3.01  Securities and money are fungible assets, which, in the words of Goode on Commercial Law, can be described as¹

assets of which one unit, in terms of an obligation owed by one party to another, indistinguishable from any other unit, so a duty to deliver one unit is considered to be performed by the delivery of an equivalent unit … Whether assets are fungibles (p. 84) depends not on their physical characteristics but upon the nature of the obligation owed with respect to them. It matters not whether the subject of the contract is grain, flour or a motor car, or whether it is tangible or intangible. In a contract for the sale of unascertained, or generic, goods, the goods are ex hypothesi fungibles, since the duty of the seller is to sell and deliver not a specific chattel identified at the time of the contract but an article (ie any article) which answers to the contract description.

3.02  The fungible nature of securities and money has allowed the financial system to transform these assets to account-based assets, that is, assets that exist as balances administered pursuant to the terms and conditions of accounts provided by banks, investment firms, and central securities depositories (CSDs). The common terms and conditions applicable to these account balances require the account provider to make equivalent property available to the account holder or a payee or transferee, if so instructed by the account holder by way of a payment order or securities settlement instruction given in accordance with the terms and conditions of the account. Ergo, fungible assets standing to the credit of the investor as account holder per the terms and conditions of the account agreement do not convey an entitlement to an in specie asset but a claim against the bank or investment firm pursuant to the account agreement with that bank or investment firm to transfer or deliver an equivalent unit of the fungible asset standing to the credit of the investor.

3.03  The corollary of the entitlement-based nature of rights arising pursuant to account-based holding structures for financial assets and money is that most arrangements expose the investor to the risk of loss due to the bank or investment firm defaulting on its payment, transfer, or delivery obligations under the account agreement as a result of insolvency. That risk, usually referred to as ‘credit’ or ‘default risk’, can be mitigated depending on whether the account holder’s entitlements are to be characterized as mere personal rights against the account provider, or whether the account agreement and the actions taken by the account provider also pass or create proprietary interests in the assets received by it in the course of the performance of its account services. A credit to the account should normally not occur unless the corresponding asset has been received. Conversely, a debit will be the result of the account provider completing a transaction for the investor that involves the transfer of an asset of the type credited to the account. Proprietary interests survive an account provider’s insolvency and therefore permit the investor to retrieve the asset in which that proprietary interest persists. Personal rights, on the other hand, are converted into a pro rata claim in the balance of liquidation proceeds that remains after preferential creditors have been satisfied. Safeguarding the creation of proprietary interests in client assets received by the account provider for credit to investor accounts, therefore, is critical to protecting the investor as the holder of an account-based asset from account provider credit risk.

3.04  The MiFID II regulatory framework, as did MiFID I,² places the protection of investors against account provider credit risk at the centre of investment firms’ regulatory duties:

1. (a)  Article 16(8) of MiFID II provides that an investment firm that holds ‘financial instruments belonging to clients, must make adequate arrangements so as to safeguard the ownership rights of clients, especially in the event of the investment firm’s insolvency, and to prevent the use of a client’s financial instruments on own account except with the client’s express consent’.

2. (b)  Article 16(9) provides that an investment firm that holds ‘funds belonging to clients, must make adequate arrangements to safeguard the rights of clients and, except in the case of credit institutions, prevent the use of client funds for its own account’.³

Accordingly, MiFID II seeks to impose credit risk-mitigating organizational requirements on investment firms that hold financial assets or money for their clients. ‘Safeguarding’ in this context means ensuring that financial assets or funds that are received by investment firms for the account of clients are properly accounted for and administered effectively and are not available to recover debts owed to other creditors of the investment firm, whether as part of an insolvency procedure or otherwise. The only funds received by a bank are exempted from the safeguarding duty on the notion that the credit risks of banks are mitigated separately under CRD IV⁴ and CRR.⁵

3.05  The legislative recognition of the account-based nature of securities and money, and the need to pass, create, or preserve proprietary interests through account arrangements are demonstrated in the detailed operational requirements that can be found in Article 2 (Safeguarding of client financial instruments and funds) of (p. 86) MiFID II Delegated Directive (EU) 2017/593. Banks and investment firms that hold client assets are required to:

3.06  Article 2(1) of MiFID II Delegated Directive (EU) 2017/593 conceptually recognizes that securities and money are fungible in nature and are typically held on a pooled basis by the institution that maintains the accounts. The provision prescribes operational arrangements that ought to (1) enhance the preservation of account holders’ ownership rights through proper identification and appropriation of the pool of assets that belongs not to the intermediary but to its clients, and (2) reduce the risk that the pool of assets that is held for clients collectively is reduced due to operational errors or misuse. Subparagraphs (a), (d), and (e) concern the proper identification of property as belonging to clients and require the firm to maintain internal records and ensure separation of accounts at delegated firms so that the firm is able to distinguish assets held for clients from its own assets, which includes records of receipts and transfers so that the balance can always be known. Subparagraphs (b) and (c) recognize that movements in and out of a fungible pool require recordkeeping and regular reconciliation with balances administered by delegated firms to ensure accuracy of the balance administered by the investment firm. Subparagraphs (d) and (e) insist that the sub-custodian or bank separate accounts at that next level in the chain into a balance representing client assets and a balance representing own assets of the outsourcing firm. Not only does operational segregation reduce the risk that the investment firm uses client assets for its own account, it also reduces the risk that in the event of default of the investment firm its clients’ proprietary interests cannot be established (p. 87) due to lack of certainty or comingling. Accordingly, sub-paragraph (f) instructs the firm to ensure that it has adequate internal control and risk management processes in place to identify, measure, monitor, and mitigate the risk that clients’ ownership rights are not adequately established or that the pool of assets in which those ownership rights purport to exist suffer shortfalls due to operational errors, misuse, or fraud.

3.07  English law classifies rights as real (in rem), personal (ad rem), and purely personal (in personam). A real right is a right in or over an identifiable asset or fund of assets that may be asserted against all persons generally. A personal right (ad rem) is a right that may be asserted against a specific person or persons to have an asset or fund of assets delivered or otherwise transferred, but does not amount to a ius in re. A purely personal right is one that does not involve the delivery or transfer of an identified asset or fund of assets but is to be satisfied in some other way, such as the payment of debt or damages from general assets.

3.08  Real rights and personal rights, the latter also known as ‘claims’, may be subject to reduction or avoidance by the exercise of equitable rights.⁶ Goode on Commercial Law describes the nature of equitable rights succinctly and effectively:⁷

An equitable right is not an existing real right in an asset but rather a personal right in one person to set aside, reduce, or extinguish to his own advantage an asset held by another. Equities are broadly of two kinds: those which entitle a person to have revested in him an asset improperly acquired by another, and those which entitle him to reduce or extinguish a personal liability to that other. Typical examples of the first kind of equity are the right to rescind a contract for fraud, misrepresentation or undue influence; [and] to have a transfer set aside for undue influence or breach of fiduciary duty. … A prime example of an equity of the second kind is the equitable right to set off against a money claim by one’s creditor a cross-claim for money due from that creditor on another account.

3.09  On insolvency of a corporate debtor, personal rights, whether ad rem or purely personal, are converted into a right to prove the value in competition with other unsecured creditors, in principle on a pari passu basis, but preferential unsecured creditors will be satisfied, first. Rights in rem, however, including any equitable interest to which that right in rem might be subjected, are as such not affected by insolvency; legal and equitable property rights, together called ‘proprietary interests’, may be exercised despite the insolvency.⁸

(p. 88) 3.10  It is important, therefore, that account agreements and holding arrangements create and preserve proprietary interests in the client assets held by the account provider to protect the investor from default risk, in particular if caused by insolvency of the account provider. Custody services documented under English law typically stipulate that the custodian must keep the client assets ‘safe’ and ‘separate’ from the custodian’s own assets, meaning that the custodian should not appropriate the client assets. Under English law, a requirement to keep assets separate is normally an indicator that the assets are impressed with a trust.⁹ Accordingly, parallel to its regulatory obligations under Article 16(8) of MiFID II and Article 2 (Safeguarding of client financial instruments and funds) of MiFID II Delegated Directive (EU) 2017/593, as implemented in the United Kingdom through the Client Assets Sourcebook (CASS),¹⁰ the account providing custodian will be under an equitable duty to identify and safeguard the securities received by it for its clients as the property of a trust, that is, to ensure the property passes to its clients as beneficiaries. Section 238(3)(a) of the Insolvency Act 1986 provides that property held by the bankrupt, that is, a natural person, on trust for any other person does not form part of the bankrupt’s estate. The Insolvency Act 1986 does not define the estate of a company, that is, a legal person, that is being wound up, but generally refers to the ‘company’s property’.¹¹ It follows that only assets that are owned absolutely by the company, meaning assets that are free of equitable or other interest, are part of the company’s estate.¹²

3.11  Following the insolvency of Lehman Brothers, it became clear that the return of client assets presented unexpected challenges caused by the complexity of the broker-dealer operating models. This resulted in delays in the identification and return of client money and assets. Subsequently, the distribution of client assets on insolvency of an investment firm was enhanced by the Investment Bank Special Administration Regulations.¹³ ‘Investment bank’ is defined in section 232 of the Banking Act 2009 to include, in summary, investment firms or banks that are incorporated in the United Kingdom and are authorized as a broker and/or dealer and to provide custody services and hold client money. Regulation 10 (Special administration objectives) sets out the three special administration objectives (p. 89) allocated to the administrator: (1) return of client assets as soon as is reasonably practicable,¹⁴ (2) engage with market infrastructure bodies, that is, most importantly central counterparties (CCPs) to facilitate the operation of default procedures and transfer client positions and assets,¹⁵ and (3) rescue the investment bank as a going concern, or wind it up in the best interests of the creditors. Regulation 10(5) makes it clear that, for the purposes of objective (1), ‘return of client assets’ means that the investment bank transfers assets to the client to the extent of ‘the client’s beneficial entitlement to those assets (where the assets in question have been held on trust by the investment bank) … having taken into account any entitlement the investment bank might have, or a third party might have, in respect of those assets, of which the administrator is aware at the time the assets are returned to the client’. The equitable character of the investor’s interest in client assets received in trust by investment firms is discussed in sections B2 and C below.

