Contracts Jim, but not as we know them: Cryptoassets and Smart Contracts
Chris Recker and Jeremy Richmond of Quadrant Chambers boldly explore the status of cryptoassets and smart contracts.
This article was first published in the Commercial Litigation Journal.
In November 2019 the UK Jurisdiction Taskforce (UKJT) published a legal statement on the status of cryptoassets and smart contracts following the issuing of its consultation paper in May 2019. As Chancellor Vos noted in the foreword to the legal statement, there is no doubt that some of the matters covered in it will in the future be the subject of judicial decision. The ever increasing use of cryptoassets and smart contracts has given rise to some difficult and subtle issues of contract, insolvency and property law, not to mention conflicts of law. This article seeks to summarise some of the main points in the legal statement with an emphasis on these areas of law. The legal statement is broadly divided into, respectively, the treatment of cryptoassets as property, and smart contracts. It is to these two areas that we now turn.
Cryptoassets as property
The UKJT noted that there is no universally accepted definition for the term ‘cryptoasset’. At a very high level of abstraction a cryptoasset is defined by reference to the rules of the system in which it exists. Functionally it is usually represented by two parameters: one public, and one private. The public parameter contains encoded information about the asset like ownership, value and transaction history, which information is available to all participants in the system or the world at large. The private parameter – the private key or keys – allows dealings in the cryptoasset to be cryptographically authenticated by digital signature. The UKJT identified some principal and novel features of cryptoassets including:
- cryptographic authentication;
- use of a transaction ledger, the function of which is to keep a reliable history of transactions and so prevent double spending;
- the distribution transaction ledger may be distributed and decentralised in the sense that it is shared over the network with no one person having a responsibility for maintaining it, or any right to do so; and
- rule by consensus in that the rules governing dealings are established by the informal consensus of participants rather than by contract or in some other legally binding way.
A pressing consideration for commercial lawyers is whether a cryptoasset is capable of being property. If the cryptoasset is capable of being property, then among other things the holder of a proprietary right over the cryptoasset may have priority over claims by creditors in an insolvency context, and may be able to create a security interest over the cryptoasset, which otherwise might not be possible.
The UKJT after an extensive analysis arrived at the conclusion that cryptoassets should be treated in principle as property because of the following:
- Cryptoassets have all the indicia of property, namely, they are definable, identifiable by third parties, capable in their nature of assumption by third parties, and have some degree of permanence or stability: per Lord Wilberforce in National Provincial Bank v Ainsworth . The conclusion that cryptoassets have the indicia of property is consistent with the relatively limited English case law on point to date: see Vorotyntseva v Money-4 Ltd  (interim proprietary injunction granted over Bitcoin and Ethereum cryptocurrency valued at £1.5m); and Robertson v Persons Unknown  (where an asset preservation order was granted over Bitcoin).
- The novel features of some cryptoassets – intangibility, cryptographic authentication, use of a distributed transaction ledger, decentralisation, rule by consensus – do not disqualify them from being property.
- Nor are cryptoassets disqualified from being property by reason of being ‘pure information’, or because they might not be classifiable either as things in possession or things in action.
Since a cryptoasset cannot be physically possessed and is, as such, virtual, the UKJT arrived at the following conclusions concerning the potential security or other interest that could be taken over them:
- As a matter of law, cryptoassets could not be the object of a bailment since bailment requires the transfer of possession. Since cryptoassets are not capable of being possessed, they could not be the subject of a bailment.
- Only some types of security could be granted over cryptoassets. The UKJT confined itself to the question of consensual security of which there are four kinds in English law: pledge, (contractual) lien, equitable charge and mortgage. It considered that pledges and liens can only be created by transferring possession of an asset. As such, since cryptoassets could not be possessed it would not be possible to create a pledge or lien over them. However, the UKJT considered that if a particular cryptoasset was property, a mortgage or equitable charge could be created over it in the same way a mortgage or equitable charge could be created over other intangible property and subject to the same requirements.
- Cryptoassets are not documents of title, documentary intangibles or negotiable instruments (though some form of negotiability could arise in future as a result of market custom).
- Cryptoassets are not instruments under the Bills of Exchange Act 1882 so that the provisions of that Act do not apply.
- Cryptoassets are not goods within the meaning of s61(1) of the Sale of Goods Act 1979 so that the provisions of that Act do not apply.
