The most consequential variable in the Modern Treasury trust charter story is a boolean. It has not been assigned yet. Somewhere in a regulator's queue โ the Office of the Comptroller of the Currency, a state banking department, or both โ sits a file whose entire meaning reduces to a single bit. Approved, or not approved. Everything else written about this event in the past week is commentary on a variable that does not yet hold a value.
I spent forty hours in late 2017 auditing a Solidity contract for a Sรฃo Paulo fintech, and I learned something that has nothing to do with Solidity. The team had already written the marketing copy. The token sale page was live. The only thing standing between them and two million dollars of user deposits was a withdrawal function I had not signed off on. The deployment was pending โ a state that in the EVM is functionally indistinguishable from blocked. Intent does not move a transaction. State does. The distance between a company's announcement and its operating reality is measured not in words but in state transitions, and Modern Treasury's announcement is a state transition that has not occurred.
This is where the reporting on the Modern Treasury trust charter bid fails. It treats "seeks a charter" as a fact with a direction. It is not a fact. It is a request. A request is an input. An input is not an output. The financial press has a structural bias toward treating inputs as outcomes, because inputs are announced loudly and outcomes arrive silently. The custody charter story is being written in the announcement phase, which is the phase with the least information and the most noise.
So let me do what I do with any contract before I discuss its economics. I read the interface. I look at what the function signatures actually promise, and โ more importantly โ at what they leave undefined.
The Interface, As Declared
Modern Treasury is not a crypto company. This is the first thing to establish, and it is the thing most crypto-native readers get wrong. Modern Treasury is a payments-operations company. Its product is a money-movement API and a ledgering layer that lets enterprise finance teams programmatically reconcile inflows and outflows across bank accounts. It sells to CFOs, not to degens. Its customers are the treasury departments of companies that already move real money through the traditional banking system.
That identity matters because it determines what a custody charter would mean in their hands. For a crypto-native firm, a trust charter is a coming-of-age ritual โ the moment a protocol or exchange grows up and accepts fiduciary duty. For a payments company, a trust charter is something else entirely. It is a vertical integration move. It is the decision to stop renting custody from a partner and to build the regulated entity that holds the keys.
Here is the core insight the reporting misses: Modern Treasury is not entering custody because custody is exciting. It is entering custody because its existing customers' money is increasingly denominated in assets that its existing infrastructure cannot hold. A treasury platform that can reconcile dollars but not stablecoins is a treasury platform with a growing blind spot. The charter is not an expansion into a new market. It is a defensive patch against a competitor who already solved the reconciliation problem for digital assets.
I have watched this pattern before. In 2020 I built a Python simulation of ten thousand ETH/USDC price paths to quantify impermanent loss for liquidity providers. What that exercise taught me was not the arithmetic of x*y=k. It was that the demand for a financial primitive is set by the constraints of the adjacent system, not by the primitive itself. Liquidity providers did not appear because automated market makers were elegant. They appeared because the order-book model could not serve long-tail assets. The primitive was a response. Custody, in 2026, is the same kind of response. The demand for regulated digital-asset custody is being manufactured by the fact that enterprise balance sheets now contain digital assets and enterprise auditors now require a custodian with a charter. The charter is downstream of an accounting problem, not upstream of a technology problem.
What a Trust Charter Actually Is
The word "charter" carries an almost mystical weight in crypto discourse, as if it were a cryptographic primitive. It is not. A trust charter is a license to perform a specific legal function โ acting as a fiduciary โ under a specific supervisory regime. That is the whole of it.
There are two flavors, and conflating them is the most common analytical error in this space. A national trust charter is granted by the OCC and permits the institution to operate as a trust company across state lines under federal supervision. A state trust charter is granted by a state banking department โ New York, South Dakota, Wyoming, Nevada โ and confines the institution's fiduciary powers to that state's framework, though it can often serve out-of-state clients through specific provisions. The difference is not cosmetic. It determines the capital regime, the permissible activities, the examination cadence, and the reciprocity of the license.
Anchorage Digital received a conditional national trust charter from the OCC in January 2021, becoming the first federally chartered digital asset bank in the United States. BitGo holds a South Dakota trust charter. Coinbase Custody operates under a New York limited-purpose trust charter. Fidelity Digital Assets holds a New York trust. Paxos and Gemini hold New York trust charters. The point of enumerating these is not to impress you with a list. It is to make a structural observation: the custody charter is not a frontier. It is a settled, competitive, well-understood licensing category, and Modern Treasury is applying to enter it roughly five years after the first mover.