3.12  If the cash account provider is authorized and regulated as a bank, Article 16(9) of MiFID II does not require it to ‘make adequate arrangements to safeguard the rights of clients’. Instead, banks and their ability to repay amounts owed to their clients are safeguarded under CRD IV and CRR. It is well established at English law that amounts expressed to be owed by a bank pursuant to a deposit account agreement are regarded as being lent to the bank by the customer. Ergo, a credit balance in a bank account constitutes a debt of the bank. The debt is incurred, owed, and payable in accordance with, and subject to, the terms of the account agreement between the account holder and the bank. Accordingly, if a broker or dealer or other firm that provides investment services accepts deposits as banker, there is no intention to create, and the investor does not acquire, any proprietary interests in relation to any monies received or held by that firm on account of, or as a result of, the sums paid to it by its client. The sums paid to a firm as banker by way of a deposit convey absolute, unencumbered ownership of the sums so transferred to that firm, whose obligation is to repay an equivalent sum.¹⁶ Common law operates to confer the banker’s lien and the banker’s right of (p. 90) set-off upon a banker, but not on other account providers, and upon the customer a right to confidentiality.¹⁷ Whether or not certain dealings between a provider of payment or deposit-taking services and the customer constitute banking business at common law is therefore of some considerable importance.¹⁸

3.13  This mantra holds true if the firm offers custody services. That a custodian receives qua banker ought to be assumed if the firm is authorized to operate as a bank,¹⁹ even if the amounts received were payments made by issuers in respect of securities held on trust as custodian.²⁰ To avoid any lingering argument that the cash credit balance standing to an account pursuant to a custody agreement may not represent a deposit, Yates and Montagu recommend that custody agreements provide express wording authorizing the custodian to take receipts as banker.²¹ The majority of standard custody agreements will do so, but in the absence of an express term, it ought to be an implied term, at least for cash accounts that are used for payment services. Obviously, in view of payment practice, parties to the custody agreement could not reasonably have intended that the custodian operate that cash account as custodian, rather than as banker.

3.14  It is easily overlooked that a bank is a service provider, not a firm that ‘keeps money safe’. The legal discourse relating to the relationship of a bank as service provider with its clients is often impeded by the terminology used to describe the account and the activities relating to it. Terms such as ‘account holder’, ‘money in the account’, ‘money deposited in an account’, and ‘taking’ or ‘transferring’ money ‘out of an account’ project images of identifiable sums of money that may be owned and traced into or out of an account to another account, as if the money that is being ‘moved’ by way of ‘debits and credits’ constitutes separately identifiable property, and as if an account were some sort of constructive bucket where the money can be ‘mixed’ with money that is already in it. Though the convenience of the prevailing terminology is not in dispute, the legal reality is quite different from the image projected by it. That which is referred to as ‘money in an account’ is the product of transactions between the bank and its account holder, not the product of the bank, or any other person, ‘moving money’. In fact, (p. 91) nothing is being moved into or out of the account, or otherwise. The bank and its account holder have a debtor–creditor relation and the concept of ‘account’ is what it is expressed to be: a statement of account by the bank to its customer for monies payable by or due to the account holder as a result of dealings that took place between the bank and its account holder, and between the bank and others, if the bank were acting for the account of the account holder.²² In legal terms, the interest in the balance in a cash account is a chose in action, a claim against the bank. Such dealings take place pursuant to the services that the bank has agreed to offer and of which the account holder has decided to take the benefit. The basis for these services is the account agreement between the account holder and the bank as account provider. Thus, the term ‘account’ is also used to describe the legal relationship between the bank as service provider and the account holder as customer, as governed by the account agreement.

3.15  The account agreement typically consists of express as well as implied terms and may be contained in several different documents.²³ It may cover banking services as well as custody services, or indeed any other services that the firm in question may legally provide, such as brokerage services and investment advice. In its most basic form, the bank undertakes to hold on deposit monies paid to it by the account holder, which usually entitles the account holder to receive interest paid by way of a credit to the account, and to withdraw the credit balance, in whole or in part, on demand during the bank’s normal operating hours in accordance with the withdrawal terms that apply to the account. In the absence of express withdrawal provisions, there is an implied condition that the balance is due on demand. In either case, the balance will only be due for the amount of the demand. Under the statutes of limitation time does not run in respect of the balance of a cash account until a demand for payment is made.²⁴ An investor may open a deposit account by way of investment in addition to cash accounts, which are used for payments to and from third parties.

3.16  If the bank has undertaken to offer payment services, at the request of the account holder it will arrange for payments to be made by funds transfer to recipient accounts held at that bank, or at other banks. Equally, the bank will undertake to accept instructions from other account holders or other banks, to make payments to the account holder. The execution by the bank of both outbound and inbound payment instructions will be subject to such legal and regulatory restrictions as (p. 92) may apply to the bank from time to time, and to such operational restrictions as the bank may stipulate from time to time. Typically, the custodian as banker will provide financing to the investor in the event that the investor requests that the custodian carry out a payment that is not fully covered by the balance of the cash account, or an incoming payment. In those circumstances, the custodian will allow the investor to incur an overdraft, at which point the positions are reversed: the custodian becomes the creditor for the amount of the overdraft, and the investor, as account holder, becomes the debtor. It is well established that an overdraft, manifested as a debit balance, is to be regarded as a loan from the bank to its customer.²⁵

3.17  The account agreement is best described as a framework agreement that covers the provision of one or more services by the bank, including loan services. The ‘opening of an account’ implies that the customer, as account holder, and the bank, as service provider, have agreed to enter into an agreement under which the bank has undertaken to make certain services available to the customer from time to time, and the customer has undertaken to make such payments to the bank as may become due pursuant to the provision of those services. Under the account agreement, the bank will be required to account to the customer for any amounts due between the customer and the bank in respect of any services that are provided by the bank under the agreement.

3.18  The account agreement creates not only a framework for the provision of services, but also a contractual netting framework. Amounts that become due between a bank and its customer because of payments made or received by the bank for the account of the customer are discharged by way of debits and credits to the account. Mutual payment obligations resulting from transactions between the bank and its customer, and by the bank for its customer, if they are within the scope of the account relationship, are netted continuously and instantaneously.²⁶ This type of cash account is therefore typically referred to as a ‘current account’ or ‘running account’.²⁷ The term ‘deposit account’ is usually reserved for accounts that do not permit third-party payment services and are commonly subject to withdrawal restrictions, but that does not mean that transactions between the bank and the holder of the deposit account (ie deposits and withdrawals) are not netted. The deposit account is also a current account.

(p. 93) 3.19  Netting as such is not a legal concept. It merely describes a legal result. Buckley LJ probably gave the most authoritative analysis when he said:²⁸

Where the relationship of banker and customer is a single relationship, such as I already mentioned, albeit embodied in a number of accounts, the situation is not, in my judgement a situation of lien at all. A lien postulates property of the debtor in possession or under control of the creditor. Nor is it a set-off situation, which postulates mutual but independent obligations between the two parties. It is an accounting situation, in which the existence and amount of one party’s liability to the other can only be ascertained by discovering the ultimate balance of their mutual dealings.

3.20  Accordingly, the current account agreement constitutes a framework agreement that permits various types of transactions to take place between the parties and to calculate amounts due, by one party or the other, by reference to the agreed accounting mechanisms. Netting in current accounts does not amount to contractual or equitable set-off or compensation, nor to novation. Even if it is assumed that a debit amounts to set-off, it could not explain a credit following a deposit, nor the emergence of a negative balance following a financing transaction. Novation does not explain the legal character of the balance adjustment either. Although the individual payment obligation is extinguished by the booking of the item in the account, the resulting balance is not a new debt that substitutes the old debt. Rather, at best, one debt has been extinguished in consideration of the variation of another, existing, debt.²⁹

3.21  The individual debts and claims that arise pursuant to the transactions carried out between the account holder and the bank within the scope of the account agreement must not be treated at any time as having a distinct identity. All transactions form part of the general banking relationship that is governed by the terms of the account agreement, which is to be regarded as a single relationship that is adjusted continually as transactions occur between the parties.³⁰ The balance constitutes a single obligation that can be discovered through an accounting effort and is due only if a demand is made, and only to the amount of the demand. It also follows that the adjustment of the balance and the discovery of the balance should be (p. 94) distinguished. The rendering of an account is not a condition precedent to effect netting in a current account.³¹

3.22  Unless the firm maintaining the cash account is authorized and regulated as a bank, Article 16(9) of MiFID II requires that the firm that holds ‘funds belonging to clients, make adequate arrangements to safeguard the rights of clients and … prevent the use of client funds for its own account’. In other words, a non-bank investment firm must hold client money as custodian and preserve proprietary interests. Even if the investment firm is a bank, an investor might not agree that the bank takes certain or all deposits as the investor’s bank, which would mean that the credit balance is owed as an in personam debt.³² Rather, the investor might request, and the bank may agree, that deposits are accepted as custodian. The position may be reversed if the investor wishes to deposit substantial amounts of foreign currencies, particularly emerging market currencies. In that case, it may be that the bank refuses to accept these currencies as banker if it does not wish to assume the risks associated with the foreign nostro account, including credit and currency exchange risks.³³

3.23  Under English law, a requirement to keep monies separate is normally an indicator that the funds are impressed with a trust.³⁴ Accordingly, if monies are received as custodian, the bank, as custodian not as banker, or investment firm will be under a fiduciary duty to identify and safeguard the amounts so received by it, as the property of a trust, that is, to ensure the property passes to the beneficiaries. In the words of Thomas and Hudson, referring to tangible assets but the principle of identification applies equally to intangible assets:³⁵

Segregation or appropriation identifies relevant assets and may be strong evidence of an intention to subject them to a trust, but in itself it is not enough. On the other hand, a failure to segregate or otherwise identify specific indistinguishable tangible assets makes it difficult, if not impossible, to ascertain which assets were intended to be irrevocably dedicated to the trust.

3.24  Usually, this will be achieved by means of payment of the amounts to an omnibus account maintained by the custodian at a third-party bank for holding client (p. 95) funds.³⁶ In practice, larger cash balances in the major currencies are typically invested in collective investment funds that invest in money market instruments denominated in the relevant currency. Such investment funds are commonly referred to as ‘money market funds’. The rate of return (net of fees and costs charged to the fund by the service providers)³⁷ offered by money market funds is typically better than the interest rate offered for the cash balance by the custodian. Thus, rather than a claim on the custodian, the investor will hold units or shares in the money market fund.³⁸ Practice and regulation align in this setting. Article 4 (Depositing client funds) of MiFID II Delegated Directive (EU) 2017/593 requires that investment firms that receive client funds, ‘promptly’ deposit those funds ‘into one or more accounts’ opened with a European Economic Area (EEA) bank, a non-EEA bank, a central bank, or a ‘qualifying money market fund’.³⁹

3.25  The MiFID II safeguarding requirement is implemented in the United Kingdom through Chapter 7 of the Client Assets Sourcebook (CASS) of the Financial Conduct Authority’s (FCA’s) Handbook of Rules and Guidance. In relation to client money, the FCA has the power to make rules that result in client money being held on statutory trust in accordance with those rules.⁴⁰ Accordingly, CASS 7.17.1G provides that CASS 7.17 ‘creates a fiduciary relationship between the firm and its client under which client money is in the legal ownership of the firm but remains in the beneficial ownership of the client. In the event of failure of the (p. 96) firm, costs relating to the distribution of client money may have to be borne by the trust.’ CASS 7.17.2R provides that a ‘firm receives and holds client money as trustee’ on the terms set out in CASS. Accordingly, funds received that qualify as ‘client money’ in accordance with the rules set out in CASS are held by the recipient firm under a statutory trust.

