This article from Lloyds Maritime and Commercial Law Quarterly ([2025] LMCLQ 82) is reproduced here by kind permission of the publishers.

The lex situs , as a long-established concept of English private international law, may hardly seem to need a defence. But the difficulties in applying existing principles to digital assets, a variety of recent judgments involving digital assets and the current Law Commission consultation on private international law for digital assets have all put this concept in the spotlight and, in some cases, subjected it to criticism. In this article, we redress the balance. We argue that the lex situs is supported by both principle and practicality, and that a vital distinction between observers and actors reveals how the lex situs should be applied to digital assets.

I. INTRODUCTION

The lex situs is the shorthand Latin tag for the law of the place where an asset is situated. By extension, it is also used for the principle that this law should be applied to certain questions concerning that asset, such as proprietary issues.

The lex situs is long established as a concept of English private international law. It has been applied to both existing categories of property: choses in possession and choses in action.1 Digital assets have recently been recognised as property and, although the details are still being developed, are likely to be treated as a new third category distinct from choses in possession and choses in action.2 This immediately raises the question whether, and how, the lex situs should be applied to digital assets.3 The Law Commission, in its final report on digital assets, expressly reserved issues of private international law to a later consultation and has now issued a call for evidence on this topic.4

Our aim is to show that the lex situs is the optimal connecting factor for proprietary issues relating to digital assets and that its limitations when applied to choses in action should not obscure its utility for other assets. It is a distinct concept that intermediates between the legal system and the real world and which maximises the necessary qualities of certainty, discoverability and enforceability. Exploring how the lex situs applies to existing assets, valuable insight is gained into the artificiality of its application to certain choses in action. This reveals that that digital assets share characteristics with tangible assets that suit them both for the lex situs as a connecting factor. In discussing different types of digital assets, we explain how the lex situs may be determined for each of them. Finally, we mention jurisdiction briefly.

II. THE LEX SITUS FOR EXISTING CATEGORIES OF ASSETS

A. The lex situs in principle

Property, to be more precise than is customary, refers not to an object but to a relationship between a person and an object.5 Property rights are one form of legal rights, that is, rights that exist in and are created by virtue of a legal system, and they are unique in that they relate to objects. For tangible assets, these objects exist in the real world and have attributes—such as colour, size and location—that subsist independently of any legal system. For the legal system to apply legal rights of property in a consistent and sensible way to tangible objects, these real-world properties must be taken account of: for instance, if the asset disappears, the legal system must recognise that it no longer exists and nobody owns it. This is not a negotiation. If there is a mismatch between reality and the legal system, reality wins. It is incumbent on a legal system simply to recognise and apply physical facts, and changes to them, when determining property rights.

Legal systems achieve this aim and avoid mismatches with reality through the use of separate concepts that intermediate between the legal right and the physical facts. For tangible assets, one such concept is possession.6 Possession provides a legal analogue that takes account of the relevant facts applicable to a tangible asset so that rights in relation to that asset are consistent with reality. An intermediary concept such as possession promotes a smooth coupling between legal concepts and real world facts—complexity and uncertainty in the real world are isolated to and dealt with by the intermediary concept, which encapsulates this messiness and prevents inconsistency.

The lex situs rule arises naturally in relation to tangible assets. To determine the applicable legal system for a proprietary right in respect of a tangible asset, one seeks not only to use the nature and extent of those legal rights but also to connect those legal rights to the physical facts relating to that asset, so that any reality mismatch is avoided. Among these physical facts, the principal one is location. The lex situs is an intermediary concept applying facts about a tangible asset to determine its location. Legal rules, such as those determining applicable law, apply the lex situs, thereby abstracting away factual details and limiting the risk of inconsistency with the facts.

Note that this reasoning does not apply to choses in action. Lex situs is applied to choses in action by extension of the rule for choses in possession, but there is no danger of inconsistency with the real world whatever rule is chosen. If the legal system decrees that a debt exists, it exists. If it determines that a debt does not exist, it does not exist. The legal system may choose to create or alter choses in action dependent on actions or circumstances in the real world, but in the case of any mismatch, it is the legal system that determines the correct answer—nothing in the real world logically requires any different conclusion. This leads to artificiality in applying the lex situs to certain choses in action.

Crucially, this artificiality is not applicable to digital assets. A key distinction in property is between assets that exist by virtue of a legal system and assets that exist independently of any legal system.7 Digital assets, unlike choses in action, fall into the latter category. If a court decrees that a bitcoin does not exist, that does not cause it to disappear.8 Digital assets exist by virtue of facts in the real world, independently of legal systems.

But what does it mean to say that digital assets “exist” if they are not tangible objects in the real world or entities existing in legal systems? The Law Commission defines digital objects in terms of “the combination of the active operation of software by a network of participants and network-instantiated data”9 and other definitions similarly emphasise the composite nature of digital assets.10 While accurate, these definitions can be misleading, as they unduly elevate the external manifestations of digital assets such as the distributed ledger or the network of computers running the relevant software.11 These manifestations attract attention, as they form the novel technological wizardry that underpins digital assets.12 But it is what they underpin that is the essential element of existence: the shared view of the digital asset world contained in the minds of the people that use it.

At first it may seem startling to assert the existence of something based simply on a shared belief in it. Beliefs appear fragile, contingent and ephemeral. But, in fact, shared beliefs constitute some of the most stable and enduring objects created by humanity. For instance, “the United Kingdom”13 exists due to a shared belief in that nation state. If everybody in the world simultaneously ceased to believe that this country existed, it would not. And yet it is ludicrous to imagine this happening or to assert that the United Kingdom does not exist because this could happen.

The paradigm example of an object constituted through a shared belief is money14 —not physical notes and coins but the value inherent in a currency. And this was one motivation for the creation of the first cryptocurrency, Bitcoin.15 The aim of Bitcoin was to create a system of mutual trust—that is, a shared belief—that required no intermediaries. The technical means—the distributed ledger, the miners, the software—are simply there to enable a community of people to maintain a shared belief without the intervention of an intermediary. It is the shared belief that is fundamental, although it continues only while the system is maintained. Nobody is required to subscribe to a shared belief in a digital asset but, if they do, they can only believe in the specific model of that digital asset world promulgated by the relevant software. That is, any participant in the system sees the same world of bitcoin balances or digital asset attributes.

Digital assets “exist” as shared beliefs held by a community of people, in the same way that money, nation states and legal systems exist. Unlike, say, legal systems, they do not require intermediaries such as lawyers and judges to determine existence. Instead, this happens automatically through the operation of the system. The technology ensures that everyone subscribing to this shared belief must see precisely the same virtual world, the same pattern of account balances and digital asset attributes.

Substantially, this chain of reasoning was adopted by the Singapore High Court in ByBit v Ho Kai Xin,16 where Jeyaretnam J described digital assets as a “social construct”, similar to money: “it is only because people generally accept the exchange value of shells or beads or differently printed paper notes that they become currency. Money is accepted by virtue of a collective act of mutual faith”.17 This leads him to consider, as we also do here, how “crypto assets manifest themselves in the physical world”.18 But then he faced the obstacle of the common law distinction between choses in possession and choses in action being exhaustive and so, while recognising an element of circularity in his reasoning, held that digital assets were choses in action and that they consisted of the right to spend the digital asset.19 This simply fails the test given above: if a court decided that the right to spend a particular digital asset did not exist, the asset would not disappear and the owner would still be able to spend it. The fallacy here is not just the circularity of defining an object (properly the property right attached to an object) as the right to enforce it in court. Rather, it is the confusion of two different social constructs: the argument goes that, because a digital asset is a social construct constituted by its acceptance as a matter of belief and legal rights are also a social construct, it is valid to say a digital asset is a legal right. But acceptance by a digital asset community of the existence of a digital asset is a different social construct within a different normative system to acceptance by the legal community of a legal right (in general, via the judgment of a court). They are two independent events. We agree with Jeyaretnam J’s insights into the nature of digital assets that lead him to see digital assets as property, but not the classification forced upon him by the exhaustive division of property into choses in action and choses in possession.

