow data.
These discussions brought back memories of the red solo cup discourse from around 5-10 years ago. In essence, this was when Facebook photos with red solo cups were tested and used as an indicator of risk. At the time artificial intelligence (AI) and machine learning (ML) techniques had become immensely popular and lenders started to see a path to start underwriting based on personal demographic information. A notion that sounds ridiculous in hindsight, the red solo cup phenomenon serves as an important example of how leveraging just one set of data points on its own can more likely leave you with an incomplete profile or more problematic, an incorrect profile.
Lenders need to gather information to capture the bigger picture. A need that has manifested in the discussions around using alternative data – things like rent and subscriptions, among others. Alternative data sources make for valuable resources, enabling lenders to create a more detailed picture of someone’s ability to pay. But the question remains, what other data is it paired with? And how is that data ranked?
Take for instance, subscription data. This data can prove useful if a consumer has, say, a Netflix subscription and their utility bill, but no other major payments. This serves as a sign for lenders that they are paying on time. With that information, lenders gain insight that can show where a consumer does well. But that picture is complicated if there are other payment responsibilities involved – i.e., does this person have a mortgage or car loan? If so, their ability to pay for a monthly subscription like Netflix will not hold as much weight. Alternative data itself is nothing new for leaders but put in the context of comparing now to 2008, the difference is practically night and day. Today, we’re unquestionably in a better position – one of the biggest shifts coming with receptivity to new tools and new data.
Underwriting with Cash Flow
As a long-time member of this industry, it’s fair to say I’m skeptical that there is some truly revolutionary piece of data that will emerge, like, the color of the flowers you buy, that will ultimately be of any use in indicating risk. The logical flow here is that how you paid your bill yesterday is likely how you will pay your bill tomorrow. And before that, how much money you make and how you decide to pay your bills. This logic forms the basis for cash flow underwriting and the use of bank transaction data to evaluate a consumer’s risk profile.
Cash data and cash flow analytics offer insights into consumer risk in a way that the credit system, as it stands, simply isn’t optimized to handle. And in a recession, credit score-based underwriting policy falls short of what’s really needed. COVID-19 added a new layer of complication, with consumers receiving stimulus payments, meaning more people were paying off debts. And in many cases, data that calibrates scores are still coming with an artificially inflated track record of making payments. With altered circumstances like this, true risk assessment is more difficult to nail down. We are currently at the brink of another time of uncertainty, and as we move forward the pressure to accurately calculate risk will grow rapidly.
Nova Credit’s Cash Atlas™ gives lenders a suite of cash flow attributes to measure and translate affordability dynamics on consumers’ bank accounts. The suite takes all this uncertainty into account, looking at trends in volatility, income and expenses associated with that consumer, helping generate a more complete risk profile across the entire credit spectrum.
As evidenced in a recent case study – we’ve seen that overlaying bank data for near-prime consumers in terms of risk attributes in addition to a score, results in lenders getting roughly 20% more “bad” or defaulting consumers in their bottom 20% – this is about 5% more for prime-consumers. These numbers represent a tremendous shift and in terms of financial inclusion, 88% of consumers are able to be scored with bank data. Effectively utilizing cash flow underwriting leads directly to greater loss reduction and originations expansion among a broad range of added value.
What Does the Future Hold for Credit?
Today, over 60 million U.S. consumers are credit-excluded, all because they aren’t using credit in conventional ways. In the same way that the credit landscape is evolving for consumers, the shift is affecting lenders as well. The consumer base is growing rapidly, but lenders focused on traditional means of gauging credit aren’t positioned to keep up.
In the wake of a years-long global pandemic and the prospects of economic downturn, lenders and consumers face one of the most complex economies we’ve ever seen. And in an environment like this, lenders need to capture as broad an audience as possible and move away from risky, not fully vetted, evaluation practices. The mandate is clear – adapt or end up left behind.
With that in mind, it comes as no surprise that lenders are questioning traditional practices, gravitating toward solutions that can benefit operations for the long-term. As lenders navigate disruption in the marketplace, cash and bank data have emerged as a real opportunity for growth without exposing them to unnecessary risk. As the use of cash flow and bank data continues to grow and gain mainstream credibility, we can expect to see a new lending environment take shape – one that is ultimately safer than ever before.