For most of fintech's history, being a bank was the thing you couldn't have. Banks had cheap deposits, a regulatory moat, and the nerve to charge you for service you were certain you could improve. Every founder I met between 2012 and 2022 wanted in. The regulators said no. So the industry built around them: sponsor banks, program managers, BaaS.

Then, this year, fintech got in.

American Banker counts more than 30 fintechs, crypto companies, and lenders through the federal charter process in 2026 alone. Erebor was the first new national bank the OCC had chartered in four years. Mission Lane filed the first credit card bank application the OCC had seen in about two decades. Comptroller Jonathan Gould said his agency would "no longer have a de facto no policy," and he meant it.

Be careful what you wish for. A charter can get you cheaper money, more control, and a moat. It also gets you capital rules, examiners, audits, and limits on what the company that owns the bank is allowed to do. And after all that, direct access to the Federal Reserve is a separate application to a separate institution that can take years and may never come.

So when you see the blizzard of "X gets a bank charter" headlines, it's easy to assume they're all doing the same thing. They aren't. Circle's bank can't take a deposit. Block's bank can, but can't give you a checking account. Mercury's bank can do both, and the price is that Mercury the company becomes a bank holding company.

Not all charters are created equal. To read the wave, start with what each company wants.

It usually starts with a customer request. Can you support RTP? FedNow? Push-to-card? Stablecoins? Then comes the next one. Because new ways to move money don’t replace what came before. ACH, wires, and even checks are still part of the mix.

Without the right foundation, every new rail means another vendor, integration, and more time before you can ship.

Modern Treasury changes that. With one unified API, you can build across today’s payment methods and be ready for whatever comes next. So as the way your customers move money evolves, your product can evolve with them.

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

Start building today: moderntreasury.com

Everyone applying wants one of three things. Sometimes two. Few want all three badly enough to pay for them by becoming a bank holding company.

Each type of charter gets you a different bundle of those three, with different consequences.

Two things are missing from that picture.

The GENIUS Act created a brand new federal license built for stablecoin issuers. It's not on the tree because it isn't open yet (the OCC's implementing rule was proposed in February and is still a proposal), and because every company you'd expect to want it has applied for a trust bank charter instead. (More on that later)

The Fed is missing too. Direct access to Fedwire and FedNow is the thing everyone wants most, because it massively improves payments economics, but doesn’t come out of the box with your freshly minted charter. It's a separate application with its own queue, and right now that queue is paused for the uninsured trust banks.

What is it? A full national bank charter from the OCC, plus FDIC insurance, plus membership of the Federal Reserve. Deposits, loans, cards, payments, custody. Everything a bank can do, you can do.

Who's doing it. Mercury got conditional approval on 27 April, and its bank is headquartered in Salt Lake City. Erebor has been chartered, insured and open since February, and had roughly $4bn of deposits by the end of Q2. Augustus got its conditional approval in May, Nubank in January, Upstart in July. Revolut applied in March. Bunq applied in January and got a public no in August.

What's the business case? Control, first. Mercury spent years as the best business bank in America running on somebody else's charter. The approval lets it put Zelle in accounts, expand lending to businesses and individuals, and build its own payments infrastructure, without asking a partner. Erebor's whole thesis is that defense, AI and crypto founders can't get banked, so it banks them. Augustus wants to be the clearing bank for dollars. I wrote up Augustus and Erebor in July, so I won't repeat it, except for the line that matters for this essay: getting closer to the metal cuts the cost permanently.

Then the money. A full bank gets insured deposits to lend against, and it's the cleanest route to a real Fed master account, because insured banks sit in the Fed's lowest-risk tier. More on that below.

What are the trade-offs? One big one.

The Bank Holding Company Act. If a company controls a bank (roughly, 25% of the voting shares, control of the board, or a "controlling influence" over it, with stakes above 10% starting to trigger notices under Regulation Y), that company is a bank holding company.

Three things follow that designation.

Block owns TIDAL, a music streaming service, and Proto, which builds bitcoin mining rigs. Both are in its latest 10-K. Neither is closely related to banking. If Block Inc. became a bank holding company, they'd have to go. The same control test applies to every fund that owns a big slice of Mercury.

TL;DR:

Most founders want the bank in a corner. Ten percent of the group is regulated like a bank; ninety percent ships like a software company. The BHCA doesn't let you draw that line, and that's why most companies stop here. It's also why Mercury going the whole way tells you something. Its approval joins the national charter, FDIC insurance, Fed membership and a parent applying to become a bank holding company.

 Mercury wants to be a bank. The regulation is the point.

Upstart's FDIC and holding company applications are still pending, which is a useful reminder that "conditional approval" is permission to finish building. The bank isn't open yet.

But what if you don't want to be a bank, and you just want your loans to be cheaper?

