Good faith acquisition of financial assets

Presenting my research project

Recently, I presented my research project and its preliminary results to the committee that supervises my PhD research at KU Leuven. The title of the research project is ‘Bezit, geld, titel: verkrijging te goeder trouw van financiële activa’, which can be translated (without the wordplay, sadly) as ‘good faith acquisition of financial assets’.

Because of the corona virus, I recorded the presentation and circulated the video to the members of the supervisory committee. Afterwards, a session was held via videoconferencing to discuss the presentation with the supervisory committee, three respondents and an audience of interested persons. It was an unusual format, but it worked well. You can find the video below. The presentation is in Dutch.

The video also appeared on Corporate Finance Lab.


Bitcoin Case Law

Recent judgements from France and England

In the video below I present two recent judgements regarding the proprietary status of bitcoins, one from France and one from England. 

Click here if you have trouble viewing the video.

This presentation is given in the context of the seminar ‘FinTech, Governance and Sustainability: Legal Obstacles and Regulatory Challenges‘, with discussion sessions scheduled on Wednesday 8 April 2020.

The full text of the judgements discussed in the presentation can be downloaded here:

The LawTech Delivery Panel’s ‘Legal statement on cryptoassets and smart contracts’, referred to in the presentation, can be downloaded here

Trigonometry For Lawyers

Trigonometry can be fun if you get rid of all the mathematics.

Three’s a party

When registers are used for registering real rights in assets, we see an interplay between three notions: ‘register’, ‘law’ and ‘reality’. These notions form the three tips of a triangle (see figure 1 below). This simple triangle may incite a better understanding of the phenomenon of registering the ownership of assets in a public register.

In what follows, I will elaborate on the interdependence of ‘register’, ‘law’ and ‘reality’ by using the example of a land registry system. After that, I will discuss how Bitcoin seemingly seems to work outside any legal framework. I conclude this post by zooming in on the interaction between physical reality and legal reality.

Land registry

Simply put, a land registry is designed to facilitate conveyance of real estate and to enhance certainty as to who the owner of a plot of land is.

One can assume that, in order to bootstrap a land registry, a very thorough state of affairs must be drawn up of who owns which tract of land. The question of ‘ownership’ is a question of ‘legal reality’, which is created and perpetuated by the law. (Ownership is, of course, not a matter of physical reality. One can live in a house in physical reality, i.e., maintain a physical presence there. The question of whether this habitation occurs within the framework of, e.g., a leasehold or a freehold is, however, solely a matter of the law.) Thus, at the outset, the law determines legal reality, which in turn determines the contents of the register.

Nonetheless, physical reality also has its role to play when bootstrapping a land registry system, especially in systems where the physical characteristics of a tract of land and the ownership of this tract of land are included in the same register. The physical reality of hedges, fences and other property demarcations (“This …“) can inform the legal reality (“… is my land“).[1]

Conversely, once the register is up and running, the register can have a certain influence on legal reality. The extent of this influence depends on the type of system that is used. There are many different types of land registry systems. Three of them can be briefly mentioned here to illustrate this point:[2]

  1. Torrens system — Land registry systems that provide for a registration of the title itself have the strongest influence over legal reality. These systems are also called ‘Torrens’ systems (after sir Torrens, who introduced it in Southern Australia in 1857). In many ways, the register is the legal reality, because the registration implies that whoever is registered as the owner, is legally recognised as such.
  2. Fides publica system — One step lower on the influence ladder, we find the systems which Professor Storme refers to as ‘fides publica systems’. In fides publica systems as well as in evidence systems, discussed below, it is not the title itself which is registered, but the legal facts which form the basis for the title, such as (parts of) the deed of sale. In essence, a fides publica system provides for a protection of good faith acquirers of real property against any defects in the powers of disposal of the transferor. Good faith acquirers can assume that the register completely and accurately reflects the real rights relating to the property they wish to acquire.

For example: Suppose Alice is registered as the owner of a plot of land in the land registry. Alice then transfers the plot to Bob, but this transfer is not recorded in the register. Alice subsequently transfers the plot to Carol, who is unaware of the prior agreement between Alice and Bob. The transfer from Alice to Carol is recorded in the register. The outcome of this situation will in principle be that Carol will have lawfully acquired the plot. Bob will not be able to recover the plot from Carol. He can solely claim compensation from Alice.

