Nine Mega Banks Still Block Crypto Accounts. The OCC Just Wrote It Down.

Nine of the largest banks in America have been turning away crypto businesses, and as of this week there's a government record that says it out loud.

The Office of the Comptroller of the Currency logged crypto-related account restrictions across nine mega banks in its latest findings, and the detail that matters isn't the number nine. It's what the restrictions were built on. Not custody rules. Not capital rules. Anti-money-laundering programs, sanctions screening, and operational risk assessments. That's where the doors close.

How We Got Here

Rewind to March 2023. Silvergate wound down, Signature Bank was seized, and Silicon Valley Bank collapsed inside of a week. Whatever you think caused those failures, the practical effect was immediate. Banks with crypto exposure got nervous, and nervous banks stop opening accounts.

Through 2023 and 2024, founders swapped stories about closed accounts and unreturned calls. They called it Operation Choke Point 2.0. Regulators called it risk management. Nobody had paper.

Then the paper showed up. In January 2025, the FDIC released a batch of internal letters showing it had asked banks to pause crypto activity while it studied the issue. Not a rule. A request. That distinction is everything, because a request from your examiner carries the weight of a rule without any of the due process.

Two months later, the OCC moved the other way. It struck reputation risk from its supervisory handbook, which had been the quiet justification for every debanking decision nobody wanted to write down. Then came Interpretive Letter 1183, which confirmed that national banks can custody crypto assets, and 1184, which cleared a path for stablecoin activities. On paper, the door reopened.

Which brings us to the current findings. Nine banks. Still restricting. After the rules changed.

The Gap Between Rules and Behavior

Here's what I think is actually happening, and it's less sinister than the loudest voices on either side suggest.

Custody rules are easy to change. AML and sanctions compliance isn't. If you're a bank compliance officer, custodying bitcoin is a legal question, and legal questions have answers. Screening a crypto customer for sanctions evasion is an operational question, and operational questions have consequences.

Can your system trace a transaction two hops out from a mixer? Can you explain to an examiner why a customer in a high-risk jurisdiction passed your controls? Can you prove your blockchain analytics vendor isn't feeding you garbage? If the answer is no, you don't take the account. Not because you hate crypto. Because you can't defend the decision in an exam.

That's the asymmetry nobody talks about. The OCC can say yes. The bank's compliance department can still say no, and it usually wins.

So who loses here? Small and mid-sized crypto businesses, mostly. An exchange with a legal team and an audit trail gets a meeting. A ten-person DeFi startup with offshore entities and a founder who's been public about privacy tools gets silence. Granted, that's not fair. It's also how banking has always worked for anyone a bank finds inconvenient.

Who wins? The big players, obviously, and the custody banks that have spent four years building the compliance plumbing everyone else now needs. Also the analytics vendors. Every one of those nine banks that eventually wants to say yes has to buy a tool that lets it say yes. That's a real revenue line, and it's growing whether or not the headlines are good.

There's a market structure consequence too. When nine of the biggest depositories in the country won't bank a sector, capital formation migrates. It goes to state-chartered institutions, to trust companies, to offshore venues, to non-bank lenders. That migration doesn't reverse overnight just because a federal agency changes its tone.

The question worth asking: if the rules are permissive and the behavior hasn't changed, was the rule ever the binding constraint?

What to Watch

The next real signal isn't another interpretive letter. It's the exam cycle. Watch whether the OCC starts publishing guidance specifically on AML expectations for digital assets, because that's the actual bottleneck. A bank can't comply with a standard that doesn't exist, and right now most of the standards live in enforcement actions and informal conversations.

Second, watch call report data over the next two quarters. If crypto-related deposits start climbing at the banks that already have programs, that's evidence the logjam is breaking. If they stay flat while the rules stay permissive, the problem was never the rules.

Third, watch the enforcement side. One well-publicized AML penalty against a bank with a crypto program would undo two years of deregulation in a week. Compliance officers remember penalties the way traders remember drawdowns.

Color me skeptical, but I don't think nine banks restricting accounts is a scandal. I think it's a lag. Deregulation at the top moves faster than risk tolerance in the middle, and the middle is where accounts actually get opened. That gap usually closes in 12 to 18 months, or it doesn't close at all and the activity simply moves to institutions that were never subject to OCC supervision in the first place.

Time will tell, though. Meanwhile, if you're a founder who got a no in 2023 and a maybe in 2025, nothing about your business changed. Only the paperwork did.