Seven UK banks test tokenised deposits for mortgages and shopping

UK Finance announced on 24 September 2026 that seven banks had completed the first live customer transactions under the Great British Tokenised Deposit (GBTD) initiative: two remortgage completions and one marketplace payment transaction, each using tokenised sterling deposits. The seven participants are Barclays, HSBC UK, Lloyds Banking Group, Monzo, Nationwide, NatWest and Santander. UK Finance has not said which banks were involved in which transaction, so participation in the wider initiative should not be read as proof that all seven handled live customer money on this occasion.

This matters to UK account holders because it is the first retail test of a mechanism that locks money inside a customer's own bank account and releases it automatically when an external condition is met, rather than moving the money into a third party's hands while a transaction is pending. That is a different proposition from faster payments. The pilots show the mechanism working on three occasions. They do not yet show it working faster, cheaper, or more safely than the solicitor accounts, escrow services and card protections that already handle the same jobs.

What happened in the two remortgages

In the remortgage cases, funds were locked in the customer's account and released automatically at completion. UK Finance said the design can allow customers to continue earning interest on the locked money until that point. Whether both pilot customers actually received interest, and at what rate, was not disclosed.

The pilots also explored how a future digital connection to HM Land Registry might improve the process. UK Finance has not said that HM Land Registry was connected to, or supplied data to, the transactions that were completed. The exploration was described as forward-looking, not as part of the transactions UK Finance announced on 24 September 2026.

No source gives a transaction value, a completion time, a cost figure or a comparison with how long the equivalent conventional remortgage would have taken. A reader cannot yet tell from the published material whether the tokenised version was quicker, slower or materially similar to the solicitor-controlled process that remortgage completions ordinarily use.

How the marketplace payment was meant to work

For the marketplace case, UK Finance said money was locked in the buyer's account and released only once the goods were successfully exchanged. This is the clearest consumer-facing use case: a payment that sits ring-fenced until a condition unrelated to the payment system itself — delivery of goods — is satisfied.

Here the record is disputed. UK Finance describes the set of three as live customer transactions. The British Retail Consortium, in its own commentary on the initiative, describes the online marketplace purchase as simulated. FinTechPulse has not been able to resolve this conflict from the published material, and neither UK Finance, Quant nor a participating bank has issued a clarification distinguishing "live" money from a "live" exchange of goods. Readers should treat the claim that an actual item changed hands between an actual buyer and seller as unconfirmed.

Several operational questions are also unanswered: who or what confirmed that the goods had been exchanged, what happens if the buyer and seller dispute delivery, and whether a locked payment can be reversed before release. None of this was set out in the material UK Finance has published.

What a tokenised deposit actually is

A tokenised deposit is a digital representation of money the customer already holds at a deposit-taking bank. It remains a liability of that bank — the same kind of claim as an ordinary current account balance — but expressed through programmable technology that can attach conditions to when and how it moves. The Prudential Regulation Authority (PRA), in a letter published on 18 May 2026, sets out this distinction and discusses both conventional and transferable tokenised-deposit models, noting that the precise legal and operational structure can vary between implementations.

This differs from a qualifying stablecoin, which is a cryptoasset that seeks to maintain a fixed value against a fiat currency through a pool of backing assets rather than through a bank's balance sheet. Both can support programmable, conditional payments. The underlying claim, the issuer, and the protection that applies if something goes wrong are not the same.

The Bank of England's policy for sterling-denominated systemic stablecoins, published on 22 June 2026, uses a different protection model built on backing-asset rules, redemption rights, capital, safeguarding and failure arrangements, rather than deposit insurance. This is a UK-specific regime; it does not extend FSCS cover to stablecoins.

What protection applies to the money

The PRA expects retail tokenised-deposit arrangements to meet its existing depositor-protection rules and associated operational requirements, in the same way as any other deposit. That expectation is conditional: it applies where the arrangement is legally a deposit held by an eligible depositor at an authorised firm, not automatically to every structure that uses token technology. UK Finance's own description of tokenised deposits as carrying conventional protections should be read with that qualification rather than taken as a blanket guarantee.

Where it does apply, the Financial Services Compensation Scheme (FSCS) protects eligible deposits up to £120,000 per eligible person, per UK-authorised firm, for failures from 1 December 2025. Balances held under different brands that share the same banking licence are added together for this limit, which matters for customers who hold both an ordinary account and a tokenised-deposit product at the same institution.

A separate, higher limit can apply temporarily. Qualifying temporary high balances — which can include certain proceeds connected with the sale of a main home — are protected up to £1.4m for six months from 1 December 2025. This is not an automatic feature of every mortgage-related balance: eligibility depends on specified circumstances and evidence, and a remortgage completion does not, by itself, establish that a balance qualifies.

The PRA has also warned, separately, that new digital-money products of this kind can introduce higher liquidity risk, operational uncertainty and reliance on new or potentially untested payment rails. None of the published GBTD material discloses which interbank settlement rail was used, whether it was Faster Payments, CHAPS or something built into the shared platform, or what would happen to a locked balance if the platform itself were unavailable.

Does programmability beat what already exists?

The pilots demonstrate that money can be locked against a condition and released automatically when that condition is met, without the customer's funds leaving their own bank. That is a genuine technical capability. Whether it is a practical improvement on existing arrangements is a separate question that the published evidence does not yet answer.

UK mortgage completions already run through solicitor client accounts, which hold and release funds according to legal instruction. Online marketplace purchases are already protected, in various forms, by card chargeback rights, payment authorisation holds and platform-specific buyer-protection schemes. None of these existing systems were benchmarked against the GBTD pilots on cost, speed, failure rate or customer outcome.

Giving oral evidence to a parliamentary committee on 25 February 2026, UK Finance's Jana Mackintosh addressed the suggestion that this is unnecessary because UK payments are already close to instant. Her response, as recorded in that evidence, was that the intended benefit is not raw payment speed but the automation of multi-party, multi-stage processes — the kind of conditional coordination between buyer, seller, bank and a verifying event that a single fast payment does not, by itself, provide. That is a plausible argument for where value might sit. It is not, on the evidence published so far, a demonstrated saving in time, cost or risk.

What the pilots do not establish

No source discloses the value of any of the three transactions, the identities of the banks involved in each one, the completion time achieved, any cost comparison with existing processes, a fraud outcome, or a customer sample size beyond the three cases described. The legal structure of the pilot tokens — including whether they were transferable between customers of different banks, and at what point a recipient became a depositor of the receiving bank — has not been published, and nor has the underlying ledger technology. Liability for a wrong automatic release, caused by faulty external data or a dispute over delivery, has not been addressed in any of the material reviewed for this article.

UK Finance has said that further pilots are planned, covering digital-asset settlement, digital debt instruments and coupon payments made in tokenised deposits. No firm dates have been given for when these will run or when results will be published.

What to watch next

The clearest next step is a clarification of whether the marketplace transaction involved a real exchange of goods between real customers, given the direct conflict between UK Finance's and the British Retail Consortium's accounts. Beyond that, readers with an interest in how this might affect their own mortgage or online purchases should look for published performance data — timing, cost, failure rates and fraud figures — against which the claimed benefits of automation can actually be tested, rather than taking the technical success of the first three transactions as evidence of a wider advantage. The PRA's expectations letter and the FSCS's own deposit-protection pages remain the primary places to check how any specific product is protected, rather than relying on a provider's general description of it as a bank deposit.