  1. Evidence system — Certain land registry systems may solely have evidentiary value that the legal act that is registered in the registry took place on a certain date. This sort of register has a meagre influence on legal reality, given that title can transfer independently from the land registry.

For example: Suppose Alice is registered as the owner of a plot of land in the land registry. Alice then transfers the plot to Bob, but this transfer is not recorded in the register. Alice subsequently transfers the plot to Carol, who is unaware of the prior agreement between Alice and Bob. The transfer from Alice to Carol is recorded in the register. The outcome of this situation will in principle be that Bob will be the lawful owner. Carol cannot recover the plot from Bob in spite of the fact that she is mentioned as the owner in the register. It suffices for Bob to rebut the register by proving that he had a prior agreement with Alice. The register solely has evidentiary value, it does not confer title nor make the transfer effective against third parties.

In other words, a public register structures legal reality. In its strongest form (e.g., a Torrens land registry system), it determines who owns what. In weaker forms, it determines which party is more likely to prevail in the event of a conflict over title.

However, the register cannot play its part without the legitimisation of the law. Parties can register all they want in whatever ledger they see fit; as long as the law does not attach legal force to the contents of the register, they nor third parties are bound by it.

Bitcoin and the law

If that is true, you might ask, how come Bitcoin seems to work just fine without the law? Bitcoin seems to have a coordinating effect on participants in the system without legal support. How can this be reconciled with the statement that the register cannot play its part without the legitimisation of the law?

Bitcoin has been analysed as a multi-party agreement (see here).[3] By participating in the Bitcoin system, participants enter into an implicit agreement with all other participants that wealth will be expressed in bitcoin. Moreover, each participant expects the other participants to accept bitcoin as a means to fulfil any payment obligations the former may have towards the latter.[4] From this point of view, Bitcoin implicitly relies on the law for its functioning. More specifically, it relies on contract law. A contract is only legally enforceable if the law acknowledges its enforceability. Moreover, the Code Napoléon reminds us that a lawful contract has force of law between the parties.[5]

However, the relationship between Bitcoin and contract law is distorted by the intervention of code. The Bitcoin software is designed to automatically and irrevocably enforce the multilateral contract between Bitcoin users. Suppose Alice, a Bitcoin user, was fraudulently induced into making a transfer of Bitcoin to Bob. From a legal point of view, this would render Alice’s consent defective, meaning that a judge can render the transaction null and void. It will, however, be difficult to enforce such a judgement if, due to the pseudonymous nature of the system, Bob cannot be identified.

The characteristics of the Bitcoin software thus offer a sort of ‘perfect negotiability by design’. A comparison can be made with the way bank notes are transferred. It is said that ‘money has no earmark’, meaning that bank notes cannot be recovered from a person who obtained them in good faith.[6] This protection of the acquirer against title defects is created by the law. In the case of Bitcoin, on the other hand, bitcoins do have ‘an earmark’. I am referring to the ‘audit trail’, for which blockchain technology is often praised, which results from the use of hash pointers. In spite of this audit trail, the identification of the acquirer is prevented by the pseudonymity that is inherent in the Bitcoin system. The acquirer is therefore factually protected against title defects, simply because he cannot be located. The protection of the acquirer against title defects is thus created by the software, not by the law. It must therefore be conceded that regulation in the Bitcoin system is for a large part imposed by code, not law.[7]

Regulation by code can work if you are dealing with a more or less autonomous system (such as Bitcoin). As soon as your distributed ledger needs to be embedded in the ‘meat space’, it will need support from the legal framework. Suppose, for instance, that you would like to replace your company’s paper share register by a blockchain alternative. You will first have to verify whether the laws that apply to your company recognise an electronic entry in a blockchain as the legal equivalent of an entry in a paper share register. If not, you may end up with an invalid share register, and share ownership registered in the blockchain may not be effective against third parties or the company. That is why jurisdictions such as Delaware, France and Wyoming have enacted legislation to provide an express legal basis for blockchain stock ledgers (see here and here).

Cookie cutters

I already very briefly mentioned the interplay between physical reality and legal reality when I talked about bootstrapping a land registry. In what follows, I will discuss some further examples of interfaces between physical reality and legal reality.

Let us look at another triangle: the ‘reality triangle’. This one represents the interaction of physical reality, the law and legal reality.

