Replay: Fintech Office Hours — Sept 2026 Edition
Fintech doesn't sleep, which is why Office Hours is here to keep you ahead.
Alex and Kiah return for a live Fintech Office Hours session covering the fintech news of note for September. Chime is acquiring Stride, one of its own partner banks, for $590M in cash. Alex and Kiah unpack what that says about BaaS exits and question the OCC's appetite for stablecoin trust charters. Then they turn to Amazon's standoff with Meta's AI assistant (and why banks should be nervous about it) before closing on the Cash App score and Ramp's habit of doing everything at once.
Key takeawaysCopy anchor linkCopied
Charter Watch 1 (Chime buys Stride Bank): Copy anchor linkCopied
- ✅ A bank, bought instead of built. Alex sees the deal as a faster path than applying for a de novo national bank charter, with potentially only a couple years left in this regulatory window. The combined company will hold about $7B in assets. Chime plans to stay under the $10B Durbin threshold for the foreseeable future, which protects the debit interchange it has historically made most of its money on. Chime expects $100 in synergies, which Alex guesses may come from lower-cost deposits funding larger loans, and the rest from sponsor bank fees Chime no longer pays. The central tension: a company that spent years proudly insisting it isn't a bank will now have a bank's business model, and it has to avoid getting sucked into a bank's incentives.
- ✅ BaaS didn't make banks harder to buy. Kiah thought BaaS was supposed to guarantee a bank wouldn't get acquired, yet Stride agreed to sell and Lincoln Savings sold for just barely over book without mentioning BaaS in its investor presentation (Alex adds Blue Ridge, which sold with no premium for its time in BaaS). Alex reads BaaS as a Hail Mary for most banks. Kiah's view is that BaaS is very expensive to do right, and a bank that isn't Bancorp or Coastal can't just throw money at it and get the halo.
Charter Watch 2 (3 stablecoin companies get conditional national trust charters):Copy anchor linkCopied
- ✅ Three charters, one unanswered question. Bastion issues its own stablecoin and lets brands white-label it, Agora manages segregated stablecoins for each client, and Catena builds stablecoins for AI agents. How many of these do we need when nobody knows how big the stablecoin market really is? Alex senses a lack of curiosity within the OCC that reminds him of the World Liberty Financial approval.
- ✅ The OCC's old de novo model assumed thousands of communities each needed their own banks. But in today's flattened competitive environment, Alex thinks you can argue we don't need nearly as many. Kiah asks whether this is BaaS from 2018 to 2020 all over again. She warns that a lot of companies entering one business model at once is "never a good sign."
Block opens up the Cash App score:Copy anchor linkCopied
- ✅ Cash App users can see their score. Block made its proprietary credit score visible to Cash App users with specific suggestions for improving it, and pilot users came back to check it more often. Nova Credit, already a consumer reporting agency regulated under FCRA, is the first partner distributing the score to other lenders, and Cash App says its testing shows the score outperforms FICO and bureau data for its younger, thin-file customers. Kiah suspects banks never tried this because the credit teams who probably own this information focus on protecting the downside, and nobody there is incentivized to care about customer experience.
- ✅ First-party scores fracture the bureau model. Block has said publicly it won't furnish Afterpay BNPL data to the credit bureaus, and Alex reads the score as Block opting out of furnishing raw cash flow data through Plaid as well, which makes the score the only way to get Cash App data. He pictures Nova Credit holding a library of first-party scores from different fintechs and lenders, which raises his question for consumers: what's my credit score? Kiah's rule is that the right amount of anything is one, so more scores empower consumers and confuse them a little at the same time.
Next upCopy anchor linkCopied
The Science (and Fiction) of Friction · Sept 30, 1:00 PM ET, Virtual