That late entry is the single most under-discussed fact in the coverage. In a novel regulatory category, an early charter is a moat. In a settled regulatory category, a charter is a table stake. The moat has already been filled in by the firms that arrived first, and what remains is a commodity licensing regime where the differentiator is not the charter but what you do on top of it.
The relevant technical question, then, is not "will they get the charter." It is "what custody architecture will sit behind it." A charter without a custody stack is a legal wrapper around nothing. A custody stack without a charter is a technical capability that institutional clients cannot buy. The two have to be evaluated together, and the reporting on Modern Treasury evaluates neither.
The Custody Stack That Nobody Is Discussing
Here is where my audit background forces me to slow down, because this is the part of the story that contains actual engineering, and it is the part that is entirely absent from the announcement.
Digital asset custody at institutional scale is not a database of private keys. It is a key-management architecture, and the architecture is a security model. There are three families, and each one encodes a different trust assumption.
Single-key with HSM. A hardware security module stores one key, or one shard of a key, in tamper-resistant silicon. The trust assumption is that the HSM vendor's silicon is not backdoored and that the operator's physical security is adequate. This is the oldest model and the least resilient, because a single point of compromise drains everything. It is used almost nowhere at institutional scale anymore, but it persists in the compliance narratives of firms that want to say "hardware security module" without saying anything else.
Multisignature. An on-chain m-of-n scheme, typically an M-of-N threshold over distinct keys. The trust assumption is that the signers are geographically and organizationally separated, and that the quorum policy cannot be coerced or socially engineered. Multisig is transparent โ the policy is auditable on-chain โ but it leaks metadata about signers and quorums, and it is awkward for high-frequency operations. I have seen multisig deployments that were mathematically sound and operationally catastrophic, because five keys were held by five people who all reported to one person.
Multi-Party Computation (MPC). The key never exists in full. It is split into shares, and signatures are produced through a distributed protocol where no single machine ever reconstructs the whole. The trust assumption is that the protocol is correctly implemented and that the share-distribution policy is sound. MPC hides the quorum from the chain โ a meaningful operational security gain โ but it introduces a new attack surface: the MPC protocol implementation itself, and the orchestration layer that decides which shares to combine.
The choice among these is not a preference. It is a statement about which failure modes the custodian has decided to accept. A charter does not select the model. The company selects the model, and the company has said nothing.
I can construct the questions that any competent auditor would ask before signing off on a custody deployment, and I can show you that the announcement answers none of them:
- What is the quorum policy, and who controls the shares? If a single executive can assemble a quorum, the MPC is theater.
- Is there a cold-storage tier, and what is the withdrawal latency from cold to hot? Institutional clients care about the
withdrawalfunction's gas cost and its human cost. - What is the key-generation ceremony, and was it witnessed? A key ceremony with one witness is not a ceremony. It is a formality.
- Is there insurance, and against what โ theft, loss, internal collusion? An uninsured custodian is a leveraged bet on its own operational discipline.
- Is there a slashing-like condition โ a scenario in which the custodian's own capital is at risk for an operational failure? If not, the custodian's incentives are misaligned with the client's.
Logic is binary; intent is often ambiguous. The announcement's intent is legible โ enter custody, capture the flow, defend the treasury franchise. The logic is absent. There is no architecture, no quorum policy, no ceremony, no insurance disclosure, no capital line. A reader who treats the announcement as a technical event has been handed an empty struct.
This is not a rhetorical complaint. It is the same defect I found in fifteen ERC-721 minting contracts in Brazil in early 2021. Two of them had open minting โ no access control on the mint function, so anyone could mint the entire supply. One used block.timestamp as a randomness source, which meant a miner could front-run the reveal. The projects' marketing was impeccable. The contracts were public functions with no onlyOwner modifier. The gap between the interface a project presents and the interface a project implements is where every material vulnerability lives. Modern Treasury's charter announcement is an interface. The custody architecture is the implementation. I will not evaluate the implementation until I can read it.
The State Machine of a Charter Application
Because the announcement gives us no technical detail, the only rigorous analysis available is procedural. A charter application is a state machine, and the states are not what most people assume.
The naive model is binary: applied, then approved. The real model has more states, and each transition has a cost and a failure probability.
