This piece begins as a case note on Re Lehman Brothers, one of the most significant judgments on insolvency and proprietary interests of the last decade. It then considers how the legal principles outlined by the minority judges would apply to the facts of the infamous Mt Gox cryptocurrency exchange insolvency proceedings, had the case taken place in England rather than Japan. In doing so, the piece intends to reinforce the central importance of the English private law of trusts to commercial activity, whether that be highly traditional investment banking or extremely radical crypto-trading.
Re Lehman Brothers International (Europe)  UKSC 6
Grappling with this complicated case is an important part of understanding the law of equity, and in particular of the interplay between trusts and proprietary claims. As in most insolvency disputes, the context was that there was not enough money in the client account to satisfy all the clients’ claims in full. Any claimant who could not establish a proprietary interest in their money would be left to join the general creditors and was unlikely to recoup any of their money. As such, a result which was to the benefit of one group of claimants would be to the detriment of another.
The case concerned the insolvency of Lehman Brothers investment bank on 12 September 2008 and specifically whether clients of Lehman Brother International (Europe) (‘LBIE’) had a proprietary interest in the funds they deposited with the bank pending investment. These funds amounted to several billion US dollars, but the banks internal failings in its handling of the money were significant. The claimants in this case fell into one of three categories:
1 – Those whose money had been deposited into a segregated client account. Their money was held under a statutory trust giving them a proprietary claim against their funds in the event of a failure of the bank. They would rank in priority to the claims of the many general creditors of LBIE.
2 – Those whose money was held in LBIE’s ‘house accounts’ at the time of the insolvency. The money was due to be moved into the client accounts at the beginning of the next working day, and this routine was permitted under the relevant UK legislation (CASS 7). This was described as the ‘alternative approach’.
3 – Those clients (mainly LBIE affiliates) whose money was mistakenly treated by LBIE not as client money but as ‘house money’.
The law at issue was chapter 7 of the Client Assets Sourcebook issued by the Financial Services Authority (CASS 7), the UK’s implementation of the EU Markets in Financial Instruments Directive 2004 (‘MiFID’). This statutory scheme intended that client money paid to an investment bank would be held in bank accounts which were segregated from those which the bank used for conducting its own business (either business operations or ‘proprietary trading’). The money was to be held on trust and, in the event of the failure of the bank, all the money was to be pooled and distributed to the clients who were entitled under the trust in proportion to their interest. The aim of the scheme was that those static client funds would always be safely guarded for the client, who would be repaid in full if the bank became insolvent.
Briggs J summed up the present predicament:
In the imperfect and hugely complex real world occupied by LBIE and its numerous clients, there has on the facts which I am invited to assume for present purposes been a falling short in the achievement of both of those objectives on a truly spectacular scale.
The CASS 7 rules were found by all the judges to be inadequately drafted to provide clear guidance in the event of banking mismanagement such as that which had occurred in LBIE. Particular difficulty was posed by the ‘alternative approach’ which LBIE, lawfully, had adopted to the receipt of client funds. Furthermore, the rules did not provide any clarity as to the status of client money if the bank became insolvent after receipt but before crediting the client account.
The questions for the Supreme Court had been whittled down from the initial 26 dealt with by the High Court to three:
- when did the trust, under which the client money was held, arise?
- do the pooling arrangements, triggered by a bank failure, include only the money in the client accounts, or money received from clients but still held in the house accounts?
- is participation in the distribution of the pool money dependant upon actual segregation, or are clients whose money was yet to be pooled entitled to a share of the pooled amount?
The Supreme Court was unanimous in its answer to the first question, but sharply divided in responding to the second and third. To assist with navigating the lengthy judgment:
|Lord Hope||Lord Walker *||Lord Clarke||Lord Dyson||Lord Collins|
* Lord Walker’s judgment would be described as the leading judgment, were he not in the minority. It quotes the facts from the High Court judgment of Briggs J and is the reference point from which the arguments of the other Lords develop their arguments.
The first question
The trust was described as a ‘statutory trust’, both because it was resulted from public law rather than an agreement between the parties and because it was a trust for the purposes set out in CASS 7.1-7.9. In summary, these CASS purposes were the protection of the beneficial interest of the client and, in the event of a bank failure, the distribution of the money in proportion to each client’s interest.
