3 Fintech News Stories
#1: FICO Jumps Into the Shallow End of the BNPL Pool
What happened?
FICO announced that it will be launching two new versions of its credit score, specifically focused on addressing the growth of BNPL:
The company will offer the products — FICO Score 10 BNPL and FICO Score 10 T BNPL — later this year, it said in a statement Monday. The scores will help lenders make more accurate decisions about consumer creditworthiness, especially for those who borrow for the first time using the payment plans, FICO said.
So what?
I read the FICO press release and stories covering the news from many major news outlets (Wall Street Journal, Bloomberg, Axios), and I feel like I’m losing my mind.
Let’s just use the Wall Street Journal as our example.
The Wall Street Journal reports that Affirm is furnishing repayment data to Experian, and other providers are expected to follow:
Because these loans haven’t appeared on credit reports, lenders were often blind to how much debt a borrower had already taken on. In April, Affirm began reporting all new loans—including its “pay in four” plans—to Experian, a move other providers are expected to follow.
Affirm is indeed furnishing data to Experian (and to TransUnion … not sure why the WSJ didn’t mention that, but whatever). However, there is no evidence to support the statement that “other providers are expected to follow”. I understand why FICO and the credit bureaus might want people to believe that, but I don’t understand why the Wall Street Journal is taking that claim at face value.
Klarna does not furnish pay-in-4 loan repayment data to any of the credit bureaus. Neither does PayPal for its pay-in-4 product. Afterpay does not furnish data on any of its loans to the credit bureaus. Nor does Zip.
Now, of course, this could change. The credit bureaus are working hard to get all the BNPL providers onboard. However, I think it’s naive to assume that Klarna, Afterpay, PayPal, and Zip will simply follow Affirm down this path.
Affirm is the premium provider in this space. Its typical customer is generally very creditworthy. It wins by offering consumers better products (embedded installment loans, a BNPL/debit card hybrid, etc.), not by extending credit where other lenders won’t. It isn’t risking much by furnishing its data to the bureaus.
By contrast, I think Klarna views its repayment data as highly proprietary. The company spends a lot of money to acquire high-risk and new-to-credit customers and using the short-term, small-dollar structure of pay-in-4 BNPL loans to sort the good risks from the bad risks (this is why Klarna’s loss rates are always higher when it first enters a new market … it’s generating performance data to use in fine tuning its underwriting model). Why would it want to give that data away for free to its competitors?
However, if the other providers don’t furnish data, the value of FICO’s new scores is dramatically reduced.
Indeed, without representative samples from all of the BNPL providers’ portfolios (especially the higher risk ones like Klarna), how can FICO even build a representative scoring model in the first place?
That would have been a great question for the Wall Street Journal to ask! Instead, they reported this:
The new scores were trained on a sample of more than 500,000 BNPL users in a joint study with Affirm.
Arrrrgggghhhhhh!!!
One last point: given that most lenders in the U.S. are still using older versions of the FICO Score (FICO 8, most commonly), wouldn’t the best approach be to work with the BNPL providers and the credit bureaus to furnish the data into the core credit file in a way that is backward compatible with those older scoring models?
I think it would. But is that what FICO is doing?
No, no, it is not:
Factoring BNPL data into scores requires more than just feeding the numbers into the existing algorithm. Current scoring models penalize consumers for opening several new lines of credit in a short period.
So FICO is rolling out the new credit-scoring model and offering it to alongside its current one, starting this fall. Banks and credit card-companies will evaluate the new data but don’t have to rely on it.
The background here is that FICO wanted a way to extract predictive insights from BNPL repayment data without compromising consumers’ existing credit scores due to the loans’ short-term, small-dollar structure.
My prediction (back when FICO and Affirm first announced the results of their joint study) was that FICO would try to get the credit bureaus to solve this problem on their end:
FICO’s innovative approach is to combine a bunch of individual pay-in-4 BNPL loans into a single personal loan. My guess is that FICO’s preferred implementation for this approach would be for the credit bureaus to aggregate and store the BNPL data in that fashion. This would enable older versions of the FICO Score (which are well-accustomed to personal loans) to be compatible with it.