The recognition of a digital asset as an object constituted by a shared belief is useful in tackling criticism of digital assets as property. Professor Stevens has argued that digital assets should not attract any proprietary rights at all.20 He starts by characterising digital assets as a special type of rivalrous information and pointing out that information is accorded property status only in specific circumstances. We have seen that this is to misunderstand the essential nature of digital assets: to say that digital assets consist of information is like saying that US dollars consist of paper because that is what banknotes are made of. Still, although this weakens his position, it leaves his central argument intact. He points out that it is an error to conclude simply from the fact that an object exists—whether in the real or the virtual world—that it is deserving of property rights. Many physical objects—for instance, human organs or the moon—do not attract property rights. This is correct. Whether to apply property rights at all is ultimately a normative question. It is precisely the independence of digital assets from the legal system that gives this question meaning: a legal system may be internally consistent whether or not it accords property rights to a physical object or a digital asset. The decision whether to do so is normative: is this something that should attract property rights?

The problem with his argument is not that it misunderstands digital assets. On the contrary, it expressly acknowledges that they are objects with an independent existence. The problem is that the normative position it advocates is deeply unattractive. It would follow that the legal system would assert that, for instance, anybody who says they own a bitcoin is mistaken and accord them no proprietary protection. The entire digital asset ecosystem would be built on objects that are not recognised by the law. Professor Stevens shows how peripheral contractual or tortious duties may give satisfactory remedies in some circumstances. But the overall assertion that the legal system is able, but ought not, to treat any digital assets as attracting property rights seems scarcely tenable in a world of digital asset exchange traded funds (ETFs), electronic trade documents21 and decentralised finance (DeFi).

Professors Watts and Low make the same argument, although they appear ultimately to accept that digital assets are too entrenched to deny them property rights.22 Writing shortly after market turmoil in 2022, they reserved the option to revisit this conclusion if digital assets collapsed in value, although as of 2025 the reverse has been the case. But their principal argument concerning the nature of private keys is important not only to the existence of digital assets but also to their location, and so we now consider it in detail.

B. It’s the public key, not the private key

Watts and Low argue that digital assets are not rivalrous because the private key is simply information which may be shared, copied, stolen or otherwise disclosed. Anyone with the private key can exercise control over the asset. It follows that knowledge of the private key and exercise of control is not sufficient to characterise digital assets as rivalrous objects that should attract property rights.

To meet this argument, we consider more closely the attributes of a digital asset. We have argued that the key component of the technological and human system that constitutes the digital asset is the shared belief in the single consistent model of the virtual world. That model contains the digital assets themselves and certain attributes of those assets. Those attributes vary: some digital assets have a notional “colour”23 or other data that may be attached to them. Certain attributes are fundamental, though, and these include the account—designated in this virtual world by a public key—and the quantity of digital assets in an account.

Note that the private key, as opposed to the public key, is not one of these attributes. What makes digital assets work is that every person sees the same view of the relevant virtual world even though nobody is in charge. This apparent paradox is known to computer scientists by the shorthand “the double spend problem” and it was the innovation of the Bitcoin system to solve this problem—that is, to find a way to maintain a single view of the virtual world for everyone without anyone having to trust anybody else.24 Private keys are an integral part of how this problem are solved. But it is not relevant to this solution whether private keys are copied, shared or lost. Private keys touch the system only at the moment of a particular transfer of a digital asset. At this precise moment, the private key ensures that everyone agrees that the digital asset has been transferred and where it has gone. At that moment, only a single person is using that private key. Whether it has previously been shared, copied or distributed is irrelevant. The distributed ledger underlying a digital asset system does not contain any private keys and private keys are not in the virtual world. Watts and Low locate property rights as supporting the private key and criticise this because private keys are subject to copying and distribution, like all information. But property rights do not attach to the private key—they attach to the public key. It is the public key that is present on the blockchain and that is an attribute of the virtual world. The shared belief in the virtual world is that a digital asset exists and it is located in the account designated by a public key.

Watts and Low do not impugn the action of the private key at the moment of a transfer. That action maintains the coherence of the virtual world and the rivalrousness of the digital assets. Their criticism goes to “excludability”: the practical ability to exclude others from exercising control over the asset, which may be affected in the real world by what happens to a private key.25 Here, the Law Commission’s use of the phrase “fragile rivalrousness” to describe digital assets is perhaps not helpful, in that it implies that an attack on the real world practicalities of the private key affects the fundamental viability of the digital asset.

Watts and Low’s criticism does not affect whether digital assets exist, because it concerns only certain real world messiness in dealing with the private key and not the rivalrousness of digital assets or the integrity of their attributes in the virtual world.26 Watt and Low’s discussion of the private key is valuable not in determining whether a digital asset exists, but where it exists, and we return to it in this context below.

C. Summary

Digital assets exist as a shared belief in a common virtual realm, independent of any legal system. To avoid mismatches with reality, legal rights must be linked to the attributes of those assets in the virtual realm and their manifestation in the real world. In this, they resemble tangible assets but are more complex as tangible assets have attributes only in the real world. As with tangible assets, the coupling between legal rights and these attributes is best accomplished via distinct concepts that intermediate between the real world and legal rights.

For the purposes of determining applicable law and jurisdiction, the important attributes are those that touch both the virtual world and the real world. The public key is a virtual world attribute that touches the real world in two ways: usage of a private key at the moment of transfer and control over the public key. While the common law is still developing in this area, control over a public key is generally taken to be the analogue of possession of a tangible asset.27 We shall defi ne the lex situs as the intermediary concept that ascribes a physical location based on attributes of the digital assets, primarily control of public keys. Whereas the application of the lex situs to choses in action is convenient but artificial, for digital assets it reflects a principled need to maintain consistency between the real world and legal rights. Attacks on the lex situs as unsuitable for choses in action do not apply to digital assets, which are not choses in action but a new third category of property.

The utility of the lex situs for enforceability and certainty considered below is not an accident—it follows from the lex situs being an appropriate intermediate concept between rights in relation to assets and facts in the real world. Limitations in its utility in relation to choses in action reflect the artificiality of extending the lex situs beyond its principled foundation to assets that exist only in legal systems.

III. THE LEX SITUS IN PRACTICE

In this Part, we evaluate the reasons for applying the lex situs to proprietary questions concerning an asset, by reference to the existing categories of property. This will guide adoption of a rule for digital assets.

Certainty and enforceability are generally cited as the justifications for adopting the lex situs.28 The argument runs that the courts where an asset is located will be best able to enforce any judgment in respect of it and the location of an asset, being unique in time and space, provides a means of linking it to a legal system that is independent and certain. For immovables, this argument is unimpeachable: clearly, the jurisdiction where the land or similar asset is located is uniquely able to enforce judgments in respect of it and any person dealing with the asset would naturally assume that this is also the law applicable to it.

Similar arguments apply to some existing intangible assets, although these are classed as movables for private international law purposes.29 For many intangible assets, there is a clear choice of lex situs that is effectively fixed: the place of incorporation of a company or the location of its share register does not generally change location.30 Certainty is thereby established according to easily discoverable criteria that also correspond to the location where enforcement is most convenient.

Looking at the lex situs for contractual rights and debts is less helpful, in that the lex situs is an artificial label for the rules which are formulated to provide a connecting factor, so any attempt to use those rules to justify the lex situs would be circular.31 We briefly set out the rules here not to adjudicate between rival interpretations but to illustrate the difficulty in determining a physical location for an asset that has a purely legal existence. The point of doing so is to make clear that digital assets do not suffer from this defect and so applying the lex situs to them is not artificial in this way.