3.26  The provisions of CASS left room for interpretation and the insolvency of Lehman Brothers International (Europe) (LBIE) the UK subsidiary of Lehman Brothers Holdings Inc, an insolvent US financial holding company, brought that into the spot light. In Lehman Brothers International (Europe) (in administration) v CRC Credit Fund Ltd and others,⁴¹ the Supreme Court considered the interpretation of CASS 7.7.2R and 7.9.6R.⁴² Briggs J held in the first instance that the statutory trust imposed on client funds by CASS 7 arose as soon as a firm received them, so as better to achieve protection of such funds as required by, inter alia, Article 13 of MiFID I and Article 16 of MiFID I Implementing Directive 2006/73/EC.⁴³ Briggs J’s view on the interpretation of the statutory intention to create a trust and when it was to take effect did not solicit much pause for judicial review. The determination of the subject matter of the trust and its purpose, that is, the intended beneficiaries and the scope of their entitlement, however, did give reason to debate.⁴⁴

3.27  LBIE had been subject to client money rules set out in Chapter 7 of CASS as issued at the time by the FSA.⁴⁵ CASS 7.7.2R provided for client money to be held on trust for the purposes of the client money rules including, in the event of a firm’s failure, the client money distribution rules in CASS 7.9. Under CASS 7.4 a firm was permitted to adopt one of two approaches to dealing with client money which it received: it could either pay it immediately into a segregated client (p. 97) account or, under what was described as the ‘alternative’ approach, pay it into the firm’s own house account(s) and then segregate it into client accounts each day according to a reconciliation conducted as at close of the preceding day’s business. LBIE had adopted the alternative approach. On 15 September 2008 LBIE failed and went into administration. LBIE’s failure constituted a ‘primary pooling event’ for the purposes of CASS 7 that crystallized the ‘client money pool’ (CMP) pursuant to CASS 7.9.6R(1), that is, the funds in each ‘client money account’ of the firm were to be treated as pooled and then, pursuant to CASS 7.9.6R(2), distributed so that each client received a sum rateable to their ‘client money entitlement’. Large amounts of client money had been collected in but not been moved out of LBIE’s house accounts not only because of the operation of the alternative approach in the period between the last internal client money reconciliation and the primary pooling event, but also because of significant non-compliance with the requirements of CASS 7 by LBIE over a long period of time.⁴⁶

3.28  The key question became what money was to be treated as pooled, that is, how the term ‘client money account of the firm’ in CASS 7.9.6R(1) was to be interpreted. It could justifiably be construed narrowly to refer only to money credited in special, separate client money accounts that were used exclusively for collection of client money, and not any other account of LBIE, and therefore exclude client money collected in LBIE’s house accounts from the CMP that formed the subject matter of the CASS statutory trust, even if it was identifiable as client money. And indeed, that was Briggs J’s decision in first instance. The further, connected question was which clients could participate in the distribution of the trust property, that is, the CMP. The term ‘client money entitlement’ in CASS 7.9.6R(2) could be read to refer to a client’s entitlement to funds that had been placed in segregated accounts so that the CMP fell to be distributed only among those clients whose funds met that criterion, leaving clients whose funds had remained unsegregated in LBIE’s house accounts as mere general creditors.

3.29  Lord Dyson gave the majority judgment. He phrased the question as follows: whether participation in the CMP was to ‘be based on (i) the amount of client money which has actually been segregated at the date of the primary pooling event (“PPE”) (the so-called “contributions basis” for participation) or (ii) the amount which ought to have been segregated at that date (the so-called “claims basis” for participation)’.⁴⁷ The majority view supported the claims basis, that is, all clients that had credit balances in their cash accounts were entitled to participate in a CMP comprised of all identifiable client money, whether segregated from LBIE’s house account into separate client money accounts or not. The decision was based on a purposive interpretation of MiFID I’s overriding (p. 98) objective to safeguard the assets of all clients and to provide all clients with a high degree of protection. Lord Dyson summarized the view as follows:⁴⁸

To summarise, for the reasons that I have given, the language of the relevant provisions of CASS 7 tends to support the claims basis for participation in the CMP. I accept, however, that the linguistic points are not conclusively supportive of this interpretation. That is why it is necessary to stand back from the detail and ask which interpretation better promotes the purpose of CASS 7. In my view, a purposive interpretation clearly supports the claims basis for participation. This basis better reflects the fact that all client money is subject to the statutory trust and that CASS 7 is intended to give effect to the Directives whose overriding purpose is to safeguard the assets of all clients and to provide all clients with a high degree of protection. I should add that we heard detailed submissions about the complexities of the process that the claims basis would entail and the inevitable costs and delay that it would occasion. The judge was impressed by these points: see, for example, para 152 above. I have little doubt that distribution on the claims basis in this case would be complex and would take a long time to complete. That is because of the extraordinary circumstances of this case. In other cases, the position might well be very different. But it has not been shown that, in a typical case, the complexity of the claims basis will necessarily be greater than that of the contributions basis. Still less has it been shown that, in a typical case, the complexity of the claims basis will be so much greater than that of the contributions basis that the draftsman could not have intended the former. I do not think that it would be right to allow the scale of the exercise that would be required in this case to lead to a solution which, for the reasons that I have given, would defeat the underlying purpose of CASS 7.

3.30  The decision has received criticism, including on the argument that clients with ‘mere contractual claims’ have no basis to participate in the pool of segregated client money under the CASS 7 rules, or general law.⁴⁹ However, that consideration does not appear to recognize that client money claims must by their very nature arise under the account agreement and result from the receipt of client money by the account-providing investment firm pursuant to the terms and conditions of that account agreement.⁵⁰ There is no obvious argument to support the conclusion that the Supreme Court’s decision is not within the parameters of basic principles of trust law, if it is accepted that the collective clients as account holders take identifiable receipts as equitable co-owners at the time of receipt. On that approach, there is certainty of intention, certainty of subject matter, and certainty of purpose.

3.31  A contribution-based entitlement would necessarily leave the decision as to whether client money received is or is not part of the trust property entirely at the discretion of the trustee, or even the arbitrary effect of its operational structure. The purposive interpretation supporting a claim-based entitlement to identifiable (p. 99) receipts does not leave room for a trustee or its operations to subvert the trust, as was the case for LBIE, whose operations fundamentally and consistently breached the requirements of CASS 7. It is difficult to defend that receipts that are clearly identifiable as not meant to be owned by the recipient, but by the recipient’s client, should be the recipient’s and not the trust’s property if there is an agreement between the client and the recipient that is clear in its purpose to impress a trust on such receipt. As Lord Dyson observes poignantly in the context of the client protection policy of CASS and MiFID (emphasis added):⁵¹

To exclude identifiable client money in house accounts from the distribution regime runs counter to this policy. It creates what was referred to in argument as a ‘bifurcated’ scheme which provides clients with different levels of protection, namely a right to claim in the CMP under the CASS 7 rules for those whose money is held in segregated client accounts but no right (other than a right to trace in equity) to those whose money is held in the firm’s house accounts. The purpose of the scheme [in CASS] (as required by [MiFID I]) is to provide a high level of protection to all clients and in respect of client money held in each money account of the firm. That purpose would be frustrated if the protection were restricted in this way. As Mr Miles and Mr Crow point out, a bifurcated scheme would provide clients with different levels of protection based on the happenstance of whether the firm has segregated money on behalf of that client. That is an arbitrary basis for a scheme which is intended to provide protection to all clients who entrust their money to a firm. It is unlikely that the draftsman of CASS 7 intended the scheme to have this effect. It is improbable that the draftsman contemplated that there would be two regimes substantially in operation for the distribution of client money (one under the CASS 7 rules set up for the purpose and one under equitable tracing principles and outside CASS 7).

3.32  The LBIE client money decisions show that, notwithstanding the decision in MacJordan Construction Ltd v Brookmount Erostin Ltd,⁵² if receipts of money can be identified, the fact that these are commingled into a house account does not prevent the trust from being impressed. The FCA has since amended its CASS distribution rules to put beyond doubt that the CMP consists of all identifiable client money, that is, a common fund, whether segregated in client money accounts or commingled in an investment firm’s trading account.⁵³ In MacJordan, the claimant sought to establish an individual trust over its insolvent debtor’s bank account since funds that were destined to be paid to the claimant by the insolvent debtor had been paid to that account. The action failed because, for the trust to be established for the benefit of the claimant individually, the property would have needed to be segregated into individual lots. The LBIE cases⁵⁴ accepted that the concept of a common fund implies that the beneficiaries have (p. 100) a common intention to share equally in a shortfall, as is provided for expressly in CASS 7.17.1G (emphasis added):

This section creates a fiduciary relationship between the firm and its client under which client money is in the legal ownership of the firm but remains in the beneficial ownership of the client. In the event of failure of the firm, costs relating to the distribution of client money may have to be borne by the trust.

3.33  Equity and fixed-income securities are usually issued either in registered form or in bearer form.⁵⁵ Until, broadly, the 1960s, the delivery obligations on a contract for the sale and purchase of securities made in the financial markets would require the seller to transfer securities of a certain type, description, and quantity to the buyer. Thus, if a security is issued in registered form, in most legal systems, a registered security is transferred by a written transfer instrument that is presented to the issuer’s registrar after which the registrar will amend the register so that the transferee will be recognized as the legal owner. If a security is issued in bearer form, in most jurisdictions, it will typically be characterized as a negotiable instrument that embodies the entitlement of its holder as against the issuer.⁵⁶ Legal title to a bearer security is transferred by delivering possession of the certificate to the transferee. Negotiability entails that the transferee will acquire legal title even if the transferor’s title is defective, so long as the transfer is made for value and the transferee acts in good faith without notice of the defect.

3.34  During the 1960s growth in the number of securities trades caused congestion in the traditional ‘paper-based’ settlement processes. In response, most markets gradually organized the holding and settlement of securities around CSDs that function as a hub in an account-based, that is, computerized, security holding, and transfer, system that consists of layers of securities accounts, the balances of which ultimately can all be traced to the original securities that are held by, or recorded on a register kept by, the CSD.⁵⁷ Typically, most notably in the case of equity securities, the CSD will be established in the jurisdiction of domicile of the issuer.

