Cryptoasset Recovery in England & Wales: Exchanges, Custody Structures and PSP Duties (Asset Tracing After D’Aloia and Jones)

In typical crypto recovery disputes, the claim against an intermediary is driven by the practical reality that the exchange or Payment Service Provider (PSP) is identifiable, while the fraudster is hidden offshore or otherwise unknown. Recent cases demonstrate the increase in, and solidification of, causes of action being deployed against intermediaries, with the type of claim depending on whether the target is an exchange/custodian (typically for asset recovery and disclosure claims) or a PSP/Electric Money Institution (EMI) (typically claims relating to mandate/authority and “on inquiry” claims).

A. Claims against custodians/exchanges

D’Aloia v Persons Unknown and others [2024] EWHC 2342 (Ch) – evidential challenges

Claims for proprietary relief are typically the primary cause of action advanced against exchanges and usually arise where a claimant’s analytics point to a deposit address associated with an exchange and the claimant then seeks proprietary injunctions and delivery up orders (often coupled with a constructive trust claim). D’Aloia v Persons Unknown and others [2024] EWHC 2342 (Ch) is a recent example of why this this form of relief is commonly pursued but exposes evidential limits and the conflicts between the current law and the factual reality of blockchain mechanics.

D’Aloia tested an exchange liability case on the facts, including an assessment of whether USDT were capable of attracting property rights and an attempt to trace USDT into Bitkub’s custodial environment. Unlike earlier cases, which were resolved at an interlocutory stage, this case proceeded through a full tracing exercise and hearing. The claim ultimately failed on the facts, therefore all comments by Judge Farnhill are obiter, however they are likely to be instructive in future cases (particularly in view of the enactment of the Act).

Mr D’Aloia was the victim of an online investment scam and was induced to transfer approximately £2.5 million in cryptocurrency (primarily USDT) to fraudsters operating a fake trading platform. The stolen crypto was moved through multiple blockchain wallets and partly routed via accounts at a number of exchanges; the case focused on whether a portion traceable to a Bitkub exchange wallet could be recovered from the exchange following the fraud.

The judgment usefully defined the legal characterisation of USDT (Tether, a stablecoin), noting that “USDT attract property rights under English law… neither a chose in action nor a chose in possession, but rather a distinct form of property not premised on an underlying legal right” at [5]. This follows prior decisions made at the interlocutory stage and is compatible with the Act.

While Judge Farnhill confirmed that, in principle, a number of actions may be pursued against exchanges, including constructive trust, unjust enrichment and proprietary remedies, the ruling also highlighted that the claimant must be able to prove that the misappropriated assets (or their traceable proceeds) are actually held within the exchange’s custody in an identifiable way. The court tested the tracing methodology and the assumptions underpinning the expert evidence, and treated deficiencies in the analysis or pleadings as fatal rather than something that could be “fixed” after the fact.

A key theme running through D’Aloia is the disconnect between how claimants often frame “wallets” and how exchanges actually hold assets. Many exchanges operate pooled custodial wallets with internal ledgering, rather than segregated on chain accounts. Assets may be mixed, swept between wallets, or represented only as account balances in internal systems. The court made clear that these features are central to the analysis of whether a claim could be made out. For example, Mr D’Aloia was unable to prove the second element of an unjust enrichment claim (i.e. that the enrichment was at his expense) due to the restrictions in his ability to trace fund through his Bitkub wallet and issues with the expert tracing methodology.

Judge Farnhill stated that common law tracing through mixed funds is not available: “In my view it is not. [...] [T]racing is only available to the Claimant in respect of his equitable claims, where tracing through a mixed fund is possible.” [at 7] This is one reason exchange facing recovery claims tend to be pleaded in equitable/proprietary terms and then stand or fall on custody mechanics and the evidence linking the claimant’s asset to what is held within that structure. However, it is to be noted that a) these comments are obiter, and b) the fast development of technology and tracing within the proprietary systems of exchanges/custodians may allow for tracing where the factual circumstances allow for it. This case is key in showing how much development is needed of evidential support for recovery claims for cryptoassets, protection needed for those investing and continued advancement of regulation to enhance this protection (and allow certainty for exchanges/custodians). 

In practical terms, D’Aloia shows that it is not sufficient to assert that assets were traceable “to” an exchange. Claimants must deploy blockchain analytics that can withstand cross examination and plead a case that aligns with how the exchange really holds assets, including how pooling, sweeping and mixed funds are addressed. Even where an exchange has received assets linked to a fraud, the claimant must still establish that the enrichment was at the claimant’s expense, and that requirement pulls the analysis back to custody mechanics and evidential proof. In practice, if the claimant cannot show that its crypto, or its traceable proceeds, actually reached the exchange in an identifiable way, unjust enrichment does not function as a shortcut around tracing shortcomings; it fails for much the same reasons as the proprietary claim.

For exchanges and custodians, the judgment provides a clear roadmap for resisting over reach by forcing claimants to articulate, with precision, what asset is said to be held on trust, where it is said to be held, and on what evidential basis. It also supports pushing for early disposal or narrowing of claims, including strike out or summary judgment where the pleaded tracing case and custody assumptions cannot, even on their own case, establish an identifiable asset within the exchange’s custodial structure (or a coherent “at the expense of” pathway), and where the claimant’s expert methodology is not capable of sustaining the required chain of attribution.

