There is now an emerging consensus that cryptoassets are property – judges have so held in cases in England and other jurisdictions and it is the unanimous conclusion of independent reports on the topic.1 But what sort of property? The category of property into which cryptoassets are placed has wide implications in areas such as insolvency, security and tax. Two rival schools of thought have developed and the ongoing confusion has led to legal uncertainty. In this article, we argue for the correct categorisation based on a principled explanation of why different categories of property exist and how they apply to cryptoassets.
By a cryptoasset, we mean a representation of data on a distributed ledger cryptographically secured and updated subject to a decentralised consensus mechanism. The definition is evolving and we eschew precise definitional debates as they do not affect the argument. What counts is that the cryptoasset is constituted by the data as it exists in the ledger itself and it is not simply a tokenisation of something that exists in the real world, as in a property register. The paradigm example of a cryptoasset is Bitcoin. We similarly avoid defining property, as would be necessary to determine the crucial issue whether cryptoassets constitute property – instead, we simply assume that cryptoassets are property. We are concerned instead with the question of what sort of property is a cryptoasset.
In English law and related systems of common law, personal property falls into two categories: things in possession and things in action. Things in possession (TIP) are things that can actually be possessed – that is, objects with a physical existence. Things in action (TIA) are rights that can be enforced by a court action – that is, they are legal rights against another person. This distinction immediately highlights the problem. Cryptoassets are clearly not physical objects – there is nothing that can be physically possessed. But nor are they rights against a person – there is nobody to sue by virtue simply of ownership of a cryptoasset.
English law has been quite insistent that these two categories are exhaustive, although the cases inevitably fail to appreciate the possible existence of a creature quite like a cryptoasset.2 This has led to the following
- Creation of a third category of property. In one sense, this is a tautology – once we accept that cryptoassets are property and given that they have distinct attributes, they require their own new rules and that means they are implicitly in their own category. Accordingly, this does not resolve whether we should treat them as analogous to TIP or to TIA, so rather than ending the debate, this really just postpones it.
- Treating TIA as a category that includes any property that is intangible, so that cryptoassets are TIA.
- Treating cryptoassets as subject to a sort of “intangible possession”, so that they can be assimilated to TIP.
We examine the distinction between TIA and TIP through the creation and destruction of property and then, which is slightly more problematical, through its transfer. First, let us consider a TIA. This can be created or destroyed in only one way – through the application of a legal rule. The relevant legal rules may be set out in legislation or case law and will be applied by the courts to determine in what circumstances a TIA has been created or destroyed. For instance, if a debt has been repaid a court will normally find that it has ceased to exist (although doctrines such as mistake could negate this conclusion in
particular circumstances). But, although this decision is due to a change in the facts, it is still the legal rule itself (expressed in a court decision) that determines creation or destruction, not the fact of repayment.
That is, the court could decide that the debt still exists and then it would exist, notwithstanding any repayment. This may seem like labouring the obvious, but we stress the point because of the different treatment of TIP. A court does not have the same power in relation to a TIP. Take a physical object such as a car, for instance. A court is not able to pronounce that a car does not exist, in the way it can pronounce that a debt does not exist. It can pronounce on who owns the car and hence transfer property interests in the car, but that is a different point and we deal with transferability below.
TIP cannot be called into or out of existence by legal decisions – there is something beyond the reach of the courts. Hitherto, this something has, of course, been the physical existence of the TIP. The legal system must recognise and deal with this. It cannot decree what shall happen by fiat; it must take account of what exists in the real world. This is generally accomplished by the mediation of the concept of possession. A physical object is capable of being possessed. Possession is a legal concept rather than a purely physical object, but it is very useful in enabling courts to create systems of rules that deal consistently with TIP.3
This analysis suggests that the fundamental distinction is not between tangible and intangible objects, but between objects that have a purely legal existence and those that have an existence independent of the law. The dividing line is not physical/non-physical, it is legal/nonlegal. The concept of possession applied to a TIP is a means of accommodating the non-legal facts that apply to a TIP: for instance, that it is somewhere in the world and somebody has it. No such mediation is needed for a TIA – any legal rules can apply to it directly.