State 0 โ Intent. The company decides to pursue a charter. This is the state the public announcement describes. It requires no capital, no approval, and no commitment. It is the cheapest state to occupy and the easiest to announce. A press release can move a company into State 0 without moving a single dollar or filing a single page.
State 1 โ Pre-filing engagement. The company meets with the regulator informally, often through counsel, to gauge receptivity and scope. Nothing is public. Most applications that die, die here โ quietly, before anyone knows they existed. There is no public record of this state, which is precisely why an announcement in State 0 tells you nothing about whether State 1 has begun.
State 2 โ Formal filing. The application is submitted with a business plan, capital plan, management roster, compliance program, and โ for a trust company โ a description of the fiduciary activities. This is a public event in most jurisdictions, and it is the first point at which an outside observer can verify that the company is serious.
State 3 โ Conditional approval. The OCC, in particular, grants charters conditionally. Anchorage's 2021 charter was conditional. Conditional approval is not approval. It is a promise to satisfy pre-opening conditions โ capital, personnel, systems โ before the charter becomes operative. Many conditional approvals never reach the next state.
State 4 โ Pre-opening conditions satisfied. The capital is raised and deposited, the compliance officer is hired, the systems are built, the examiners are satisfied. This is where the real work happens and where most of the time is spent. It is invisible from outside.
State 5 โ Operating. The charter is live. The entity can act as a fiduciary. Only now does custody become a product.
Read the announcement again with this state machine in mind. It does not specify the state. It could be State 0. It could be State 1. It could be State 2. The announcement is written to be consistent with all of them, which is exactly what you would expect from a communication designed to maximize the appearance of progress while minimizing the commitment to any specific claim.
I have no way to verify which state Modern Treasury occupies, and neither does anyone reading the coverage. That is not a failure of my analysis. It is a property of the input. When the input is a single fact and two opinions, the output is a probability distribution, not a point estimate.
What I can say with confidence is that the historical base rate matters. Crypto-adjacent trust charter applications are not fast. Anchorage's path from engagement to operative charter spanned years and multiple regulatory postures. The OCC's receptivity to digital asset charters has oscillated with each administration โ expansive under Brian Brooks in 2020, restrained under subsequent leadership, and reopened by the interpretive guidance that clarified banks may hold crypto in a custody capacity. An applicant's success probability is a function of the regulatory posture at the moment of decision, not the moment of announcement. A company that announces in a friendly posture may be decided in an unfriendly one, and the announcement cannot hedge against that.
The Economics of Custody, Stripped of Narrative
Custody is sold as a trust product. It is priced as a commodity service. This contradiction is the engine of the entire custody business, and it is worth dissecting carefully.
Institutional custody fees are quoted in basis points of assets under custody, typically single-digit to low-double-digit basis points per annum for large balances, with tiering that compresses toward single digits as balances grow. A custodian holding ten billion dollars at five basis points earns five million dollars a year. That is not a large business. It is a small business with a large balance sheet and a regulatory cost structure.
The economics only work at scale, and the scale is achieved through two mechanisms: the accumulation of assets under custody, and the cross-sale of adjacent services. This is why the custody charter is a strategic move for a payments company and a strategic necessity for a pure custodian. For a pure custodian, the basis-point fee is the whole business. For a payments company, the custody charter is a loss leader that protects the fee-bearing payment rail. If custody keeps the customer's digital assets inside Modern Treasury's reconciliation engine, the payments business retains the customer. If custody leaks to a competitor, the payments business loses the customer too.
The charter is not a revenue line. It is a customer-retention mechanism disguised as a revenue line. This is the economic reading the announcement invites and the coverage ignores. A company does not spend years and millions pursuing a low-margin fiduciary license unless the license is defending something with a higher margin.
There is a second economic layer that almost no one discusses. A trust charter imposes capital requirements. The institution must hold a certain amount of regulatory capital against its fiduciary activities, and that capital has a cost. The cost is the opportunity cost of the capital, plus the cost of the compliance apparatus โ examiners, auditors, legal, insurance. For a well-capitalized fintech, this is manageable. For a thin-margin one, it is a drag. The announcement does not disclose the capital plan, which means an outside observer cannot compute whether the custody business is accretive or dilutive to the parent. You cannot evaluate a regulated business without its capital line, and the capital line is missing.
I have seen the same omission destroy a project from the inside. In the Uniswap V2 analysis I ran in 2020, the constant-product formula was transparent โ anyone could read x*y=k and compute the fee revenue against the impermanent loss. The protocol's economics were auditable because the mechanism was public. The custody charter is the opposite. The mechanism is a private application, the capital is a private plan, and the fee schedule is a private negotiation. Opacity is not a bug in the custody business model. It is a feature, because opacity is what allows a commoditized service to be priced as a differentiated one.