3.35  As a rule, securities are centralized either by dematerialization or by immobilization.⁵⁸ In dematerialization, the physical certificates or other documents or (p. 101) records of title are replaced by an electronic record maintained by the CSD. The electronic record performs the legal function previously performed by the company’s register, so that the account holder is recognized as such by the issuer based on an entry made to the electronic record maintained by the CSD. The position of the holder of dematerialized bearer securities resembles that of the holder of dematerialized registered securities,⁵⁹ and the issuer will recognize the account holder as the holder of record. In centralization by immobilization, title to the original securities is transferred to the CSD. This typically occurs at the initial public offering of the securities. Under a statutory or consensual scheme, the CSD undertakes to its account holders, in the case of registered securities, to be recorded as the registered owner on the shareholders’ register, or, in the case of bearer securities, to take possession as the legal owner of the bearer instruments, and to treat the account holders as entitled to the benefits of the centralized securities. Thus, the characteristic feature of the creation of a primary account structure at a CSD by immobilization is that the CSD acquires the original securities for the benefit of the account holders on a pooled, undivided basis. The entitlements of the account holders as against the CSD and in respect of the pool of original securities are represented by the balances of their primary securities accounts, which can be manipulated by simple variation to the record, rather than by complying with the more extensive traditional security-transfer formalities.⁶⁰

3.36  The account-based holding and settlement system created ‘book-entry’, ‘digitized’, or ‘computerized’ securities and permitted a change in the way a debtor of a securities transfer obligation may discharge his/her as to duty to transfer securities. Instead of passing title in the original securities to the transferee, transferors instruct their securities account provider to arrange for the transferee to receive a credit to his/her securities account which corresponds to the quantity and description of the type of original security that is the subject of the transfer obligation. It follows that investors do not normally hold original securities in their investment portfolios. Instead, an investor will hold the balance of an account that is maintained with a CSD or a custodian who has access to an account at a CSD. The premise underlying a securities-holding and settlement system is the notion that the centralized securities are fungible goods that need not be in direct possession (p. 102) or control to enjoy the fruits. The corollary of this is that the right to require the CSD to cause the account holder to become the holder of original securities can ultimately be dispensed with. Jurisdictions that permit immobilization under a statutory scheme may stipulate that account holders cannot require the CSD to transfer original securities to the account holder. Blockchain technology may yet evolve to create a de-centralized ledger system that could dispense with the central administrative hub function of a CSD.

3.37  The development of centralized securities holding and transfer systems has remained fragmented at national level. The EU legislator considered that the ‘divergent measures likely to be taken at national level will have a direct negative impact on the safety, efficiency and competition in the settlement markets in the Union’ and therefore, that it ‘is necessary to remove those significant obstacles in the functioning of the internal market and avoid distortions of competition and to prevent such obstacles and distortions from arising in the future’.⁶¹ The CSD Regulation aims to create an integrated market for securities settlement that does not distinguish between national and cross-border securities transactions. To that end, the CSD Regulation requires that all securities issued by an EEA issuer and admitted to trading on a trading venue regulated under MiFID II or MiFIR must be established in book-entry form by way of immobilization or dematerialization through a CSD.⁶² Each firm that provides CSD services depositories must be authorized and regulated under the CSD Regulation.⁶³ Recital (11) of the CSD Regulation provides a succinct outline of the requirement either to immobilize the original securities in the hands of a CSD, or to issue original securities in immaterialized form through the CSD, both with a view to permit the holding and transfer of digitized interests in original securities by way of book-entry (emphasis added):

The recording of securities in book-entry form is an important step towards increasing the efficiency of settlement and ensuring the integrity of a securities issue, especially in a context of increasing complexity of holding and transfer methods. For reasons of safety, this Regulation provides for the recording in book-entry form of all transferable securities admitted to trading or traded on the trading venues regulated by [MiFID II] and [MiFIR]. This Regulation should not impose one (p. 103) particular method for the initial book-entry recording, which should be able to take the form of immobilisation or of immediate dematerialisation. This Regulation should not impose the type of institution that is to record securities in book-entry form upon issuance but, rather, should permit different actors, including registrars, to perform that function. However, once transactions in such securities are executed on trading venues regulated by [MiFID II] and [MiFIR] or provided as collateral under the conditions laid down in [the Collateral Directive], such securities should be recorded in a CSD book-entry system in order to ensure, inter alia, that all such securities can be settled in a securities settlement system. Immobilisation and dematerialisation should not imply any loss of rights for the holders of securities and should be achieved in a way that ensures that holders of securities can verify their rights.

3.38  Importantly, Recital (11) recognizes that national laws should preserve the bundle of rights that attach to the original securities at least at the first intermediate account level, and it could be argued that legislative expectation should hold true throughout the chain of custody in EEA jurisdictions based on the operation of the effet utile.⁶⁴ Recital (4) of the CSD Regulation specifically mentions ‘cross-border settlements as a consequence of the development of link agreements between CSDs’ and the need to ensure ‘the resilience’ of the system, which ‘creates additional risks and costs for cross-border settlement’.

3.39  Not all issued securities are transferred to a CSD to be recorded in a securities holding and settlement system. Units or shares in collective investment schemes are still predominantly held directly.⁶⁵ The reason is that these shares or units, with the obvious exception of exchange-traded funds, are not publicly traded. The main purpose of a collective investment scheme is to pool investor contributions so that these can be invested as a single common fund. The investor will not usually intend to trade the holding in the fund but redeem funds or securities by requesting a cancellation of the unit or share if it concerns an open-ended fund, which may take place in the form of a buy-back by the scheme’s operator. In the case of closed-ended funds, the investor will usually hold to maturity, although may seek to assign or transfer interests prior to that time in a private transaction.

3.40  The CSD is the hub and the participant custodians are the spokes, each forming a custody chain in the account-based securities holding network. Figure 3.1 shows a simplified schematic overview.(p. 104)

• View full-sized figure

Figure 3.1  Basic intermediated holding structure

3.41  The protection of proprietary rights in the context of an account-based securities holding structure is not as straightforward as it perhaps ought to be. Remarkably, the securities industry originally changed to account-based holding and settlement methods without much regard to the legal uncertainty that was created by the new structures. CR Reitz observes in relation to the US practice:⁶⁶

With these and related changes, the securities industry and its customers ‘solved’ the ‘paperwork’ crisis. The new efficiencies enabled the continued upward trajectory in the daily volume of securities trading at greater speed and vastly reduced transaction costs to firms in the industry and to ultimate investors. … The industry effected these dramatic moves without any significant change in the legal platform regarding the issuance or trading of securities. … Looking back, it is difficult to comprehend how the securities industry and the securities markets could have effected these changes without seeing the legal uncertainty and the enormity of the legal risks that had been created. No one could know what, exactly, was created by the credit to a securities account. … The novel and uncertain relationship between an intermediary and its account holders presented a sizeable class of concerns, which collectively came to be called ‘intermediary risks’.

3.42  The uncertainty prompted a legislative response, particularly in the continental European jurisdictions, during the 1960s and 1970s. English practice, except where there was a need to cater for the wish to dematerialize securities, continued to rely on general principles of property law and equity, that is, the law of trust. Accordingly, some uncertainties and deficiencies exist in English law, particularly in the absence of specific agreement and proper structuring of the holdings.⁶⁷

(p. 105) 3.43  The US response has been much more comprehensive. The 1994 revision of Article 8 of the US Uniform Commercial Code is a very sophisticated statutory scheme aimed at providing a legal framework for account-based holding and transfer of securities, including for collateral purposes. Its main conceptual achievement is that it puts beyond doubt that all assets⁶⁸ acquired by a securities account provider⁶⁹ for the benefit of its account holders form a single common fund in which the account holders acquire co-ownership interests pro rata.⁷⁰ Furthermore it puts beyond doubt that the interest arises pursuant to a credit, or a pending credit, to the account,⁷¹ regardless of actual receipt by the account provider of the corresponding financial asset.⁷² Article 8 is modelled on trust law, but it eliminates some of the questions that arise as a result of market practices, such as the claim versus contribution questions that bedevilled the Supreme Court in LBIE v CRC,⁷³ in particular, without limitation, questions around the subject matter of the trust and timing of allocation, the content and conditions of the rights and proprietary interests that arise pursuant to a credit to an account,⁷⁴ possible set-off rights that arise in relation to the accounts, and the liability of the account provider. It is beyond doubt that the revision of Article 8 has greatly enhanced legal certainty around the operating conditions of US-based securities holding and settlement systems.

3.44  Investors based in an Organisation for Economic Cooperation and Development (OECD) member country are not restricted to transactions in domestically issued securities and, accordingly, the volume of cross-border investment transactions is substantial. Notwithstanding legislative efforts to improve the legal basis for centralization of securities in the hands of CSDs, these responses have been predominantly domestic in nature and uncertainty persists in the financial markets, particularly in cross-border transactions where interests in relation to the original securities would have to be exercised through a chain of intermediaries.⁷⁵ This (p. 106) prompted UNIDROIT to engage in an important study on harmonized substantive rules regarding securities held with an intermediary.⁷⁶ The UNIDROIT Convention on Substantive Rules for Intermediated Securities was adopted in Geneva, Switzerland on 9 October 2009 by the diplomatic Conference to adopt a Convention on Substantive Rules regarding Intermediated Securities. The conference decided pursuant to Resolution 1 that the convention be known as the ‘Geneva Securities Convention’ (GSC). The GSC is designed to promote harmonization of the legal frameworks underlying holding and disposal of securities in securities account and settlement systems to enhance the internal stability of national financial markets and their cross-border compatibility in these respects, and accordingly, to promote capital formation. The explanatory text observes that the GSC:

[I]n particular, within the context of holding securities with an intermediary, … describes the rights resulting from the credit of securities to a securities account; establishes different methods to transfer securities and to establish security and other limited interest therein; clarifies the rules regarding the irrevocability of instructions to make book entries and the finality of the resulting book entries; precludes ‘upper-tier attachment’; establishes a priority ranking among competing interests with respect to securities; protects the innocent (‘good faith’) acquirer of securities from adverse claims; sets out the rights of the account holder and the responsibilities of the intermediary in the event of insolvency; establishes a regime for loss allocation; and defines the legal relationship between collateral providers and collateral takers where securities are provided as collateral.