In short, the question was whether the trust arose as soon as LBIE received the money, or only once it had been segregated in the client account. The respondents, major clients and affiliates whose money had been segregated and so who stood to lose out if unsegregated clients were also able to participate in the ‘pool’, argued the latter, stating that the crux of the ‘alternative approach’ was that LBIE could use the money received for its own purposes and was obliged only to top up the client account with its own money at the start of the next day. Money which LBIE freely used could not, in practice, be said to be held on trust for the client (Henry v Hammond  and Paragon Finance plc v DB Thakerar & Co  applied).
This approach was rejected by every judge involved with the case. Firstly, Lord Walker noted at para 189 that a statutory trust does not necessarily import all the duties of a private law trust. Specifically, here, money held on trust for the purposes of CASS 7.1-7.9 did not necessarily give the client a proprietary claim against that money. However, neither did it necessarily allow LBIE to use the money for its own benefit – the statutory trust required LBIE to use the money for the client’s benefit. Secondly, the applicants’ argument would result in the temporary disappearance of the client’s beneficial interest in the money, followed by its unexplained re-emergence once the money had been transferred to the client account. There was nothing in the wording of CASS 7 or MiFID on which it was based to suggest such a conclusion. Furthermore, Lord Clarke cited Re Kayford  as an example of when a trust over money can exist without an obligation to hold it in a segregated account, but also without causing conception consternation.
For analytical clarity, the Supreme Court chose to address the third question before the second, and so the author follows likewise:
The third question
The third question essentially asked which clients were entitled to a share of the money in the ‘pool’. It would decide whether the ‘pool’ would be made up only of the money segregated in the client account or that which was traceable by clients from any account, meaning the third question largely answered the second question, which focused on whether only segregated clients or all clients could participate in the distribution. The arguments considered by the judges were both textual and legal in nature, but for the purpose of understanding the case’s significance to the law of equity the latter will be highlighted over the former.
The majority of the Court began with the premise that the statutory trust took its inspiration from MiFID and so adopted a purposive approach to the interpretation of the pooling requirement. They found that the purpose was to provide a high level of protection to all clients and in respect of client money held in each money account of the firm. As such, it would be arbitrary and unfair to distinguish clients with a proprietary claim from those without based on whether LBIE had successfully met its regulatory obligations and properly segregated the money.
It was common ground amongst the majority judges that money received after the final update of the client accounts (on the morning of Friday 12 September 2008) but before the bank failure (the morning of Monday 15 September 2008) automatically fell into the ‘pool’ under CASS 7. The clients who provided that money were therefore entitled to participate in the distribution of the ‘pool’. It seemed illogical to suppose that clients who sent money just moments before would not be treated equally. The majority favoured this ‘claims basis’, on which those clients who had a claim to client money, held under a contractual obligation to segregate, were entitled to participate, regardless of whether the money actually had been segregated.
The minority judgments of Lords Walker and Hope are, in the opinion of the author, much more interesting. In contrast to the majority, the minority began from the premise that the statutory trust invoked, first and foremost, the English law of trusts. They also recognised the potential for bad law, and were loath to forgive the drafting of CASS 7 based upon the Directive’s recitals, which are not directly effective, or considerations of fairness which, given the mammoth complexity of the facts, could not accurately be judged.
Lord Hope summarised the relevant English law as follows:
Under English law the mere segregation of money into separate bank accounts is not sufficient to establish a proprietary interest in those funds in anyone other than the account holder. A declaration of trust over the balances standing to the credit of the segregated accounts is needed to protect those funds in the event of the firm’s insolvency. Segregation on its own is not enough to provide that protection. Nor is a declaration of trust, in a case where the client’s money has been so mixed in with the firm’s money that it cannot be traced. So segregation is a necessary part of the system. When both elements are present they work together to give the complete protection against the risk of the firm’s insolvency that the client requires.
In response to Lord Dyson’s supposition that the Financial Services Authority would not choose to treat some clients differently to others without expressly stating their intention, Lord Hope argued that the FSA would not have chosen to abandon one limb of the English law of trusts without giving express guidance. He held it to be fundamental to the fiduciary nature of a trust that the beneficiary’s interest was over distinct property which could be neither inflated nor deflated by the interests of other claimants. This is described as the ‘contributions basis’ for distribution. Lord Hope observed that those clients whose assets were unsegregated still had a proprietary claim against the money which was traceable into the house accounts (as much of it was). Indeed, the author notes that most of the law of following and tracing would not exist if unsegregated assets were subject to a proprietary claim purely on the basis of a declaration trust.