The Regulatory Posture Nobody Wants to Name
The coverage frames the charter bid as evidence of a friendly regulatory environment. This is half right, and the half that is wrong is the important half.
US federal regulatory posture toward digital asset custody has been a pendulum, not a trend line. The OCC's 2020-era guidance opened the door to national trust charters for digital asset firms. Subsequent guidance narrowed and clarified. The interpretive letters that followed confirmed that national banks may provide crypto custody services โ a clarification that cut both ways, because it authorized custody as a permitted activity for the entire national banking system, not just for specialized crypto custodians.
That last point is the one the crypto press consistently underweights. When the OCC clarifies that any national bank may custody crypto, it does not merely legitimize crypto custody. It floods the category with the most well-capitalized, most politically connected, most boring institutions in finance. A specialized crypto custodian with a national trust charter competes against Anchorage and BitGo. A payments company with a national trust charter competes against every regional bank that decides custody is a fee product it should offer its wealth-management clients. The addressable market for "crypto-native custody" is small. The addressable market for "custody as a line item in a bank's wealth platform" is enormous, and it is not contested by crypto firms at all.
This is why I am skeptical of the "compliance is a moat" thesis that dominates the commentary. Compliance is not a moat when the entire banking system is being invited to comply. It is a moat only when the compliance regime is narrow enough that few can enter, and the direction of US policy has been to widen the regime, not narrow it. A license is a moat only if it is scarce. A license that the OCC is handing to the whole banking system is not scarce. It is a standardized entry permit.
I want to be precise about what I am claiming. I am not claiming the charter is worthless. I am claiming the charter is a table stake, not a differentiator, and that the commentary is pricing a table stake as a differentiator. That is a mispricing, and mispricings in narrative are the ones that hurt retail readers most, because retail readers are the last to learn that the moat was a commodity.
The Blind Spot: Custody Is Not Decentralization
Here is the contrarian angle, and it is the one that the compliance-first framing actively suppresses.
Digital asset custody is, by construction, a re-centralization of the asset. The entire premise of a self-custodied asset is that the holder controls the key and no intermediary can freeze, seize, or deny access. A custody charter inverts this. It places the key โ or the shares of the key โ inside a regulated fiduciary, and a regulated fiduciary has obligations that run to the regulator and to the law, not only to the client.
This is not a hypothetical. It is the live, demonstrated behavior of the most compliance-forward stablecoin issuer in the market. USDC's issuer can freeze any address holding its token within a defined operational window, and it has done so. The freeze function is not a bug; it is a compliance feature, and it is disclosed. When a regulated custodian holds digital assets, it inherits the same structural capability: the ability to restrict access under legal compulsion. A trust charter does not remove that capability. It formalizes it.
So the reader should hold two facts simultaneously. First, a trust charter is a genuine upgrade in consumer protection โ it imposes fiduciary duty, capital requirements, and examination on an entity that would otherwise be unregulated. Second, a trust charter is a genuine transfer of control from the asset holder to a regulated intermediary, and that intermediary answers to the state. Both are true. The commentary presents only the first and treats the second as paranoid.
I am not a maximalist. I do not think self-custody is right for every institution, and I have written enough about slashing conditions and validator mechanics to know that institutional staking and institutional custody are real needs with real security properties. But I have also spent three weeks in 2022 dissecting the difference between liquid staking derivatives and native staking, and what that exercise taught me is that the trust assumptions of a wrapped or custodied asset are strictly larger than the trust assumptions of the underlying asset. stETH is not ETH. It is a claim on ETH, intermediated by a set of node operators and a smart contract. Custodied BTC is not BTC. It is a claim on BTC, intermediated by a fiduciary and a key-management architecture.
The trust-minimization cost is the price of the charter. That price is not disclosed in the announcement, and it never is, because disclosing it would undercut the compliance narrative that the charter is pure upside. Logic is binary; intent is often ambiguous. The logic of custody is a transfer of trust. The intent of the announcement is to present that transfer as a service. Those are the same event described from two directions, and only one direction is being marketed.
The Layer Nobody Is Building, and Why It Matters
I spent most of 2024 testing Data Availability Sampling on a custom node in Sรฃo Paulo, measuring the latency and cost of Celestia's blob space against monolithic execution. The finding that stuck with me was not the ninety-percent data-cost reduction. It was a structural observation about where value accrues in a modular stack.