The UKJT expressed the view that since cryptoassets can be property at common law, they could be property for the purposes of the Insolvency Act 1986. In an interesting observation, the UKJT also considered that where a particular cryptoasset was not property at common law, it might still be property within the meaning of the Insolvency Act 1986. This was because s436(1) of the Insolvency Act 1986 defined property to include:
… money, goods, things in action, land and every description of property wherever situated and also obligations and every description of interest, whether present or future or vested or contingent, arising out of, or incidental to, property.
A cryptoasset that was not property at common law could still be property under the Insolvency Act 1986 insofar as it could be considered an ‘obligation [or a] description of interest… arising out of, or incidental to, property’.
The UKJT also considered some difficult choice of law problems that arise in the treatment of cryptoassets. Generally, English law treats questions of:
- how property is to be classified; and
- how and when a transfer of property affects third parties,
as questions of the lex situs, that is to say the law of the country where the property is situated at the relevant time. Since cryptoassets are intangible property it is not altogether obvious how the lex situs would be identified or indeed whether the general rule would (or should) apply at all. Other candidate choice of law rules might include the country where the cryptoassets have some sort of central control; or the law which governs the transaction between the parties (as in cases where a tangible asset is in transit so that its location is really a matter of chance or is unknown). The UKJT considered that these complex conflicts of law issues should be resolved by legislation most likely following international cooperation.
The UKJT’s analysis is welcome and comes at an important time; smart contracts are becoming increasingly utilised and embraced by businesses. The significance and potential applications for smart contracts are diverse; in the construction industry it is thought that smart contracts could be implemented t simplify greatly and automate the payment of contractual obligations. In the insurance industry, smart contracts are being explored as mechanisms for both the purchase of policies, and the fast resolution and settlement of claims. The key theme is efficiency:
Before considering those issues, it is worth describing why smart contracts warrant consideration. In summary, it is because they improve the efficiency, speed and performance of contracts. Efficiency is improved because of the automation of contractual actions, which reduces the need for human involvement and, as a result, the potential for human error. Speed is improved as actions can occur in real time as information is collected and verified. Performance is improved as the terms are unambiguous and results predictable and auditable.
(Farrell, Machin and Hinchliffe, ‘Lost and found in smart contract translation – considerations in transitioning to automation in legal architecture’.)
In contrast, issues have also been raised about the extent to which too much trust is being placed in the software rather than the other party to a contract. The concern is that no code is perfect and, therefore, the more complex it is, the greater the risk:
A similar situation took place in 2017 and concerned the already mentioned QuadrigaCX. The company lost US$14 million in cryptocurrency [Ethereum (ETH)] because of the defective smart contract which processed the transactions involving ETH. Apparently, an update of the algorithms caused an error in the code, which disabled the correct acceptance of payments in ETH. The defect was noticed after a couple of days, when it was found that ETH was ‘blocked’ in the smart contract.
(Hulicki, ‘Cryptocurrencies – a legal dilemma of the digital age’).
The UKJT's position
The UKJT highlights that contract law ‘is concerned with the enforcement of promises’ and addresses the argument that a smart contract’s code will do ‘what it is programmed to do’. This is where the primary risk sits; the quality of the code will in principle dictate the quality of the performance of the aspects governed by the smart contract itself. The UKJT, however, also makes the valid point that there could be external factors which affect the performance of the code. A party can, therefore, on a case-by-case basis, argue that a smart contract is or is not legally binding.
The UKJT points out that the concept of a contract under English law is not defined, and there is generally no particular ‘form’ that a contract must exist in. The courts will enforce a promise provided that the constituent common law requirements are met (offer and acceptance, certainty, an intention to be legally bound by the agreement and consideration). The court would, of course, also take into account what the UKJT refers to as ‘vitiating factors (such as duress, misrepresentation or illegality)’.
The UKJT highlights that ‘the precise role played by the software in a smart contract can vary’. This is important. The interpretation of the contractual relationship will be governed on its merits. Parties to a contract may wish for a smart contract to define all of their obligations, merely enforce them or use some other hybrid arrangement. Other parties may not wish to utilise any code at all. The enforceability and scope of the contract, as the UKJT points out, will be a matter of analysing the words and conduct of the parties to determine objectively what was agreed.
A few examples are explored, and the theme raised by the UKJT is that the further away the parties move from the ‘conventional’ methods of contracting with each other, the less straightforward the position:
- A natural language contract with certain performance obligations executed with code. In these circumstances, the contractual interpretation should not be controversial; any differences in relation to performance (including the operation of the code) can be determined by reference to the contract.