3.45  In other words, the GSC conceptually recognizes the asset-backed nature of account-based claims and seeks to safeguard the pools of assets that are held against the securities balances through the custody chain to the accounts maintained at the CSD. However, despite the critical importance to financial stability of the matters addressed in the GSC, the GSC has not seen rapid adoption and legal uncertainty in the cross-border chain of custody persists,⁷⁷ although the finality and conflict of laws rules provided for in the Settlement Finality Directive and the Collateral Directive have done much to reduce uncertainty in the (p. 107) European context.⁷⁸ Micheler observes the difficulty that needs to be overcome when an international convention is seeking consensus on matters of national property law:⁷⁹

There is, however, also evidence that some national delegations and representatives were approaching their contributions with a view to preserving their own system rather than working out what concerns there were in terms of legal and systemic risk or in relations to market efficiency and what changes are required to address these concerns. Representatives of the French and German banking sector, for example, expressed in very clear terms that they would not want their national approach to change. It is worth noting that their respective and joint interventions were not substantiated by reference to legal risk, systemic risk or market efficiency.

3.46  The account-providing firm will normally maintain client assets on a commingled, fungible (ie pooled) basis. Therefore, if a proprietary interest is to persist pursuant to the account agreements, it can usually only persist in the form of legal or equitable co-ownership in the undivided pool of client assets received and available for distribution. The critical question concerns the allocation of the pool, that is, that there must be certainty about the size of the pool and its intended division.

3.47  There is robust judicial precedent supporting a rule that contributors to a mixture of tangible, fungible goods can take pro rata as legal co-owners, more precisely as tenants in common, provided that the ‘bulk’ has been allocated properly to determine pro rata claims.⁸⁰ Consider for instance Re Stapylton Fletcher Ltd,⁸¹ which concerned the application of commercial law to the sale of goods.⁸² It is recognized that if the goods have been paid for, an agreement to buy 10 units from a (p. 108) bulk of 100 equates to an agreement to buy a 10 per cent interest in the entire bulk.⁸³

3.48  The situation is similar if the co-owned property concerns intangible assets. Consider for instance Hunter v Moss.⁸⁴ The Court of Appeal held that a trust could validly arise in relation to 50 shares that were issued in registered form, where the issued and paid-up share capital of the company consisted of 1,000 shares and the settlor held 950. Although the subject of much criticism, the decision appears entirely sound,⁸⁵ so long as it is recognized that the body corporate, as represented by its share capital, is a single asset and, accordingly, there is no lack of certainty as to the subject matter of the co-ownership interests.⁸⁶ The allocation question is limited to certainty about the pro rata division, and in the circumstances, it is irrelevant whether the allocation is expressed as 50 shares of 950 shares held by the settlor in a total share capital of 1,000, or as a 50/950 portion of 950 shares held by the settlor, or as a 50/1,000 portion of the total share capital. As long as it is clear what pro rata interests have been allocated.

3.49  Based on the principles that operate on the allocation of a bulk or intangible share capital, there ought to be no objection in principle based on certainty of subject matter if the intention is to impress a trust on a proportion of a fungible fund of account-based assets, as opposed to in specie indistinguishable components of that fund.⁸⁷ Like share capital, an account-based asset, that is, the balance standing to the credit of an account, is a single intangible asset. Unlike share capital but like a tangible bulk, a credit balance is divisible. Briggs J discusses Hunter in Pearson v Lehman Brothers Finance SA and embraces the co-ownership analysis of the equitable interests in trust property if there is more than one beneficiary:⁸⁸

The difficulty with applying the Court of Appeal’s judgment in Hunter v. Moss to any case not on almost identical facts lies in the absence of any clearly expressed rationale as to how such a trust works in practice. There has not been unanimity among those courts which have followed Hunter v. Moss, nor among the many academics who have commented upon it, as to the correct approach. The analysis (p. 109) which I have found the most persuasive is that such a trust works by creating a beneficial co-ownership share in the identified fund, rather than in the conceptually much more difficult notion of seeking to identify a particular part of that fund which the beneficiary owns outright. A principal academic advocate for the co-ownership approach is Professor Roy Goode: see for example ‘Are Intangible Assets Fungible?’ [2003] LMCLQ 379. Among the judicial commentators I have found the analysis of Campbell J in White v. Shortall (supra) at paragraph 212 to be the most persuasive. My own preference for the co-ownership analysis may be observed in LBIE v. RAB Market Cycles [2009] EWHC 2545 (Ch), at paragraph 56. I propose to adopt it for the purposes of the analysis which follows.

3.50  Hunter concerned co-ownership of a single asset. The pools of account-based client assets held by banks and investment firms as account providers against the balances administered in their clients’ accounts, however, are comprised of multiple component parts, which are held or controlled by the bank or investment firm as intermediary. The fungibility of the pool of account-based client assets coupled with the complexity of the operational reality may raise questions about certainty of the subject matter of the co-ownership interest of the collective account holders. In LBIE v RAB Market Cycles,⁸⁹ Briggs J had to consider rights of securities account holders that had entered into a prime brokerage agreement (PBA) that appointed LBIE as custodian and, common to prime broker service models, provided for extensive rights of use of the custodied assets by LBIE.⁹⁰

3.51  The question turned on whether the balance in the custody account, given the operating model and the terms and conditions of the PBA between RAB Market Cycles (Master) Fund Limited as the client and LBIE as the prime broker, created proprietary interests for the account holders in the pool of assets held by LBIE as custodian, particularly, whether the terms created a trust. The decision provides a delightful account of arguments contra made by counsel for the applicant, LBIE’s administrators. Counsel pointed to the fact that pursuant to the PBA, LBIE was entitled to hold the client’s securities in fungible accounts and mix them with securities of other customers and to commingle client securities in accounts with LBIE’s own securities. Further, counsel pointed to the primary obligation of LBIE as being to deliver only equivalent securities rather than securities in specie, and to LBIE’s rights of use, whereby LBIE could do, in effect, what it liked with the counterparty’s securities without even an obligation to give prior notice. Counsel asked, ‘under what sort of trust known to English law … could the trustee be free to use the trust property for its own purposes, free to ignore the beneficiary’s instructions to sell or deliver them, and be freed from any proprietary obligations in relation to them upon termination of the contract?’⁹¹

(p. 110) 3.52  Briggs J did not spill much ink on the response. Noting that although he considered counsel’s arguments ‘cogent submissions, and amply sufficient to demonstrate that if any trust is created by the [PBA], it is a most unusual type of trust’, he was not persuaded that the effect of the PBA ‘in relation to securities is that it creates no trust known to English law’.⁹² He subsequently made a number of helpful observations:

1. a)  The fact that the PBA uses the words ‘custody’ and ‘custodian’ and the phrases ‘belong to the client’ and ‘do not belong to LBIE’ is considered to be ‘the clearest language to display their intention that securities held by [LBIE] for the time being continue to belong beneficially to the [client], subject of course to the charge in favour of [LBIE]’.⁹³

2. b)  The existence of a right in the holder ‘to mix the fungible property of one beneficiary with the fungible property of another beneficiary in a single fund has never been a powerful contra-indication against the existence of a relationship of trustee and beneficiary’.⁹⁴ There is (emphasis added)

   nothing incompatible with the recognition of a proprietary [client] interest in securities in the provisions whereby they may be held in omnibus fungible accounts with equivalent securities of other [clients]. While it may be that the consequence is that the proprietary interest of any particular [client] is to a rateable share in the fungible account rather than in particular securities in that account, there is no reason in my judgment why that interest should not be recognised as proprietary, or that the obligations of the account holder (be it as custodian or sub-custodian) are those of a trustee.⁹⁵

3. c)  The ‘primary obligation of [LBIE] under … the [PBA], to deliver, upon reasonable request, equivalent securities to the [client]’ is not ‘pointing away from the recognition of a proprietary interest of the [client] in securities’.⁹⁶

4. d)  The ‘quid pro quo for any exercise by [LBIE] of the Right of Use is an obligation to deliver equivalent securities to the [client], on the basis that the equivalent securities are thereupon held upon the same terms, as between [LBIE] and [the client], as the original securities’.⁹⁷ On that basis, the right of use is ‘in substance, a right in [LBIE] to swap trust property for equivalent property of its own. By equivalent I mean, of course, not equivalent in value, but equivalent within the detailed meaning of that word in the [PBA]’.⁹⁸

3.53  In summary, account agreements can create equitable co-ownership interests in a fungible pool of underlying account-based assets, that is, in an undivided (p. 111) common fund, and an investment firm’s right of use or replacement does not per se result in a failure of the trust. Exercising the right of use does not constitute some form of appropriation that could result in the trust failing for want of certainty, which appears to be the implied argument in the reasoning deployed by LBIE’s administrators’ counsel. Briggs J’s analysis and decisions recognize that the fungible pools of underlying assets form the operating basis of account-based financial asset-holding and transfer systems. The approach taken, that is, equitable co-ownership interests persist in the fungible pool held pursuant to the custody arrangements, does much to safeguard investors’ proprietary rights in securities and money that are received and held by the account provider on a fungible basis for the account of its clients and accounted for in the form of a credit balance standing to the credit of clients pursuant to the terms of account agreements between the account provider and each client.⁹⁹

3.54  The constitution of a trust of personal property such as securities requires no formalities,¹⁰⁰ but there must be certainty of subject matter, of intention, and of object. The terms and conditions of custody agreements will leave no doubt as to the intention to safeguard the property of the investor. Common to most if not all custody account agreements are references to the fact that assets credited to the securities account are received and held for the account of, or for the benefit of, the account holder, and must be separated from the custodian’s property. Account agreements will typically authorize the custodian to pool assets held for the benefit of a customer with assets held for the benefit of other customers, or, as did the LBIE PBA, with the custodian’s own assets. A court should have little difficulty in construing account terms that contain these references as a common intention of the account holders and the custodian that the custodian will not take the beneficial interest of the securities received for the benefit of its account holders, but hold the assets that form the subject of the account agreements as a single common fund upon trust for the account holders from time to time, pari passu and in proportion to the number of securities credited to the accounts.¹⁰¹ Indeed, Briggs J came to that conclusion on the terms of the PBA in LBIE v RAB,¹⁰² and neither Briggs J nor Lord Dyson struggled with this point in deciding the LBIE client money cases discussed paras 3.26 and 3.29 above.

(p. 112) 3.55  The LBIE cases also make clear that the requirement of certainty of subject matter at least conceptually, de jure, does not disable the impression of a trust on the common fund or funds of fungible account-based client assets, such as book-entry securities and cash in bank accounts, held by the custodian. Briggs J had further occasion to consider this point in respect of securities received by LBIE as settlement agent for the account of affiliates in another LBIE decision, Pearson v Lehman Brothers Finance SA, commonly known as the ‘Rascals’ case.¹⁰³ He, again, very helpfully summarized what he believed to be the operative principles in the context of certainty of subject matter:¹⁰⁴

3.56  Principle i) and ii) were, and presumably are, uncontested. Counsel for LBIE’s administrators submitted in relation to principle iii) that:¹⁰⁵

the probability that on any given day the supposed trust fund—constituted by LBIE’s un-segregated house depot account for a particular description of security—would fall well short of the aggregate of the affiliates’ supposed beneficial interests raised such difficulties of apportionment between them as to render the supposed trust uncertain as to its terms, as well as to its subject matter.

