Crypto trust charters and community banks: the binding decision sits in the core

The week

On December 12, 2025, the Office of the Comptroller of the Currency, the federal agency that charters and supervises national banks, conditionally approved five national trust bank charters for digital-asset firms. A trust bank is a specialized kind of bank whose job is to hold and look after assets on behalf of others; nobody opens a checking account at one.

For a community bank chief executive, the consequences could arrive within days: a fintech partner asking whether the bank will “support stablecoins,” a business customer who imports goods wanting to know if supplier payments could settle in digital dollars, a core technology provider hinting at a new module.

Two of the five were brand-new charters: First National Digital Currency Bank, owned by Circle, and Ripple National Trust Bank. Three were conversions of existing state trust companies into national ones: BitGo, Fidelity Digital Assets and Paxos. The activities each firm was allowed to carry out differed, and every one of them still had conditions to meet before receiving final approval. “Conditional” mattered: it was a green light with strings attached, short of the finish line.

One limit stands out. National trust banks cannot take deposits or offer accounts insured by the FDIC, the federal agency that guarantees ordinary bank deposits up to a limit. Whatever a customer held with one of these firms, it was not an insured deposit.

Not everyone welcomed the move. Bank trade groups had urged the OCC to reject the applications, arguing that the charters gave nonbanks a lighter route into the banking system without the full obligations traditional banks carry.

The timing also mattered. The approvals came five months after Congress passed the GENIUS Act, which created a federal framework for payment stablecoins. A stablecoin is a digital token designed to hold a steady value, usually one US dollar, and backed by reserves such as cash and short-term government debt; “payment stablecoins” are the ones meant to be used for moving money. When the charters were approved, regulators were still drafting the detailed rules needed to put that law into practice. So the week delivered new, federally supervised players into a market whose rulebook was only half written.

For the largest banks, the December approvals were a competitive signal: new rivals with federal credentials had arrived in a business the giants were already weighing. For community banks and credit unions, the meaning was more practical. Federally supervised counterparties for digital-asset custody, and for some of these firms stablecoin activity, were about to exist. That meant fintech partners, business customers and core technology providers would soon be asking what the institution planned to do.

Bank crypto questions tend to arrive disguised as custody questions: a customer asking the bank to hold something, rather than a strategy paper on entering crypto. The charters made custody more ordinary. They left a smaller institution with three decisions, which work best taken in order: what scope to accept, who holds the assets, and how any of it connects to the core. The third is the one that binds, and it gets the most attention below.

Ledgers, tokens, keys and the core

The pieces are less exotic than the headlines suggest.

A digital asset is a record of ownership kept on a blockchain. A blockchain is a shared ledger, copied across many computers, where new entries are added in batches and, once accepted, are extremely hard to change: if one copy is rewritten, all the others disagree with it.

A stablecoin is a digital asset designed to stay worth a fixed amount, typically one dollar, because the issuer holds reserves to back it. The promise is a dollar that can move at any hour, on any day, directly between digital wallets.

A wallet is not a container of coins so much as a set of keys. Each wallet has a public address, like an account number anyone can send to, and a private key, a long secret number that authorizes spending. Whoever controls the private key controls the asset. Lose the key and the asset is effectively gone; leak it and someone else can move the asset in seconds.

Custody, then, is mostly the business of protecting keys: generating them securely, storing them in specialized hardware, requiring several people to approve a transfer, and keeping careful records of who owns what.

The core is the bank’s central system of record, the software that holds every customer’s accounts, posts every transaction, calculates interest and feeds the general ledger. The general ledger is the master set of accounting books from which financial statements are produced. Most community banks and credit unions do not build their cores; they rent them from a small number of large core processors. Everything customers see, from the mobile app to the monthly statement, depends on it.

Scope: which activities, if any

A trust charter does not make a stablecoin a deposit. That single fact shapes how a community institution should frame the whole question. Holding a payment stablecoin for a customer differs from offering an insured account, and the disclosures, risk appetite and board approvals differ accordingly. A customer who assumes otherwise is a complaint, and possibly a lawsuit, waiting to happen.

The options form a ladder:

1. Do nothing, and say so clearly when asked.
2. Accept stablecoin-funded transfers from a partner, so money that started as a stablecoin arrives in the bank as ordinary dollars.
3. Settle some business payments on a stablecoin rail, meaning the bank itself sends or receives stablecoins for selected clients.
4. Let customers hold tokens through a custodian, with the bank presenting those holdings as part of the relationship.

Each step up adds obligations. Sanctions screening and anti-money-laundering monitoring must cover blockchain addresses as well as names; it is no longer enough to check that “Acme Imports” is not on a sanctions list if the money came from a wallet linked to a sanctioned party. Liquidity planning must account for redemptions outside banking hours, because blockchains do not close at 5 p.m. on Friday. Complaint handling must cover transfers that cannot be reversed: there is no chargeback on a blockchain, so a customer who sends tokens to the wrong address may simply lose them.

