Oh look! It’s my sponsors again 👇. Fintech Nerds, I want you to know these guys are the masters of moving all of the payment types: checks, ACH, wires, FedNow, RTP, even stablecoins- all of them. If you’re building and moving money. Make sure you speak to these folks first.

Stablecoins are here.

Checks, ACH, and wires aren’t going away.

RTP, FedNow, and new payment rails keep coming.

The challenge isn’t predicting which one wins. It’s supporting all of them without slowing down.

Build faster. Launch sooner. Scale with confidence.

Trusted by Procore, Navan, Morse, and companies of all sizes. Backed by $600B+ in payments.

Start building today: moderntreasury.com

1. FDIC floats a standard-setting body for banks' fintech partners

The FDIC is working with the ABA, ICBA, Bank Policy Institute, FTA, American Fintech Council and CFES on an independent standard-setting body for fintechs and third parties that partner with banks (Bloomberg). It would set baseline risk management standards and certify firms against them, with independent assessors doing the reviews. The standards are voluntary, there's no safe harbor, the FDIC is expected to provide seed funding, and the OCC is expected to join within days.

🧠 If you found Brainfood during the BaaS era, this is the third act of that story. In March 2023 I wrote BaaS is Dead. BaaS had an accountability gap (cue the spiderman meme), nobody could say who owned the regulatory risk. Then Synapse collapsed in April 2024 and locked end users out of up to $90m. It got so bad the CFPB paid $46m from its civil penalty fund because the reconciliation could never be finished.

🧠 CFES was an initiative to plug this gap. FS Vector and Sima Gandhi built the Coalition for Financial Ecosystem Standards (CFES) in 2024, writing guidelines across BSA/AML, marketing, complaints and operational risk. The STARC framework now covers 54 core standards across six compliance areas, rated on a CAMELS-style 1 to 5 maturity scale drawn from interagency guidance and OCC handbooks. Stripe, Block, Brex, Mercury and Relay signed up as members, and assessor firms (CrossCheck, FS Vector, FTI Consulting, Wipfli) are now trained and certifying against it. CFES + 4

🧠 Banks were dubious when this started. I remember pitching the idea to a room full of banks in Vegas. The doubts were loud. Fintechs joined first because they were the ones whose banks were landing under consent orders. Those consent orders meant no new customer accounts could open

🧠 Two years on, the FDIC may seed-fund the thing. The regulator that answered Synapse with a recordkeeping rule everyone hated is now backing certification instead.

🧠 My favourite detail: Jelena McWilliams floated this idea as FDIC Chair back in 2020. It went nowhere. Then she became the Chapter 11 trustee in the Synapse bankruptcy, personally tallying the missing money a standard like this might have caught. Few people alive have better proof it's needed.

🧠 What I love is how this has united the houses of westeros. ABA, ICBA, Bank Policy Institute and the fintech trades FTA and AFC are usually at loggerheads not collaborating. This is the cleanup that comes after the mess. The hard work that rarely gets applause. Well, lets fix that. Well done Sima, CFES, Kayce, Dan and everyone behind the scenes that worked to make this happen.

🧠 I suspect crypto and prediction markets will go through this cycle. Stablecoin linked cards will blow up or have a giant AML issue. Prediction markets will have some giant public outcry. And then, and only then, they’ll get fixed too. That’s what operators do.

2. Block processed its first Square transaction on its own bank rails

Block reported Q2 on August 5th. Gross profit was $3.17bn, up 25%, with Cash App at $1.97bn (+31%) and Square at $1.16bn (+13%). Adjusted operating income came in at $864m on a 27% margin, up 57%. In June, Square Financial Services, "processed its first Square acquiring transaction," and that Block expects "to shift more Square and Cash App acquiring volume to SFS" over the coming years. Square did $72.8bn of GPV in the quarter.

🧠 The company raised guidance based on its AI investments paying off. Block took a lot of heat for their layoffs. But the company says engineers are shipping 150% more code than 6 months ago, and 3x more features than a year ago. Shipping velocity helps grow revenue. That shipping velocity also helps you bring things in house…

🧠 Today four third-party payment processors handle Square volume. Today those processors are holding 44%, 14%, 14% and 11% of settlements receivable.

🧠 Block is going down the stack to bring that in house. They have their own captive bank to process transactions through and now the beginnings of their own processor. That all means better unit economics over time.

🧠 It's the Cash App Borrow play again. When Borrow originations moved into Square Financial Services, the product stayed exactly the same and Block got to own more of the economics. The Utah charter now does double duty, making the lending cheaper and the processing cheaper.

