Moven's Plan B after buyout bid stalls: Expansion into lending
Obstacles remain in Moven’s path to a charter in the U.S., even as its private-label banking offering expands across Africa and Asia.
Marek Forysiak, Moven's chief executive, acknowledges it is still searching for a small bank to buy and is continuing its discussions with the Office of the Comptroller of the Currency about its plans.
Still, the mobile-first digital bank plans to expand beyond transactional banking and financial advice into lending with Moven 4.0 later this year.
Moven will use the analytics it is already employing with Stash — a calculator that shows users how much money they can spend while still paying the bills — to help customers know when to borrow without tapping themselves out.
“We’re going to figure out what you’re comfortable borrowing, then look for opportunities in your existing spending,” Forysiak said. “We want to find those residual resources that allow you to borrow in a way that doesn’t overextend you.”
While Moven plans to extend the credit through a partner bank, MovenBank will keep the majority of its loans on its balance sheet until maturity. Forysiak notes underwriting will consider alternative data points, like mobile handset data and psychoanalytic data.
Forysiak sat down with American Banker to discuss its continuing effort to acquire a bank, what goes into building Moven 4.0 and how alternative data will factor into its underwriting.
Is the acquisition of a small bank still an interest?
MAREK FORYSIAK: It is. I had a meeting with the OCC, about two to four weeks ago, precisely on the challenges around acquiring an institution. I spent the better part of 2018 working on that undertaking, entering into various negotiations and due diligence. None of those proved successful.
We still have it on our radar screen. We still want to operate on our own charter, whether or not that comes by acquiring a small institution. We thought that would be the quicker path as opposed to applying for a charter, seeing that Varo Money has been trying to do that for a long time.
What has been the issue?
If you're looking to buy a challenged bank operating under a consent order and I come into recapitalize that bank, you would think that that consent order would be delisted, right? The consent order remains with that institution, even though I've come in and provided all the requisite capital. Unfortunately, what happens is if you have an institution that has certain challenges and you have an acquisition by a strong institution, your strong institution becomes a challenged institution.
An OCC consent order means they're required to operate under enhanced supervision of the OCC. That means that every operating decision needs to be cleared by the OCC. I'm not looking to get into an investment where every decision I want to make operationally requires OCC approval; that means the OCC is running the bank and not me.
What does Moven 4.0 look like?
Moven 4.0 is built around realizing a smart bank account that provides everything from savings to credit. It is underpinned by the ability for users to transact and move money, as well as being underpinned by our advanced analytics process, the advice piece. We’ve been working on this development for about nine months.
Moven 3.0 was built around contextualized savings where customers have an ability to identify wants and needs. By way of spending controls and insights, we are able to prompt, nudge and advise clients when it's best to save and when it's safe to spend and what behaviors are ongoing in a meaningful manner that allows for our customers to save more.
Contextualized savings, real-time prompting, the real-time insights, all underpinned by our back-end advisory layer has resulted in MySpend customers at [Moven customer] TD Bank saving on average between 6-8% more than non-TD MySpend users.
Version 1.0 and 2.0 were all around spending insights and controlling spending habits. 3.0 was largely linked to stash, as we call it. 4.0 is where we're going to be introducing contextualized credit. As we think about that we truly, really allow for us to realize smart banking.
Can you explain what contextualized credit is?
What we try to do is contextualize the whole idea of spending insights, spending controls, so that allows for you to start thinking about wants and needs. As customers gain better control of spending, that control allows them to stash money based on whatever their wants and needs are. Within that same context, we would now be able to responsibly deliver credit.
If we've helped you save $150 and $200 more per month and you have wants and needs, we may be able to help you attain them by getting you access to credit. We’d be utilizing the same resources that you've found through better control of spending. The customer would reallocate that spending to borrow responsibly, to accelerate acquisition of those wants and needs.
It's done in a way that we're not looking to get our customers further into a position that negatively affects their ability to pay or put them into a more difficult cash-flow position. Typically, when you look at the at-risk and vulnerable-customer segment, they're overextending themselves, which is why they find themselves at risk or vulnerable. That's what we totally want to avoid, which is why we started with savings. We only have had a debit card, never a credit card product.
What does the underwriting process look like for Moven? Any alternative data or credit models?
All of the above. Our chief risk officer [Carl Spilker] comes with deep expertise in credit from GE [and] Dollar Financial. He's been with traditional lenders to nontraditional lenders.
As a data scientist, he's built his own scoring models at an earlier part of his career. He's been assembling a fairly comprehensive capability to assess credit using traditional modeling sources and traditional data as well as partnering with other third parties for us to be able to leverage alternative data.
Mobile handset data is one example. We're already seeing 30 to 40 million data points we're capturing from mobile handsets. How we interpret that data and understand the correlation and the dependency of that data, we'll figure out over time. We're working with a third party to effectively capture the data, mine the data and analyze the data in a meaningful way.
What data from the mobile handset is useful for underwriting?
How often you spend time on your mobile handset is a data point that allows for us to identify certain user characteristics and behaviors, not to mention some of the most basic things in regards to identity and location verification.
The metadata that comes off of that can be incredibly rich data to assess your customers’ willingness to pay. Using it to assess customer's ability to pay — we're not 100% convinced yet.
What other kinds of alternative data are going into this model?
That's primarily what we're focusing on today. We're also using the traditional data elements used for making credit determinations. We're looking at things like psychometric scores. We're not immediately convinced about the benefits of psychometric analysis. It's around certain questions that they would ask about user behaviors. Like, how frequently do you go on vacation? When you go on vacation, do you travel alone? Do you travel with friends?
What's shaky about that metric?
We haven't seen strong consistent predictive power. The correlation is random.
What sparked your initial curiosity?
We want to approach vulnerable markets here. What we see consistently with those at-risk U.S. consumers is that they also have thin files. That's what led us to evaluate everything else that's out there. There's a group out there called Lendo, a bunch of people out of [Massachusetts Institute of Technology] that are using psychometric scores to be predictive on creditworthiness.
We naturally gravitated to the mobile handset data because we're a mobile-first financial product. All of our customers have smartphones. For handset data, we're using CredoLabs, ZoroLabs and GDS Link. They all specialize in consumer lending and have deep expertise on credit.