The leading authority is Hardy v India,32 where the issue arose in the context of a third-party debt order. The problem faced by Peter MacDonald Eggers QC, sitting as a Deputy Judge of the High Court, was that the previous authorities had defi ned the lex situs of a debt in terms of both its location and the law governing it. He held that “[t]he situs of the debt is necessarily linked to the law governing debts as recognised by generally accepted principles of private international law”33 and that this is the country where the debt is “recoverable”.34 The presumption is that this is the country where the debtor is located. Where there is not yet a judgment establishing the right to a debt—which will generally be applicable for contractual rights—“the debt may be determined as properly recoverable only by a court of a competent jurisdiction to determine whether the relevant debt is properly recoverable or has been discharged”.35 This may be the residence or domicile of the debtor, but it may also be a country chosen by an exclusive jurisdiction agreement or another country as determined by contract or statute.

For contractual rights, there is a specific rule that applies to assignment set out in Art.14(2) of the Rome I Regulation,36 which applies the choice of the law governing a
contractual right to certain proprietary questions relating to its transfer. The scope of this rule is contested and, while it clearly overlaps with the principles set out in Hardy , it is
unclear whether the two are entirely consistent.37 On top of this, there is also the rule in Gibbs,38 which states that, as far as English law contracts are concerned, only the English courts are able unilaterally to modify or cancel rights and obligations created under them.

Why is there such an embarrassment of overlapping rules and exceptions here? Again, our aim is not to resolve them into a coherent whole but to explain why this has arisen. The third-party debt order in Hardy puts the problem in focus: in granting a third-party debt order, the court will order the debtor to pay a person other than the original creditor; if this is not respected by another legal system to which the debtor is subject, they may be forced to pay the debt twice. For a contractual right, there is no nexus in the real world that legal systems are forced to respect that can prevent this—there are only rights and liabilities as found by legal systems. The legal system that created the right, the legal system of the courts that have jurisdiction over the dispute and the legal system with jurisdiction over the debtor may all seek power over the right and, whichever is successful, no inherent contradiction arises. Contrast the situation with a tangible asset: for the court to find that it has ceased to exist when it does in fact exist would simply be nonsensical.

The lack of real world nexus lies behind the rule in Gibbs and the rule covering assignment. Because a contractual right has only a legal existence, it can be transferred only by a legal process; hence the need for assignment and the conflicts of law rule addressing it. Contrast again the situation with a tangible asset: if it has physically moved from one place to another, it is the role of the court not to decide whether legally it has moved from one place to another, but simply to elucidate the consequences of its having moved.

These finely textured overlapping rules are all in pursuit of the legal system that best fits the particular fact-situation, without creating absurdity or unfairness. Another way of looking at them is to see a balance between enforceability, which leans towards the location of the debtor, and certainty, which favours the law which can discharge or modify
the debt and which generally does not change throughout the life of the asset.39 Ultimately, certainty appears to be the more important factor—avoiding double liability of the debtor is paramount.

The conclusion is that, for choses in action that exist only by virtue of legal systems and where different legal systems may compete to control them, the lex situs is a bundle of rules and exceptions that seeks to balance certainty and enforceability rather than a principled concept intermediating with the real world. But this criticism does not apply to the lex situs for digital assets—they do benefit from an intermediating concept based on location.

Tangible assets have a definite location but raise a new problem, due to the temporal aspect: their location can change. This creates an additional tension between certainty and
enforcement. These values may be directly in conflict with each other. For enforcement, one considers the time when the dispute arises; for certainty, one considers the time when the relevant event involving the asset took place. A consistent, multilateralist approach to private international law requires certainty to be preeminent over enforcement and this is embodied in Dicey, Morris & Collins’s Rules 141 and 142,40 which state that proprietary rights in respect of a transfer are governed by the lex situs at the time of the transfer. For example, in Winkworth v Christie, 41 a painting was stolen in England, taken to Italy, where it was sold in good faith, and then transported back to England. The transfer in Italy was valid according to Italian law but not English law. The English courts applied Italian law and held that the buyer was entitled to the painting. The original owner of the painting was in England throughout and the painting was in England at the time of the claim—enforcement pointed to England but the certainty of the lex situs prevailed. Note the significance of the correct approach being split into two separate Rules: Rule 141 applies the lex situs at a given moment and Rule 142 forbids that application from being displaced retrospectively. Although they appear to overlap, Rule 141 creates a certain and predictable rule and Rule 142 guards it against abuse.42 These are the two desiderata of a connecting factor: certainty in application and resistance to manipulation.

Certainty, then, requires a connecting factor that operates at the time of the relevant event. As property rights necessarily affect third parties, this factor must be discoverable. To avoid abuse, it must also be an effective control mechanism—not something that can be arbitrarily manipulated by the parties to the original transaction. This suggests using some fact related to the physical manifestation of the asset. Where there is no physical manifestation, as with contractual rights, a rule can still be manufactured that maximises certainty by reference to the legal system creating the right, although it may be somewhat artificial. These considerations lead to three possible connecting factors:

  1. The law governing the interaction between the parties to the relevant asset, for instance, the law applicable to the sale contract under which the asset was transferred. This may not be discoverable by any third party and it is easily manipulated by the parties to the event. It would allow the thief in Winkworth v Christie validly to transfer the painting simply by selling it subject to a governing law that would have allowed the transfer, even if all the parties and the painting were still in England.
  2. The location of the transferor at the time of the event. This provides certainty and relates to a physical fact but its connection to the asset itself is tenuous and may not be discoverable to third parties at the time of the relevant event.43
  3. The location of the asset at the time of the event, ie, the lex situs: this is a physical manifestation that both provides certainty and is most likely to be discoverable by anybody affected by the transfer of the asset at the relevant time.

The lex situs is clearly the winner on certainty and discoverability.

In addition, the importance of enforceability in relation to tangible assets is largely illusory. In almost all situations, claims will be satisfied by payment of a sum of money
rather than any physical movement of the relevant asset.44 The default remedy for a claim in conversion is payment of the value of the asset, not return of the asset.45 Even proprietary claims involving digital assets have generally sought remedies such as knowing receipt, which is a personal claim available where the defendant does not still possess the asset.46

The recent Supreme Court decision in Byers v Saudi National Bank47 illustrates the policy choices behind the lex situs . Shares registered in Saudi Arabia were transferred in breach of an English law trust to the defendant, a Saudi bank. The defendant was aware of the trust and so was not, in English terms, a bona fi de purchaser of the legal estate for value without notice. Saudi law did not recognise the trust and treated the transactions as an effective legal transfer of the shares. The Supreme Court held that a knowing receipt claim against the defendant failed because the claimant had no continuing proprietary interest in the shares after the transfer, its legal effect being governed by Saudi law, as it had determined in a previous judgment.48 Enforcement was not the problem: the English law governed claim was being brought in the English courts in respect of an English law trust. But certainty and discoverability required Saudi law to apply to the proprietary issues.

Imagine a different rule applied to the converse to the Byers v Saudi National Bank situation: transfer of an English registered share in breach of trust to a recipient who is not a bona fi de purchaser. If proprietary consequences followed the governing law of the transfer agreement, the parties could simply arrange for this to be Saudi law: English trust law would be instantly and comprehensively undermined. Proprietary consequences following the location of the transferor or the asset (if it is a tangible movable rather than a registered share) can be manipulated only by moving things—assets or people—in the real world. And, of these, the location of the asset is the circumstance most closely associated with that asset and likely to be the most discoverable and predictable. An even worse solution would be a rule that refers to the time that the dispute arises rather than the time of the relevant event—this would enable abuse by participants even after the fact, turning an invalid transfer into a valid transfer in defi ance of legal common sense. Considering the practicalities of hypothetical alternatives leads us back to the lex situs.

Summary

We set out the following principles for determining a connecting factor between an asset and a legal system to determine proprietary questions in respect of that asset:

  1. Rules that favour enforcement and certainty/discoverability are in conflict whenever the asset can change location. In that situation, certainty and discoverability are paramount.
  2. For certainty, the applicable law must be tied to a particular moment in time, which is the moment that the relevant event in relation to the asset (generally, a transfer or purported transfer) occurred.
  3. For an asset that is not created by virtue of a legal system, the connecting factor should be a physical manifestation of the asset that is independent of legal systems, resistant to abuse and discoverable.49
  4. The physical location of that asset at the time of the relevant event is the connecting factor that best satisfies these criteria.