- A natural language contract but with some terms in code (such code likely being embedded in the contractual documentation, and/or on an execution platform). Here, the question is of the constituent parts; has an agreement been reached, was there sufficient intention and has there been sufficient consideration? It is not envisaged that there is significant material difference to the category appearing immediately above given that there is some form of underlying natural language contract.
- A contract with no natural language is most problematic, given that this would mean that the entire agreement existed in code. The terms of the agreement will be enshrined within the source code, and any consideration ought to be identifiable by an examination of that code or considering the code’s behaviour. The challenge is assessing whether or not an agreement has been reached at all, and whether there was an intention to create legal relations. The analysis of those aspects will need to be explored by reference to the code itself, and any applicable evidence (such as written correspondence) which may indicate what the parties intended.
- A party utilises code on a distributed ledger platform, where in exchange for payment or transfer of value, the platform performs a service (such as algorithmic investment). A third party who is not known to the party deploying the code could potentially form a unilateral contract with the other party. The underlying analysis would be very fact specific, and would evidentially depend on the nature of the platform and the arena in which it operated (among other things).
- A party deploys code to set up a decentralised autonomous organisation (DAO), but has no intention to participate or enter into contracts with other parties. Entirely separate third parties will contract with each other in accordance with the terms of the smart contract running the DAO. The UKJT points to the challenges with identifying who the parties are to those contracts, and establishing whether any legally binding agreement has been reached. Helpfully, the UKJT compares this situation with transacting with an unincorporated association, whereby each member of that association contracts with the membership as a whole and intends to be bound by the applicable rules of that association (even though each member does not necessarily know the identities of the others).
The English court's application
The rules of contractual interpretation in England and Wales are well known; the court will determine what the reasonable reader would consider the parties objectively intended (which will call for an analysis of the language utilised in the agreement). The court generally relies on what was stated in an agreement, unless the language is unclear or ambiguous.
The UKJT points out that these rules do not easily reconcile with a contract containing terms exclusively in code. However, to the contrary, the UKJT suggests that a contract of that nature could be interpreted as nothing more than:
… an extreme example of a contract whose language is clear, with the result that there is no justification to depart from it.
It still, however, begs the question as to what happens if the code itself is not clear (and which may give rise to different meanings or interpretations).
The UKJT envisages that these issues would be resolved by either:
- reference to other parts of the code which are clear; or
- looking ‘beyond the four corners of the code to interpret it’.
The court’s role in that scenario is to assess all of the admissible evidence and make findings relating to whether or not the code actually defined the obligations, or was intended to merely implement them. This will of course dictate the nature of the evidence (including expert evidence) required.
The parties to a dispute may, therefore, ask the court to consider specific issues. The UKJT points out that, by way of example, one of the parties may argue that the code needs to be rectified, which would require analysis beyond the outcome or operation of the code itself. The UKJT also considers that the usual rules relating to the court’s interference in circumstances of duress, fraud and misrepresentation would continue to apply.
The UKJT clarifies that there is no requirement under English law for parties to a contract to know each other’s real identities and, therefore, sees no issue with smart contracts being utilised in this manner. The issue of knowing the contracting party is, however, an important one for enforcing the terms of a contract following a breach.
Smart contracts and writing
Another issue raised relates to the use of a private key as a mechanism for satisfying a statutory signature requirement. The UKJT acknowledges that electronic signature can generally satisfy these requirements (depending on the precise statutory instrument) and that a private key is a type of electronic signature when used in this way. The key issue the UKJT has raised:
… is not what the signature looks like, but whether or not it is clear that the party intended to authenticate the full terms of the document.
The UKJT takes this further and considers whether or not the statutory ‘in writing’ requirement could be met by a smart contract composed partly or wholly of computer code. The UKJT points to the wide definition of ‘writing’ in the Interpretation Act 1978, and on balance considers that it is likely the statutory requirement would be satisfied given that:
… the fact that the code might not be comprehensible to an uninitiated English speaker without an expert translator is irrelevant.
The legal statement provides a comprehensive overview of smart contracts but for understandable reasons does not purport to analyse comprehensively every possible legal issue that may arise in the area. We tentatively suggest that two technical areas that may merit further consideration are:
- whether or not it is possible to obtain a court order securing the rectification of a blockchain ledger; and
- the extent to which a judgment could be obtained in a virtual currency.
In summary, the UKJT has helpfully clarified that smart contracts are in principle capable of satisfying the constituent parts of contractual common law. As with the status of cryptoassets, we expect that the legal status of each smart contract will need to be considered on a case-by-case basis. As a result, the law around cryptoassets and smart contracts will no doubt develop and be refined over time.