When data availability becomes cheap and commoditized, the value migrates to the layers that are expensive to replicate โ the verification layer, the settlement layer, the trust layer. Custody is a trust layer. It is expensive to replicate because it is gated by capital and license, not by code. That is why custody is attractive to a payments company and why it is uninteresting to a protocol engineer. A protocol engineer can fork a rollup in an afternoon. A protocol engineer cannot fork a charter.
But here is the twist that the compliance narrative hides. The thing that makes custody valuable โ that it is gated by a license โ is also the thing that makes it fragile. A license is a permission, and permissions are revocable. A charter granted by a regulator can be conditioned, restricted, suspended, or interpreted out of existence by the next regulator. A business whose core asset is a revocable permission is a business with a policy risk it cannot engineer away. No amount of MPC, no quorum policy, no key ceremony can hedge a regulatory reversal. The custody stack is robust. The license is not.
This is the deepest reason I am unmoved by the announcement. The custody charter is being sold as a hardening of the business. Structurally, it is a concentration of the business's fate into a single variable โ the boolean โ that the business does not control. Before the charter, Modern Treasury's fate was distributed across product quality, customer retention, and engineering execution. After the charter, an increasing fraction of its fate is concentrated into a regulator's queue. Diversification is risk reduction. The charter is diversification in reverse, and the announcement frames it as a milestone.
What the RWA Story Has Been Hiding
I have been watching the tokenized real-world asset narrative for three years, and I have a position that is unpopular in the rooms where that narrative is sold. RWA on-chain has been a storytelling exercise, and the reason it has not shipped at scale is not technological. It is that traditional institutions do not need your public chain.
This is directly relevant to the Modern Treasury charter, and it is the connection nobody is drawing. If the thesis of RWA is that traditional assets migrate onto public blockchains, then the custody story should be a story about on-chain settlement, about tokenized treasuries settling in blockspace, about composability. It is not. The custody story is a story about a trust charter, a fiduciary, a capital line, and a reconciliation API. The custody story is the admission that the institutional version of this industry runs on regulated rails, not on public chains.
The institutions do not want the chain. They want the ledger, the API, and the charter. They want the reconciliation, the audit trail, and the fiduciary duty. They want the parts of crypto that look like banking and the parts of banking that look like crypto, and they want both of them inside a supervised entity. The public chain is a settlement substrate that they will use if it is cheaper and avoid if it is not. It is not a value system. It is a cost input.
This is why the Modern Treasury charter is a more honest signal about the industry's direction than any RWA announcement. It reveals where the institutional demand actually lives: in the plumbing, in the compliance layer, in the reconciliation of digital and fiat balances. The charter is not a bet on the chain. It is a bet on the plumbing. And the plumbing is where the money has always been, which is why the crypto-native firms who spend their energy on narrative rather than plumbing keep getting out-competed by boring financial technology companies that nobody in the industry has heard of.
The Hong Kong Analogy, Applied
I want to bring in a comparison that will annoy some readers, because the structural logic is identical and the industry refuses to see it.
Hong Kong's virtual asset licensing regime is not, primarily, an embrace of innovation. It is a competitive move in a regional contest for financial primacy, and its most important target is not a crypto user โ it is Singapore. The licensing framework is a mechanism for attracting the institutional flow, the family offices, the exchanges, and the fund structures that Singapore has been accumulating. The regulatory posture is the product. The innovation is the packaging.
The Modern Treasury charter is the same structure at the national level. The US is not pursuing digital asset custody charters because it believes in custody. It is pursuing them because custody is a financial activity, financial activities generate fees and tax revenue, and a jurisdiction that hosts them captures the flow. The charter is a jurisdictional bid dressed as a technology story. The commentary treats it as a signal of technological maturity. It is a signal of competitive positioning in a race between financial centers, and the positioning has very little to do with the chain.
When I read the coverage of the Modern Treasury bid, I do not see a company entering a technology market. I see a company entering a regulatory market, in a jurisdiction that is trying to win a share of a global custody flow, using a license as the instrument. The technology โ MPC, HSM, multisig, whatever they choose โ is downstream and largely interchangeable. The differentiator is the license and the jurisdiction that issues it. That is the actual competition, and it is not a competition that a better key-ceremony design can win.