The operational complexity may make it problematic to identify de facto rather than de jure what is or is not part of the pool. A professional custodian (p. 113) will usually operate globally, have many clients, and will maintain many securities accounts with many sub-custodians and CSDs. All these accounts will see constant activity, which means that the recordkeeping systems of a custodian need to adhere to high standards to always ensure that the true position can be determined without too much delay. The matter is complicated further by the fact that in some securities settlement systems transfers will occur on a net basis so that both transfers in and out may be included. In addition, the custodian might be engaging in a securities-trading business so that there may be operational challenges to identifying promptly what property belongs to the custodian and what property is held for the benefit of its clients. However, none of the operational complexities ought to be a bar to the trust arising, as the LBIE client money cases demonstrate and as Briggs J confirmed in Pearson v Lehman Brothers Finance SA.¹⁰⁶ The balances credited to the accounts of its clients are usually known accurately and without much delay and constitute the basis for each client’s claim against the account provider and, consequently, the co-ownership interest in the assets held by the account provider in receipt for its account holders. The fact that the client assets are received in multiple accounts held with sub-custodians or CSDs does not somehow create uncertainty of subject matter. The constitution of the balances is known and can be impressed with a trust based on the internal administration of the trustee/account provider, even though the trustee/account provider’s own property may be commingled in the common fund. Equally, net transfers to or from the pool ought not to present an obstacle. Net receipts in an undivided pool redistribute value in the same manner as gross receipts, if the items included in the net transfer can be determined from the trustee/account provider’s records.¹⁰⁷

(p. 114) 3.57  Briggs J’s response to counsel’s submission that a shortfall causes the trust to fail for lack of certainty follows the co-ownership analysis in LBIE v RAB Market Cycles.¹⁰⁸ He said on the question of certainty of subject matter and object:¹⁰⁹

Subject matter certainty requires not only that the identity of the shared fund is certain, but also that the proportionate amount of each alleged beneficiary’s share is also certain. Thus, an uncertainty as to the terms upon which LBIE and its affiliates were to bear any shortfall would be destructive of the necessary certainty of each affiliate’s beneficial share in the securities remaining in LBIE’s depot account. In truth, uncertainty as to terms in this context means, or inevitably leads to, uncertainty as to beneficial subject matter.

But he concluded—in line with his operative principle iii)—that there was no uncertainty as to the proportion in which the beneficiaries would share in the common fund if there is a common intention to share the shortfalls:¹¹⁰

But I see no reason why, on the assumption that LBIE and its affiliates had a common intention which authorised LBIE to operate its house depot accounts in a matter which included lending to the street and the making good of short positions, the consequential day to day shortfalls thereby created in the depot accounts should not be shared on a pari passu basis. Bearing in mind that the evidence shows that LBIE’s operation of its house depot accounts in that way was an aspect of the Group settlements practice rather than something done idiosyncratically by LBIE behind the backs of one or more of its affiliates, there is no basis in fact for concluding that the manner in which LBIE operated those accounts before the implementation of Rascals was other than consensual. If all those interested in the fund consented to the arising of the shortfall inherent in the way it was managed, then I can see no reason why they should not be taken to have agreed to bear the consequences of the shortfall equally.

Briggs J’s decisions on the questions of certainty of subject matter were affirmed by the Court of Appeal.¹¹¹

3.58  Accordingly, the fact that a custodian has pooled its own property with the property held for the benefit of the account holders, or that the pooled property may experience a shortfall, might cause a trust to fail, but not in all circumstances. The LBIE client money cases have shown, notwithstanding the decision in MacJordan,¹¹² that if receipts for credit to a client’s cash account can be identified, the fact that these receipts are commingled into a house account does not prevent the trust from being impressed. The LBIE cases support the concept of a common fund that operates subject to a common intention of the beneficiaries and the custodian as the trustee who has commingled its own assets with the assets of the beneficiaries to share equally in a shortfall. Consequently, it ought to be accepted (p. 115) that a trust may arise in respect of the balance of the house account which comprises the traceable proceeds of a receipt of an asset for credit to a client’s account to the extent necessary to protect the interest of the beneficiary.¹¹³

3.59  Conceptually, therefore, a pool of account-based assets should not be treated as a commixtio because individual receipts into individual accounts can be identified. Rather, the pool of account-based assets should be treated as a confusio.¹¹⁴ Receipt of a client asset conceptually constitutes a confusio—a single asset created out of the mixing of contributing elements—with all other client and house assets held under the fungible custody arrangement and the property in the common fund may be co-owned, and all co-owners share pro rata both the gains and the losses. That is the correct approach to be applied to the holding of account-based assets for the benefit of underlying account holders. Unlike commingled, tangible assets that are considered to be a commixtio rather than a confusio, a pool of account-based client assets may be treated as allocated so long as the proportion of the co-ownership shares of the underlying account holders and claimants can be determined.¹¹⁵

3.60  The decisions in the LBIE cases discussed above—a credit to an account establishes co-ownership entitlements in identifiable receipts of client assets and account holders share pro rata in shortfalls—are in line with the model that was adopted by Revised Article 8 UCC. §8-501(b) UCC provides that a person acquires a security entitlement if a securities intermediary: (1) indicates by book-entry that a financial asset has been credited to the person’s securities account; (2) receives a financial asset from the person or acquires a financial asset for the (p. 116) person and, in either case, accepts it for credit to the person’s securities account; or (3) becomes obligated under other law, regulation, or rule to credit a financial asset to the person’s securities account. The Official Comment notes: ‘the point of paragraph (1) is that once an intermediary has acknowledged that it is carrying a position … the customer … has a security entitlement. The precise form in which the intermediary manifests that acknowledgement is left to private ordering.’ It further notes that ‘paragraph (2) sets out a different operational test, turning not on the intermediaries accounting system but on the facts that the systems are supposed to represent. Under paragraph (2) a person has an entitlement if the intermediary has received and accepted a financial asset for credit to the account’. It is not immediately clear how in practice manifestation of the acknowledgement in paragraph (1) and the acceptance in paragraph (2) can be distinguished, but the point is that it is not a mere accounting effort which constitutes the securities entitlement. §8-503(b) UCC provides that an account holder’s interest in the pool is an interest pro rata without regard to the time the account holder acquired the entitlement, or the custodian acquired the asset that accrued to the pool.

3.61  Shortfalls do raise the question whether a claim may have perished or been reduced through depletion of the pool, which is a matter of methodology. The rule in Clayton’s Case¹¹⁶ suggests that a first-in first-out approach may be appropriate. This has been criticized as rigid. The equitable approach would be to apply pro rata apportionment, so that debits are applied against the credits pro rata rather than chronologically.¹¹⁷ Indeed, if it is accepted that the account holders take as equitable tenants in common, in the absence of contrary intentions or traceable fault, it would be difficult to see how losses could be allocated in any manner other than on a pro rata claim basis.¹¹⁸ This still leaves to be determined, however, the time and the basis on which the pro rata allocation should be calculated. Three possible methods appear to compete for prevalence:¹¹⁹

1. (a)  First, there is judicial precedent for the ‘rolling charge’ method, which allocates a shortfall only among those persons that were account holders at the time that the improper transaction occurred, applied by the Ontario Court of Appeal in Re Ontario Securities Commission and Greymac Corp¹²⁰ and considered by the Court of Appeal in Barlow Clowes International (in liquidation) v Vaughan.¹²¹

2. (p. 117) (b)  Next, the basic sharing method has been used, which disregards the order of dealings and allocates the loss based on the size of each holding on a particular date, which, following the LBIE decisions, appears to be the accepted rule for commingled client assets held by an investment firm. The Court of Appeal adopted this approach in Barlow Clowes and the basic sharing method set out in CASS 7 was accepted in LBIE v CRC.¹²² Lord Dyson summarized that:

   (i) client money is held on the statutory trust imposed by CASS 7.7 from the time of receipt by a firm; (ii) the money treated as pooled at the [primary pooling event] should be distributed to clients in accordance with their respective client money entitlements under CASS 7 construed in accordance with this judgment; and (iii) the pooling at the [primary pooling event] includes all client money identifiable in any account of LBIE into which client money has been received and is not limited to client money in the firm’s segregated accounts’

   Ergo, the moment a primary pooling event, in casu the failure of LBIE, occurred served as the time at which the pool and the entitlements in it would be determined, without regard to the time at which a client’s account was credited for the amount. The Revised Article 8 UCC adopted the same principle. §8-503(b) UCC provides that

   [A]n entitlement holder’s property interest with respect to a particular financial asset under subsection (a) is a pro rata entitlement in all interests in that financial asset held by the securities intermediary, without regard to the time the entitlement holder acquired the security entitlement or the time the securities intermediary acquired the interest in that financial asset.

   This means in practice that under Revised Article 8 UCC the applicable distribution rules will depend on the lex concursus, the law of the insolvency forum.¹²³

3. (c)  Finally, there is support for the ‘lowest intermediate balance rule’, which overlays the basic sharing method with a ceiling so that an account holder cannot claim in excess of the lowest balance that was held in the account during the period in which it was an account holder.

3.62  The Law Commission argued in 2008 that English law is likely to apply the basic pro rata sharing method measured at the date of insolvency of the account provider.¹²⁴ That was the right call, as has now been demonstrated in the LBIE cases, but had arguably already been accepted in Sinclair v Brougham,¹²⁵ which decided that the calculation should be based on claims as at the date of commencement of the winding-up. The position can also be argued based on the decision in Barlow Clowes, although decided in the context of tracing into mixtures, that the process (p. 118) of calculating separate entitlements on a time-value basis is too complicated and therefore inequitable.¹²⁶ The administrator appointed in the insolvency of an investment bank under the Investment Bank Special Administration Regulations,¹²⁷ has the power to allocate shortfalls in client securities pro rata in an omnibus account.¹²⁸

3.63  Subject to the terms of the account agreement, individual account holders in their capacity as third-party beneficiary will not be entitled to exercise rights directly against the issuer of the original securities held by the CSD nor,¹²⁹ if held in an intermediate account, against an underlying account provider, neither in respect of the rights that arise as a matter of the balance of the account,¹³⁰ nor as a matter of a breach of a duty arising as a matter of the account relationship.¹³¹ The trust structure ensures that the account holders’ rights are rights that must in the first instance be exercised against their account provider. They cannot assert rights directly in the property held for the benefit of the accounts by that account provider.¹³² It follows that, in these circumstances, neither can (secured) creditors of the third-party beneficiaries exercise any rights in respect of the subject matter of the equitable interests that they seek recourse against. In Pearson v Lehman Brothers Finance SA, Briggs J commented on the custody chain:¹³³

It is common ground that a trust may exist not merely between legal owner and ultimate beneficial owner, but at each stage of a chain between them, so that, for example, A may hold on trust for X, X on trust for Y and Y on trust for B. The only true trust of the property itself (i.e. of the legal rights) is that of A for X. At each lower stage in the chain, the intermediate trustee holds on trust only his interest in the property held on trust for him. That is how the holding of intermediated securities works under English law, wherever a proprietary interest is to be conferred on the ultimate investor. In practice, especially in relation to dematerialised securities, there may be several links in that chain.