3. Chime posted its second straight quarter of GAAP profit

Chime reported Q2 on August 5th. Revenue was $670m, up 27%, at an 89% gross margin. Net income came in at $28m on $102m of adjusted EBITDA. Active Members reached 10.4m, up 20%, running $38bn of purchase volume, with ARPAM of $260. MyPay, the payroll advance product, originated $4.5bn in the quarter at a 0.9% loss rate and threw off $73m of transaction profit, more than triple a year ago. Chime raised full year guidance to $2.73bn of revenue and $470m of adjusted EBITDA, and acquisition payback is down to 5 to 6 quarters at 9x LTV to CAC.

🧠 A branch bank spends $250 to $400 a year just keeping a checking account open. Chime's entire revenue per member is $260. For legacy banks this segment is unprofitable. So they made that up with minimum balances, maintenance fees, $35 overdrafts. The segment paid its way through punishment and hated its bank for it.

🧠 Chime genuinely disrupted that. Every swipe earns interchange, and because its partner banks sit under the asset threshold they're exempt from Durbin's caps, so Chime earns more per debit swipe than JPMorgan is legally allowed to. The merchant funds the account, one transaction at a time.

🧠 For Chime more transaction is more revenue, and they do REALLY well here. The average Chime member transacts 53 times a month. Cash App, the nearest fintech, is at 24. Cash App has 59m monthly transacting actives to Chime's 10.4m and earns about $134 of gross profit from each a year. Chime earns $229. It’s hard to argue Chime is anything other than the “top of wallet” for their members.

🧠 They've now found a model for lending too. MyPay advanced $4.5bn against paychecks last quarter and lost 0.9% of it. Chime watches the payroll land, so it underwrites live cash flow. "Subprime" describes the credit file. The repayment behavior reads prime.

🧠 It turns out the unprofitable customer was a design choice all along.

🧠 Will Chime go Prime? Its fastest-growing segment earns $75k+. If that holds at scale, the next fight is for the customers banks actively want to keep, which is a segment banks would fight harder to keep. I wonder how hard they’ll push there, or if they can stay in their line and squeeze out more value?

4. Polymarket is raising at a $20bn+ valuation

Bloomberg reports Polymarket is in early talks to raise roughly $1bn at a valuation north of $20bn, which would be its third repricing in ten months after $9bn last October and $15bn in April, when ICE, the owner of the New York Stock Exchange, wrote a $600m cheque. The company turned trading fees on in January, having charged nothing to trade for the five years before, and by late June it told CNBC annualized revenue had passed $1bn. Rival Kalshi raised at $22bn in May and is chasing $40bn next per the FT.

🧠 Zero to $1bn of annualized revenue in six months of charging fees. The volume was always there; Polymarket just gave it away for five years to win liquidity. It was also timed with the US launch. Polymarket was offshore for a long time. Go on-shore vs a competitor that’s charging; it's logical that you would too. And now that revenue justifies a valuation.

🧠 ICE is buying into a new data business. ICE built its business by buying exchanges and selling the data that comes off them, and Interactive Data, Ellie Mae and Black Knight are the receipts. It backs Polymarket for the same reason. Live odds on everything become reference data, like index levels or the VIX.

🧠 Regulation is still very uncertain. The CFTC fined Polymarket $1.4m in 2022 and forced it to block American users. Three years later it licensed the company's US exchange, which went fully live in May, and it now reportedly has a probe open into Polymarket's social media promotion. Several state gaming regulators have already sent cease and desist letters over sports contracts, and consumer advocates see betting products that escaped the gambling perimeter.

🧠 Wall St clearly believes prediction markets are better markets. The states and consumer advocates don't yet. Prediction markets settle in real time, price in public and let anyone hedge anything, which is exactly what an exchange group wants to own and exactly what a gaming regulator wants to supervise. But with this level of growth, who's gonna stop them?

📎 If you want to know more about what is happening in prediction markets… check out this 70 page report for free from the team at Predicted. It’s killer. Read for free here!

5. The CFTC warned against displaying American style gambling odds.

First reported by Bloomberg, the CFTC told all prediction markets they “need to comply with US laws governing derivative trades and not use ‘deceptive’ practices to list, solicit or advertise the products.” The letter “called out as prohibited the use of American-style gambling odds to market a wager,” where venues use a plus or minus sign followed by a number that represents the potential winnings based on betting a certain amount.

🧠 The CFTC is aggressively suing states that claim event contracts are prediction markets, NOT gambling. So when someone blurs the line, they have to be the cop on the beat making sure that doesn’t happen.