The desire to structure physical reality by coordinating the behaviour of physical persons leads to the emergence of laws. These laws determine the legal reality. Legal reality has a coordinating effect on the behaviour of physical persons. The (coordinated/altered) behaviour of physical persons leads to more law making, and so on. Thus, there is a constant interaction between these three elements.

In order to be a clear and transparent manager of physical reality, the law must from time to time disregard physical reality’s messy complexities. Legal certainty sometimes requires that the law simplifies physical reality in a predefined formal manner, in order to enhance the predictability of the outcome. The law thus resembles cookie cutters, and physical reality resembles cookie dough. The cookie cutters are pressed into the dough, and the excess dough is removed.

Let me give some examples to clarify this point:

  1. Regulated evidence: In civil law jurisdictions such as Belgium, the way evidence is to be furnished in private law proceedings is regulated, e., subject to rules that must be followed in order to furnish a valid proof. This is unlike criminal law proceedings, where the furnishing of evidence is free, i.e., not subject to rules. The legal reality established in private law proceedings may therefore differ from physical reality to a greater extent than the legal reality established in criminal law proceedings.[8]
  2. Irrefutable presumptions: An irrefutable presumption in civil law jurisdictions is a presumption that cannot be refuted (no s***, Sherlock). If the presumption (the legal reality) does not reflect the physical reality, a discrepancy will occur between the legal reality and the physical reality (because the presumption cannot be refuted, Sherlock). An example is black lung disease in coal workers. Belgian law provides for an irrefutable presumption that, if a coal worker suffers from black lung disease, this disease is caused by working in a coal mine. This implies that the coal worker is entitled to benefits from the occupational diseases fund. The fund cannot refute this presumption in order to avoid having to pay the benefits.
  3. Statutes of limitations: One last example is statutes of limitations. Suppose Alice commits a crime that goes unnoticed for 30 years, which is also the maximum term provided by the statute of limitations. If after 30 years someone finds out about the crime, Alice can no longer be prosecuted. While Alice committed the crime in physical reality, she will forever be innocent in legal reality due to the presumption of innocence.

Now let us revisit the notion of a public register. If we incorporate figure 3 into the reality triangle, we get a ‘reality rectangle’.

As explained under figure 4 above, the law directly determines legal reality. As set out under figure 3 above, the law also has an indirect influence on legal reality via the register. The law legitimises the register, and the register partly structures legal reality. The register can thus be regarded as an extension of the law. If the law would be Batman, the register would be Robin, the law’s trusted sidekick.

In case of a land registry, there is a clearly discernible physical reality to which the register relates (i.e., the land). Sometimes, however, the register solely reflects legal reality, not physical reality. One example are book entry securities, such as exchange-traded shares and bonds. Book entry securities, just like the company or government that issued them, are “figments of our collective imagination.”[9] This absence of an interface with physical reality makes it easier to use the register for keeping track of legal reality (so long as we can agree on the laws that are to govern legal reality and determine the legal meaning of the register).

Things may get troublesome when the register is expressly designed to ‘track’ physical reality. Examples, for which blockchain technology is sometimes presented as a solution, are supply chain tracking and the ‘tokenisation’ of physical assets. While a distributed ledger may be a good means of underpinning and enforcing agreements between participants in the system, it is in itself insufficient for proving that a shipment of crude oil left the dock at a certain time, or that a diamond that is registered in the ledger as belonging to Alice is 20 carats.

[1] Cf. Tim Hanstad, ‘Designing Land Registration Systems for Developing Countries’ [1997] Am. U. Int’l L. Rev. 647, 656.

[2] See Matthias E. Storme, ‘Het grondboek: de vereiste hervorming van de regels betreffende de verkrijging van onroerende zakelijke rechten’, s.d., (accessed 20 November 2018).

[3] Simon Geiregat, ‘Eigendom op bitcoins’ [2018] Rechtskundig Weekblad 1043, 1045-1046, paras. 11-14.

[4] Simon Geiregat, ‘Eigendom op bitcoins’ [2018] Rechtskundig Weekblad 1043, 1047, para. 18.

[5] Article 1103 of the French Civil Code: « Les contrats légalement formés tiennent lieu de loi à ceux qui les ont faits. » Article 1134, first paragraph of the Belgian Civil Code : « Les conventions légalement formées tiennent lieu de loi à ceux qui les ont faites. »

[6] Austin W. Scott, ‘The Right to Follow Money Wrongfully Mingled with Other Money’ [1913] Harvard Law Review, 125, 125 footnote 1 in fine.

