CrossDesk is not a regulated benchmark administrator.
Benchmark administration is a licensed activity in the United Kingdom and the European Union. CrossDesk is not authorised, registered, recognised or endorsed as a benchmark administrator in those jurisdictions or in any other, and nothing on this site, in the methodology, or in the software is a claim otherwise.
What follows from that. A UK- or EU-supervised entity that wishes to reference a CrossDesk fixing in a regulated product must resolve the recognition or endorsement route first. Until then a CrossDesk fixing is a published number with an open methodology and an on-ledger audit trail — not a benchmark within the meaning of the UK or EU Benchmarks Regulation.
Nothing on this page is legal advice. It is a plain statement of status and a factual mapping the administrator can support from its own published documents.
Where the methodology stands against the nineteen principles.
The IOSCO Principles (July 2013) are the international reference for benchmark governance, quality and accountability, and the basis of the UK and EU regimes. The table below maps each principle to what the CrossDesk methodology and signer protocol actually specify. It claims only what those documents support. "Aligned" means the published rule addresses the principle; "partial" means it addresses part of it, or by policy rather than by code; "not addressed" means there is nothing to point to yet. This is a self-assessment, not an audit.
| # | Principle | Status | What CrossDesk can point to |
|---|---|---|---|
| Governance | |||
| 1 | Overall responsibility of the administrator | partial | CrossDesk is named as administrator in the methodology and publishes it under its own name. No separate legal entity, board or accountability framework exists yet. |
| 2 | Oversight of third parties | partial | The signer protocol defines what each seat asserts and how a checker must behave (P4); the API refuses a condition that does not belong to the seat claiming it. No contractual oversight of signers is in place. |
| 3 | Conflicts of interest | partial — by disclosure, not exclusion | The administrator does not trade the instruments it prices (§7). Signers are chosen because they have positions, composed to oppose one another, with no single interest holding K. The methodology states this is a mitigated conflict, not an absent one (P6). |
| 4 | Control framework | partial | The ledger enforces the venue's traded range and the wrapper-mark arithmetic; the API enforces seat-condition matching; a restatement needs the same K of N. No independent control function or written control framework beyond the methodology. |
| 5 | Internal oversight | not addressed as IOSCO intends | The committee is the oversight function (§7) but it is composed of interested parties, which is the opposite of the independent oversight function the principle describes. |
| Quality of the benchmark | |||
| 6 | Benchmark design | aligned | §2 defines what the fixing measures and what it is not; the official/indicative split is published with the drift between them. |
| 7 | Data sufficiency | aligned | Tier 1 minimum quality conditions (at least two orders from two distinct parties, crossed quantity above zero) are enforced as module constants; a missing input yields a gap, never an estimate (§5). |
| 8 | Hierarchy of data inputs | aligned | §3 is an ordered waterfall, and the tier used is published with every fixing. |
| 9 | Transparency of benchmark determinations | aligned | Every fixing carries its identifier, strike time, value, tier, signers, each signer's verified conditions and protocol version, drift versus indicative, and carry-forward age (§10, P3). |
| 10 | Periodic review | partial | Three consecutive carried-forward strikes trigger a cessation review (§3, §8). No scheduled periodic review of the methodology itself is specified. |
| Quality of the methodology | |||
| 11 | Content of the methodology | aligned | Published, versioned, with definitions, the calculation, inputs, the restatement rule, and a per-section statement of what is implemented (§12, P7). |
| 12 | Changes to the methodology | aligned | Material changes require 30 days' notice; every version is retained; a fixing is interpreted under the version in force at its strike (§9). |
| 13 | Transition (cessation) | aligned | A cessation notice of no less than 60 calendar days naming the final strike and any successor, enforced on-ledger (§8). |
| 14 | Submitter code of conduct | partial | The signer protocol is a per-seat code: named conditions, no opinion on the price, checker rules that forbid auto-confirming a failed condition or widening tolerances (P2, P4). It is not yet a signed agreement with any submitter. |
| 15 | Internal controls over data collection | partial | Sealed inputs are protected by the ledger's signatory/observer model rather than an API filter (§5); external market data is never an input to an official fixing. No documented controls over the administrator's own proposal computation beyond publishing every input. |
| Accountability | |||
| 16 | Complaints procedures | partial | Anyone may report a suspected error to the administrator, and a correction needs K of N (§6). No formal complaints procedure with timelines is published. |
| 17 | Audits | not addressed | No external audit of the administrator or the methodology has been performed. |
| 18 | Audit trail | aligned | Every fixing, attestation, refusal and restatement is an immutable ledger record; a superseded fixing is never archived; who signed what, under which protocol version, is permanent and attributable. |
| 19 | Cooperation with regulatory authorities | not addressed | No arrangement exists. The audit trail is designed to be readable by a regulator; no regulator has been engaged. |
Section references are to the fixing methodology (§) and the signer protocol (P). The count and headings of the principles follow the IOSCO final report of July 2013.
The gap between an open methodology and an authorised administrator.
Under the UK and EU Benchmarks Regulations, an administrator located in the jurisdiction must be authorised or registered by the national competent authority before a supervised entity may use its benchmarks; an administrator located outside the jurisdiction reaches supervised users through equivalence, recognition or endorsement. Whichever route applied to CrossDesk, the items below are what it does not have today.
A legal entity and a jurisdiction
An administrator is a person, not a repository. The entity, where it is established, and which authority supervises it come before any application.
An application, or a recognition or endorsement arrangement
Authorisation or registration where the entity is located; for supervised users elsewhere, recognition by that jurisdiction's authority or endorsement by an administrator already authorised there.
An independent oversight function
The regimes expect an oversight function independent of parties with positions. The CrossDesk committee is deliberately the opposite. Recognition would require an independent function alongside it, or a different governance answer for the conflict the committee mitigates rather than removes.
A control framework, complaints procedure and external audit
Written controls over input data and determination; a published complaints procedure; and an independent audit of adherence to the methodology — none of which exists today.
A benchmark statement and submitter agreements
A published benchmark statement per identifier, and a signed code of conduct with each contributing seat rather than a protocol they follow informally.
Regulatory advice before anyone contracts against a number
Until there is an entity and, where required, an authorisation, the repository licence states that nothing in it is a price, a valuation, or a benchmark — and the administrator will obtain regulatory advice before publishing a number anyone settles against.
The cost of authorisation is not treated as a tax. An administrator that has done this work is much harder to replace than one that has not.
What the numbers on this site are.
The values shown on this site come from a hosted sandbox: an in-process Canton ledger running the CrossDesk Daml packages, seeded with a demonstration book, that resets when the hosting instance recycles. It is not a Canton Network participant and is not connected to DevNet, TestNet or MainNet. It proves the software, not the network integration.
No fixing has been published for commercial use. No committee has been convened; the seats shown on benchmark pages describe the protocol, not occupants. Instruments in the demonstration basket are issued for demonstration and are not the production tokens. A separate, earlier claim of 4.16 cBTC through a CIP-56 registry flow was made on a shared hackathon participant that is no longer online; it is not what the sandbox holds today.
Figures cited on the licensing page carry their sources. The repository is source-available under a proprietary licence whose terms state that nothing in it is a price, a valuation, a benchmark, investment advice, or an offer of any financial product or service.