I will be presenting my thoughts on the subject above in the upcoming Congress of the UNCITRAL for the celebration of its 50th anniversary (4-6 July 2017).

My paper currently on the Congress website is a version which I sent to the UNCITRAL Secretariat some months ago and which no longer represents my latest thinking in some significant respects. I am asking the Secretariat to replace it with the latest version, to which I make a link from here. 

The paper gives a particular emphasis on the topic of proprietary restitution of blockchain-based tokens as an area which calls for a globally unified solution.

Postscript: As from 16 June, the Congress website carries the latest version. Many thanks to the Secretariat for swiftly acting on my request. 

The above phrase is the title of the note at http://sclbc.zehuti.co.uk/site.aspx?i=ed47875. I guess what the authors meant was who owns "bitcoins" while they also rightly point out relevance to other blockchain-based assets. It is nice to see this and other writings appearing in the course of last year which deal with the issues I have been discussing in this blog since 2015. 

The note acknowledges that questions of choice of law could arise (See my previous post), it only examines English law. It does not diminish the value of the note in view of the preeminence of English law in the practice of international commercial transactions. Any global standard which might emerge in the future might indeed mirror the framework of English law.

What should attract the attention of civil law lawyers is the meaning of "ownership." The note explains that English law recognises that there can be multiple individuals each entitled to assert 'ownership' rights to the same property. The note further casts doubt on the availability of a proprietary restitutionary claim, a cause of action which seems to come closest to rei vindicatio in the civil law legal systems.

Like my previous post, the note draws an inspiration from the legal treatment of emissions credits. It observes similarity with Bitcoin but is careful to point out an important difference: a Bitcoin is mere information and does not come with a bundle of rights conferred by statute.

The note finds that an unjust enrichment claim is the most promising approach in English law to the recovery of a misappropriated bitcoin. Among the "property-like" remedies, however, the note observes that a claim for knowing receipt founded by a constructive trust is the most promising avenue. As my previous post has pointed out (at page 4 of the slide), this remedy constitutes an alternative avenue for recovery in some legal systems; it can indeed be the most potent one as in English law.

Attached here is the powerpoint file I used in my presentation at the Cyberspace 2016 conference on 26 Nov. 2016.

I examined the Tokyo District Court's decision in a suit filed after the opening of Mt.Gox's bankruptcy proceedings. The suit was filed outside the bankruptcy proceedings to obtain a full recovery of the bitcoins of which the claimant had a contractual right to return from the bankrupt exchange. To do that, he sought rei vindicatio of the bitcoins over which he asserted ownership rather than making a contractual claim to join other creditors in the bankruptcy proceedings.

I considered three questions noted at p. 4 of the slides. On the first question, the Court relied on the Japanese law concept of "shoyuken" to deny that bitcoins could be an object of ownership. But if we leave aside the technicality of the Japanese law concept and understand the concept of ownership more broadly to mean a right to monopolize the exploitation of objects, it may be possible to say that bitcoins are fit to be an object of ownership. This would open the possibility of rei vindicatio, if other prerequisites are satisfied. The other two questions I considered relate to those other prerequisites. In most of the cases, a customer of a bitcoin exchange would have difficulties meeting those other prerequisites but the possibility of a successful claim is not foreclosed in all cases.

As I mentioned in my earlier post, the Tokyo District Court in its judgment on 5 August 2015 denied the ownership (more precisely, "shoyûken" in Japanese) of bitcoins. In that post, I have also noted that translating "shoyûken" into the English word "ownership" may be misleading.

In this post, I will look at the choice-of-law aspects. The judgment is based on the assumption that Japanese law was the applicable law without giving any reasons. Nor did either party discuss choice-of-law issues (according to the Court's summary of their arguments).

The scope of the bankruptcy estate of MTGOX is surely a matter for Japanese law since the bankruptcy proceedings opened in Japan. The effect of Japanese bankruptcy proceedings extends to all assets of the bankrupt wherever in the world they are situated. This follows from the repeal in 2000 of the then Article 3(1) of the Bankruptcy Act which provided:

The bankruptcy which is declared in Japan shall only have effect on the asset of the bankrupt situated in Japan.

The bankruptcy estate of MTGOX, therefore, covers all bitcoin units wherever situated. Here, the borderless nature of the blockchain poses no problem.

On the other hand, it is not as obvious that Japanese law was the governing law of the plaintiff's ownership-based claim to recover the bitcoin units. On the face of it, the application of Japanese law may seem uncontroversial since there were no strong foreign elements in that case: the plaintiff was an individual residing in Kyoto and the defendant was the bankruptcy representative of the bankrupt MTGOX, a Tokyo-based company, appointed in the Japanese bankruptcy proceedings. But as I noted in my earlier post, choice-of-law rules for determining the ownership of intangible property are not well-established and, depending on the connecting factor to be adopted, the borderless nature of the blockchain may make it difficult to localise bitcoin units in a particular country. If the plaintiff had been a foreigner residing abroad (which would not have been a remote possibility since a majority of the creditors of MTGOX are such persons), the parties and the court might have felt it necessary to address the choice-of-law question.