A one-page position, approved by the board, that names the chosen step and the reasons is enough at this stage. The charters made the higher steps more feasible. They did not make them necessary. Doing nothing, deliberately and in writing, is a legitimate strategy. Drifting into step three because a partner switched something on is the outcome to avoid.

Scope also sets the compliance build. Even the second step, accepting stablecoin-funded transfers, means the institution needs some way to screen the originating wallet. That usually means a blockchain analytics provider, a firm that traces the history of wallet addresses and flags links to known bad actors, or a partner that provides screening under contract. Someone in the BSA function, the team responsible for Bank Secrecy Act compliance, has to understand the alerts well enough to file a suspicious activity report, the confidential filing banks make to federal authorities when they suspect financial crime. Examiners will ask how the board reached its decision and how the controls were tested.

Each step up the ladder adds a vendor, a policy and a training need, and those costs are easier to count before the first customer asks than after. A useful exercise is to price each rung: what new contracts, what new staff skills, what new monitoring, what new reporting. Boards are often surprised that the second rung, which feels modest, carries most of the compliance cost of the third.

Custody: who holds the assets

Before December, a community bank that wanted to offer digital-asset services usually depended on state-chartered trust companies or less regulated providers. That was a hard sell to a board and to examiners. A national trust bank is a more familiar counterparty. It is examined by the OCC, it owes fiduciary duties, meaning a legal obligation to act in its clients’ interests, and it has a clearer resolution path than a technology vendor if it fails.

That familiarity lets custody run through a process the institution already has: the 2023 interagency guidance on third-party risk management, which sets out how banks should choose, contract with and oversee outside providers. Rather than inventing a special crypto process, the institution can apply its normal vendor discipline, and press hard on a handful of points:

• Segregation of client assets. Are each client’s holdings kept separately from the custodian’s own, and from other clients’, in a way that can be proven?
• Key management and who can authorize transfers. How are private keys generated and stored, how many people must approve a movement, and what stops one insider from acting alone?
• Insurance and its exclusions. What is covered, up to what amount, and, just as important, what is not? Policies often exclude the very scenarios boards worry about.
• Audit rights. Can the institution, or its auditors, verify what the custodian says it holds?
• What happens if a conditional approval is never finalized. A conditional charter is a promise rather than a certainty. The contract should say what happens to the service, and the assets, if final approval never comes.

Institutions should also ask counsel how customer assets would be treated in an insolvency. The answer depends on account structure and documents as well as on the charter. Two custodians with identical charters can produce very different outcomes for customers if one holds assets in clearly titled individual accounts and the other pools them under its own name.

The core: how it connects

This is the decision most institutions underestimate, and it is where good intentions meet operational reality. Most community banks and credit unions run on a small number of core processors. Anything they offer has to appear in online banking, statements, the general ledger, contact-center screens and regulatory reports. A product that lives outside those systems is, in practice, a product the bank cannot see, explain or report on.

Digital assets strain cores in three ways. Balances move at any hour, while most cores were designed around a nightly batch, a single overnight run that posts the day’s activity. Prices change by the minute for some tokens, while cores are built for amounts that stay put. And transaction identifiers look nothing like card or ACH references; ACH is the network that carries direct deposits and bill payments, and its references fit neatly into existing fields in a way that long blockchain transaction codes do not.

There are three broad patterns for connecting a custodian to the core:

1. Link out. The partner holds everything and the institution shows customers a link to the partner’s own app or website. The core stays untouched, which is cheap and fast. But the customer relationship weakens, and the institution sees little of the activity it is still accountable for.
2. Memo post or sub-ledger. Balances come into the core as a memo post, an informational line that shows a figure without moving money, or as a sub-ledger, a separate set of accounts that rolls up into the main books. Customers get one view of their relationship. But the institution depends on its core vendor’s roadmap, pricing and timing.
3. Thin integration layer. The institution builds a small piece of software of its own that pulls data from the partner, usually through an API, reconciles it and feeds the ledger and digital channels. An API, or application programming interface, is a structured way for one system to request data from another. This option gives the most control and requires engineering capacity many small institutions lack.

The first month on a memo balance

Suppose a community bank with a handful of commercial clients who import goods chooses the second rung of the ladder and the second integration pattern. On paper, it is simple: the custodian handles the tokens, the core shows a memo balance. In practice, the first month surfaces matters nobody wrote down: the time of day the balance is “as of,” what the contact center says when a client calls on a Sunday about a transfer that settled on Saturday, and which general ledger account a fee charged in a stablecoin lands in. Each is easy to settle, and each needs an owner.

