# Benchmark Statement — ETP Foundry Canton Series **Version 0.1 DRAFT · 22 September 2026 · ETP Foundry (formerly CrossDesk)** > **No benchmark described here is published, live, or in use.** This statement is written > in advance of first publication so that the rules exist before the numbers do. It becomes > effective on the date of the first published value and not before. A benchmark statement is what a licensee's risk and compliance team reads instead of the methodology: what the number measures, what it does *not* measure, who should use it, when it publishes, and what happens if it stops. It follows the structure UK/EU BMR requires of authorised administrators (Art. 27) and the shape used by CF Benchmarks. --- ## 1. Version history | Version | Date | Changes | |---|---|---| | 0.1 | 22 Sep 2026 | Initial draft. Not effective. | ## 2. Introduction **Administrator:** ETP Foundry (formerly CrossDesk). Operating from Ontario, Canada. **Contact:** s.borjas@lucilla.ca · etpfoundry.com **Regulatory status — stated plainly:** ETP Foundry is **not an authorised or registered benchmark administrator in any jurisdiction**. It is not supervised by the FCA, ESMA or any other authority. Nothing published is a benchmark for regulatory purposes. See §9 and `6-RULEBOOK.md` §12. **Series covered:** the ETP Foundry Canton Series — fixings and indices for tokenised assets recorded on the Canton Network, and for baskets of them. ## 3. Benchmark description and aims **What these benchmarks seek to measure:** the price at which a tokenised asset, or a defined basket of tokenised assets, could be exchanged at a defined moment — observed from transactions on constituent venues, and, where no observable price exists, attested by a committee of parties holding opposing economic positions against the mark. **The underlying economic reality:** for a single wrapped asset (e.g. cBTC), it is the exchange of that token for its quote asset on venues where it actually trades. For a basket, it is the cost of assembling the constituent tokens in their stated weights. **We price the token, not the asset it wraps or represents** — the wrapper's redemption friction and reserve condition are reflected through the wrapper factor (`6-RULEBOOK.md` §5.5), never silently. **The gap these exist to fill:** a basket of tokenised assets has no observable price even when every constituent does, and a tokenised asset whose home market is closed has no reference price at all for roughly 80% of each week. Both are states in which an oracle has nothing to relay. ## 4. Methodology summary Full rules: `6-RULEBOOK.md`. In brief: - **Indices** are computed from published rules and observable inputs. No committee. - **Fixings** are attested by **K of N** signers drawn from four seats — issuer, custodian, risk taker, market-facing party — who already hold positions against the mark. Recommended N=5, K=3. **The issuer may never reach quorum using only parties it controls.** - **OFFICIAL fixings** use a 60-minute observation window, partitioned into 12 × 5-minute intervals; the volume-weighted median of each partition is computed across all constituent venues, and the fixing is the equally-weighted mean of those 12 medians. This is the construction used by the CME CF Bitcoin Reference Rate. - **Minimum two independent constituent venues.** Below that, no value is published. - Every value publishes an **uncertainty band**, the inputs used, and which were stale. ## 5. General disclosure — limitations **This section is the honest one and is not to be softened.** 1. **No track record.** As of this version, no value has been published. A benchmark's reliability is demonstrated by an unbroken history, and ours does not yet exist. 2. **Thin markets.** Canton-native tokenised assets trade on young venues with modest volume. A benchmark can be no more robust than the market it observes. 3. **Wrapper risk.** For wrapped assets, the value depends on a reserve we do not hold and an attestation we do not produce. A reserve shortfall is a price event we can report but not prevent. 4. **Committee dependence.** A fixing requires K signatures. If signers are unavailable, no value is struck and the prior value is carried forward, flagged and aged. 5. **Judgement, where permitted** (illiquid and spread classes only), is a stated opinion, not an observation, and is labelled as such. 6. **Concentration.** With few venues, a single venue may represent a large share of observed volume. The venue-level screen (`6-RULEBOOK.md` §5.2b) limits, but does not eliminate, the effect. 7. **Not a valuation of the underlying.** A cBTC fixing is not a bitcoin price. A tokenised equity fixing is not the share price. Users who need the underlying should use a benchmark for the underlying. 8. **Not suitable, yet, to underlie a regulated financial product.** See `9-SUITABILITY-ANALYSIS.md`. ## 6. Usage of the benchmark | Type | May be used for | May **not** be used for | |---|---|---| | `INDEX` | information, analysis, product design | settlement, NAV | | `INDICATIVE` fixing | margin, haircuts, health factors, collateral monitoring | **NAV, creation/redemption, audited records** | | `OFFICIAL` fixing | NAV, creation/redemption, reporting, contract settlement | — | **Prohibited representation [FIXED]:** no licensee may present an `INDICATIVE` value as a NAV. This is a term of the licence. **Licence classes:** use/reference, redistribution, display. Reading the value through an oracle or data vendor does not substitute for a use licence. ## 7. Publication timings Per `6-RULEBOOK.md` §5.1: OFFICIAL at the asset's home-market close (Canton-native assets: 16:00 UTC daily); INDICATIVE at 00:00 / 08:00 / 16:00 UTC every day including weekends. Times are declared once and moved only as a methodology change under §8. Values are published to the ledger with their full fixing record, and to the public tape. ## 8. Changes to, and cessation of, the benchmarks **Changes:** proposed on-ledger, require the same K of N, and take effect after 5 business days' notice; material changes require 30 calendar days and written notice to every licensee. The rulebook is versioned and every value carries the version that produced it. **No change is ever retroactive.** **Cessation:** minimum 90 days' notice before ceasing any OFFICIAL benchmark, to licensees and publicly. Each factsheet names a recommended fallback. The methodology and history may be transferred to a successor administrator. Records survive for 7 years. Full policy: `6-RULEBOOK.md` §9. ## 9. Oversight The committee is the oversight body until an independent one exists. It meets quarterly on a fixed agenda (`1-operations-runbook.md` §4), reviews every restatement, carry-forward, judgement use and venue change, and publishes summary minutes to signers and licensees. **Conflicts:** any interest of the administrator or a signer in a priced asset is disclosed in the conflicts register and, where material, disqualifies that party from signing. **Complaints:** any licensee or signer may challenge a value in writing. Acknowledgement within 2 business days, substantive response within 10, escalation to the quarterly meeting if unresolved, minuted either way. ## 10. Updates to this statement Reviewed at least annually and on any material methodology change. Superseded versions remain published.