IV. DIGITAL ASSETS

We now seek to apply these principles to digital assets. This entails an examination of digital asset platforms and, in particular, a legal classification of participants into observers and actors.

A. Permissionless blockchains

We start with the paradigm asset: the fully permissionless distributed ledger, such as Bitcoin or Ethereum.

In this type of platform, transactions are recorded in a distributed ledger verified by a consensus mechanism. The ledger is a database containing a record of the current state of the system, via a complete list of transactions or a set of account balances or some equivalent data. There is no privileged single version of this ledger; nor is there any entity that can confirm its veracity. Rather, the consensus mechanism enables anyone to obtain a copy of the ledger and to verify that it is accurate.

Anybody wishing to add a new transaction does not do so by amending the ledger directly—they need not have a copy of the ledger at all. Rather, they will interact with the consensus mechanism, generally by sending a message in some electronic format to the system operating that mechanism. One universal feature of these platforms is the use
of cryptographic means to link a person with an account or an asset—knowledge of the relevant private key is the only means to effect a transfer of that asset.

New transactions can be added but old transactions cannot be removed. That sometimes leads to the assertion that rectification and rescission are not available for distributed
ledgers. This is wrong. Rectification of contracts on the physical world does not require the participants to go back in time and reverse the original transfers of value—instead, future transfers are made to align the world with the rectified contract. Similarly, rectification and rescission of distributed ledger arrangements are accomplished by adding suitable new transactions.50

This error is an example of a wider misconception about distributed ledgers. The ledger is not a means of making changes in the way that an electronic database is both a record and a means of making changes to its contents. A distributed ledger is a fixed view of the world that is available to any person downloading software that is compatible with it. Those who possess a copy of the distributed ledger are not thereby participants in the system, in the sense that they can actively influence it, but are only observers. Even an individual involved in updating the ledger via the consensus mechanism—in Bitcoin, a miner—does not have the ability to influence the system. They are performing a purely mechanical task that allows the system as a whole to operate in return for some personal reward.

Participants in the distributed ledger must, in respect of any particular transaction, be divided into observers and actors. The actors are those who are involved in the particular
transaction: undertaking actions to bring it about (other than as part of running the system as a whole to bring about all transactions) or whose individual accounts or balances are
changed by that transaction. Observers are all other participants in the system.

Observers may be located throughout the entire world, but it does not follow that digital assets thereby exist everywhere in the world at once: the idea of “omniterritoriality”.51
A physical object might be observable from anywhere in the world—perhaps if an image of it is broadcast on television or is particularly large or if distant observers use
telescopes to see it—but it does not thereby exist everywhere in the world. A physical object exists in just one place, which is where people can actually interact with it and move it to another place. Similarly, the lex situs of a digital asset should be based not on all participants in the system, but only on those who interact with it, those who are actors
in the relevant transaction. The distributed ledger, miners and software developers being located everywhere does not affect the calculation of lex situs —only those who are actors in the relevant transaction need to be considered.

The recent Court of Appeal decision in Tulip Trading52 demonstrates this point. Curiously, it does it by arguing the exact opposite. The claimant, Tulip Trading, claimed
that it owned a large amount of bitcoin contained in a wallet to which it no longer had the private key and that the developers of bitcoin owed it a duty to create and promulgate
changes to the underlying bitcoin software that would restore its control over that bitcoin. The Court of Appeal held that it was arguable that the developers of a given network were a sufficiently well-defi ned group, that they had undertaken a role which involved making discretionary decisions and exercising powers for and on behalf of other people, in relation to property owned by those people and entrusted to the care of the fiduciaries, and that therefore the developers were fiduciaries with a duty to act in the interests of owners, even by introducing new code to the Bitcoin software. Whether this argument would succeed would depend on the facts, to be decided later at trial.

On the face of it, this model of Bitcoin governance contradicts what we have set out here. The essence of Tulip Trading’s claim is that “the decentralised governance of
bitcoin really is a myth”53 —that bitcoin does not exist independently of a small group of developers who can create or destroy it purely according to their interests. But there is a glaring problem with this reasoning. If it is correct, then it appears to follow that bitcoin is not property at all—it does not satisfy the existing, or any proposed criteria, to constitute property.54 It might be thought that this will ultimately be an impediment to Tulip Trading’s claim, based, as it is, on proprietary remedies in respect of its alleged
bitcoin, and legal scholars have been watching with keen interest any attempt by Tulip to put its case in any subsequent trial.55 But for present purposes it suffices to point out that Tulip Trading shows that attempting to resile from the generally accepted facts as to the operation of the Bitcoin platform leads to a situation where the law of property, including the lex situs, is not relevant at all.

Bringing together all this reasoning, we have a digital asset on a permissionless blockchain that exists as a shared belief in certain attributes, including the account denoted by a public key associated with that asset. The applicable law requires a connecting factor that ascribes a location in the real world. To be consistent with the digital realm, the correct question must demand that:

  1. there is a location that connects the digital asset’s public key with the real world;
  2. the real world correlate of the virtual public key attribute is control of the public key in the real world;
  3. this correlate is determined at the time of the relevant event; and
  4. it is determined by reference to the very limited group of actors (usually just two) in respect of that event (as opposed to the wider universe of participants who may
    observe that event.

The lex situs is the legal concept that intermediates between legal rights and digital assets by posing this question. By asking the question in this way, consistency with principle and the nature of digital assets is ensured. Considerations of certainty and discoverability can then be employed to help provide useful answers—that is, answers
that are consistent with the real world. The real world will pose difficulties in providing the answer in complex fact-situations and these difficulties will be addressed by glosses
on the lex situs rule that maximise certainty and discoverability and, above all, resistance to abuse. This maintains a conceptual separation between the existence of digital assets
as property and the attributes of those assets which are independent of the legal system in order to determine the precise question that constitutes lex situs and, conversely, the
application of real-world fact situations applying values of certainty and discoverability in order to determine the answer to that question. In this way, much of the objections of Watts and Low and others to the limitations of the private key may be seen as querying only the answer in particular situations and not the theoretical applicability of the question of lex situs or the status of digital assets.56

In practice, defining the lex situs as the law of the residence (or possibly domicile) of the owner of the digital asset at the moment of the relevant event is relatively stable
and resistant to abuse.57 It has the requisite degree of certainty and discoverability. Conceptually, it provides a close analogy to the location of a physical asset: the owner of a digital asset is generally the person with the private key who is physically able to exercise control over the asset. Practically, it is the fact about the digital asset most likely to be ascertainable, as the private key must be used by the owner at the moment of the relevant event. Where there is more than one person with access to the private key, the relevant location is that of the person who actually does use it—just as the person in possession of tangible property is the one who actually takes hold of it, even if several other people might have been able to do so.58

Other permutations of facts may emerge to complicate the determination of the lex situs. We do not attempt here to predict all the details that the courts may develop, but only the general principle that should be applied. This is that the essential attribute to be physically located is the public key associated with an account, as refl ected in the real
world by control over that public key and, when considering a particular transfer, the usage of a private key to effect that transfer.

What if it is difficult to determine the owner or, as in Tulip, the owner is disputed? This concern largely disappears when focusing on applicable law rather than jurisdiction. Current rules on jurisdiction may lead to a relevant time, which is an arbitrary point after several disputed or uncertain events have taken place.59 But for the applicable law, the
dispute will concern a particular event involving a digital asset with the actors in respect of that event recorded (possibly pseudonymously) on the platform and the relevant time is when that event occurred.

