Kaustav Das is the Chief Risk and Analytics Officer at Petal, an alternative credit card fintech. Previously, he served as the Chief Risk Officer at Kabbage, and spent 15 years in credit and fraud risk at American Express.
We spoke with Kaustav about how risk is evolving during COVID-19, the experiences of small business lenders, and the future of the SMB lending landscape.
The words of a top bank’s CEO summed it up when he said that he wished there was a handbook for a pandemic. Everyone prepared for a recession, but no one prepared for a pandemic, or a recession that's driven by a pandemic. We have recession readiness and then we have recession planning. The pandemic is going to result in a lot of changes in the risk landscape and will change the risk playbook for the entire customer life cycle.
As an example, understanding risk by industry has always been important, but now people will pay special attention to industry subcategories. Knowing whether a business is a restaurant is no longer enough, you need to know if they deliver or if they partner with delivery service. If a business is a retailer you need to know if they have a brick and mortar presence or if they’re only online.
Also data freshness becomes increasingly important. You cannot rely only on bureaus, you need more real-time data like bank data, processing data, or other alternative data as that captures recency.
I don't think the pandemic is going to cause a big tectonic shift in terms of the way SMB lending is done, but there will be a major shift in the players that will remain in the market.
I think we will see consolidation, but remember, consolidation can happen in multiple ways. Consolidation can happen in the form of someone going out of business, or consolidation can happen when a company becomes a very attractive target for acquisition.
Some of these fintechs have great products, great teams. Larger competitors might want to use this opportunity to get a good bargain and acquire them. Each and every one of these bigger financial institutions will be keeping a close watch. The question is how willing are these institutions to part with cash right now?
Also, the PPP (Paycheck Protection Program)* can be meaningful and make the difference in whether a SMB fintech survives or doesn’t. If a fintech does manage to get a chunk of the PPP commision money, it is going to help them survive. If they don't, there could be a high likelihood that they might regress far behind where they were before this recession.
Another variable in terms of a fintech’s survival is funding — when they were funded, who they were funded by, and how well funded they are.
Not to mention, an additional key factor would be the default rates. Not every institution has the same default rates, and the extent of that impact would play a very important part.
My thinking around SMB risk attributes and data points has evolved. Need of alternative data or alternative bureau data, along with smart capabilities, has taken the center stage in this pandemic-driven recession. How does someone assess the extent of revenue loss, or when the PPP check or stimulus is getting deposited? Every institution suddenly has started appreciating the need of having robust KYC/KYB capability for PPP. Document fraud suddenly has come to the forefront.
I'll give you a specific example of a data attribute that’s evolving: Industry. Some financial institutions will have a knee jerk reaction and will stop lending to restaurants as they are highly impacted. Other institutions that have access to better data, would figure that, restaurants are impacted, but some restaurants’ business is only down 25-30%, not even 50%. And guess why? The amount of online orders have shot up.
So, yes, people are not doing fine dining, but these people - or a portion of them - are ordering in. So as that dine-in portion of the revenue is going down, take out orders have significantly gone up to offset some of the lost revenue.
This just goes to show that even if you've identified an industry, such as the restaurant industry, that categorization alone is not enough. The next level of detailed categorizations, such as does a restaurant deliver or is it enrolled with any delivery partners, becomes important. The data - or more accurately, the granularity of data - becomes important.
The time to act is now. For existing customers, think of smartly reducing contingent liability. For new customers, weigh risk reward with a new lens. You need to quickly change your risk policies, and procedures. Models that worked perfectly before may not be as accurate or viable. As a new model cannot be changed, tested, approved, and implemented so quickly, you have to resort to smart risk policy and strategy changes.
*Editor's note: As of 4/16, the initial $349B in funding allocated for PPP rescue loans had run out. As of 4/22, another infusion of funding is expected to be approved in the next few days.