Continuing with topics that might arise at the May 2022 Abrigo Think Big conference (in San Antonio) where SMARTER risk management, LLC. is pleased to be a sponsor, I’ll touch on the topic of Alternative Data in Credit Decisioning.

For a long time, there has been a common set of attributes used in the art of recognizing the credit risk involved when determining the creditworthiness of the consumer. Regardless of your specific location, this has almost always meant the same set of assessments, based around the 5 Cs: character, capacity, capital, collateral, and conditions. In slightly more words, character is basically credit history, conditions include principal/rate, capacity is a debt-to-income ratio, etc.

These have been consistent for a long period of time (decades or even a century) as the true indicators of a business or consumer’s ability to repay a loan, though many other factors obviously go into it (otherwise, would there be a need for risk management?!). Recently, however, the sheer amount data available online has led many industry leaders to the decision of looking for new, unique criteria for evaluating credit risk when some person or business entity is seeing a loan.

Gartner, a technological research and consulting firm, has been speaking with executives about the ways they’re evaluating credit risk in the modern era. Their answers show a vision, especially when it comes to areas that toe the line between what we’ve always done and what the future might hold.

The Nature of the Paradigm Shift

For industry leaders, the last couple of years has been a real paradigm shift in the credit landscape. As Gartner notes, the pandemic and the associated financial vulnerability that came with it have really shattered many of the “traditional” credit risk metrics. Financial vulnerability across various social groups and demographics has increased, almost across the board. Thanks to the effect of COVID-19, and the associated Great Resignation, income and balance sheets have also become stressed in regard to whether they’re good indicators of risk.

Another significant change is that advances in the automation of both credit and the determination of credit risk have resulted in a world where the results of a credit risk analysis can be determined near instantaneously with factors and tools not even able to be considered before. For example, there’s AI/ML risk models taking thousands upon thousands of variables and modeling them in ways that are nigh impossible for humans to come up with or understand, and that’s just the tip of the iceberg.

The sum of all those changes is a fundamental adjustment to our overall perspective. We can datamine thousands of pieces of data we didn’t have access to before and analyze it so fast, giving us potential access to all sorts of information we may not have had before. With the shift in landscape comes a need for new holistic and even non-financial factors for the process of assessing credit risk.

What Gartner Says Risk Industry Leaders Are Saying

So, what are these new factors for credit risk? First and foremost, let’s start with a disclaimer: if you’re a financial institution considering using these metrics, you should work with your risk and regulatory teams to make sure that you are covering your bases when it comes to addressing regulatory risk. That disclaimer aside, Gartner has explained that many industry leaders are examining various other metrics on both the business and individual side of lending.

On the individual side, they’re looking at social media profiles and self-assessed customer health. The former is being used to analyze traits like “brand loyalty, reliability, conscientiousness, [and] work ethic.” While Gartner doesn’t elaborate on exactly how those intangible traits are going to be used in credit risk models, it’s not exactly unclear how some of these might translate into reliable lending practices. The same kind of idea applies to self-assessed customer health, although this is a much more difficult aspect to legally incorporate due to the intense regulations around health information. Some firms used client surveys during the pandemic as a method to understand how consumers and businesses were handling the crisis, whether in financial stability or actions taken, which may be good indicators of risk in some scenarios.

On the business side, they’re looking at management quality and operational data. Management teams are an important indicator of a business’s quality, as a more educated and experienced manager is more likely to be able to handle navigation of crises and other economic downturns. With the pandemic somewhat in the rearview in the US, this may not be of current crucial importance, but were something bad to happen, it’s more likely experience will pay off, reducing long-term risk. Operational data, like regulatory filings and government records, have a clear and direct relation with credit risk, so examining those would likely be beneficial.

How Much Is Credit Risk Actually Going to Change?

It’s hard to say exactly how much these new ideas will be incorporated into credit risk models or approaches in the near or distant future. As mentioned earlier, creditworthiness determinations have been relatively stable in the sorts of data they’ve analyzed for a long time, though new metrics and criteria do often come into the equation when they’re discovered and vetted. It’s possible that some of the methodologies that these industry leaders are discussing may even become part of standard credit risk modeling practices in the years to come!

For now, though, it seems as though the traditional methods will continue to be the path most financial institutions use for assessing creditworthiness. In our path toward more robustly identifying and mitigating risk, though, finding these new metrics could mean the difference between a default and a repayment. Even if those discussed in this article don’t become widespread, it’s likely that some other new indicator will, but until then, at least we’ve got the 5 Cs.