Even where a digital asset has a disputed history, the ownership of the asset at the time of a particular event is unlikely to be disputed, if a peculiar feature of one current legal
understanding of the nature of digital assets is accepted. This feature is that transfers of digital assets take place through the extinction of one asset and its replacement by an
identical copy in another place. Accordingly, proprietary claims against a digital asset will proceed by tracing that asset through the platform, rather than following it. So, if a digital asset is stolen and transferred into a different account and then the corresponding digital asset is transferred again, the owner of the asset at the time of the second transfer will be the person with control over the private key for the second account and not the victim of the thief. The victim will have various remedies, including proprietary remedies based on tracing the digital asset into its replacement in the second account, but that does not affect the ownership of the replacement asset at the time of the second transfer. Assuming this analysis prevails, concerns about identifying the owner for the purposes of the lex situs are misplaced.60

One genuine difficulty in applying this lex situs rule is reconciling off-chain transactions involving a digital asset. For instance, consider the situation where the owner of a digital asset enters into a written contract of sale with a third party and then effects a transaction on the blockchain that transfers the digital asset to a different person at a time after the transfer of title specifi ed in the contract. The legal effect of this chain of events is currently unclear—yet this is a weakness not of conflicts of law and lex situs but of the substantive law generally relating to digital assets. It may well be that legal title to the digital asset is not transferred to the buyer until a corresponding entry is made in the digital ledger. Until this situation is clarified, the application of the lex situs to it cannot be finalised.

Overall, principle leads us to use an intermediary concept that is linked to the physical manifestation of a digital asset and practicality requires that this concept is certain and
discoverable. This concept is the lex situs and the situs is defi ned as the residence of the owner at the time of the relevant event—if the event is a transfer, the residence of the transferor.

B. Permissioned and governed blockchains

Many uses of distributed ledger technology involve permissioned blockchains: platforms that can be accessed only by users following a predetermined process such as entering into an agreement. For these platforms, the consensus mechanism may be the confirmation of a single person or small group of people. Some of these platforms are simply centralised databases that happen to employ blockchain technology and should be analysed in purely contractual terms, with no new property being created. Other permissioned platforms may satisfy the criteria to create digital assets.

Alternatively, the platform may be a governed blockchain: a permissionless blockchain where the underlying protocol requires every transaction to be signed with a hashed
reference to an agreement governing use of the platform.61 In both of these situations, the relevant contractual framework may contain a choice of law that is intended to apply to proprietary as well as contractual questions.

The two options here for a choice of law are the lex situs , as for permissionless blockchains (that is, ignoring the choice of law for proprietary questions), or the choice of law expressed in the platform. Unlike permissionless blockchains, the argument here is finely balanced. If the situation is seen as analogous to creation of a registered asset, then the chosen law would be apposite, so there is conceptual support for this position. Certainty and discoverability also point to the chosen law.62

The obstacle to this option is the possibility of abuse. Proprietary choice of law is often seen as distinct to contractual, in that it is a mandatory choice, determined by rules of
confl icts of law and not the parties. One key policy behind this distinction is to prevent abuse. Proprietary rules affect third parties who are not involved in any choice and may be disadvantaged by it. They apply exactly when the interests of the parties to a transaction may be distinct from the interests of the third party.

Governed and permissioned blockchains have some support against this argument, because any party involved in the platform will have to agree to the relevant law. Any third
party will have expressly opted for the chosen law, and so should be taken to have agreed to the consequences, provided that they are a participant in the platform. Unfortunately,
not all third parties fall into this category.

We illustrate the obstacle with a hypothetical blockchain called DiceyCoin. DiceyCoin is a conventional permissionless distributed ledger that is used to create a digital asset
known as a diceycoin. DiceyCoin is also a governed blockchain, meaning that every transaction contains a hashed reference to a governing agreement, a constitution. The constitution contains dispute resolution and choice of law provisions. It chooses a governing law for proprietary issues that disregards the validity of trusts and gives full and immediate effect to any legal transfer of an asset irrespective of any beneficial interests in that asset under any applicable law.63 This chosen governing law respects the validity of contracts concluded electronically and via private key cryptography and—by hypothesis—allows an explicit choice of the law applicable to proprietary questions.

Typical users of DiceyCoin are trustees who wish to realise trust assets. The process is straightforward. Consider a trustee located in England and an English law-governed trust
containing English assets. The trustee sells one of those assets for an amount of diceycoin. The diceycoin thereby becomes part of the trust estate. The trustee then transfers this
diceycoin to another person in exchange for fiat currency. Now the beneficiary, also located in England, enters the scene in pursuit of their share of the trust assets. The fiat currency having, astonishingly, been squandered and the trustee, equally astonishingly, being impecunious or missing, they seek compensation from the transferee of the trust assets—the person who received the diceycoin trust assets in exchange for fi at currency. This is when the beneficiary discovers that there is no proprietary claim against the diceycoin, because proprietary questions are governed by the chosen law, which has wiped out their proprietary beneficial interest. Even a personal claim against the transferee of the diceycoin is problematical: the transferee points out, quite reasonably, that DiceyCoin exists precisely so that no checks on beneficial interests are necessary and to create a quick, uncomplicated way of transferring assets, so there can be no negative inferences about the transferee’s state of knowledge.64

The situation may be even worse. Assume that the trustee swaps diceycoin with other users before exchanging it for fi at currency. Then it is not even clear that there is a proprietary claim to the trustee’s fi at currency. If the currency is not simply squandered but traceable to another person, the beneficiary may have no claim against it.65

DiceyCoin demonstrates that the third parties affected by a voluntary selection of proprietary governing law are not only those who come after the relevant transaction—transferees, takers of security—but also those who come before—such as trust beneficiaries. While third parties who are involved afterwards might be expected to fi nd out about the implications of the proprietary governing law before they engage with the asset, those involved before have no such opportunity.

Party choice of proprietary governing law in relation to digital assets is too easy: there is no control mechanism, such as requiring the parties to be in a certain jurisdiction or to move the relevant asset to that jurisdiction, that can adequately protect third parties. Such a straightforward way to circumvent English trust law would threaten to undermine its efficacy: even if all the parties are located in England, dealing with English assets and currency and even if all the servers running DiceyCoin are also in England, English law would not apply to protect the beneficiary.

The problem is not just that a party choice rule is susceptible to abuse, but that there are no anti-avoidance rules that protect against such abuse. This is not surprising, as the legal rule implementing party choice which would require such anti-avoidance limitations does not yet exist. Compare the rule permitting voluntary choice of applicable law for contracts.66 Although contract law does not affect third parties in the same way as property law and so there is less scope for abuse, voluntary choice is still limited by anti-avoidance rules that apply provisions of the law of another country in specified circumstances.67  An effective choice of proprietary law would contain similar, probably more extensive, limitations. Delivery of this entire package as a single unit is likely to be beyond the

incremental approach of the common law—it could only be done by statute.

Our, somewhat hesitant, conclusion is that a voluntary proprietary rule for permissioned blockchains—that is, a voluntary lex situs —would be feasible, but only if it was created by statute so as to include all the anti-abuse provisions necessary to protect third parties.68 The hesitation arises from the fact that, once limitations and exceptions are admitted to the rule, it becomes less appealing. Instead of a simple, universally applicable rule, there would be a qualified rule subject to exceptions. Whether this would create more problems than it solved would depend on the details of its statutory implementation.

C. Crypto exchanges

Many of the judgments on digital assets in English courts have concerned crypto exchanges. Generally, the issues raised by these judgments concern contract rather than property. Following a series of decisions based on interim applications, the situation was clarified by Piroozzadeh v Persons Unknown,69 where Trower J accepted the uncontested evidence of the defendant crypto exchange that clients of the exchange retained no property in digital assets deposited with them. Rather, the arrangement was similar to a banking deposit. The underlying assets were swept into a general account owned by the exchange and the client was allocated a credit balance in an account with the exchange. This balance was simply a contractual right against the exchange—again, just as with a deposit held by a bank. Tracing of the digital assets was not feasible. Leaving aside any regulatory implications, in this sort of arrangement, proprietary issues will in general not be relevant as between clients and exchanges.