(p. 119) 3.64  Briggs J’s summary and comment capture, as he himself notes, the account-based chain of custody on which the modern securities holding and settlement infrastructure is premised. The principle underlying the maxim nemo dat quod non habet implies that the custodian can only confer an equitable interest on its account holders on the interest that it owns itself. For instance, if the interest held by the custodian is itself an equitable co-ownership interest, as it does not seek to assign its own equitable rights, it can only transfer the interest by establishing a sub-trust.¹³⁴ The same analysis applies to balances held by a sub-custodian. If the sub-trust is established in favour of the intermediate account holders,¹³⁵ a further question arises whether the intermediate account provider disposes of an equitable interest by doing this.¹³⁶ For that to be true, it would probably have to be argued that the primary trust and the sub-trust collapse. It is difficult to see how that position could be maintained. The intermediate account holder does not have the right to instruct the primary account provider. According to Briggs J, the better view appears to be that the establishment of a sub-trust amounts to the creation of a new and distinct, albeit subsidiary, equitable interest, not to the transfer of that interest.¹³⁷ Nevertheless, it could be argued that the creation of a subsidiary equitable interest limits the rights of the custodian as primary beneficiary in a way that amounts to a part-disposal. If the conclusion is that the intermediate account provider transfers an equitable interest once the sub-trust is created, then the formalities of section 53(1)(c) of the Law of Property Act 1925 must be complied with and the ‘disposition’ must be in writing. It is conceivable that the intermediate account provider’s records will suffice for that purpose.¹³⁸

3.65  Even a beneficiary that is absolutely entitled to the whole of the possible equitable interest in an asset must first terminate the trust and direct the trustee to transfer the trust property if that beneficiary seeks to exercise rights directly in relation to (p. 120) the trust property.¹³⁹ Until termination, it can merely seek to compel the trustee to act in accordance with the trust. Immobilization under a trust, unless there are specific provisions in the account terms, therefore, entails the account holder not being able to enforce directly any rights that a holder of the original securities has against the issuer.¹⁴⁰ It demonstrates that centralization of securities severs the link between the issuer and the investor. The equitable co-ownership interest in the legal title to the original securities or any book-entry securities that arises under the relationship between the account holder and the account provider is an asset that is separate and legally distinct from the issued security.

3.66  As a rule, an inter vivos express fixed trust settles if a transferee accepts a transfer of property subject to a condition that s/he holds as trustee, in which case the transferor is the settlor, or if a holder or recipient of property declares itself a trustee, in which case the trustee is the settlor. The operating reality is that usually, an investment firm as account provider will already hold most assets of the description that a new account holder may seek to acquire, which means that most property received by an account provider for credit to the new account holder will become part of a pre-existing pool of assets of the same description. Note Briggs J’s observations in relation to the operating practices of LBIE in Pearson v Lehman Brothers Finance SA:¹⁴¹

The starting point is that it is misleading to think of LBIE’s house depot accounts as if they were simply one big omnibus account into which all its holdings of securities of any type were deposited, like some single bank account used for all its payments and receipts of cash. However they were recorded, the reality is that LBIE held a great multiplicity of positions on house depot accounts, one position for each type of security, by which I mean not one for equities and one for fixed income, but one for ICI ordinary shares, one for BP ordinary shares, separate positions for any different class of shares of the same issuer, and further separate positions for different classes of fixed income securities, again, separately for each issuer. The question of certainty must be addressed by looking, conceptually at least, at each position individually, even if they were all recorded in one account. LBIE’s depot account position for ICI ordinary shares is a convenient (if now purely historical) example. It would consist at any given moment in time of a chose in action against LBIE’s depository in relation to a specified number of ordinary shares, and would consist of a beneficial co-ownership interest in the depository’s total holding of shares of that type.

3.67  If the new account holder is treated as a settlor, any transfers to the account provider for its benefit should accrue to the trust property of a previous settlement and be held on the trust terms thereof. The previous settlement, in those (p. 121) circumstances, is the first transfer of securities of a particular description into the hands of the account provider.¹⁴² It could be argued that, as the account terms clarify that the account provider is authorized to pool the received property with property of other, existing account holders, the collective account terms operate to the effect that, at each new contribution to the pool of assets of a certain description, that pool is treated as a new mixture in which all contributing account holders acquire a fresh co-ownership share.

3.68  Support for a purposive interpretation of the collection of account holders’ terms as constituting the terms and conditions for the holding of collective client assets of the same description may be found in the operative principles for the recognition of a proprietary interest of B as beneficiary of a trust impressed on property where A has the legal or superior title, outlined by Briggs J in Pearson v Lehman Brothers Finance SA:¹⁴³

1. iv)  A trust does not fail for want of certainty merely because its subject matter is at present uncertain, if the terms of the trust are sufficient to identify its subject matter in the future.

2. v)  Subject to the issue of certainty, the question whether B has a proprietary interest in the property acquired by A for B’s account depends upon their mutual intention, to be ascertained by an objective assessment of the terms of the agreement or relationship between A and B with reference to that property.

3. vi)  The words used by the parties such as ‘trust’, ‘custody’, ‘belonging’, ‘ownership’, ‘title’, may be persuasive, but they are not conclusive in favour of the recognition of B’s proprietary interest in the property, if the terms of the agreement or relationship, viewed objectively, compel a different conclusion.

4. vii)  The identification of a relationship in which A is B’s agent or broker is not conclusive of a conclusion that A is, in relation to the property, B’s trustee, although it may be a pointer towards that conclusion.

5. viii)  A relationship which absolves A from one or more of the basic duties of trusteeship towards B is not thereby rendered incapable of being a trustee beneficiary relationship, but may be a pointer towards a conclusion that it is not.

6. ix)  Special care is needed in a business or commercial context. Thus:

   1. (a)  The law should not confine the recognition and operation of a trust to circumstances which resemble a traditional family trust, where the fulfilment of the parties’ commercial objective calls for the recognition of a proprietary interest in B.

   2. (b)  The law should not unthinkingly impose a trust where purely personal rights between A and B sufficiently achieve their commercial objective.

(p. 122) 3.69  Principles v), vi), and ix) recognize that the context and purpose of the account provider’s business should be considered in determining the effect of the terms and conditions of the account agreements pursuant to which the account provider receives, holds and administers, and transfers client assets. As noted in the context of the statutory trusts impressed on client money pools,¹⁴⁴ it is difficult to defend that receipts that are clearly identifiable as not meant to be owned by the recipient, but by the recipient’s client, should be the recipient’s and not the client’s property if there is an agreement between the client and the recipient that is clear in its purpose to impress a trust on such receipt. In line with Lord Dyson’s observations in the context of the client protection policy of CASS and MiFID, to exclude identifiable client assets from the trust based on ‘the happenstance’ of operational decisions by the trustee could be considered ‘arbitrary’.¹⁴⁵

3.70  If the custodian is considered to be the settlor and is bound as trustee pursuant to a covenant to declare a trust, any property received after the trust is first settled constitutes after-acquired property, which means that the account holder will not automatically acquire an interest in that property, but only after the custodian has also declared the trust in relation to that property.¹⁴⁶ This may be accomplished by any appropriate manifestation. A credit to an account should certainly suffice. It follows that each declaration must relate to the whole of the pooled property and each account holder’s share in it, so that upon the declaration, the after-acquired property merges with the existing pool and all account holders acquire a new share. Each transferee account holder, in turn, acquires an interest in the existing fund pro rata to its new share. The account provider’s covenant to declare a trust in respect of receipts should be enforceable under the account terms as the account holder is not a volunteer but a paying customer. An issue could arise in the event of insolvency if the account holder cannot seek specific performance of the covenant. That could be remedied if it is accepted that the account terms constitute a trust of the benefit of the covenant itself which allows for specific performance notwithstanding the account provider having become the subject of insolvency procedures.¹⁴⁷

3.71  It may be concluded that an account holder will acquire several different rights pursuant to the terms and conditions of a securities account, which are both personal and proprietary in nature. For this reason, it is common to describe the custody account holder’s rights as a ‘bundle’ of rights that is hybrid in nature.¹⁴⁸(p. 123) Nevertheless, this does not explain the nature of the bundle of rights. The rights in personam and the proprietary rights co-exist, as the term ‘bundle’ seems to suggest, but this proposition is troublesome. Proprietary rights cannot exist or arise independently from the rights in personam. In fact, the proprietary rights are predicated on the existence and the scope of the claim in personam against the account provider, which arise upon a credit pursuant to the terms and conditions of the account agreement. A claims-based entitlement in the client asset pool means that the account holder must establish that claim to exercise proprietary rights in respect of the pool, while the size of the share in the pool is determined by reference to the scope of the right in personam against the account provider.

3.72  The predication may best be demonstrated by the account holder’s position in the event of shortfalls. Due to operational complexity and mistakes, the records and subsequent statements of account might reflect the position inaccurately. Consequently, an account provider might give notice of a credit to the transferee without having received the asset. It must be assumed that contractual liability arises for the account provider if the erroneous entries cannot be validly reversed under the applicable terms and conditions, and are otherwise caused by a breach of duty by the account provider. On those facts, the account provider will also become liable to the collective account holders to cover any shortfalls in the pool. Nevertheless, in the event of account provider insolvency, the question of protection will turn on whether the erroneous and irreversible entry affected the account holder’s co-ownership share in the pool so that the shortfall is spread among all co-owners. Logic dictates that the loss should be borne by the account holder if the error can be traced back to a specific account. If not, provided the account holder received in good faith without notice of the error, the rules suggest that all account holders ought to share equally in the shortfall, pro rata to their co-ownership share. If that is true, then it must be accepted that the erroneous notice of credit to an innocent account holder creates property rights and redistributes co-ownership shares in the pool, much in the way that an in-house transfer does. It would appear to be the position in the LBIE cases discussed above.

3.73  Clearly, due to the operational framework of the financial system built around the fungible nature of securities, the rights created pursuant to a balance standing to the credit of an account in accordance with the terms and conditions of the account agreement are, typically, of a different order than the rights in the original securities, even if the account provider has received the original securities. Exercise of account holder rights depends on the ability to compel the account provider to perform its duties under the account agreement. The size of the share in the pool is not determined absolutely, but as a function of the scope of the rights in personam against the account provider, as was borne out in the LBIE cases discussed above and has been the system in the US since the revision of Article 8 UCC.