---

Jones v Persons Unknown [2025] EWHC 1823 (Comm): tracing errors can create satellite disputes

If D’Aloia demonstrates the importance of robust tracing for a successful proprietary or unjust enrichment claim, Jones v Persons Unknown illustrates the risks of getting that tracing wrong, especially in a pooled environment where compliance choices can affect third parties.

In Jones, the claimant was the victim of a cryptocurrency investment fraud carried out between January 2019 and January 2020. Unknown fraudsters induced Mr. Jones to transfer approximately 89.6 Bitcoin to a fake investment platform, believing his assets were being legitimately invested. In reality, his Bitcoin was misappropriated and dissipated across the blockchain. The Bitcoin was valued at approximately £1.5-£1.75 million at the relevant time. 

During 2022 court proceedings, Mr. Jones relied on expert tracing evidence as the basis for allegations that the misappropriated Bitcoin had been traced to a wallet (the “tHEL” wallet) hosted by the cryptocurrency exchange Huobi Global Ltd. Mr Jones obtained without notice relief, including worldwide freezing injunctions (both proprietary and personal) against the persons unknown; and proprietary injunction and disclosure orders against Huobi. 

The relevant order proceeded on the basis that a specific “Exchange Wallet” under Huobi’s control “has been used to store fraudulently obtained Bitcoin including those belonging to [Mr Jones].” Huobi complied with the order by paying out BTC from a different wallet and “reimbursing” itself by debiting BTC held in the tHEL wallet. It later emerged that the assets in the tHEL wallet did not derive from the claimant’s fraud but rather belonged to another exchange (Kyrrex) and other third party customers. 

When Kyrrex sought to unwind the outcome via an application under CPR 40.9, ³the court dismissed the application on the basis that Kyrrex was not “directly affected” by the judgment itself. Rather, Kyrrex’s loss arose from Huobi’s operational decision as to how to satisfy the judgment, namely, paying out from another wallet and then debiting the tHEL wallet to “reimburse” itself. In addition, the application failed because it was brought too late, more than two years after the judgment had been satisfied. Ultimately, as recognised by the court, an innocent party would be suffering loss.

This set of proceedings highlights a number of key factors to bear in mind. In particular, it reinforces the importance of accuracy in tracing, particularly when using experts. While technology is developing fast, it is key that parties are able to properly explain how tracing methods are used and provide clear, accurate evidence to demonstrate where cryptoassets were ultimately moved to ensure that relief is not granted that impacts innocent third parties in the face of fraud. Intermediaries will be key to assist with this process if they are able to properly track and locate cryptoassets. Huobi chose not to engage in the proceedings at an early stage and it may be that future intermediaries will now seek to engage early on in order to a) avoid large judgments against them, and b) protect their reputation in respect of possible fraudulent activity through the platform and the impact of that on third parties. However, this is not just a case of accuracy in tracing but also about how exchanges satisfy court orders in pooled custody systems, and how operational compliance choices can transfer fraud risk to innocent third parties. The litigation outcome was shaped less by blockchain attribution and more by how Huobi chose to comply with an order under time pressure.

A critical element of the 2025 judgment is that a third party whose assets were wrongly taken indirectly, due to an exchange’s compliance error, was not “directly affected” for the purposes of CPR 40.9 and therefore lacked standing to set aside the original judgment. Remedies may lie against the exchange, not via reopening the fraud judgment. Jones also underscores how uncertain CPR 40.9 can be in practice. The court quoting the White Book commentary observed that r.40.9 “gives no clue as to when such an application may succeed.” With the time pressure on intermediaries to comply with court orders, and the risks of inaccurate tracing, it is key that such intermediaries are careful in how they comply with judgments and that they engage early to ensure that their own proprietary information can be used to challenge inaccurate tracing.

Finally, the court emphasised the importance of prompt action. Delay by the third party applicant, Kyrrex, was fatal to its claims. Exchanges and custodians must investigate suspected misattribution immediately and engage at the earliest procedural stage to preserve procedural rights and enhance the opportunity of a successful claim.

As with Jones, in many cases, the immediate battleground is not final liability but interim relief and disclosure. Claimants need freezing/proprietary injunctions to prevent dissipation and disclosure orders to identify account holders and transaction pathways. Those applications often land on exchanges because exchanges have the customer records and the control points. Earlier authorities such as AA v Persons Unknown [2019] EWHC 3556 (Comm) remain the reference point for proprietary injunctions and related relief in crypto cases.

Exchanges and custodians should be prepared to be served with urgent orders seeking disclosure and operational steps on short timelines. Where wallet attribution is disputed, Jones shows what can go wrong if compliance decisions are made on assumptions and in a pooled environment.