As I mentioned in my earlier post, the Tokyo District Court in its judgment on 5 August 2015 denied the ownership (more precisely, "shoyûken" in Japanese) of bitcoins. In that post, I have also noted that translating "shoyuken" into the English word "ownership" may be misleading.
Until recently, the only way to examine the actual text of the judgment was to go in person to the Court pursuant to Article 91(1) of the Code of Civil Procedure. But it is now reported on Westlaw Japan (accessible by subscribers only). Reading it through, I have picked up some noteworthy points from the court's summary of the parties' arguments and its own reasoning. 

1. The plaintiff relied on Article 62 of the Bankruptcy Act which provides: 

The commencement of bankruptcy proceedings shall not affect the right to recover, from the bankruptcy estate, property that does not belong to the bankrupt. 

2. The plaintiff claimed the return of bitcoin units as appearing in his or her account with MTGOX. The plaintiff admitted that MTGOX had merged those with units from other customers and held them in addresses which were not associated with any specific customers and for which only MTGOX kept private keys. But the plaintiff argued that those units constituted a commingled deposit (“konzô kitaku”) over which each depositor retained “shoyûken” in proportion to his or her share. 

3. The court reasoned deductively from the following provisions of the Japanese Civil Code to reach its conclusion that bitcoin, being intangible, could not be an object of shoyûken. 

Article 85 The term "thing" (“butsu”) as used in this Code shall mean tangible property. 

Article 206 The shoyûken holder of a thing (“butsu”) shall have the right to freely use, profit from and dispose of it, subject to the restrictions prescribed by law. 

My comment: This conclusion is hardly controversial, as noted in my earlier post. 

4. The court disposed of the case, holding that the plaintiff could not obtain recovery, under Article 62 of the Bankruptcy Act, of the bitcoin units based on his or her shoyûken over them. 

My comment: It is true that most claims under Article 62 in practice are for the recovery of tangible property and are based on shoyûken. It is, however, also possible to recover intangible property, such as chose in action and intellectual property right, under this provision. In such cases, shoyûken, by definition, cannot be the basis of the claim. But the claim can be based on the proof that the property belongs to the claimant. Thus, “ownership” in its usual sense of the English word, as opposed to a narrower concept of “shoyûken” under Japanese law (See my earlier post), can be the basis of a claim under Article 62. The Court did not consider the possibility of recovery of the bitcoin units under that provision based on the argument that they were owned by (namely, belonged to, rather than an object of "shoyûken" of) the claimant. Nor did the plaintiff’s submission (as summarized by the Court) contain this argument. The consideration of this argument would have made it necessary for the Court to step into a theoretically difficult territory: it would have had to determine who owned the bitcoin units which were deposited and commingled with other units.

When we consider whether it is possible in law to obtain the recovery of stolen cryptocurrency units, what matters is normative, rather than technical, traceability of the stolen units. This is so whether the claim for recovery is based on the ownership of the units or the restitution of unjust enrichment.
There may be no technical traceability where stolen units are mixed up with other units in the address they have been forwarded to unless the units had been dyed prior to being stolen. In this sense, Patrick Murck seems right to observe in his presentation that while transactions are traceable, coins are less so. On the other hand, to affirm normative traceability, it might be enough to be able to say that the attacker or the persons further down the line could be deemed to hold all or part of the stolen units or their value.
Let us suppose that Alice had 70 units in her address (i.e. an address for which she holds the private key). Bob has stolen them through a phishing attack and transferred them to his address in which they have been mixed up with the 30 units he had held there. Unless the stolen 70 units had been colored, it may be technically impossible to say which of the now 100 units Bob holds in his address is originally Alice's. It is, however, possible to say that Bob holds the stolen units.
It is true that even a normative tracing becomes harder as the stolen units are forwarded down the line. Thus, let us suppose that Bob has thereafter transferred 40 units to the address of Carol. Again, unless the stolen 70 units had been colored, it may be technically impossible to say which, or even how much, of the 40 units Carol has received is originally Alice's. But it does not foreclose the possibility of a normative assessment that Carol holds part of the stolen units or their value. The normative assessment may take into account the circumstances surrounding Carol's acquisition including how much, if at all, she knew of Bob's theft. It will also be part of the normative assessment how much of the 40 units is deemed to derive from the 70 stolen units: (i) 28 units representing the proportion of the 70 units among the 100 units Bob held in his address or (ii) 10 units on the assumption that all 30 units Bob held legitimately has been transferred to Carol, or (iii) other amounts based on other calculations.
Some legal systems might opt for a simple solution of equating normative traceability with technical traceability. But other legal systems might differentiate them. It is also conceivable that different tests for normative traceability be applied between proprietary recovery and restitutionary recovery. While a test aligned with technical traceability may be preferred for proprietary recovery in order to ensure the specific identification of stolen units, it is not inevitable. Since it is ultimately a matter for legal policy whether to grant recovery merely as restitutionary relief or as proprietary relief (See my earlier post), it is not unimaginable to allow recovery of the stolen value (as opposed to the specific stolen units) as proprietary relief. The blockchain technology being a new invention, I guess that the rules are currently uncertain under most legal systems.