That, predictably, did not happen (it would have created a lot of downstream problems for the bureaus), so, instead, FICO took an unprecedented step:
It is the first time FICO has ever introduced a score to account for a type of loan, according to Ethan Dornhelm, vice president of scores and predictive analytics at FICO.
And now it is offering its new BNPL scores, free of charge, alongside FICO 10/10 T, in the hopes of encouraging lender adoption.
I’m not optimistic.
#2: The Lonely Mountain
What happened?
Palmer Luckey, CEO of the defense contractor Anduril, is reportedly merging a stablecoin company that he is investing in with a banking startup that he has also invested in?
I don’t know. It’s confusing.
We’re learning more about Atticus, the stealthy stablecoin startup that we recently reported was raising funding from Anduril CEO Palmer Luckey and others at a $2.25 billion valuation.
The agreement will see Atticus (or at least its team) effectively merge into something called Erebor, recently formed by Luckey with plans to get a banking license.
Crypto startups and founders say they’ve been victims of debanking, and this may be Luckey’s way to ensure it doesn’t happen to him or fellow founders in the future.
So what?
Atticus has been the subject of much speculation in fintech and crypto circles for the last few weeks (including for Simon Taylor and me in our most recent NFIA podcast), so it’s good to get a little clarity.
Erebor has indeed filed an application for a national bank charter with the OCC. And according to reporting from the New York Post, the idea is to create a narrow-ish bank, focused on serving tech startups:
Talks about Erebor began soon after the sudden collapse in 2023 of Silicon Valley Bank — once the go-to bank for tech startups and venture firms, sources said.
Offering stablecoin deposits and support will allow Erebor to offer constant service even during bank holidays – providing flexibility for nimble startups, sources said.
Luckey and his allies also are seeking an alternative to traditional fractional reserve banking – where banks deploy most of their client deposits while keeping only a relatively small amount in reserve at any given time.
The group wants a conservative balance sheet and have considered imposing a maximum limit on loan-to-debt ratio, potentially in the ballpark of 50%, sources familiar with internal discussions said.
One concept under consideration is to offer a bank account with covenanted one-to-one deposits – meaning any assets stored at Erebor would be kept there untouched, the sources added.
All of this seems very achievable, in the current moment.
The OCC has clearly signaled that it is open for business when it comes to bank charters for fintech companies and other non-traditional business models. The GENIUS Act just passed the Senate and is on track for passage in the House and signature by the President. And if ever there was a time to argue for the merits of narrow banking as an alternative for tech startups, it is certainly now (though that doesn’t guarantee that all regulators will be onboard … the Fed has rejected applications for Fed Master Accounts from other narrow banks before).
The real question is whether Erebor is going further than necessary.
Even if it is approved by the OCC, Erebor will still be subject to the standard three-year trial period for de novo banks, during which it will be subject to increased supervisory scrutiny and will be unable to significantly deviate from the business plan approved by the OCC. That’s not a lot of fun, especially if you view yourself as a tech company that wants to innovate and move fast. Plus, if we assume that Erebor would attempt to obtain FDIC insurance (likely, given that it is reportedly not planning to be a completely narrow bank), the holding company would be required to maintain significant liquidity (ring-fenced reserves for the stablecoin + Basel capital for the bank), which would be quite capitally intensive.
Wouldn’t it just be better to be a non-bank stablecoin issuer under the GENIUS Act regulatory framework? If stablecoin companies can find ways around the need for partner banks to get access to Fed Master Accounts and payment card issuing (as Rain seems to be doing), is there really any need for traditional bank charters anymore? Will narrow stablecoin “banks” become the new platform upon which innovators build?
(Editor’s Note — I really need the nerds who are founding companies in the defense industry to stop stealing names from Tolkien. I know you like his works, but you’re ruining the books for me. Plus, it’s not like Tolkien was a huge fan of war or industrialization.)
#3: The Psychology of Money
What happened?
A couple of interesting B2C fintech pieces of news!
First, Frich announced a new feature:
Called Frich Scoop, the feature lets users upload screenshots of anyone’s LinkedIn and Instagram profiles to receive a speculative “scoop” revealing estimated income, lifestyle costs, and financial red and green flags.It uses AI image analysis to combine professional and personal online personas, providing users with estimated annual income, monthly lifestyle burn, and projected years to millionaire status – all presented in a shareable, Spotify Wrapped-style format.Frich claims that it is taking on the growing disconnect between what people see on social media and financial reality in an effort to encourage healthier conversations about money.
