Blockchain is finally entering the US rules for transfer agents, but not as a register with no master. In a 421-page proposal issued on 1 September, the Securities and Exchange Commission explicitly accommodates recordkeeping on distributed ledgers while retaining a conventional requirement: for each securities issue, one agent remains responsible for the official securityholder file and must keep exclusive control over it.
That detail changes how the text should be read. The SEC is neither approving a particular blockchain nor declaring that possession of a token is enough to establish ownership of a share. It is adapting a regulated function designed around paper in the 1970s and 1980s to electronic systems, cloud providers, smart contracts and distributed ledgers. GatherHub compared the current framework, the proposed text and the questions posed by the Commission. The overhaul opens the technology; it does not decentralise legal accountability.
The record can be distributed; responsibility cannot
A transfer agent maintains the official list of a security’s owners, records issuances, cancellations and transfers, distributes certain payments and handles restrictions. An error therefore does more than produce the wrong screen: it can deprive an investor of rights, create an overissue or break communications with the issuer.
The proposal defines the “master securityholder file” as the official list of individual accounts maintained by the registered agent. The list would have to be electronic and could combine multiple linked files or systems. The technology would remain at the agent’s discretion—a conventional database, cloud infrastructure or distributed ledger—subject to two decisive conditions: the agent must retain exclusive control at all times, and only one recordkeeping agent may maintain the file for a given issue.
That is the first result of GatherHub’s comparison. The SEC accepts that a blockchain may form all or part of the register, but it rejects the idea that technical consensus should make the entity accountable for an entry impossible to identify. Even on a shared network, the point of responsibility stays centralised.
Three concrete changes for tokenization
The first change is visibility. Annual Form TA-2 would ask how many issues use distributed-ledger technology for their master file. It would also separate issuer-sponsored tokenized securities from those sponsored by a third party. Agents would name their tokenization providers and distributed-ledger platforms alongside banks, recordkeeping systems and call centres.
The second is regulatory access. When a third party preserves records, the agent would need to consult them without that provider’s intervention, allow regulators to examine them and promptly supply legible, accurate and current copies. The proposal says a distributed ledger can meet this test if the agent genuinely has that independent access. “Public” or “immutable” does not automatically mean “compliant”.
The third is operational risk. New Rule 17ad-12 would require procedures addressing custody, operations, cybersecurity and other material risks. Issuer or securityholder funds would have to be segregated in a “for the benefit of” bank account. A business-continuity plan would need to be tested, reviewed and updated at least annually.
| Question | Current framework | Proposal | What does not change | |---|---|---|---| | Record technology | Language inherited from paper and older electronic systems | DLT and linked systems accommodated neutrally | An official file must remain identifiable | | Accountable party | Registered transfer agent | One recordkeeping agent with exclusive control | Blockchain does not replace regulatory responsibility | | Transparency | Little tokenization-specific data | Volumes by model, platforms and providers disclosed | Reporting usage is not approval of the product | | Resilience | Rules silent on several digital risks | Cybersecurity, continuity and independent access formalised | The agent still answers for accuracy and safeguarding |
An onchain transaction is still not automatically a legal transfer
The overhaul complements, rather than reverses, the distinction GatherHub set out in our analysis of legal transfers of tokenized securities. A token may itself constitute the security, form part of the official register, send an instruction to an offchain register, represent a claim against a custodian, or merely track a financial return. The movement shown in a wallet does not by itself reveal which architecture applies.
The SEC wants issuer-sponsored and third-party-sponsored models reported separately precisely because their risks differ. In the latter case, a holder may depend on the custody, reconciliation and solvency of a company that is not the issuer of the underlying share. Modernising transfer agents does not remove that chain of rights, or the separate rules governing offering, trading, custody and customer identification.
Registration does not become a commercial endorsement either. When Injective obtained transfer-agent status, it had disclosed no client or asset administered under that status. If the proposal is adopted, firms using similar models would have to make their infrastructure more visible to regulators and document their controls; registration would still prove neither usage, liquidity nor the quality of a tokenized security.
Control, not the word “blockchain”, is the real test
Two passages show that the policy is unsettled. The Commission asks how to treat a master file kept exclusively on an immutable blockchain where the agent does not alone control the infrastructure. It also asks whether a wallet address and onchain quantity can be linked to offchain information about the holder’s name and address, so that transferring the token also updates the master file.
Those questions isolate the technical problem. A network may make history observable and difficult to alter without giving the agent power to correct an error, stop a legally disputed transaction, restore access after a compromise or reconcile a divergence. Conversely, a database controlled by one operator may create clear accountability while introducing a single point of failure. The proposal does not eliminate that tension; it requires the agent to demonstrate how it manages it.
Updated processing deadlines point in the same direction. Entries would need to be posted within the shorter of one business day or the applicable settlement cycle. Co-transfer agents would send debit and credit records within one business day. Chain speed is useful only if the legal register, controls and corrections move with it.
A proposal, not yet a rule
The SEC release provides for a 60-day comment period after publication in the Federal Register. As of 3 September, this is still a proposal. The duties, forms and implementation timetable may change; no firm can present the text as permission already granted.
Decrypt and Bloomberg Law corroborate the scale of the first substantive overhaul in roughly four decades and the new prominence given to tokenization. GatherHub’s own conclusion is narrower: the proposal does not endorse “blockchain recordkeeping” in the abstract. It establishes an accountability test. Who controls the official file, who can show it to regulators, who corrects discrepancies, and who bears the failure? Until those four answers are documented, “onchain” describes a technology, not a legal protection.