It is needless to say that in the cases where it is considered normatively that the stolen units are traceable, the possibility of recovery will depend on other conditions which the applicable law puts in place to protect legitimate holders in due course.

In my earlier post, I have noted the difficulty of localising cryptocurrency for the purpose of choice of law for proprietary issues.

The difficulty of localisation in choice of law is not unique to blockchain. The high seas and outer space, too, present the same difficulty since no nation exercise sovereignty over such space. The difficulty does not solely concern proprietary issues but could also arise with respect to other issues such as tort, for which the applicable choice-of-law rules may specify the law of the place of the harmful event. I will discuss three approaches to get around this difficulty below.

One approach is to come up with an alternative connecting factor. Thus, where a ship is involved, its flag may be used as a connecting factor. In a case involving a collision of ships on the high seas, the Sendai District Court in its judgment on 19 March 2009 cumulatively applied the laws of the flag states to a tort claim for damages. It is not, however, easy to conceive of similar connecting factors for cryptocurrencies since their units are stateless by nature.

Another approach is to apply the law of the country with which the issue in question is most closely connected. This approach was taken by the Tokyo High Court in its judgment on 28 February 2013 when it determined the law applicable to a tort claim for damages caused by dangerous cargoes on board a ship while the ship was in transit on the high seas. Since a major (if not the most important) goal of choice of law rules is to ascertain the law of the country with which the issue is most closely connected, this approach pursues this goal directly without relying on other more concrete concept as a connecting factor. This approach could also be taken to determine the proprietary issues of cryptocurrency. But a drawback of this approach is the lack of certainty and predictability since all the relevant factors must be taken into account on a case-by-case basis.
A third approach is to unify the substantive rules of national legal systems. The unification of substantive rules, to the extent it is achieved, dispenses with the need for choice of law. The Cape Town Convention and its Space Protocol would offer many lessons when we consider proprietary issues of cryptocurrencies.

Arvind Narayanan writes in his blog post that since banks can reverse fraudulent transactions and law enforcement of digital financial crimes is relatively competent, the risk of getting caught combined with the diminished ability to cash in on attacks skew the economics against attackers. He contrasts this with bitcoin whose design puts the entire onus on preventive measures. This insightful observation is based on the assumption (in his words) that "[i]f an attacker breaks into a server containing private keys, he can steal the bitcoins ... irreversibly." This assumption seems widely accepted. It seems to rest largely on technical grounds but I think it also depends on legal issues which are unsettled yet. Narayanan comments on the technical aspect saying, "[w]hile there’s been talk of taint-tracking mechanisms to prevent thieves from cashing out, these haven’t materialized and there are fundamental technical and political difficulties with such proposals." But he also adds a footnote referring to a paper which argues that it currently is difficult for thieves to launder large sums of bitcoins.
I do not have expertise to comment on the technical feasibility of tracking. But the point I want to make here is that should there be cases in which it is possible to trace the movement of stolen bitcoins, it will raise a legal question whether a recovery should be granted to the original owner. This question has not been tested before courts under any legal system to my knowledge. It would not be inconceivable that thieves and third parties who have acquired stolen bitcoins in bad faith are denied legal ownership and held liable to make restitution of the stolen bitcoins or their value in fiat currencies.

With respect to most tangible items, ownership is not determined by possession. Thus, if I hold in my possession a bicycle which I have rented, it does not make me the owner of the bicycle.
With respect to traditional coins and notes, Japanese law makes an exception to this principle. According to an established line of case law, the ownership of coins and notes depends on their possession (e.g. Supreme Court decision on 24 January 1964). There is a good reason behind this treatment. Coins and notes are different from other tangible items in the sense that their financial worth is derived not so much from the material (such as metal and paper) they are made of but from the monetary value (e.g. 10,000 yen) they represent. Accordingly, the ownership of coins and notes is in essence the ownership of monetary value. An analogy could therefore be drawn from it when we consider the ownership of cryptocurrency.
What should, then, be understood as the possession of cryptocurrency? Since the holder of the private key for the address in which cryptocurrency units are held has control over them, holding the private key could be equated with the possession of the cryptocurrency units. It would then mean that the holder of the private key owns the units.