This is only one example of an arrangement with a crypto exchange. It is notable in being the first case in the English courts where the exchange itself has set out its own understanding, but it is possible that other exchanges operate in a different way that might attract proprietary consequences.70

V. JURISDICTION

Rules for jurisdiction are shotguns, not lasers. Whereas a rule for applicable law seeks the single legal system that is suitable above all others to determine a particular legal issue, rules for jurisdiction may apply a number of different, mutually inconsistent rationales that permit a variety of different fora. In English private international law, order is maintained through the use of the concept of “appropriate forum”, which allows the English court to determine on the facts of a particular case whether the possible forum indicated by a jurisdictional rule should actually be permitted to proceed with the dispute.71

Several rationales are evident in the current rules for jurisdiction.72 These include: practical ability of the court to enforce any resulting judgment against a party to the dispute;73 ability of the court to enforce any judgment affecting disputed property;74 consistency with international norms so as to enhance the possibility that any judgment is enforceable abroad;75 and expertise of the English courts in dealing with questions of English contract and tort law.76 The lex situs of a digital asset, chosen to determine applicable law in relation to a proprietary issue, will add one more possible forum to be made available by the jurisdictional rules.77 Accordingly, fashioning the lex situs of a digital asset to suit a particular jurisdictional basis rather than the appropriate applicable law is to let the tail wag the dog.

There is, however, one aspect of the lex situs that raises an issue peculiar to jurisdiction. This is the temporal aspect. As discussed above, applicable law is determined in the context of a particular event, generally a transfer of the asset. A transfer requires a transferor and a transferee and this gives some assurance that a lex situs based on the asset’s location at the time of the transfer is discoverable. But, for jurisdiction, the lex situs may be determined at the time of the proceedings. This moment has nothing to do with any action in relation to the asset. The lex situs at this arbitrary time may be harder to discover. Accordingly, we suggest a gloss to the lex situs rule when it is used for jurisdiction: to allow the lex situs of the digital asset either at the time of issue of proceedings or, if that cannot be determined, at the time of the most recent event in respect of that asset.

VI. CONCLUSION

The lex situs is justified by principle and practicality. For assets that exist by virtue of a manifestation in the real world, including tangible assets and digital assets, principle requires a connecting factor that intermediates between the real world and rights relating to that asset in the legal system. The lex situs achieves that intermediation, defining the question of location in a way that is conceptually consistent with digital assets and providing answers that encapsulate all necessary real world complexities.

Applying this reasoning to permissionless digital assets and recognising that it is the actors in any event—as opposed to observers—who are relevant to that event, we are led to a lex situs based on the physical location of those actors at the time of that event. This smoothly links the real world and the appropriate legal system while maximising certainty and discoverability.

Footnotes

  1. 1

    See Dicey, Morris & Collins, [24R.068] (Rule 140) (immovables); [25R.001] (Rule 141) and [25R.020] (Rule 142) (tangible movables); [23R.023] (Rule 136) and [25R.057] (Rule 143) (intangibles).

  2. 2

    We use “digital assets” as an umbrella term but avoid trying to give a precise definition. We follow the same nomenclature as Law Com No 412 and the Law Commission’s proposed legislation contained in “Digital assets as personal property: Short consultation on draft clauses” (February 2024). A wider approach is suggested in the UNIDROIT Principles on Digital Assets and Private Law, Art.5 (UNIDROIT 2023 Study LXXII – WG8 – Doc 6) that may include, for instance, entities such as password-protected Word fi les. The discussion here is centred around a narrower conception of digital assets—the paradigm example is bitcoin. We also avoid giving a description of the technology underlying digital assets, except for aspects that are relevant to the discussion. Explanations are available, for example, in C Kerrigan (ed), Crypto and Digital Assets Law and Regulation (London, 2023) and S Green and A Sanitt, “Smart Contracts”, in PS Davies and M Raczynska (eds), Contents of Commercial Contracts (London, 2020). English cases where digital assets have been recognised as property include AA v Persons Unknown, Re Bitcoin [2019] EWHC 3556 (Comm); [2020] Lloyd’s Rep FC 127 ; [2020] 4 WLR 35; LMN v Bitflyer Holdings Inc [2022] EWHC 2954 (Comm); Osbourne v Persons Unknown [2023] EWHC 39 (KB) and [2023] EWHC 340 (KB); Tulip Trading Ltd v Van der Laan [2023] EWCA Civ 83; [2023] 4 WLR 16; and Piroozzadeh v Persons Unknown [2023] EWHC 1024 (Ch). See also UK Jurisdiction Taskforce, Legal statement on cryptoassets and smart contracts (November 2019).

  3. 3

    For contrasting views on this question, see eg M Ng, “Choice of law for property issues regarding Bitcoin under English law ” (2019) 15 JPIL 315; A Held, “Crypto Assets and Decentralised Ledgers: Does Situs Actually Matter?”, in A Bonomi, M Lehmann and S Lalani (eds), Blockchain and Private International Law (2023); K Takahashi, “ Law applicable to proprietary issues of crypto-assets” (2022) 18 JPIL 339; F Guillaume, “Aspects of private international law related to blockchain transactions”, in D Kraus, T Obrist and O Hari (eds), Blockchains, Smart Contracts, Decentralised Autonomous Organisations and the Law (2019), A Dickinson, “Cryptocurrencies and the Conflict of Laws”, in D Fox and S Green (eds), Cryptocurrencies in Public and Private Law (Oxford, 2019); and A Dickinson, “ Electronic trade documents and the conflict of laws in the United Kingdom” [2024] LMCLQ 55 .

  4. 4

    Law Commission, “ Digital assets and ETDs in private international law: Which court, which law? Call for evidence” (February 2024).

  5. 5

    See eg K Gray, “Equitable Property” (1994) 47 CLP 157, 160. This level of precision is normally not required and “property” is used as a shorthand for both the asset and the relationship, including elsewhere in this article.

  6. 6

    The English common law has embroidered this concept over time to the point where it is now classified as “legal possession” and “physical possession”. This amply demonstrates its intermediary position. It is mainly physical possession with which we are concerned here.

  7. 7

    A Sanitt, “What sort of property is a cryptoasset?” [2021] 2 JIBFL 83; Law Com 412, [3.54].

  8. 8

    For an application of this test to distinguish choses in action from digital assets, see the discussion of ByBit v Ho Kai [2023] SGHC 199 in the text immediately following..

  9. 9

    Law Com No 412, [4.15].

  10. 10

    See eg D Fox, “Digital Assets as Transactional Power” [2022] 1 JIBFL 3, 3 (“an ideational thing containing different components”). Although note that we do not follow him to the conclusion that digital assets exist as rights to impose changes to the distributed ledger.

  11. 11

    This is shown by, eg, RH Stevens, “Crypto is not Property ” (2023) 139 LQR 615, where digital assets are seen as a type of information (and see the discussion in the text immediately following) and in Tulip Trading v Van der Laan [2023] EWCA Civ 83, [72] (Birss LJ: “ all there is, is software”).

  12. 12

    Further discussion of the underlying technology is below under “Permissionless blockchains” ( post , Pt IV(A)). We defer it until that point, as the emphasis here is not on the details of that technology but its effects.

  13. 13

    As in the nation-state, not the geography in which it is located.

  14. 14

    One popular description asserts that currencies created by people are “the figments of their collective imagination”: Yuval Noal Harari , Sapiens: A brief history of humankind (2011), 201.

  15. 15

    The Bitcoin genesis block (a part of the original development of Bitcoin inserted by its pseudonymous creator Satoshi Nakamoto) famously includes the message “The Times 03/Jan/2009 Chancellor on brink of second bailout for banks”, which, apart from serving as a time stamp proving it was not created before that date, also shows that the inspiration for Bitcoin was money as a shared belief that was vulnerable to a deficiency in trust.

  16. 16

    ByBit v Ho Kai [2023] SGHC 199.

  17. 17

    Ibid , [36].