B. Claims against PSPs: mandate/authority and "on inquiry" duties

Breach of mandate / Quincecare‑type duties (more common against PSPs, increasingly relevant to crypto‑linked wallet products)

On the PSP side, the recent cases point to a slightly different risk profile. The dispute is often framed less as “you are holding my stolen asset” and more as “you executed instructions when you should have been on inquiry”. Hamblin v Moorwand is a good illustration of how these arguments are being run in a crypto‑linked wallet context. Moorwand Ltd was an FCA-regulated electronic money institution offering an electronic wallet product capable of holding and transacting in multiple currencies and bitcoin. The Hamblins were induced to pay £160,000 to a company (RND Global Ltd) whose wallet was held with Moorwand, and the funds were then dissipated through instructions given by a fraudster purporting to act for the company. The appeal succeeded on the basis that Moorwand should have made inquiries before executing those agent instructions because it was “on inquiry”. Hamblin also contains a helpful description of “push” fraud and why these cases are being reframed: “A ‘push’ fraud is to be contrasted with a ‘pull’ fraud… the real difference… lies in the identity of the victim.” In a push‑fraud scenario, it is the customer who is deceived into authorising the payment, whereas in an agent‑instruction case the relevant victim is the account holder whose mandate is being abused, which is why the focus shifts to whether the PSP ought to have questioned the authority of the instruction.

The claim was brought via a derivative route on behalf of the corporate customer, which is one reason the case has attracted attention. This distinguishes it from Philipp v Barclays Bank UK PLC [2023] UKSC 25 where the customer personally gave the payment instructions, leading the Supreme Court to narrow the application of Quincecare‑type duties in cases of authorised push payment fraud.

Hamblin shows claimants seeking to reframe cases through an agent-instruction analysis, including via derivative mechanics. The result is that PSPs providing crypto-linked wallet products face an increased likelihood of “on inquiry” allegations tied to onboarding issues and transaction behaviour. The pleaded duty formulation quoted by the court is a useful way to express the risk: the PSP was said to owe a duty “not to execute a payment instruction… when… put on inquiry”, assessed by the standard of “the reasonable payment service provider”.

In practical terms, this places the focus on operational decision‑making and the contemporaneous record supporting it. Where a PSP is said to have been “on inquiry”, the court will examine what the PSP actually knew at the time and how it responded: what information was available at the time, what internal triggers were engaged, who reviewed the transaction, what enquiries (if any) were made, and why the payment was ultimately processed or stopped. A PSP that can produce a coherent, time‑stamped audit trail, showing defined red‑flag criteria, escalation steps and the rationale for the outcome, is typically in a stronger position than a PSP that relies on general policies without being able to show how they were applied in the circumstances.

Conversion: increasingly not the answer

Finally, pleading strategy has shifted following Yuen v Li. Conversion is often attractive in theft scenarios because it is a strict-liability tort and can provide a direct route to a money judgment assessed by reference to value at the relevant date. In Yuen v Li, the High Court held that conversion was not available for cryptoassets as intangible property, removing that route for claimants and pushing recovery claims towards unjust enrichment, proprietary restitution/constructive trust, and fraud-based causes of action, typically supported by proprietary injunctions and delivery-up relief. 

As Judge Cotter put it, OBG v Allan was “a clear block” to extending conversion to digital assets, however that does not mean that victims are left without remedy: 

Sometimes a tort such as deceit will apply – for example, where a transfer is obtained by deception – and it may be that in practice the economic torts may protect the victim. However, it now seems that there may be a tort, or tort-like, remedy against a person who interferes with the claimant's rights in an item of intangible property. In Armstrong DLW GmbH -v- Winnington Networks Ltd [2012] EWHC 10 (Ch) the claimant was deprived of certain EU emissions trading permits as a result of a fraud: the permits passed through the hands of the defendant and from it to a legitimate holder. Julian Knowles J was satisfied that the claimant had a cause of action against the defendant, which he referred to as a proprietary restitutionary claim, less the latter were innocent of knowledge of wrongdoing, which in the circumstances they were not, having had at least a suspicion that the permits had been stolen. If this is correct, then there is possibly now a general cause of action for loss suffered available at common law against a person who with guilty knowledge or notice appropriates or deals with intangible property belonging to another. This could be significant in the case not only of the slightly unusual asset in Armstrong DLW, but of such things as computer files, NFTs, cryptocurrency, and possibly domain names. (at [75]).

It is somewhat unfortunate that Judge Cotter struck out the claim for conversion rather than taking the opportunity to develop the case law, as was contemplated in the Law Commission in their Reports on Digital Assets. This judgment currently leaves a gap in available remedies, particularly where no parties to the dispute have access to the wallets.

For exchanges and custodians, this case is likely to increase the weight placed on tracing and custody mechanics, because the viable claims are the ones that require the claimant to show a defensible link between the misappropriated assets and what is held within the custodial structure. It also means exchange‑facing disputes are likely to be fought more often on identification, attribution, and the operational feasibility of compliance (freezes, delivery‑up, segregation in pooled systems), rather than on strict‑liability tort framing. For claimants, the consequence is that the case likely has to be pleaded and evidenced as an equitable/restitutionary claim from the outset, with the tracing methodology and custody assumptions capable of standing up to scrutiny.