[7] Cf. Lawrence Lessig, Code version 2.0 (Basic Books, 2006), 24.

[8] Henri De Page, Traité élémentaire de droit civil belge, III/2, Les obligations (Bruylant, 1967), 699-700, para. 709; Walter Van Gerven, Beginselen van Belgisch privaatrecht, I, Algemeen deel (E.Story-Scientia, 1987), 358, para. 116.

[9] Yuval Noah Harari, Sapiens. A Brief History of Humankind (Vintage, 2014), 32.

ECB Uncovers Philosopher’s Stone

Crypto-assets do not represent financial claims or proprietary rights, ECB says

Last month, the ECB Crypto-Assets Task Force released a paper entitled “Crypto-Assets: Implications for financial stability, monetary policy and payments and market infrastructures.” One of the goals of the paper is to put forward a common definition of the term ‘crypto-assets’ that can serve as a basis for the consistent analysis of this phenomenon.

A ‘crypto-asset’ is defined in the paper as denoting “any asset recorded in digital form that is not and does not represent either a financial claim on, or a financial liability of, any natural or legal person, and which does not embody a proprietary right against an entity.” 

The distinctive feature of crypto-assets,” the task force continues, “from which they derive their specific risk profile, is the lack of an underlying claim/liability. Units of a crypto-asset may be used as a means of exchange and are de-facto considered by their users as assets, in the sense of ‘something of value’, although they do not correspond to the liability of, and claim on, any party. As a consequence, crypto-assets are fundamentally different from various forms of financial claims and/or their digital representation using the technology and possibly the infrastructure that underpin crypto-assets.” 

Crypto-assets fall outside the scope of PSD2, EMD2 and MiFID II

Building on its analysis and definition of crypto-assets, the task force notes the following:

  • Crypto-assets are not electronic money within the meaning of EMD2.
  • Nor are crypto-assets scriptural money in the form of commercial bank money or central bank money.
  • Therefore, crypto-assets do not fall within the scope of PSD2.
  • Furthermore, crypto-assets in itself are not ‘financial instruments’ within the meaning of MiFID II.


Taming the Intangible: MtGox Judgment Translated into English

A Japanese judgment relating to the MtGox bankruptcy proceedings has been translated into English. The judgment addresses property rights in bitcoins. Under Japanese law, only tangible things can be the object of ownership. This bars bitcoin holders from proprietary protection. In civil law jurisdictions that do not have this corporeality requirement, part of the solution may lie in the analysis of Bitcoin as a multi-party agreement.

MtGox, once the world’s largest bitcoin exchange, filed for bankruptcy protection in Tokyo on 28 February 2014. Bankruptcy proceedings began on 16 April 2014. One unhappy MtGox user brought a lawsuit against MtGox and its bankruptcy trustee in order to retrieve more than 458 bitcoins from the exchange. On 5 August 2015, a lower court in Tokyo, Japan, passed a judgment in this case. In its judgment, the Japanese court takes a stand on the nature of Bitcoin as property under Japanese law. Academics at Oxford University’s Commercial Law Centre at Harris Manchester College have now made available an annotated unofficial English translation of the judgment.

The plaintiff relied on the right of segregation of Article 62 of the Japanese Bankruptcy Act, which, freely translated, provides as follows: “The commencement of bankruptcy proceedings shall not affect a right to segregate, from the bankruptcy estate, property that does not belong to the bankrupt […]”. The plaintiff alleged that the relevant bitcoins, of which MtGox has taken possession, do not fall within the bankruptcy estate, because they are owned by him. He therefore requested that the 458+ bitcoins be transferred to him.

Touching the Intangible

A major hurdle that had to be overcome for this claim to succeed, was to convince the judge that bitcoins can be the object of property rights under Japanese law. Unfortunately for the plaintiff, Japanese law only recognises property interests in tangible things (as per Article 85 of the Japanese Civil Code).

The plaintiff’s reasoning for arguing that a bitcoin is in fact a tangible thing, is paraphrased in the judgement as follows: “The electronic record held on a number of electronic computers embodies the bitcoin and is not merely a record of it, so that bitcoin has an existence; and therefore it is possible to subject it to exclusive control, thus it is the object of ownership, corresponding to a “thing”, that is, a tangible thing, under Article 85 of the Civil Code.” (emphasis mine)

The court is not convinced by the ’embodiment’ argument. It has a strict reading of the corporeality requirement, according to which even things that can be physically observed, such as electricity, are excluded from being a ‘tangible thing’. Referring to Bitcoin’s digital and internet-based nature, the court holds that “it is obvious that bitcoin has no corporeality which occupies space.”