What is it? An industrial loan company (ILC, or industrial bank) is a state-chartered, FDIC-insured bank. Utah has most of them, Nevada a few. It can take insured deposits, and it can lend. But what makes them most interesting to corporates who want to lend is a rule from 1987. The Competitive Equality Banking Act exempted ILCs from the definition of "bank" in the Bank Holding Company Act. The bank subsidiary is regulated like a bank by the FDIC and the state. The parent is not a bank holding company.

Since 1987, there has been an explosion of this type of charter, sometimes with pushback.

In 1987, there were 11 of them in Utah, averaging under $45m in assets. Senator Jake Garn wrote the exemption for industrial companies with a lending arm, which is how BMW, Toyota, Harley-Davidson, and GE ended up owning banks. By 2005, there were 58 with $213bn of assets.

Then Walmart applied. The FDIC got more than 12,600 comment letters, most of them opposed, and froze all new applications in July 2006. Walmart withdrew in March 2007 before Dodd-Frank froze them again from 2010 to 2013. After that, the FDIC approved nobody for six more years, until Square and Nelnet on 17 March 2020. Critics argue that tech companies could introduce solvency risk into banking, that big companies threaten community banking, and that regulation is unequal.

In retrospect, this now seems quaint given we have full-on de novo banks that are VC-backed. And for the time being at least, the door does appear open to new ILCs.

Who's doing it? Block has had Square Financial Services since 2021. PayPal applied in Utah in December. Affirm applied in Nevada on 23 January. Klarna applied in Utah on 6 July. And the car companies are back: GM and Ford approved in January, Stellantis in May, with Nissan currently pending.

What's the business case? Cost of funding. Go back to those Affirm numbers. Roughly 6% on warehouse lines and securitizations against a bank funding itself at under 2%. Affirm has $7bn of loans held for investment and paid $425m in funding costs last year. You don't have to move all of that onto deposits to change the P&L. You have to move some of it.

The second reason is more art than science: the market values Block, Affirm and Klarna as technology (or at least payments) companies. A bank inside the group, with the group outside the BHCA, lets them keep telling that story. And, while you could argue the P2P lenders did something similar only to see their multiples crushed, it's fair to say Affirm, Klarna, and Block are very different-scale businesses.

There’s also another critical nuance.

None of these companies stop using partner banks. Block still routes volume through other banks and other processors. Four third-party processors hold 44%, 14%, 14% and 11% of Square's settlements receivable. In June, Square Financial Services processed its first Square acquiring transaction, and Block says it expects to shift more Square and Cash App acquiring volume there over the coming years.

When Cash App Borrow originations moved into SFS, the product didn't change at all. Block just kept more of the economics. Klarna runs on WebBank today and will until Klarna Bank USA opens. Affirm says its bank "would complement" its existing bank partnerships.

The charter doesn't force Block to own everything. It gives Block the choice. Sell the loans you don't want on the balance sheet. Keep the ones you do. Move acquiring volume when the numbers say so, one processor at a time, instead of a big-bang migration that could take Square down for a weekend.

What are the trade-offs? Three.

First, an ILC with more than $100m of assets can't offer demand deposits. No checking accounts. They use NOW accounts and savings accounts instead, which feel identical to a customer but not to a lawyer.

Second, the FDIC still reaches the parent. Capital and liquidity commitments, a source-of-strength agreement, and the right to examine how the parent affects the bank. It's an exemption from the Bank Holding Company Act, and from nothing else. Ask anyone who works at Square Financial Services.

Third, the bank lobby hates it. The ICBA called the ILC a "regulatory loophole" the day it approved Stellantis. Congress gave the exemption in 1987 and froze it twice. It can freeze it again.

So lenders want cheap money.

The stablecoin companies want something else.

What is it? The OCC charters a national trust bank. It can hold and manage assets for other people (custody and fiduciary services), settle, and, under GENIUS, issue stablecoins. It can't take deposits. It can't lend. It isn't FDIC-insured. So "Circle becomes a bank" means Circle becomes a federally supervised custodian and issuer. That is a different animal from Mercury.

So why this when Congress wrote a purpose-built stablecoin license? Read the definition in 12 USC 5901(11). A "federally qualified payment stablecoin issuer" is one of three things: (A) a nonbank approved by the OCC, (B) an uninsured national bank chartered by the OCC and approved to issue, or (C) a federal branch of a foreign bank.

A national trust bank is an uninsured national bank. Option B is the trust bank.

So Circle, Bridge and Paxos didn't skip the “GENIUS license.” They picked the version of it that comes with custody powers, fiduciary powers, qualified custodian status, and the right to ask the Fed for an account. Option A, the "nonbank" route, gets you issuance and reserve management and nothing else, and it isn't open yet.

I'm genuinely curious whether anyone chooses option A once it opens. Why would you?

Who's doing it? Anchorage was first, in January 2021. Then, on 12 December 2025, the OCC approved five in one day: Circle, Ripple, Paxos, BitGo, and Fidelity Digital Assets. Since then, Bridge (Stripe), Crypto.com, Coinbase, Nomura's Laser Digital, Morgan Stanley, Sony and World Liberty have all had conditional approvals. Circle National Trust went live on 24 July. Kraken, Payoneer, Catena, Agora, Dakota and Zerohash have applications in.