How does this distinction apply to a cryptoasset? Crucially, no legal rule can create or destroy a cryptoasset independent of the underlying facts; no court decision can effectively say that it has ceased to exist. The permanence and stability of a cryptoasset are inherent in the fact of the distributed consensus mechanism and the underlying cryptographically secure ledger. Contrast this with an online bank account – something that may seem superficially similar to a cryptocurrency account. There is nothing to stop a court holding that a balance in an online bank account is zero; that is, that the debt owed by the bank should cease to exist. This works independent of any password held by the bank customer or the security applied by the bank to its own records. The bank account is a TIA (a debt owed by the bank to the customer, if the account is in credit) and a legal rule can extinguish it. If the bank fails to change its records to reflect the decision, then its records are wrong. The same court decision applied to a cryptocurrency would not be effective. If the cryptocurrency account still shows the amount in its records, then that amount exists irrespective of any court decision (again, this is a separate question from who legally owns that amount, which we
Legal decisions can affect cryptoassets, but they must take account of these realities. The physical reality of a TIP is reflected in the concept of possession – a state of affairs that can only apply to a physical object. What is needed for a cryptoasset is an analogous legal concept, that reflects the physical reality that somebody can affect the cryptoasset just as somebody with possession of a physical asset can affect it. Although it captures the analogy with possession, the suggestion of “intangible possession” as a name for the concept is subject to the reasonable objection that it is an oxymoron, because if something is intangible it cannot be possessed. “Control” has also been mooted, but this risks confusion with the existing legal concept of control, which can apply more widely than the law of property. While emphasising that this concept is closely linked to possession in its purpose and effect, we suggest that a completely new name is preferable. To complement things in action and things in possession we contribute “things in command” (TIC). That is, “command” is the analogue of “possession” for cryptoassets, reflecting the inescapable physical fact that dealing with a cryptoasset depends on the ability to give commands that will be accepted by the distributed consensus and lead to changes in the distributed ledger. “Command” also nods to the computational origins of cryptoassets and the process by which commands are sent to a computer system to effect changes in them.
Irrespective of whether the category of TIC is seen as part of TIP or a separate third category distinct from both TIP and TIA, what is important is that in behaviour it is assimilated to TIP. TIA contains property which is purely legal in existence – it can be created or annihilated by a court decision or legislation without creating any inconsistency with the real world. TIP and TIC contains property which has an ineradicable real world existence; it is not just legal.
Finally, we deal with the argument that a cryptoasset resembles a TIA rather than a TIP because a cryptoasset cannot be transferred to another person – it can only be created or destroyed, not moved. One weakness of this argument is that “transfer” is not a term of art and can be mapped to different legal concepts.4 But more fundamentally, the relationship between the type of property and how it is transferred is actually much more fluid than it first appears and cannot support a categorisation of cryptoassets as similar to TIA.
The most basic method of transferring possession of a physical object is to move it from one person to another person. The law places numerous glosses on this, of course, relating to legal control, agency and so on, but all these rules must fundamentally allow for the reality that physical objects can move from place to place. TIA can also move from place to place, but only to the extent that the law specifically allows them
to. For instance, s 136 Law of Property Act 1925 allows contractual rights to be legally transferred through assignment. Absent compliance with that rule (or other rules on equitable assignment), a contractual right cannot be transferred, irrespective of any facts in the real world. By contrast, a cryptoasset can be moved from one account to another account, irrespective of legal rights. Whether this is characterised in law as a deletion followed by the creation of a new asset or a transfer, legal rules need to cater for the physical reality that the change in account balances has taken place. That is similar to TIP, where questions of legal possession, ownership and so on must in some way be reconciled with the physical location of the asset. Dwelling on whether a cryptoasset has moved or been deleted and recreated misses the point – the law must cater for the fact that a cryptoasset is in a different place, however it got there and whether or not its identity is the same. In fact, rather than trying to use transference as a legal concept to characterise property, we
should recognise that it is a factual concept. To illustrate: if it ever became possible to teleport a physical object, it would not be the role of the law to decide whether the object had moved or whether it had disappeared and been replaced by a new object. That would be a factual, perhaps a scientific, question. The role of the law would simply be to incorporate those facts into its legal rules. In the same way, deciding whether a cryptoasset has moved or been replaced is a factual question. It is the fact that a cryptoasset can move or be replaced, without legal intervention, that suggests it is better seen as a type of TIP than a TIA.
In conclusion, we suggest that the relevant division in the classification of property is between legal assets and non-legal assets, where a non-legal asset has some inescapable connection to the real world that must be accommodated by the law. Hitherto, that connection has been expressed by concepts of possession and the corresponding category of “thing in possession”. A cryptoasset also has an inescapable connection with the real world; it cannot be created or deleted by legal rule, and so it should be seen as a non-legal asset, similar to a “thing in possession”. The phrase “thing in command” is a legally neutral but expressive term for the appropriate category.