The Execution Risk Nobody Prices
Every analysis of a charter bid prices the regulatory risk. Almost none price the execution risk, and the execution risk is larger.
A payments company entering custody is entering a domain with a fundamentally different failure mode than payments. A payments system fails by being slow, expensive, or wrong. A custody system fails by losing keys. The first failure mode is recoverable, embarrassing, and priced. The second is terminal, existential, and underpriced.
Modern Treasury's core competence is API reliability and reconciliation accuracy. Those are excellent competences, and they are orthogonal to custody security. A team that is superb at idempotent API design is not, by virtue of that skill, superb at key ceremonies, quorum policies, physical security, and insider-threat modeling. Those are different disciplines with different hiring profiles and different failure cultures. The competence that makes a company a great payments company is not the competence that makes it a safe custodian, and the announcement conflates the two by placing them under one brand.
This is the pattern I have watched repeatedly in the NFT audits of 2021 and in the cross-chain infrastructure that followed. A team with a strong reputation in one domain extends into an adjacent domain, and the reputation travels faster than the competence. The market prices the brand. The attacker prices the implementation. And the attacker is patient.
I am not predicting a breach. I am identifying a structural asymmetry. The custody business is a business where a single operational failure destroys the franchise, and the franchise is being extended into custody by a company whose demonstrated competence is in a domain where failures are survivable. A survivable-failure culture entering a terminal-failure domain is a mismatch that no charter can correct.
The Question the Charter Cannot Answer
Let me return to the boolean, because the boolean is where this analysis has to end.
Suppose the charter is granted. Suppose the custody stack is world-class. Suppose the capital line is comfortable and the compliance apparatus is immaculate. Even then, the charter cannot answer the question that determines the business's fate: does the market want regulated custody at the price regulated custody costs?
Regulated custody is expensive. It carries capital, insurance, examination, and fiduciary duty. Self-custody is free, and it is technically available to any institution with the discipline to manage its own keys. The demand for regulated custody exists because most institutions do not have that discipline, and because their auditors and boards require a counterparty with a charter. That demand is real. It is also elastic, and it is elastic against the price of self-custody, which is falling.
Every advance in key-management tooling โ better MPC, better HSM, better multi-signature standards, better social recovery โ reduces the operational burden of self-custody and, therefore, reduces the premium that a regulated custodian can charge. The charter is a bet that the compliance premium is durable. The technology trend is a bet that the compliance premium compresses. A company whose strategy is to capture the compliance premium is making a bet against the direction of key-management tooling, and the direction of key-management tooling has been downward-cost for a decade.
I want to be fair. There will always be institutions that prefer a chartered custodian for reasons that have nothing to do with cost โ board comfort, audit requirements, regulatory mandates, insurance availability. The premium will not go to zero. But it will compress, and a business built on the premium is a business built on a compressing spread. The charter does not change that. The charter is a fixed cost against a compressing spread.
Takeaway
The Modern Treasury trust charter bid is a request, not an event. It is an input, not an output. It occupies a state in a regulatory state machine, and the state is not disclosed, which means the announcement is consistent with every possible outcome from "we have not filed" to "we are weeks from operating."
What the bid actually encodes is not a technology story. It is a defensive vertical integration by a payments company whose customers' money is migrating into assets its infrastructure cannot hold. The charter is a table stake in a settled, competitive licensing category that a first mover occupied five years ago. The custody stack behind it is undisclosed, and the capital line is undisclosed, and the quorum policy is undisclosed, and until those are public, the technical analysis has no input. Logic is binary; intent is often ambiguous โ and in this case, the logic is not even available, while the intent is being marketed as though it were the logic.
What I will watch, and what I would advise any reader to watch, is not the announcement. It is the filing. The filing is a public event in most jurisdictions, and the filing is where the interface stops being a press release and starts being a contract. Watch for the state transition from announcement to filing. Watch for the capital line. Watch for the custody architecture disclosure, if it ever comes. Watch the examination record, if it is ever published.
And watch the boolean. It is still unassigned. Everything written about this story before that variable is set is a prediction wearing the costume of a report, and the costume fits poorly.
A charter is a permission. Permissions are granted by states, and states grant permissions for reasons that are political before they are technological. The custody charter is not a milestone on the road to decentralization. It is a way station on the road to institutionalization, and institutionalization is a road that leads away from the properties that made this asset class interesting in the first place. Whether that is progress or regression depends on a variable that has not been assigned, and will not be assigned by any amount of code.