Can anyone really own virtual tokens, such as Bitcoin, Ether, and Facebook’s anticipated token – the Libra?

The advantages offered by these cryptocurrencies stem from their underlying technology – a decentralized ledger, which automatically records any transfer of tokens between individuals, without any involvement from a centralized authority. Transfers can only occur when token holders use their ‘private key’ (comparable to a username and password) to match the token’s ‘public key’ (comparable to an account number), and once they are completed – all transactions are grouped in immutable inter-connected blocks (yielding the technology’s common name, ‘blockchain’). Yet, this decentralized trait creates a conceptual problem: do tokens constitute a property that can be owned? And, if so, how should the rights of an owner be protected? 

The US Internal revenue service has given a positive answer to the first question: cryptocurrencies should be viewed as property for tax purposes (See Notice 2014-21, IRB 2014-16). Courts around the world seem to concur, with cases in Canada, Singapore, and the UK willing to recognize that cryptotokens are (possibly) eligible to be considered as property.

However, the courts starkly diverge on the type of remedy that is applicable when rights in cryptocurrencies are infringed. Specifically, some judicial decisions impose an in rem property rule that enables to recover misappropriated tokens (eg due to theft or scams) from the hands of any person, including third parties. Other decisions adopt an in personam liability rule that grants victims only with monetary damages. However, these various decisions make no distinction regarding the token’s type, such that a one-fit-all (property or liability) rule is implicitly adopted for all tokens.

Such a one-fit-all rule, however, seems inefficient from an economic perspective. In particular, economic policy typically seeks to maximize allocative efficiency, where goods are allocated to whomever values them the most. The economic analysis of law (aka ‘law and economics’) hence offers a simple criterion to decide whether property or liability rules should be adopted: Property rules should be adopted when parties can bargain at zero-cost. Otherwise, liability rules should be adopted. The rationale is that if parties do not incur any transaction costs, they will voluntarily agree to assign the goods efficiently, irrespective of who has an official ‘property right’, as this creates a surplus that can be split by the parties. Respectively, if transaction costs exceed the potential surplus, the parties might be unable to reach an agreement, so that a property rule may be inferior (for details, see Guido Calabresi & Douglas A Melamed, 'Property rules, liability rules, and inalienability: One view of the cathedral' (1972) 85(6) Harv L Rev 1089, which builds on the canonical ‘Coase Theorem’).

Applying this insight to cryptocurrencies reveals that different tokens entail very different transaction costs. For instance, transaction costs likely depend on factors such as whether tokens can be traded online via a crypto exchange (so that they are liquid); whether the parties require the assistance of a legal or technological expert to design their agreement; and whether parties must invest resources to determine the token’s value. Thus, there are good economic arguments for imposing property rules for some tokens, and liability rules for others.

While every token is different, a case-by-case distinction might be too costly to implement. However, a proxy in the form of token groups can be used – where the commonly used taxonomy, which classifies tokens into ‘currency’, ‘security’ and ‘utility’, seems helpful.  In a nutshell:

  • Utility tokens provide a specific utility, such as granting access to a product or a service.
  • Security tokens provide financial benefits, such as revenue-sharing or dividend.
  • Currency tokens are used as a substitute for traditional currencies, ie as means of payment.

Albeit it may be challenging to classify some tokens, the taxonomy can assist in identifying the relevant transaction costs. For instance, in negotiations surrounding security tokens, the parties must evaluate the underlying project from which revenues are expected – making the trade complex and costly. Thus, property rules are plausibly better for security tokens. Conversely, for currency tokens, parties need not spend time overthinking the token’s value, as it is merely a means of payment. Thus, liability rules might be better for currency tokens. For utility tokens, a more intricate evaluation may be inevitable.

Designing an optimal regime requires careful thought about the aforementioned economic functions, as well as the incentives of the parties prior to the trade (eg whether efficient precautions are taken to prevent thefts) and other considerations (distributional effects or fairness). However, acknowledging the problematic nature of a one-fit-all rule is a first step towards improvement.

The post is based on my recent article, ‘Your Bitcoin Is Mine: What Does Law and Economics Have to Say about Property Rights in Cryptocurrencies?’, available here.

Roee Sarel is a post-doctoral research associate at the Institute of Law & Economics at the University of Hamburg.