The corporeality requirement in Japanese law seems to be a trace of the influence German law had on the codification and ‘westernisation’ of Japanese law in the late nineteenth century.1 The German Civil Code provides, in its § 90, that “Sachen im Sinne des Gesetzes sind nur körperliche Gegenstände.” It is also interesting to see how closely the Japanese court’s understanding of a tangible thing (“A tangible thing shall be defined as an object which occupies space such as liquid, gas, substance”) resembles German commentators’ understanding (“Körperlichkeit bedeutet das räumliche Zutagetreten von Materie (fest, flüssig, gasförmig) in beherrschungsfähiger Einheit.”).2

While bitcoins are not securities,3 similar discussions can be found in the literature on property in securities. To civil lawyers such as myself, the bone of contention summarized in the MtGox judgment may be reminiscent of the French ‘théorie de la scripturalisation’ which was developed in the 1990s in respect of book-entry securities. French legal scholars sought to extend the protection of bona fide purchasers provided for by Articles 2276-2277 of the French Civil Code to book-entry securities. The problem was that, much like in the MtGox judgment, these rules applied only to tangible things. Therefore, scholars argued that a credit entry in a securities account in fact ‘reifies’ the securities; the credit entry ‘embodies’ the security (cf. the plaintiff’s reasoning above), such that the credit entry must be considered a tangible thing.4

Taming the Intangible

The ‘scripturalisation’ theory may be the closest one will get to touching the intangible. Yet, without the support of express statutory provisions, it is unlikely to defeat rigid corporeality requirements such as the one found in the Japanese Civil Code.

In other jurisdictions, for instance in England and Wales, ownership can also be acquired in intangibles, such as choses in action. In commenting on the judgment, Professor Gullifer, Professor Hara and Professor Mooney give the example of a bank account: “The bitcoin blockchain is merely a ledger, and what is recorded are just transactions relating to bitcoin. In English law, we might think of this as analogous to the ledger of transactions involving a bank account. However (in English law though not, technically, in Japanese law) the bank account itself can be owned since it is a debt owed by the bank to the account holder, that is, a chose in action which is classified as intangible property. The ledger is just a record of transactions concerning this chose in action.” They are of the view, however, that “Bitcoin is definitely not a chose in action. Thus, the bank account analogy does not work, and a different analysis is needed for it to be the object of ownership even in English law.”

In civil law jurisdictions, where the concept of ‘choses in action’ does not exist, part of the solution for providing proprietary protection to a holder of bitcoins may be at hand. One of my colleagues at Ghent University, Simon Geiregat, analysed Bitcoin from a Belgian law perspective.5 He found that bitcoins can be regarded as personal claims (schuldvorderingen, créances, Forderungen). Bitcoin can be regarded as a multi-party agreement. Whoever has a Bitcoin wallet accedes to this multi-party agreement. The Bitcoin user thus becomes the holder of a contractual claim against anyone and everyone else who enters into the agreement. The holder of the Bitcoin wallet can rightly expect that these counterparties will accept her bitcoins if she has a payment obligation towards one of them, unless expressly agreed otherwise. At the same time, she has an identical obligation towards any and all other Bitcoin users. In other words, by acceding to the multi-party Bitcoin agreement, the Bitcoin user acquires a personal claim against (and at the same time becomes obligor of) all other Bitcoin users.6

Moreover—at least under Belgian law—personal claims can be the object of ownership, and proprietary protection of personal claims can be enforced through revendication.7 The only way in which the claim of a Bitcoin user can usefully be redeemed, consists of knowing the private key to a public key on which a so-called ‘unspent transaction output’ is registered.8 In other words: the bitcoins have to be ‘in your account’ for you to be able to spend them. Thus, a revendication would mean that the bitcoins (i.e., the ‘unspent transactions outputs’) have to be transferred to a public key designated by the plaintiff. This is the result that the plaintiff in the MtGox case sought to obtain by bringing the lawsuit. From this perspective, the intangible nature of bitcoins should not prevent the proprietary protection of the holder.