  18. 18

    Ibid , [31]; and see the discussion in the following section (“It’s the public key, not the private key”).

  19. 19

    ByBit v Ho Kai [2023] SGHC 199, [31] See also K Low, “Trusts of Cryptoassets” (2021) 34 TruLI 191.

  20. 20

    In Stevens (2023) 139 LQR 615.

  21. 21

    Given statutory recognition in the Electronic Trade Documents Act 2023.

  22. 22

    Watts & Low.

  23. 23

    Colour was added, for example, to bitcoins via a variety of methods, of which an early example is the EPOBC protocol (Enhanced Padded-Order-Based Coloring protocol).

  24. 24

    Watts & Low mention the “double spend problem”, but the examples they give are not actually relevant to that problem. This leads them to undervalue the innovative nature of digital assets and, by extension, the worth of according them property rights.

  25. 25

    Law Com No 412, [4.33].

  26. 26

    Watts and Low’s point has more impact if the essential element of the digital asset system is characterised as the power relationship enabling the asset to be transferred: see A Dickinson, “Cryptocurrencies as ‘Property’ in the conflict of laws” , in S Green and A Dickinson (eds), Cryptocurrencies in Private and Public law (2019) and D Fox, “Digital Assets as Transactional Power” [2022] 1 JIBFL 3, 3. We differ from them in this respect.

  27. 27

    Except for electronic trade documents, where the intermediate concept is decreed by statute to be possession: see the Electronic Trade Documents Act 2023, s.3(1) (“A person may possess, indorse and part with possession of an electronic trade document”). This has the benefit of allowing existing legal rules applicable to trade documents that depend on possession to apply directly to electronic trade documents.

  28. 28

    See, eg, Dicey, Morris & Collins, [25.005] and the cases cited there.

  29. 29

    Ibid , [23.010].

  30. 30

    Macmillan Inc v Bishopsgate Investment Trust Plc (No 3) [1996] 1 WLR 387 (CA) and Akers v Samba Financial Group [2017] UKSC 6; [2017] AC 424.

  31. 31

    As stated explicitly in English, Scottish and Australian Bank Ltd v IRC [1932] AC 238, in which the House of Lords confirmed that debts do have a lex situs : “But debts do, in one form or another, represent property of very considerable value in the modern world, and it appears to me it is desirable that they should possess a locality, even if they are invested with it by means of a legal fiction” (at 245, per Lord Buckmaster).

  32. 32

    Hardy Exploration & Production (India) Inc v India [2018] EWHC 1916 (Comm); [2018] 2 Lloyd’s Rep 331 ; [2019] QB 544; approved on this point in SAS Institute Inc v World Programming Ltd [2020] EWCA Civ 599; [2022] Lloyd’s Rep Plus 55 , [59–60].

  33. 33

    [2018] EWHC 1916 (Comm), [82(3)].

  34. 34

    This conclusion, and the reasoning, also being in accordance with Dicey, Morris & Collins, [23R.023] (Rule 136).

  35. 35

    Hardy Exploration & Production (India) Inc v India [2018] EWHC 1916 (Comm), [82(7)].

  36. 36

    Regulation (EC) No 593/2008.

  37. 37

    What is clear is that the scope of Art.14 is unclear. The original proposal for the Rome I Regulation included a rule on third-party effects of assignment which was not included in the final text. Instead, the European Commission was required to prepare a report on this issue. Accordingly, the European Commission asked the British Institute of International and Comparative Law (BIICL) to produce a study, which was delivered in September 2016. The Commission then proposed a Regulation harmonising conflicts of laws rules on the third-party effects of assignment of claims on 12 March 2018. This proposed Regulation has not been enacted.

  38. 38

    Antony Gibbs and Sons v La Société Industrielle et Commerciale des Métaux (1890) 25 QBD 399. See also OJSC International Bank of Azerbaijan [2018] EWCA Civ 2802; [2019] Bus LR 1130. Note that the rule does not apply if the parties have submitted to the foreign proceedings. The rule has been subjected to academic criticism for its perceived incompatibility with modifi ed universalism but retains considerable judicial and practitioner support. In July 2023, the Insolvency Service published the outcome of a Consultation on adopting the UNCITRAL Model Law on Enforcement of Insolvency-Related Judgments that acknowledged many responses had highlighted its possible incompatibility with the rule in Gibbs and proposed further development of the proposal to deal with those concerns (the Consultation is available here: www.gov.uk/government/consultations/implementation-of-two-uncitral-model-laws-on-insolvency/outcome/implementation-of-two-uncitral-model-laws-on-insolvency-summary-of-consultation-responses-and-government-response).

  39. 39

    These considerations are also set out in Dicey, Morris & Collins , [25.060].

  40. 40

    Ibid , [25R.001] (Rule 141: “The validity of a transfer of a tangible movable and its effect on the proprietary rights of the parties thereto and of those claiming under them in respect thereof are governed by the law of the country where the movable is at the time of the transfer ( lex situs )”; [25R.020] (Rule 142: “A title to a tangible movable acquired or reserved in accordance with Rule 141 will be recognised as valid in England if the movable is removed from the country where it was situated at the time when such title was acquired, unless and until such title is displaced by a new title acquired in accordance with the law of the country to which it is removed”).

  41. 41

    Winkworth v Christie Manson and Woods Ltd [1980] Ch 496.

  42. 42

    Rule 142 deals with the situation of conflit mobile: where lex situs, the connecting factor, subsequently changes location.

  43. 43

    This rule does apply in certain cases involving succession.

  44. 44

    For breach of contract: “the attitude of the courts, reflecting that of the Court of Chancery, is that specific performance of contractual obligations should ordinarily be refused where damages would be an adequate remedy” (Cavendish Square Holding BV v Makdessi [2015] UKSC 67; [2016] 1 Lloyd’s Rep 55 ; [2016] AC 1172, [30] (per Lord Neuberger).

  45. 45

    The Court retains a discretionary power to order return of the goods rather than damages: see the Torts (Interference with Goods) Act 1977, s.3. It will exercise this discretion only if damages are not an adequate remedy: “A principle on which the court has long acted is not to order delivery of goods which are ordinary articles of commerce with no special value or interest, whether to the plaintiff or others, when damages will fully compensate … the goods detained by purchasing their equivalent on the market, then the payment of damages out of which the price of the equivalent may be paid is adequate compensation to the wronged plaintiff, and there is little or no point in making an order for the delivery of the goods”: Howard Perry & Co Ltd v British Railways Board [1980] 1 WLR 1375, 1382–1383 (Sir Robert Megarry V-C).

  46. 46

    Although in theory it can still be used if the defendant still has the asset.

  47. 47

    Byers v Saudi National Bank [2023] UKSC 51; [2024] 2 WLR 237.

  48. 48

    Akers v Samba Financial Group [2017] UKSC 6; [2017] AC 424.

  49. 49

    Where the asset is created by virtue of a legal system, ie, falls within the existing category of intangible assets, the connecting factor will be a legal manifestation of the asset; that is, something determined by the legal system in which the asset subsists.

  50. 50

    For further details of exactly how and when this works, see A Sanitt, “Remedies for smart legal contracts: Rectification and rescission reconsidered”, in B Soyer (ed.), Damages, Recoveries and Remedies in Shipping Law (London, 2024). Note that there may also be considerable practical difficulties in achieving rectification and rescission, due to pseudonymity of the blockchain, cross-border issues and legal uncertainty as to the proprietary effects of different transfers. Disputes often involve expensive and detailed blockchain analytics culminating in claims against “Persons Unknown”. See T Chan and K Low, “Post-scam crypto recovery: final clarity or deceptive simplicity?” (2023) 139 LQR 379, 380.

  51. 51

    M Lehmann, “Extraterritoriality in Financial Law”, ch.24 of A Parrish and C Ryngaert (eds), Research Handbook on Extraterritoriality in International Law (2023). See also Held, supra, fn.3.

  52. 52

    Tulip Trading v Van der Laan [2023] EWCA Civ 83.