Lord Walker also dissented, preferring the contributions basis like Lord Hope. He noted the submission of Mr Zacaroli, the representative of LBIE’s fully-segregated clients, that the effect of the claims basis was that the client account had become a form of compensation fund for those clients who could not benefit from segregation. Lord Walker argued that the ‘pooling’ term which resulted from CASS 7 was an agreed contractual limitation on the proprietary claim of each segregated client against their money. In contrast, he did not accept that the segregated clients’ trust relationship with the bank could be altered as a result of the misfortune of the unsegregated clients. To do so, Lord Walker argued, was to subject all clients to the same disadvantage rather than allowing the proper operation of the law of equity to fully protect some clients and leave others which the task of tracing their assets. He concluded:
The majority’s decision makes investment banking more of a lottery than even its fiercest critics have supposed.
The second question
The majority and minority answers to the second question flowed quite naturally from their respective answers to the third. Lords Dyson, Clarke and Collins held that all client money, whether in the client or house accounts, formed part of the ‘pool’ which was to be distributed. To argue otherwise would have been illogical, given that some client money in house accounts was traceable, and against the grain of the policy considerations which had informed the majority’s answer to the second question. The minority favoured the view that the segregated clients could claim against their own segregated property and nothing else. Traceable client money in house accounts was the property of those unsegregated clients, and so not included in the ‘pool’.
Conclusion to the case note
The court in Re Lehman Brothers International (Europe) was essentially faced with a number of contests. One between fairness to the unfortunate clients whose money had been improperly treated by LBIE and fairness to the other clients, some of whom had taken extra precautions to ensure that their money was correctly treated but now found it significantly diluted. The other contest was between the influence of MiFID and the influence of the English law of trusts. Whilst a court surely cannot assume that CASS 7 had been badly drafted and that clients must as a nature consequence suffer, it must be asked why the word ‘trust’ appeared in CASS 7 if not to import the relevant private law.
Perhaps the largest question that the majority judgment raises is why segregation was ever a part of the legislative scheme. It would appear to be entirely superfluous if all traceable client money was to be treated equally regardless of which account is was held in. Fortunately the majority judges chose to distinguish the statutory trust to which their judgment applies from trusts generally, but it is easy to imagine that the terms of any private law trust could mimic the legislative scheme at which the majority arrived. For the moment, however, the general rules applying to trusts other than statutory trusts are clear: a declaration of trust and the full segregation of assets is required in order for a proprietary claim to arise in favour of clients who deposit money with a financial institution.
Mt Gox Insolvency 2014
The next part of this article considers how the general principles of equity and proprietary claims, as set out by Lords Hope and Walker in Re Lehman Brothers, might be applied to the infamous insolvency of the Mt Gox cryptocurrency exchange. The exchange was incorporated in Japan and its insolvency proceedings in Tokyo naturally fall within the jurisdiction of Japanese civil law. But whilst the court answered certain questions very differently to how an English court might, the arguments made by the claimants provide a useful context within which to consider the importance of proprietary claims over client assets.
Mt Gox was the world’s largest cryptocurrency exchange, and provided two kinds of service. Firstly, it offered a ‘wallet’ service, whereby users could open an account into which they could safely deposit their cryptocurrency, specifically bitcoin. The wallet was accessible from a platform provided by Mt Gox, which many users preferred to the blockchain software and single private key (a form of password) which otherwise was their sole means of accessing their bitcoin. The second service was an exchange. Whilst regulators today warn against the use of the term ‘exchange’, Mt Gox’s business model was in buying and selling different kinds of cryptocurrency just as a conventional exchange does with fiat currencies.
The provision of ‘wallet’ services suggests that each individual’s cryptocurrency was segregated into different accounts, as would occur in a financial institution. In reality, the ‘wallet’ function existed only within the Mt Gox platform. Behind this IT interface, all the clients’ bitcoin were held together in a single address registered on the Bitcoin blockchain. In addition to the client address (the cryptocurrency equivalent to the client accounts in LBIE), Mt Gox held its own bitcoin in operating addresses from which it would trade on a proprietary basis.
