How Innovative Banks Scale Consumer Loans

Consumer
loans can be an afterthought for most banks chasing bigger, commercial credits.
Banks typically don’t market consumer loans, let alone originate them on a
large scale, because these loans have historically been too small and
inefficient to be profitable. But over the last ten years, fintechs have
infiltrated this space in a big way. It’s time for banks to re-evaluate their
offerings.

After the financial crisis, regulatory requirements for consumer loans became more cumbersome. Unsecured loans became an untenable product for some community banks; in their absence, fintech companies stepped in to fill the void. Fintechs originated 49% of all unsecured loans in the U.S. in 2019, up from just 22% in 2015, according to Experian.

They
capitalized on high demand for consumer loans – now broadly called “personal
loans” – with technological scale, enabling them to grab significant market
share. Some fintechs are using technology to help banks to get in on the game
too.

One such bank
is First Federal Bank of Kansas City. The $819 million asset bank has focused
on mortgages since its 1934 founding, but shifting home ownership trends in
recent years prompted a closer look at the market. What executives found was
that mounting credit card debt was a major barrier to home ownership. CEO J.R.
Buckner decided to figure out a way to help customers over that hurdle.

One
attractive option was a personal debt consolidation loan that would help
customers get out of debt and begin saving, so that they could ultimately become
mortgage customers. “[T]hink of the debt consolidation loan as an entryway into
our philosophy on what it takes to experience financial wellbeing,” says
Buckner.

The Kansas City, Missouri-based bank worked with Upstart to launch an unsecured personal loan product in the spring of 2019. Upstart’s technology uses about 1,600 data points to assess creditworthiness, an eye-popping number compared to traditional lenders, which typically use just 12 data points on average. That amount of data doesn’t work with simple regression analyses and spreadsheet calculations, so Upstart uses machine learning and automation to crunch the numbers.

Buckner confirms that Upstart’s model “is approving loans that we would have traditionally denied without the extra data points that they have,” so the bank can extend credit to more customers and, hopefully, put them on the path to savings and home ownership.

While First
Federal is using personal loans to cultivate mortgage customers, Garden City,
Missouri-based Lead Bank, is using them to bring new, underserved customers
into the financial ecosystem.

A $357
million institution also in the Kansas City area market, Lead Bank already has
a strong concentration in commercial real estate. So it was up to CEO Josh
Rowland to find new “levers to pull” that would generate revenue in keeping
with the bank’s mission.

One way is
through its partnership with Self Financial. This Austin-based fintech found a
unique way to package traditional financial products in the form of “Credit
Builder” installment loans, which are backed by a certificate of deposit at
Lead Bank and help people with thin-file or no credit history build up their
scores without going into debt.

The partnership between Lead Bank and Self has been running for two years. Because the bank doesn’t advance borrowers money – the CD is funded once the customer has reached the required amount of equity – the program gives Lead a way to serve the underbanked without the risk of losses.

Lead Bank is also piloting a program of unsecured consumer loans with Helix. So far, the bank has seen “dramatic reductions” in the costs and losses associated with the Helix portfolios. Rowland says it’s too early to tell whether they’ve cracked the nut on what makes a profitable consumer loan program, but the bank has board support and is “not afraid to try.”

These partnerships aim to improve the financial lives of their respective bank’s customers. They help average people go from indebtedness to home ownership; from credit invisible to credit enabled. All warm, fuzzy benefits of financial inclusion and wellness aside, these new loans also present opportunities for both banks to acquire new customers and grow existing ones – but not without risks. Rowland estimates that unsecured consumer loan losses are around 30% for the industry; that’s a risk level a lot of banks won’t tolerate.

But Lead Bank
is betting on its own hypothesis: The bank can use technology to reach a larger
segment of borrowers, solving the scale problem inherent in consumer lending
and providing the cover of the portfolio effect to ease potential losses.

Consumer loans
are risky, and aren’t a fit for every bank. But they present a sizable
opportunity for institutions looking for new revenue levers to pull, and all
institutions must decide whether they’re in or out. Banks must take action to
get back in the personal loan game, or risk forfeiting the space to fintechs.

Potential Fintech Partners

Upstart

According to CEO Dave Girouard, Upstart’s machine learning models were shown to reduce personal loan losses by three-quarters with the same approval rating, when compared to models used by large financial institutions.

Self Financial

Helps thin-file and no-file borrowers proactively save and build credit through an installment loan backed by certificates of deposit with bank partners.

Happy Money

Partners with banks to sponsor Payoff Loans, which use psychometric data to choose borrowers that show a propensity to want to get out of debt.

CuneXus

Uses a bank’s existing customer data to keep customers “perpetually approved” for various consumer loan products. Offers appear upon the customer’s sign on to online or mobile banking, and require only a few clicks to apply.

MinuteLender

A digital platform that enables community banks to underwrite small loans for both consumers and businesses in under 5 minutes.

Learn more about the technology providers in this piece by accessing their profiles in Bank Director’s FinXTech Connect platform.