FCA Weighs Bespoke Rules for Tokenized Gold: The Prize Is the Fund Wrapper
The most informative number in tokenized gold this month is zero. Zero is the count of UK-authorised fund vehicles permitted to hold tokenized gold as an eligible asset. Every other figure — supply curves, vault attestations, quarterly bar lists — sits downstream of it.
That number may be about to move. The FCA has signalled it is weighing bespoke rules for tokenized gold, coupled with a fund exemption. Two words carry the load: bespoke and exemption. The first concedes that the existing rulebook does not fit. The second concedes that the regulator knows it does not fit and is prepared to route around its own architecture.
An unnamed industry participant told the Financial Times that UK fund-rule uncertainty is slowing tokenized gold's development and restricting investor access. The concern is credible. The attribution is not. We trace the hash to find the human error; when no hash exists, we trace the incentive. This incentive points at one door — the one that opens onto ISA and SIPP money.
Tokenized gold is the oldest RWA category that still functions. The design is boring, which is its virtue: an issuer buys LBMA Good Delivery bars, places them with a vaulting agent, publishes a bar list, and mints tokens against the holdings. Paxos holds through Brink's. Tether's XAUT sits in Swiss vaults and is redeemable in physical form above a threshold. Redemption for Paxos Gold requires a full 400-ounce Good Delivery bar, which prices out most holders and quietly defines the product as institutional at the exit and retail at the entry.
The instrument competes against a crowded, cheap, liquid incumbent set. iShares Physical Gold ETC carries an ongoing charge of 0.12%. Invesco Physical Gold, the same. SPDR Gold Shares, 0.40%. Tokenized gold issuers mostly charge no management fee and recover economics through spread, mint and redemption fees, and float. On a pure cost line, the incumbent wins for anyone who never needs the token itself.
Then regulation. In the UK, an authorised fund needs a depositary, an independent valuation regime, and eligible assets defined under the FCA's collective investment rules. Tokenized gold fits none of those boxes cleanly. The depositary role — the independent party that safekeeps and reconciles — has no obvious analogue when the register is a distributed ledger and the attestation is a monthly PDF. The EU's MiCA addresses the token, not the fund wrapper. Switzerland's FINMA and Singapore's MAS have moved further on classification. London, the vaulting capital of the world, sits behind.
Start with what a fund exemption is worth. It is worth nothing to the token. It is worth a great deal to the wrapper.
A tokenized gold product sold directly to UK retail is a taxable disposal on every trade, outside any shelter. Placed inside a UK-authorised fund that qualifies for an ISA or a SIPP, the same exposure becomes tax-advantaged, with an annual ISA allowance of £20,000 per adult and pension contributions capped only by the annual allowance. That is the prize. The bespoke rules are not about making gold tokens legal. They are about making gold tokens packagable.
This is where I have direct footing. In 2024, I worked with two institutional custodians to build a real-time data bridge between traditional settlement systems and on-chain oracle feeds — 50,000 daily records standardised to meet reporting requirements, reconciliation time down roughly 60%. The hard part was never the chain. It was agreeing on what counts as a record. Every tokenized gold fund will hit the same wall, and the wall is made of valuation methodology.
Consider the reconciliation stack:
| Layer | What an authorised fund needs | What tokenized gold currently provides |
|---|---|---|
| Asset eligibility | Listed eligible asset, defined depositary | Unlisted claim on vaulted bars |
| Valuation | Daily independent NAV with an auditable source | Issuer-published bar list, periodic attestation |
| Custody | Regulated depositary, segregated accounts | Vaulting agent plus smart contract |
| Reconciliation | Daily, deterministic, immutable trail | On-chain supply against off-chain bar list |
| Reporting | Standardised, machine-readable | Heterogeneous PDFs and dashboards |
The single point of failure is not the contract. It is the bar list. The bar list is the balance sheet. It is a schedule of serial numbers, weights, and fineness, signed by a vault and matched against token supply. If a serial number appears twice, or a bar is pledged to two counterparties, token supply is correct and collateral is wrong. That is a human error, not a code error, and it is exactly the class of failure that never surfaces in an audit of the smart contract.