🧠 The CFTC has issued 500+ pages of new regulations in the past few months. It’s clear they’re working overtime on covering any gaps that appear in events contracts.

🧠 DraftKings displays using this style. You have to wonder why? If anyone has lost volume and market share to the prediction markets, it’s probably them. Did they want this to happen?

6. Bunq denied Charter by the OCC

European Neobank Bunq, which has more than 20 million users across Europe and focuses on the global citizen user base, had its charter application denied by the OCC with two primary concerns. Firstly, the bank would not be adequately capitalized; it appeared Bunq’s founder would privately capitalize the bank with $50m, but using a dividend payment from the parent company. The OCC did not think this would be sufficient, nor did Bunq disclose a clear enough path for that funding. Secondly, the management team did not have experience with the US market and its regulations, especially given the breadth of lending products it intended to offer.

The OCC also did not believe the bank would be operated in a safe and sound manner, with its allowance for credit losses below those of credit card programs offered by banks supervised by the OCC, and the regulator believed more marketing would be required to achieve user acquisition given how competitive the US credit card market is.

🧠 You either have the right management team, or you have no charter. The management team was a major complaint in the rejection of Wise’s application for a charter too. If you look at Augustus, Erebor, and the countless trust charters, the one thing they all do is hire a former US bank CEO or similar caliber talent in all of the key roles.

🧠 The specific objection to the CEO pick is that he "plans to spend a majority of the year outside the United States." Bunq's whole proposition is borderless living. Niknam said in January that "our users are building their lives across borders." But that’s just now the OCC rolls.

🧠 This is not about nationality. Nubank, Brazilian-founded and Brazilian-controlled, got conditional approval on January 30th, 121 days after filing, and 2026 has produced roughly a dozen approvals from more than two dozen applications.

🧠 The letter says the denial "does not prohibit the filing of a de novo charter application in the future," and Bunq says it plans to address the concerns. A capital plan can be rewritten. A character and fitness finding follows the same management team into the next filing. So attempt three is a different question: does Bunq change the paperwork, or the people?

7. Cloudflare gave AI agents an identity and a wallet

Cloudflare announced Cloudflare Wallets aimed at giving AI agents the ability to spend online. The wallets allow anyone with a Cloudflare account to claim a handle at cloudflare.pay. Behind the handle sits a “Wallet” that holds stablecoins. Users can then issue Virtual Wallets to individual AI agents, who get a spending cap, a merchant allow list and a maximum transaction size. Agents pay per request over x402.

🧠 Cloudflare is making a play to be the monetization gateway for the internet. Its product, literally called Monetization Gateway, lets any site charge agents per request. This week it opened the waitlist for the wallet those agents spend from.

🧠 Getting people to claim a handle is easy. Getting merchants, AI companies and everyone else to accept those wallets is a whole other battle, and nowhere near as easy as it sounds.

🧠 Where’s the demand? Companies like Ramp and Meow already offer full accounts and virtual cards for agents. These products are not flying off the shelves. But there’s broad conviction that they one day will, and its smart to build for the future.

🧠 The wallet ships “in the coming months.” But we don’t know what it can and cannot do. The only terms published so far are a single line saying Cloudflare can reject any handle reservation, for any reason. That sounds like DNS language. But building agent identity and a rulebook for payments? That’s much harder. Maybe the leadership team has started to read one from many. Maybe not.

🧠 NET Dollar goes unmentioned. Cloudflare announced its own stablecoin in September 2025 and this launch does not mention it once. The docs default to USDC, settled through Coinbase’s facilitator, and Circle said in its results this week that 99.3% of x402 agent payment volume already settles in USDC. Whether those two things are related is the question, and we will find out soon enough.

1. Saible - The construction finance infrastructure

Saible provides embedded bank accounts and wallets for construction projects to secure and automate funding and contractor payouts. All funds are ringfenced in a controlled payment account, and it bakes in security and compliance oversight that projects and lenders need.

🧠 Construction projects lose millions to poor payment processes. They’re often overseen by project managers who have sites, contractors and other things to manage, not finance professionals. Baking in the banking controls to infrastructure makes sense. Under the hood this is quite clever too; it's a real bank account powered by Griffin, the API-first bank in the UK. We could start to see more of these where its less “embedded finance” (i.e. Neobank like structures on a small bank) and more directly managed API-first bank accounts.

2. Eversettled - The estate settlement AI

Eversettled uses AI and human support to guide bereaved families through the process of settling the estate of a deceased family member. It builds an estate checklist and helps families prioritize against deadlines and keep all of the documents together in a single place. The service can automatically scan to uncover bank accounts and close out subscriptions.