This approach would commend itself for simplicity since in a majority of cases, no further question would arise. But some people may not like this solution since it would give wallet providers the ownership of units that they are entrusted with. Furthermore, the outcome is not clear in any of the following events:

1. The holder of a private key has disclosed it to others. The others would have the same  control over the units. No analogy could be drawn with traditional coins and notes which are physically possessed.

2. A private key has been stolen.

3. The holder of a private key has transferred it to others without effecting transfer of the corresponding cryptocurrency units on the blockchain.

4. Another person has generated an address with the same private key. This is likely to happen where a simple phrase is chosen to generate an address by using a brain wallet.

Other complications would arise if a multi-signature account is used or if a trust is set up over cryptocurrency units.

It seems to me that there is no simple test which could furnish answers to all ownership questions.

In my earlier post, I have noted why legal ownership of cryptocurrency matters. The use of cryptocurrency as a collateral is also imaginable. 

The prerequisites for acquiring such proprietary rights and their effect are to be determined by legal rules. In the absence of internationally uniform rules, an applicable national legal system must be determined by choice-of-law rules of the country in which such issues are litigated or considered.

With respect to a tangible item, proprietary issues are subject to the law of the country where it is situated (lex situs or lex loci rei sitae) under the prevailing choice of law rules. With respect to intangible goods, choice-of-law rules are not as well established. In one of my articles, I have argued that the proprietary issues of an emissions quota, a species of intangible goods, should be subject to the law of the country where it is registered. 

Cryptocurrency is intangible. It is also a financially valuable data (information) as is an emissions quota. However, unlike the latter which is registrable on a national registry, cryptocurrency is registered on a blockchain which is borderless. This makes it difficult to localise cryptocurrency in a particular country for choice-of-law purposes.

What follows is a description of three of the circumstances where the legal ownership of cryptocurrency matters.

1. The person with whom cryptocurrency is deposited (such as the provider of an exchange or an online wallet) has gone bankrupt. 

The depositor would certainly have a contractual claim for the return of the deposit. But the deposit would be converted into the fiat currency which is legal tender of the country where the bankruptcy proceedings are opened. Furthermore, the depositor would have to join other creditors and could obtain only a proportional recovery.

If the depositor could alternatively claim the ownership of the units of cryptocurrency which have been deposited, he would be able to obtain a full recovery outside the bankruptcy proceedings. The Tokyo District Court case, discussed in my earlier post, arose in this context but was decided on a narrower ground that only tangible items could be an object of "shoyûken" in Japanese law. 

2. Cryptocurrency units have been stolen and then transferred to third parties. 

The original owner could certainly claim damages in tort from the thief but his effort of recovery will often end in vain. Even if a restitutionary claim is available against the third party who obtained the units in bad faith, it would not defeat the seizure of the units by a creditor of the third party nor would it lead to a full recovery in the case of bankruptcy of the third party. If, however, the original owner retains the ownership of the units, he could defeat the seizure of the units and could also obtain a full recovery in the case of the third party's bankruptcy.
3. Cryptocurrency units have been sent to a wrong address or in a wrong quantity due to an error. In the case of bankruptcy of the recipient, a full recovery could not be obtained by a restitutionary claim but could be obtained if the original owner retains the ownership of the wrongly sent units.

　田髙寛貴「金銭所有権と価値の追及」(法学教室June 2015 No.417、22頁)は、仮想通貨の物権的側面の検討に大変示唆的である。

　この論文は、現金通貨に物権的保護を認める議論を展開した上で、預金債権は、預金保護措置などにより、弁済確実性が高いことから、「現金をもっている」のと同視でき、他の一般の債権と区別されうるとする。ただ、郵便切手や有価証券，さらには商品券やプリペイドカード，電子マネー等々，多様な金銭価値の存在形態のうち，どこまでを物権的帰属保護の対象に含めてよいかは問題であるとし、「ビットコインは、通貨としての価値を担保する機関が存在しておらず、このようなものにまで物権的な帰属保護が妥当するとはいえない」と述べる。

The Tokyo District Court ruling on 5 August 2015 is reported as having denied the "ownership" ("shoyûken" in Japanese) of bitcoins on the ground that shoyûken is a concept limited to tangible things. This ruling cannot be wrong so far as the Japanese law concept of shoyûken is concerned since no one can deny that bitcoins are intangible. It is, however, misleading to translate "shoyûken" into "ownership" since the English word "ownership" is often used in a broader sense covering also intangible things. Thus, it is not uncommon to speak of the ownership of patent or copyright. On the other hand, "shoyûken" of patent or copyright is never heard of. The ruling, when reported outside Japan in the language of "ownership", seems to have caused a lot of alarm but the appreciation of difference between the Japanese law concept of "shoyûken" and the English word "ownership" should allay concerns.