Which is why the FCA's wording on custody will matter more than its wording on tokens. Watch three phrases: whether the rules require LBMA Good Delivery provenance as a condition rather than a convention; whether the depositary function can be split between a regulated entity and a technical attestation provider; and whether redemption rights must be exercisable in kind at the fund level.
I built my first checklist for this problem in 2017, auditing pre-sale contracts against whitepaper projections. Three integer overflow vulnerabilities in a Parity wallet fork taught the market that financial logic has to precede technical innovation. The lesson transfers. A tokenized gold fund is not a code problem with a legal wrapper; it is a custody problem with a code interface. The audit that matters reviews the vaulting agreement, the bar-list cadence, the insurance schedule, and the redemption mechanics — then confirms the contract does what the documents say.
The cost arithmetic is less favourable than the marketing suggests. Strip out the management fee and the token still carries mint and redemption spreads, gas on settlement, and a redemption minimum that caps the exit for small holders. For an investor who rebalances quarterly, the ETF at 0.12% plus commission is hard to beat. For an investor who wants the same asset inside a tax wrapper, the comparison inverts — which is precisely why the exemption request exists.
The second hypothesis attached to tokenized gold is that it becomes low-volatility collateral in DeFi. The arithmetic is real: a 30-day realised volatility around 12–15% annualised against 50%-plus for major crypto assets means a smaller haircut and higher capital efficiency in lending markets. But this thesis is independent of the fund wrapper, and conflating the two is a category error. A token inside a UK-authorised fund cannot simultaneously be pledged into a permissionless lending pool — the depositary and the smart contract cannot both control the same bar. The two use cases are mutually exclusive for a single unit of metal, and the FCA rules will decide which one London gets.
| Jurisdiction | Status for tokenized gold | Binding constraint |
|---|---|---|
| UK | Bespoke rules under consideration | No fund wrapper; depositary undefined |
| Switzerland | Workable classification | Limited distribution reach |
| Singapore | Published digital asset stance | Small institutional base |
| EU | MiCA covers token issuance | Fund eligibility rules remain national |
| US | Enforcement-led ambiguity | Custody and broker-dealer overlap |
There is a version of this story in which bespoke rules are framed as a remedy for liquidity fragmentation in gold markets. That framing deserves scrutiny. Gold is the most fungible asset on earth; there is no fragmentation problem in the underlying metal. What is fragmented is distribution: separate issuers, separate vaults, separate bar lists, separate legal wrappers. A carve-out that standardises the wrapper consolidates distribution. A carve-out that incentivises new issuers multiplies the fragments and calls the multiplication a market.
The consensus reading of the FCA signal is that regulatory clarity unlocks institutional flows. The data does not support that yet.
Take the jurisdictions that already have clarity. Switzerland has a workable framework. Singapore has published a stance on digital payment tokens and a functional sandbox culture. Tokenized gold supply in both markets remains a rounding error against the roughly $250 billion sitting in physically backed gold ETFs globally. Clarity did not produce scale. Something else is binding — most plausibly the fact that gold exposure is already cheap, liquid, and deeply collateralised in futures and ETF form. Regulation removes a legal obstacle. It does not manufacture demand for a second copy of an asset people already own efficiently.
Then consider the source. The warning about UK uncertainty comes from an unnamed participant. Redact the name and you redact the incentive. An issuer lobbying for an exemption is not wrong because it is self-interested; it is simply not neutral. The request is a request to be included in tax-advantaged distribution, and it should be evaluated as a distribution question, not a technology question.
What to watch, concretely: the FCA consultation paper and its treatment of the depositary function; the first bar-list attestation standard a UK-authorised fund will accept; and the supply delta in PAXG and XAUT, which will move before any prospectus does. If supply rises while redemption minimums stay high, the product is being built for wrappers, not for holders. The market corrects; the data endures.