Reconciliation: the control that matters most

Whatever the pattern, reconciliation is the control that matters most. Reconciliation means proving that two sets of records agree. Each day, the institution must show that what its systems say a customer holds matches what the custodian holds. It is the digital equivalent of counting the cash drawer at closing and checking it against the till receipts.

Doing that well needs a few things settled in advance:

• A defined cut-off time, because blockchain activity has no business day. The bank has to choose a moment, say 6 p.m. Eastern, at which it takes a snapshot of both sides and compares them.
• Pricing rules for any non-stablecoin asset, so everyone agrees which price source and which moment is used to value holdings that fluctuate.
• Break-handling with named owners. A “break” is a mismatch. Someone specific must investigate it, by a deadline, and escalate if it is not resolved.

Customer-facing documents need the same care. Statements need language that distinguishes custodied tokens from deposits, so no customer mistakes one for the other. Call report and other regulatory reporting treatment should be settled with accounting advisers before launch; the call report is the detailed quarterly financial filing every bank submits to its regulators, and an unclear line item there tends to become an exam finding.

The contracts tie all of this together. The custody agreement should specify data feeds, formats and timing that the integration layer can rely on; reconciliation is impossible if the custodian’s file arrives at unpredictable times in a format that keeps changing. The core contract may need amending to permit memo posting or new transaction types, and core vendors often price those changes separately. An institution that signs with a custodian before talking to its core provider can find the integration costs more than the product earns.

Why the order of decisions sets the cost

The sequence matters for that reason. Choosing a custodian before setting scope invites a product nobody approved. Choosing a custodian before checking the core’s capabilities locks the institution into whichever partner the core vendor happens to support. Most of the cost of getting this wrong shows up in operations, reconciliation breaks and exam findings rather than in the technology bill.

A well-prepared smaller institution has these in place:

• A board-approved, one-page scope statement exists before any vendor demonstration.
• Custodian due diligence runs through the normal third-party risk process, with the five pressure points above written into the file.
• The core provider has been asked, in writing, what it supports and at what price, before any custody contract is signed.
• Reconciliation cut-off, pricing rules and break owners are documented and tested with sample data before a single customer can hold an asset.
• Statement language and regulatory reporting treatment have been agreed with counsel and accounting advisers.

The failures are the mirror image. A partner switches on a feature and the bank discovers it is on rung three without a board vote. A custody agreement is signed on the strength of the charter alone, with no reading of the insolvency terms. The core is treated as an afterthought, and the integration bill arrives later. Or reconciliation is left to “the operations team” without anyone being named, so the first break sits unresolved for a week.

Each failure comes from governance and order of operations rather than technical sophistication, which puts them within any leadership team’s reach to avoid.

A bank ready to answer its fintech partner

If a fintech partner, a business customer or a core vendor raises digital assets tomorrow, a few questions show quickly whether the institution is ready to answer:

• Which rung of the ladder are we on today, and has the board approved it in writing?
• If a customer sent us stablecoin-funded money tomorrow, who would screen the originating wallet, and how?
• For any custodian we are considering: how are client assets segregated, who can authorize transfers, and what does the insurance exclude?
• What does our core provider support today for memo posting or sub-ledgers, and what would it charge?
• What is our daily reconciliation cut-off, which price sources do we use, and who owns a break?

An institution that can answer these in a sentence each is in a strong position, whatever it decides. One that cannot should be comfortable saying “not yet.”

With hindsight

The December approvals started a process rather than finishing one, and the months that followed showed how long the path from “conditional” to “open for business” can be.

Circle reported that it received final OCC approval for First National Digital Currency Bank on July 9, 2026, and that the bank opened on July 24 as Circle National Trust, initially offering digital asset custody to Circle and its affiliates. The conditional approval of December was followed by a long stretch of further work, and the first service offered was custody for the firm’s own group rather than a broad offering to outside banks.

The rulebook moved slowly too. The OCC proposed its rule implementing the GENIUS Act for stablecoin issuers under its jurisdiction in February 2026. The statute gave regulators one year from its July 2025 enactment to issue implementing rules. That deadline passed without a final OCC rule, and the Comptroller said in August that the agency would have one out by November. For a community bank board, that meant planning for a market whose formal rules were still not finished more than a year after the law passed.

For smaller institutions, the shape of the decision held. An institution with a written scope, custodian criteria and a view on core integration can answer a partner or business customer quickly and confidently. One without them risks letting a vendor’s roadmap decide, which is a poor way to make a strategic choice.

Two conclusions follow. First, conditional approvals and pending rules are reasons to prepare rather than to wait. Second, the preparation that pays off regardless of how the rules land is the unglamorous part: scope, vendor due diligence, core conversations and reconciliation design. With final rules still pending, the integration and reconciliation work remains the part that can sensibly be prepared now, and it is the part that will decide whether any future product runs smoothly or generates exam findings.