And second, Grifin (an investing app) announced a Series A:
Approachable investing app Grifin announced that it raised $11 million this week to help users invest where they shop. The Series A funding round, which brings the company’s total raised to $20 million, was led by Nava Ventures with participation from TTV, Draper Associates, Gaingels, Nevcaut Ventures, and Alloy Labs.
Grifin was founded in 2017 to make investing fun by allowing shoppers to invest in a portion of the brands they purchase from. The company removes complexity and fear associated with investing by building an investment portfolio based on the consumer’s purchasing habits. Grifin automatically transfers $1 for every transaction the user makes during the week, then invests the funds into their portfolio that is comprised of companies from which the user purchases. Grifin calls this approach Adaptive Investing.
With Adaptive Investing, Grifin creates a dynamic investment portfolio that is uniquely personalized to the user and their everyday habits. As the user’s shopping habits change, Grifin adapts the portfolio. The company also offers users full control on how much and in which companies they invest, allowing them to block companies and manually adjust their investment amount.
So what?
Before I get into the details of these announcements, I gotta say that I love how thoughtful B2C fintech companies are being in trying to address the psychological and emotional components of their customers’ financial needs, not merely their functional needs.
On Frich, I understand and appreciate the problem that they are trying to solve with Scoop. Social media encourages us to compare ourselves to others, and that is a surefire recipe for misery and financial nihilism (I wrote more about this here). However, the implementation of this feature feels problematic. Frich claims that the tool is fun, non-judgmental, and “kinda” accurate, but that doesn’t seem like a high enough bar when we are talking about creating “shareable, Spotify Wrapped-style” financial profiles of other people without their permission.
On Grifin, my general stance is that anything that encourages consumers to invest in stocks more is probably a net positive. From my brief research, the product appears to leverage Plaid to connect to users’ debit and credit cards, giving Grifin insight into users’ shopping habits, which it then uses to determine which investments to make (it will buy $1 of stock for every transaction made with a publicly traded company). The product also allows users to block the purchase of stock from companies that they shop at, but don’t want to own, which is absolutely hilarious to me. I would KILL to see the data on that, as Grifin scales up its service (they mention utility companies and cellular providers as examples in their FAQs). I personally would love to see them add the ability for users to short specific companies’ stock, if they have a particularly infuriating customer service interactions with them. Or, perhaps, create some type of payment for order flow-type information-sharing service that funnels negative customer experiences with public companies to activist short sellers. These are probably terrible ideas, but they would be funny.
2 Reading Recommendations
#1: The CFPB Has Entered the Chat: Is A Synapse “Bailout” In the Cards? (by Jason Mikula, Fintech Business Weekly) 📚
Big news. Huge. And very surprising given everything else the current CFPB has been doing. We’ll see if it results in relief for Synapse victims, but either way, expect a trip back to BaaS Island in the next edition of Fintech Recap.
#2: Coinbase is quietly taking on Visa and Mastercard (by Jevgenijs Kazanins, Popular Fintech) 📚
Coinbase has quietly been included in a number of different stablecoin announcements recently, all of which have related more to infrastructure (through the company’s Base Ethereum L2 network) than its better-known consumer-facing business.
Jev does a great job breaking down Coinbase’s new Commerce Payments Protocol, which will (potentially) make stablecoins more of a direct competitive threat to the card networks.
(BTW, this article pairs nicely with Chuk Okpalugo’s most recent newsletter, which covers the news that JPMorgan Chase has launched a tokenized deposit product, also on Coinbase’s Base network.)
1 Question From The Fintech Takes Network
There are a TON of interesting questions being asked in the Fintech Takes Network. I’ll share one question, sourced from the Network, each week. However, if you’d like to join the conversation, please apply to join the Fintech Takes Network.
In 10 years, will we have more tokenized deposits (e.g., JPMD) or tokenized money market funds (e.g., non-bank issued stablecoins)?
If you have any thoughts on this question, reply to this email or DM me in the Fintech Takes Network!
