Looming trade wars, slowing job growth and a wary bond market are all signs of a potential recession, and venture capital investors are warning fintechs that not all of them will survive such a scenario.
“Niche lenders and neobanks don’t have enough credit data and their models are not robust enough to adjust when a downturn comes,” Victoria Treyger, general partner at Felicis Ventures, said at a recent fintech conference.
Yet not all fintechs are worried about such an eventuality. Many say they're confident their models will stand up to the next downturn, which may even be good for them if it helps persuade more customers to turn away from banks and toward saving and financial apps.
“What tends to happen to a consumer in a recession? They’re more motivated to find savings,” said David Klein, CEO of the online lender CommonBond. “We believe it would drive more demand to our site and increase the propensity of conversion.”
A few years ago, CommonBond pulled data from the credit bureaus and ran it through its own models — one test looked at the models during a normal economy from 2005 to 2007 and the other looked at a stressed economy from 2007 to 2009.
“We didn't just look at student debt, we looked at all debt, including mortgages and unsecured loans,” Klein said. “Our delinquencies and defaults were lower than the historical rates of large public and even private low risk lenders.”
Affirm, a point-of-sale lender, has pulled credit bureau data from as far back as 2002.
“The real cost is in the data scientists we have and how sophisticated they are in terms of analyzing that data,” said Sandeep Bhandari, Affirm’s chief strategy officer and chief risk officer.
While smaller fintech companies may not have the resources to stress test their underwriting models as thoroughly as their larger competitors, their size might insulate them from large economic swings, said Shamir Karkal, one of the founders of Simple and a co-founder and CEO of the developer platform Sila.
Karkal argued that a more manageable size could make a difference in weathering a downturn. Lending startups with hundreds of thousands of customers, neobanks with millions of customers and payments companies with tens of millions of customers could be punished by heavier exposure to defaulting clients, he said.
Still, he said, not all small startups would be safe. Early-stage fintechs might find that the capital they need to advance to the next round is no longer there, Karkal added.
Karkal recommends that companies lock in as much long-term capital from warehouse lines of credit as possible, just in case. "That way, you have the cash on your balance sheet, and cash is king in a recession," he said.
While VC funding might be harder to come by in a recession, customer acquisition costs and hiring employees will become much cheaper, too.
“A recession will always take the froth out of the market,” Karkal said. “After the recession the survivors will be bigger and stronger … banks may wake up to find the biggest financial institution in the country isn’t Chase, it’s now PayPal or Stripe.”
Because of the number of startups using alternative data to better price products, banks shouldn’t be surprised if they find warnings of an imminent economic downturn coming from startups first, said Jackson Mueller, the Milken Institute’s fintech lead.
LendingClub saw this a few years ago when there was fragility in marketplace lending, said Anuj Nayar, its financial health officer.
“The first time we started talking about a recession was in Q1 of 2016,” Nayar said. “I don’t think many other people were talking about a tightening of a credit market three years ago.”
The company underwrites customers using roughly 10,000 variables from more than 100 different customer attributes, with FICO scores providing just five of those attributes.
“Cash flow is a fantastically powerful one,” Nayar said. “Often the FICO scoring is historical … the vast majority of Americans don’t have a steady monthly income.”
There are also powerful economic indicators in alternative data for small-business lending, Mueller said.
“In years past, banks used just audited financial statements and tax returns,” he said. “Now we’re moving into data sets that have never been employed before, like payroll, invoicing and accounting apps.”
