3 BIG IDEAS FROM THE PODCAST
Jason Mikula and I kicked off the 12th season of the Fintech Takes podcast with a deceptively simple question: What happens when you try to modernize a 36-year-old credit score with data the credit bureaus don’t actually have?
If you’ve been watching the credit bureaus’ pursuit of BNPL data with a mix of curiosity and confusion, this episode is for you.
We also carve out time for a quick trip to BaaS Island (is that a rescue helicopter?) and a brief discussion on Bill Pulte and the FHFA.
And read on below for my three big ideas …
1. FICO’s New BNPL Solution Has a Glaring Problem
FICO just announced it’s updating the FICO Score to incorporate BNPL data. The new versions of 10 and 10T are being hailed as progress by mainstream media (and to be fair, it’s not not progress), but no one’s asking the obvious chicken-and-egg question: how do you build a better score if you don’t have the data?
The new model was trained on 500,000 users from a study with Affirm — still the only BNPL provider furnishing pay-in-4 data to Experian and TransUnion.
Yes, you read that right. The credit bureaus only have pay-in-4 BNPL repayment data from a single large provider. That’s it. No Klarna. No Afterpay. No Zip. No PayPal. Just Affirm.
That’s like building a weather model for the Midwest and leaving out Chicago, Indianapolis, and Detroit! (Plus, Affirm isn’t your typical BNPL provider; most of its business is installment lending, and its users skew more affluent and creditworthy.)
Additionally, FICO’s new scoring models are assessing pay-in-4 BNPL loans by lumping them together into larger single installment loans. This approach is more consistent with how credit scores have traditionally worked (and how the credit bureaus have traditionally captured repayment data), but it’s arguably not the right approach if your goal is to understand how use of this novel credit product (pay-in-4 BNPL is a fundamentally new thing!) impacts consumers’ overall creditworthiness.
The credit bureaus store the data in one way, while FICO’s BNPL scoring methodology aggregates it in another way. The only way to make it work is to release a whole new score (because older versions of the score couldn’t support it) and give it away for free.
We’ll see how this one plays out, but (as I wrote last week) I’m not optimistic.
| 🎬 DIRECTOR’S COMMENTARY You know what’s weird? Equifax never comes up in these conversations. TransUnion is focused on solving this data coverage problem. So is Experian. They are doing a ton of work to chip away at the reluctance of BNPL providers (and other novel fintech lenders) to furnish. Where is Equifax? What are they doing in this area? |
2. All Signal, No Sharing
If the bureaus can’t see it, maybe open banking can.
With most BNPL tradelines missing from the credit bureaus, cash flow underwriting could become a fallback option.
In theory, bank transaction data offers a workaround: if someone’s making recurring $90 payments to Klarna, you can reasonably infer what’s going on (even if you don’t see the exact loan terms).
But that only works if the pipes are connected. And right now? The coverage is awful. Try linking your Klarna account in a PFM app.
Coverage is patchy at best, and it’s not a technical issue. It’s a business issue.
Open banking data aggregators view BNPL providers as customers, not data sources. Historically, they haven’t wanted to scrape them and risk their commercial relationship. That’s how the biggest BNPL platforms have avoided data sharing, even as they extract plenty of it themselves.
And with the CFPB’s Personal Financial Data Rights Rule on life support, there may not be a legal mechanism to compel the opening of that consumer financial data.
So we’re stuck in a standoff: underwriters want transparency. BNPL companies want control. And data aggregators don’t want to rock the boat.
In a world where every player hoards their own version of the truth, underwriting becomes guesswork. Everyone wants to de-risk, but no one wants to go first. Everyone wants the signal, but no one wants to share.
Not ideal.
3. A Rescue From BaaS Island?
I hated to force our embedded reporter on BaaS Island to file yet another report, but we had breaking news!
The CFPB (yes, this CFPB) is apparently contemplating getting involved in the Synapse mess. The bureau filed a statement of interest in the Synapse bankruptcy case, stating that the “Bureau has reason to believe that Synapse may have engaged in unfair practices by failing to properly track and reconcile consumer funds across Partner Financial Institutions.”
Interesting! And unexpected!
Jason broke the news on this development in his newsletter, so we obviously had to talk about it!
The main takeaway is that this action may lead to monetary relief for the end customers who have lost access to their funds, despite the fact that the Synapse estate has no money. As the CFPB wrote in its statement:
The Bureau has a strong interest in pursuing its claim because it has a mechanism to potentially make victims whole without payment from Debtor’s estate.
The mechanism in question is the CFPB’s Civil Penalty Fund, which is the big pot of money that all the fines assessed by the bureau go into, after companies settle enforcement actions with it.
In our conversation, Jason and I delve into two critical questions:
- Will the CFPB’s involvement in the Synapse bankruptcy result in us learning more about how this meltdown occurred in the first place?
- Would the use of the Civil Penalty Fund in this case constitute a “fintech bailout”? And if so, does that set a bad precedent?
| 🎬 DIRECTOR’S COMMENTARY Neither Jason nor I can remember how the BaaS Island metaphor started, but both of us are unwilling to let it go. It predates the Synapse collapse, but perhaps once the Synapse situation is finally, mercifully over, we can retire it. No promises, though. |
WHAT I’M LISTENING TO
#1: Tim Newell, CEO of GreenFi, on building a climate-friendly fintech (Fintech One on One) 🎧
There are some folks out there who are stubbornly trying to make the idea of a sustainability-focused neobank a viable and profitable reality.
Tim Newell is one of those people. And his journey in this space has been fascinating. And Peter Renton is one of the fintech industry’s great interviewers. And so.
#2: Robinhood’s CEO on the Plan to Tokenize Everything (Odd Lots) 🎧
I’m not a huge Robinhood fan (as you likely know), but they are undoubtedly on a tear right now, and it’s important to understand what they’re up to.
This is an interesting interview with Robinhood’s CEO, in which he talks extensively about his desire to “tokenize everything.”
WHERE I’LL BE
💻 Fintech Open House — July Edition | July 16 | Zoom
Digital wallet wars. Embedded everything. BaaS growing pains. We’re unpacking the big shifts in our Fintech Takes Open House, open to all. Register to join.