  53. 53

    Ibid, [91] (Birss LJ).

  54. 54

    See the criteria set out by Lord Wilberforce in National Provincial Bank Ltd v Ainsworth [1965] AC 1175 and the discussion in Law Com No 412.

  55. 55

    Sadly for legal scholars, in April 2024, the action was discontinued.

  56. 56

    A Dickinson, “Cryptocurrencies as ‘Property’ in the conflict of laws” , in S Green and A Dickinson (eds), Cryptocurrencies in Private and Public law (2019), takes a slightly different path to a similar conclusion—he suggests a connecting factor based on the physical location of the relevant participant in the system. Although recognising the independence of digital assets, he adds a gloss to the independent public key attribute, locating the object of property in the expectation of participants that transfers will be effected from that public key in accordance with the rules of the system. We do not consider that extra stage is necessary: the shared belief constitutes the object as the item in the public key’s account, without any expectation or claim as to what will happen to it—see the discussion above, Part II(A). Dickinson’s gloss tends to emphasise the private key as part of the system, which may interfere with the clear development of lex situs as locating the virtual world attribute, the account labelled by the public key.

  57. 57

    This is the test that has so far been adopted by the English courts. See Ion Science v Persons Unknown (21 Dec 2020) Unreported; Fetch.AI v Persons Unknown [2021] EWHC 2254 (Comm); the first instance decision in Tulip Trading v Van der Laan [2022] EWHC 667 (Ch) (where the judge preferred residence to domicile; this was not challenged on appeal).

  58. 58

    For further discussion of the alternatives and additional argument in favour of the location of the transferor, as opposed to the transferee, at the time of the transfer, as well as consideration of related fact-situations such as control of the private key being in the hands of an agent, see Dickinson [2024] LMCLQ 55.

  59. 59

    Eg PD6B, para.3.1(11) (“the subject matter of the claim relates wholly or principally to property within the jurisdiction”: in Osbourne v Persons Unknown [2023] EWHC 39 (KB), Lavender J inclined to the view that the relevant time was when the application was made for permission to serve out of the jurisdiction); PD6B, para.3.1(15) (a “claim made against the defendant as constructive trustee, or as trustee of a resulting trust, where the claim arises out of acts committed or events occurring within the jurisdiction or relates to assets within the jurisdiction”). In Denisov v Delvecchio [2022] EWHC 377 (Comm), HHJ Pelling QC, sitting as a judge of the High Court, assumed that the part of the test dealing with location of assets applied at the time the application was made for permission to serve out the jurisdiction.

  60. 60

    If it does not, the courts will develop other rules and we do not attempt to predict what will replace it or how that will affect ownership questions. Nor do we assume this analysis is necessarily correct: the point did not directly arise in Tulip Trading v Van der Laan [2023] EWCA Civ 83, but was mentioned (at [25] by Birss LJ: “some would say a fresh piece of property is created every time bitcoin is transferred, but there is no need on this appeal to get into that debate”) and it was the view of Law Com No 412. But D’Aloia v Persons Unknown [2024] EWHC 2342 (Ch) held that the features of one particular cryptocurrency, Tether, meant that it did persist through transfers. The contrary view was also supported in ByBit v Ho Kai [2023] SGHC 199, [31] (Jeyaretnam J: “This physical manifestation at the level of digital bits and bytes is not permanent, and changes with every transaction. Nonetheless, we identify what is going on as a particular digital token, somewhat like how we give a name to a river even though the water contained within its banks is constantly changing”). And, for a discussion of the difficulties arising from this view, see Chan & Low (2023) 139 LQR 379, 380. Ultimately, this is a factual not a legal question: see A Sanitt, “What sort of property is a cryptoasset?” (2021) 2 JIBFL 83.

  61. 61

    See A Sanitt and I Grigg, Legal analysis of the governed blockchain (Norton Rose Fulbright, 2018). The agreement is commonly referred to as a “constitution”, although it is likely to take effect as a series of multilateral agreements. We assume that the governing law chosen in the constitution is one in which this method of concluding an agreement is legally binding (and note that under English conflicts of law rules it is the law purportedly chosen in the agreement that would be used to determine its validity).

  62. 62

    The UNIDROIT Principles on Digital Assets and Private Law (UNIDROIT 2023 Study LXXII – WG8 – Doc 6), Art.5, proposes a “waterfall” structure that would allow choice of law for proprietary questions. Certainty is given as the justification for this approach in the Summary Report of the second preparatory meeting of the HCCH-UNIDROIT applicable law digital assets project (UNIDROIT 2023 Study LXXIIA – PM2 – Doc 3, November 2023, para.77).

  63. 63

    As demonstrated by Akers v Samba Financial Group [2017] UKSC 6; [2017] AC 424, such systems of law do exist.

  64. 64

    The personal claim might be based on dishonest assistance in a breach of trust—knowing receipt appears to be precluded by Byers v Saudi National Bank [2023] UKSC 51; [2024] 2 WLR 237, even if the transferee still owns the diceycoin.

  65. 65

    The position is unclear because, although Byers v Saudi National Bank [2023] UKSC 51; [2024] 2 WLR 237 specifies that a trust will spring up again if assets fi nd their way back to the original trustee, this does not on its face apply to new diceycoin that are economically equivalent to the original assets but are not traceable (or followable) into them, due to the eradication of beneficial interests under the proprietary governing law.

  66. 66

    Contained in the Rome I Regulation, Art.3(1). effective as part of EU retained law.

  67. 67

    These circumstances are “where all other elements relevant to the situation at the time of the choice are located in a country other than the country whose las has been chosen”, in which case non-derogable rules of that other country will apply (Rome I Regulation, Art.3(3)) and, on a discretionary basis, overriding mandatory rules of the forum or, where they would render performance illegal, the country of performance (ibid, Art.9).

  68. 68

    In addition to the potential for legal abuse mentioned here, there may also be potential for regulatory abuse: regulatory arbitrage, whereby elective lex situs is used so that digital assets are subject to a less onerous regulatory regime than would otherwise apply.

  69. 69

    Piroozzadeh v Persons Unknown [2023] EWHC 1024 (Ch).

  70. 70

    In B2C2 Ltd v Quoine Pte Ltd [2019] SGHC(I) 43, the Singapore International Commercial Court found that a crypto exchange did not hold cryptocurrency on trust for its users. In Ruscoe v Cryptopia Ltd (in liq)[2020] NZHC 728, the New Zealand High Court found that a crypto exchange did hold assets on trust for users, distinguishing Quoine on the basis that, among other things, the terms and conditions expressly referred to cryptocurrency being held “on trust for you” (at [27]).

  71. 71

    The rules are set out in PD6B, which contains a long list of possible permitted fora. The relationship between this list and the control mechanism of appropriate forum is a matter of debate: see FS Cairo (Nile Plaza) LLC v Brownlie [2021] UKSC 45; [2022] 2 Lloyd’s Rep 127 ; [2022] AC 995 (“Brownlie (No 2)”). The view implicitly taken here is that the judgment of Lord Lloyd-Jones in that case states the law (see esp [77]) and that this supersedes the narrower view of Lord Sumption in Four Seasons Holdings Inc v Brownlie [2017] UKSC 80; [2018] 1 WLR 192 (“Brownlie (No 1)”).

  72. 72

    That is, the list set out in PD6B.

  73. 73

    Eg, PD6B, para.3.1(1) (defendant domiciled within the jurisdiction)

  74. 74

    Eg, PD6B, para.3.1(11) (claim relates to property within the jurisdiction).

  75. 75

    Most of the grounds in PD6B would serve as an example: PD6B, para.3.1(1) is the paradigm.

  76. 76

    PD6B, para.3.1(6)(c) (claim relates to English law contract); PD6B, para.3.1(9)(c) (claim relates to tort governed by English law).

  77. 77

    Via PD6B para.3.1(11), which applies where the subject matter of the claim relates wholly or partly to property within the jurisdiction.

Contact