The purpose of the segregation of client assets from operating assets was that the client assets could always be returned immediately if the exchange was to fail. As in LBIE, however, financial mismanagement was extreme, and allegations of fraud by the directors continue to be investigated. It is unnecessary to rehearse the full extent of the organisational dysfunction, but suffice to say that the hack which eventually forced the exchange into insolvency took place over several months and siphoned off approximately $460m in bitcoin before the director and majority shareholder of the exchange noticed. The exchange was able to recoup 23% of the stolen cryptocurrency.
The Law in Japan
Cryptocurrency is a phenomenon which legal systems have either ignored or struggled with. It remains unclear whether common cryptocurrencies such as bitcoin and ethereum constitute commodities, securities, currencies or other kinds of property. Although Japan is regarded as years ahead of most jurisdictions in the development of its legal approach to cryptocurrencies, it remained unsurprising when the judge in the Tokyo District Court held that the stolen cryptocurrency was not ‘property’ at all. As such, the clients of Mt Gox had only a contractual claim in debt against the exchange.
The quantum of the claim was calculated according to the value of each bitcoin lost at the date of insolvency – approximately £370. As the proceedings progressed, however, the trading value of bitcoin rose exponentially until each was worth almost $20,000. The 23% recouped was sufficient not only to repay all the clients but also to pay a £1.5bn closing dividend to the director/shareholders whose incompetence and potential fraudulent activity had contributed to the collapse of the exchange.
The Common Law
In the UK, it is much more likely that cryptocurrencies constitute property. The High Court in Armstrong DLW GMBH v Winnington Networks Ltd  characterised stolen ‘carbon credits’ as ‘intangible property’ because of their tradeable value under the EU emissions trading scheme. Although yet to be tested before an English court, the analogy is attractive. As such, a UK court faced with the facts of the Mt Gox insolvency may well address the subsequent issue of which a proprietary claim existed against the bitcoin segregated into the client account.
As Lord Hope noted in the Lehman Brothers case, segregation is not enough by itself to establish a proprietary claim, at least under the general law of trusts. The existence of a trust is also required. Trusts can be created without the use of the word ‘trust’, but a clear intention to create the trust must have manifested (Re Kayford  per Megarry J). In the context of financial institutions, a trust will arise where assets are held in a segregated account by a custodian. This most often occurs in securities deposits. Custody must be distinguished from banking, in which any deposit merely alters the debt relationship between the parties. It is likely, however, that any cryptocurrency exchange which describes itself as an ‘exchange’ will be found to have held itself as a custodian and so implicitly declared its intention to hold any assets its received and segregates on trust.
There remains, however, a rather gapping question as to whether the clients of Mt Gox actually owned the bitcoin which Mt Gox held, as a matter of English law. As noted above, the ‘wallets’ which each client was given on the Mt Gox platform did not actually contain any bitcoin. Rather, they enabled the client to issue commands to Mt Gox as to whether to buy or sell bitcoin which it held on the client’s behalf. In more mechanical terms, the client could increase or decrease its interest in a fund of bitcoin which Mt Gox had acquired and to which Mt Gox held legal title. Whether Mt Gox could be described as a trustee of client bitcoin when it could not in fact deliver the bitcoin to the client, but only its monetary equivalent following a contract of exchange, is arguable. As such, a proprietary interest in cryptocurrency held in an exchange may prove elusive even under English law.
It should be noted that, following the discovery of the enormous difference between the value of a contractual claim and a proprietary claim against Mt Gox, insolvency proceedings were stayed following a petition for ‘civil rehabilitation’. Civil rehabilitation will result in Mt Gox temporarily becoming solvent for the sole purpose of repaying its creditors in bitcoin, pari passu. Each creditor will recoup several times the value of their contractual claim, crucially, without needing to establish any proprietary interest in their bitcoin.
Re Lehman Brothers demonstrated the important role which the law of trusts plays in financial services, and reinforces the criteria which must be established before a proprietary claim can be made under English private law. Both a declaration of trust and segregation are necessary to protect investors fully, but contracting parties must also engage with the possible layers of financial regulation which might sit on top of private law contracts. A comparison with the insolvency of the Mt Gox exchange demonstrates the enormous practical significance of such claims in certain circumstances, and gives an insight into the new legal challenges which private lawyers around the world now face.