🧠This competes with expensive attorneys or DIY that makes mistakes. Often with a larger estate, banks will want you to secure probate and proper documents before releasing funds. Trying to DIY is painful, slow detective work, where you’re probably missing something. AI tooling to do the discovery and guide you by the hand with expert support is a good middle ground. Plans start at $1499, which isn’t cheap and could scale higher for complex estates. And much of the value of an attorney is handing it to them and letting them get on with it for $5k to $10k. If you’re very price-conscious, this could make sense.

3. Glacis Labs - A clearing house for stablecoins and tokenized assets

Glacis Labs runs ZeroDelta, a clearing house where KYB-verified institutions settle stablecoins and tokenized assets across 40+ chains. Trades match against offsetting institutional flow, and tokens move by burn-and-mint rather than through liquidity pools, so size doesn't move the price. Issuer rules like KYC requirements, whitelists, and transfer restrictions run automatically on every transaction.

🧠 Clearing houses make sense. There are many stablecoins, many rails, and many payments; all of those are sent in full. If you have $10 to send, and $9m to receive from all counterparties, clearing means you just settle the different (net) of $1m. That is a treasury team’s dream, and one they’re often willing to give up on “instant” payments for. The promise here is solid, get instant payments and netting. The challenge is, Better Money Company and Ubyx are also doing this (with strong networks), TCH is doing it with deposit tokens too, so it’s going to be hard to win on scale or efficiency. Focusing on RWAs is a smart wedge; that’s where the efficiency is needed.

4. Risk Ledger - LinkedIn for supplier security

Risk Ledger lets security teams run third-party risk as a network instead of a questionnaire. Suppliers complete one standardized security profile, free, and share it with every client; buyers connect and see it alongside 16,000+ organizations already on the platform. Because each supplier maps its own suppliers, you can trace 4th, 5th and 6th-party dependencies and spot concentration risk.

🧠 Third-party risk tools have promised to kill the security questionnaire for a decade, and the questionnaire is winning. The clever bit here is the graph: suppliers join free and map their own suppliers, which surfaces the 4th and 5th-party concentration risk DORA, and the Bank of England now expects firms to understand. But the answers are still self-attested. Still, this is interesting, and I imagine, broadly applicable to similar issues in the US.

What it takes to run JP Morgan after Jamie

Jamie Dimon has been the CEO of Chase for 20 years; he’s now 70. This is a bank with a balance sheet of $5 trillion and $63 trillion of dividend exposure. The Bank of International Settlements ranks JP Morgan as the most complex bank on earth. Marc says he believes Jamie can walk through every line item on that balance sheet with ease. The inventor of the “fortress balance sheet” is expected to stay for at least three more years, with supposed successors having gone elsewhere in the industry. Who’s to say if that will happen to current frontrunners Doug Petno or Troy Rohrbaugh?

🧠 How do you replace the GOAT? You don’t; you build a franchise that can win. With 320,000+ staff, JP Morgan is massive; it has systems, principles, and technology that are embedded deep into the org. 

🧠 People stay in the job because they enjoy it. Jamie is obviously doing his life's work. And unlike Dalio, who’s gone into publishing and repeating principles, Jamie is on the field but providing color commentary too. I hope we never lose that. I learn from every annual letter and every interview.

If you enjoy this kind of content, I can guarantee you’ll love being in a room of 1,500 other folks who love to go deeper into where finance meets AI. That’s a huge theme for us at this year’s Nerdcon in San Diego on the 19th and 20th November. I’m bringing my audience, the operators, the people who read this newsletter. And it’s the perfect place to find your next hire, client, or just get inspired. Let’s make events awesome again.

Remember, if you're enjoying this content, please do tell all your fintech friends to check it out and hit the subscribe button :)

Want more? I also run the Tokenized podcast and newsletter.

(1) All content and views expressed here are the authors' personal opinions and do not reflect the views of any of their employers or employees.

(2) All companies or assets mentioned by the author in which the author has a personal and/or financial interest are denoted with a *. None of the above constitutes investment advice, and you should seek independent advice before making any investment decisions.

(3) Any companies mentioned are top of mind and used for illustrative purposes only.

(4) A team of researchers has not rigorously fact-checked this. Please don't take it as gospel—strong opinions weakly held

(5) Citations may be missing, and I’ve done my best to cite, but I will always aim to update and correct the live version where possible. If I cited you and got the referencing wrong, please reach out