What's the business case? Three things, and none of them is "we wanted to be a bank."

What about yield? GENIUS bans paying holders yield just for holding the coin, and it bans it for every permitted issuer, whichever option they picked. But if you are the stablecoin issuer, and you are the trust bank custodying those assets, then that yield is indeed yours. And, while the rules are still being worked on, it looks like there is at least some scope for white label issuance, or “issuing as a service” to pay yield to other corporates. (e.g., M0 creates Moneygram’s MGUSD, and Moneygram retains that yield, but doesn’t pass it to consumers).

What are the trade-offs? This is not a low bar.

Diogo Mónica, who co-founded Anchorage, told me on Tokenized that some people assumed the trust charter was easy mode because it wasn't a "full bank." He said the examinations, the audits, and the ongoing burden were anything but. It was still worth doing, because Anchorage became the one partner you could put in front of a board if you were doing anything serious in digital assets.

The receipts back him up. The OCC put Anchorage under a consent order in April 2022 for BSA/AML failures, fifteen months after chartering it. Acting Comptroller Michael Hsu at the time: "The OCC holds all nationally chartered banks to the same high standards." The order was lifted in August 2025, by which point Anchorage had spent most of its life as a chartered bank under it.

You know you're a real bank when you've been through a full enforcement cycle.

Then there's what you don't get. No deposits or lending, no FDIC insurance, and, as it turns out, the worst seat in the Fed's waiting room.

Wise explained why better than anyone.

When it connected directly to the Bank of England, Harsh Sinha said its blended cost per UK payment fell from 60p to 6p. Ten times cheaper, per payment, forever.

You’re also free from “deplatforming risk” and settlement risk, which Circle experienced a lot of in 2023 when SVB collapsed (and ironically, as many blamed “crypto” for being high risk, Circle came out unscathed in the long run despite the federally protected banks going under). If Circle had direct Fed master account access, its payments would settle without relying on any other bank.

The charter is not a guarantee. Far from it. An OCC charter makes you eligible to ask.

If you look at the Fed's own words: "by law, requests for access to Federal Reserve accounts and payment services are decided by Reserve Banks." And there are twelve of them. Each with its own president, supervision staff, and reading of the Board's guidelines. The Board in Washington writes the policy. Kansas City decided for Kraken. Salt Lake City sits in the San Francisco Fed's district, so Mercury's request lands there. Your relationships with a specific Reserve Bank matter, and that makes this closer to an art than a process.

Then the tiers.

The Fed's 2022 account access guidelines sort applicants into three.

Now map our three branches onto it.

The feature that keeps Circle's parent out of the Fed's reach is the same feature that puts Circle's bank at the back of the Fed's queue. To reach Tier 2, the parent would have to accept Fed oversight, which is the thing it chose a trust charter to avoid.

Minimum viable regulation has a price, and you pay it at the Fed.

And Tier 3 is currently frozen.

On 20 May, the Board proposed a new "payment account" and encouraged Reserve Banks to pause Tier 3 decisions until the policy is finished, which staff expect "on or before December 31, 2026." Comments closed on 27 July. This would allow institutions to clear and settle only; they’d get no interest on balances, no intraday credit, no discount window, a balance cap of up to $1bn, and a 90-day target for review once your paperwork is complete.

Exactly one crypto-native firm has got through. Kraken Financial, a Wyoming bank, got a limited account from the Kansas City Fed on 4 March. Vice Chair Bowman called it a pilot. Kraken's CEO told Congress it still isn't operational.

And the queue can freeze again. The Fed publishes a list of pending requests. A pause that "ends by December 31" can be extended by memo. A Board with different governors in 2029 can rewrite the guidelines. Custodia has been in court over a denial since 2022.

You can win the charter and still spend years waiting at the Fed, with no guarantee the answer is yes.

Brex applied for a charter in 2021 and withdrew it. Ramp has never applied. Both decided the bank is a utility and the money is in the software on top. If you're growing 100% a year, an examiner in your product roadmap is a cost you don't need yet, and "yet" is doing a lot of work in that sentence.

I’m not sure Ramp ships an accounting platform and a token router within 4 weeks of each other while going through a full charter application. You’ve also got to remember that the tides in Washington can turn. The current administration is more open to new applicants; a future one may be harsher on those that made it.

Line the three up side by side, and the pattern is hard to unsee.

The next time a company says it's becoming a bank, ask four questions.

If I want to lend, I go for an ILC. If I want to custody stablecoins, I go for a trust bank. If I want to be a better bank than the one I'm renting, and I'm willing to become a bank holding company to do it, I go the whole way. Whichever I pick, the Fed is a separate mountain.

Fintech got in. Getting in was the easy part.

Flynn got into the Grid and found out the Grid had a Master Control program running the place.

You just might get in, and what you find is a regulator.

ST.

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 18th to 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.

(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