3.74  On-exchange transactions in listed securities and derivatives, and over-the-counter (OTC) transactions in certain designated swaps and forwards are subject to central clearing through CCP structures, which means that the original trade ends up as two mirroring contracts in the hands of the CCP. After the trade clearing process has completed, the CCP will be the counterparty to each clearing member who assumed responsibility for the original trade. CCP systems significantly mitigate the default risk in respect of the payment and/or delivery obligations under the original trade. A CCP becomes counterparty to trades with market participants through novation, an open-offer system, or through an analogous legally binding arrangement.¹⁴⁹

3.75  CCP systems originated by necessity in venues that wished to organize on-exchange derivatives trading. In a derivatives market, the CCP system enables the ‘trading’ of the derivative contracts because centralization permits the closing-out of long and short positions. Where an investor has a ‘long position’ through an earlier purchase of, for example, index futures and wishes to ‘sell’ one or more contracts, the broker will not seek a buyer for the existing contracts. Settlement would require a written novation agreement regarding the transfer of the existing contract between seller and buyer. Rather, the broker will seek to sell the same number of that type of standard contract and create a ‘short position’ for the investor. If the trade is successful, the long position in the hands of the CCP can be closed out at the instruction of the broker against the short positions, as the contracts will all have the same standard terms and expiration date. Only the price will differ, which will translate into a profit or loss to the investor upon close-out. When the operators of exchange-based spot markets started focusing on reduction of counterparty, settlement, and systemic risks, CCP systems were introduced in the spot markets as well.

3.76  In case of an order-driven trading venue that relies on broker-facilitated transaction execution, buy or sell orders are executed between brokers that are admitted to trading by the market operator in accordance with the rules and procedures of the trading venue. By accepting an order, the broker undertakes to access the trading venue and execute the order for the account of the investor by entering into a transaction with another broker on behalf of the investor, subject to and in accordance with the rules of that trading venue.¹⁵⁰ This structure is common (p. 125) to the operating model of exchange-based trading systems in the spot and derivative markets. The operating rules will ordinarily seek to deal with the fact that the broker acts pursuant to an agency mandate by requiring that the brokers are personally liable for the contract made, to the exclusion of the client who gave the order. Only the broker will have rights and liabilities on the contract. The client is excluded from that contract and consequently, cannot sue, nor be sued, nor otherwise intervene on it.¹⁵¹ Any right of intervention would undermine certainty of execution, and thus market liquidity, and could potentially impair the clearing model.

3.77  Figure 3.2 shows the basic CCP clearing structure for on-exchange spot or derivatives trades where the buyer’s broker is not also a clearing member and the seller’s broker is also a clearing member.

• View full-sized figure

Figure 3.2  CCP clearing of on-exchange trades

3.78  In CCP systems framed to support exchange-based trading, each broker that is admitted to the exchange must maintain an account with the CCP as clearing member, or must have a clearing agreement with a clearing member. Under the terms of the account, the CCP will provide CCP-based clearing services. Immediately after a trade has been executed and confirmed,¹⁵² the clearing members that are responsible for clearing the trade and the CCP enter into a tri-party novation agreement under which the original contract is cancelled and replaced by two identical contracts. Through this process, the CCP becomes the counterparty to the clearing members under each of the replacement contracts. The terms of the replacement contracts are otherwise identical to the terms of the original (p. 126) contract. The process of trade confirmation and contract novation, including any netting of settlement obligations booked in the clearing accounts, is called ‘clearing’, and the CCP is typically referred to as the ‘clearing house’. Accordingly, the accounts at the CCP are called ‘clearing accounts’.

3.79  If the broker is not eligible to be admitted as a clearing member, the broker will have to enter into a clearing agreement with a firm that is admitted as clearing member and that is willing to act as clearing member for the broker. In the clearing agreement, the clearing member will undertake to become a party to each contract made by the broker, and to each subsequent novation agreement with the CCP so that, upon completion of the novation agreement, the clearing member and not the broker is the party to the substitute contract with the CCP. The clearing member becomes a party to the contracts made by the broker based on a power of attorney for the benefit of the broker that is provided for in the clearing agreement. The broker acts as agent on behalf of the clearing member when it enters into the trade and the subsequent novation agreement.

3.80  In the period 2007–08 it became clear that the nature and quantity of OTC derivative trading was not transparent to regulators. It transpired that OTC derivatives had created a complex web of bilateral contracts that was difficult to measure or control and turned out to present significant systemic risk. The lack of transparency increased uncertainty during the period of market stress and, accordingly, undermined financial stability. At their September 2009 summit in Pittsburgh, G20 leaders agreed that all standardized OTC derivative contracts should be cleared through a CCP by the end of 2012, and that OTC derivative contracts should be reported to trade repositories.¹⁵³ In June 2010, G20 leaders in Toronto reaffirmed their commitment. In response the EU introduced the European Market Infrastructure Regulation (EMIR).¹⁵⁴ Accordingly, derivatives (p. 127) that are within scope of the clearing obligation of Article 4 of EMIR must be cleared via a CCP system and reported. To be within scope, a type of derivative must have been declared to be so by the Commission in accordance with Article 5(2) of EMIR,¹⁵⁵ and must be between institutions within the meaning of Article 4(1)(a) of EMIR. If a type of derivative is not brought within the scope of the clearing obligation, qualifying parties are still subject to mandatory risk mitigation techniques prescribed by Article 11 of EMIR. The parties to the G20 in Pittsburgh also agreed to pair the clearing obligation with a trading obligation in an effort to move trading in standardized OTC derivative contracts to exchanges or electronic trading platforms where appropriate, taking relatively lower liquidity of OTC derivatives into account.¹⁵⁶

3.81  The CCP clearing system for OTC derivatives, in essence, works the same as the system used for on-exchange derivatives, that is, the CCP is interposed between the parties to the OTC derivative transaction. If one of the parties or neither party is a clearing member, the non-clearing member party must have a clearing agreement with a clearing member of an EMIR eligible CCP.¹⁵⁷ The clearing agreement or agreements will typically permit the ‘give up’ of the original transaction to the clearing member, based on agency authority embedded in the clearing agreement. That means that if the clearing member accepts the transaction, the clearing member, and not its client, is the principal to the original transaction. Between the two clearing members the transaction will be novated to become two mirroring swap transactions in the hands of the CCP, so that the CCP becomes Party A to original Party B’s clearing member and Party B to original Party A’s clearing member. This is made possible because the OTC swap and forward transactions are documented under standard market agreements such as the International Swaps and Derivatives Association (ISDA) Master Agreement, permitting the reproduction of the original transactions on identical terms under the standard master agreement. It also explains why not every OTC transaction is eligible for mandatory clearing, as more bespoke transactions cannot be replicated in the CCP system using standardized terms.

3.82  Figure 3.3 shows the basic CCP clearing structure for an OTC transaction where Party A to the OTC transaction is not also a clearing member and Party B is also a clearing member.(p. 128)

• View full-sized figure

Figure 3.3  CCP clearing of OTC transactions

3.83  It is often said that a CCP ‘guarantees’ settlement. This term is not used in a strict legal sense, as the CCP is not a surety or guarantor of another person’s obligations. The CCP assumes obligations as principal. The original trade is split into two and the CCP is interposed as riskless principal between the clearing members.

3.84  The CCP structure brings substantial credit risk mitigation benefits:

1. a)  First, the clearing members must meet, and must continue to meet, certain financial conditions to become and remain eligible for admission.¹⁵⁸

2. b)  Second, in a spot market for securities, all securities-transfer and payment obligations arising under the contracts between the clearing member and the CCP relating to the same assets and the same settlement date can be netted on a rolling basis so that only the net balance of transfers and payments will be due between the CCP and each clearing member in respect of contracts in the same category on a settlement date for those contracts.

3. c)  Third, the clearing terms in both spot and derivative markets will impose collateral and cross-guarantee obligations on clearing members. Each clearing member will be required to provide initial margin and variation margin collateral to secure payment and securities delivery obligations to the CCP.¹⁵⁹

4. d)  Fourth, clearing members may be required to give cross-guarantees to the CCP in the form of an undertaking to pay on demand any and all amounts due from any of the other clearing members to the central counterparty that remain unpaid after the central counterparty has realized and applied the collateral provided by that clearing member. The cross-guarantees will typically have to be secured by collateral payable to a general default fund.¹⁶⁰

These features insulate a default of a clearing member and mitigate the systemic risk that a default of one clearing member cascades into defaults of other clearing members due to liquidity problems.

3.85  The question should be asked what the duties of the investment firm are in relation to the safeguarding of contracts cleared through a CCP system and related collateral. As discussed in paragraph 3.04 above, art 16(8) of MiFID II provides that an investment firm that holds ‘financial instruments belonging to clients, make adequate arrangements so as to safeguard the ownership rights of clients, especially in the event of the investment firm’s insolvency, and to prevent the use of a client’s financial instruments on own account except with the client’s express consent’.

3.86  ‘Financial instruments’ as used in Article 16(8) of MiFID II, per the definition of the term in Annex I, Section C, of MiFID II, includes contract-based financial assets, such as derivatives.¹⁶¹ The question may be asked whether the term ‘financial instrument’ in Article 16(8) was consciously used or whether the drafting process mainly considered primary and secondary financial assets but not contract-based rights.¹⁶² The reasons for the suspicion that the MiFID II drafting process did not fully consider contract-based financial instruments in the context of Article 16(8) are, first, the fact that preserving ownership rights in contract-based investments is anathema to civil law jurisdictions, as neither agency law nor property law will normally permit the creation of a proprietary interest in a contract made by an intermediate acting personally, with the intention to be a party to that contract. Investment firms in civil law jurisdictions, therefore, would struggle to preserve proprietary interests in contract-based investments, at least, under the generally operative concepts of private law, unless there were special legislative intervention. Second, the substance of Article 16(8) and the Level 2 rules in Articles 2ff of Delegated Directive (EU) 2017/593, as well as the restrictions on the use of title transfer collateral arrangements (TTCAs), undeniably aim at financial instruments that are securities. Finally, and most persuasively, as is the case in relation to the exemption for funds deposited with banks that are subject to authorization and regulation under CRD IV and CRR or equivalent regulation, MiFID II’s aim to safeguard investor’s financial assets is achieved through the client asset and position segregation, and default portability arrangements set out in Articles 39 and 48 of EMIR.

3.87  Article 39 (Segregation and portability) of EMIR provides in sub-paragraph (1) that a CCP must be able, ‘at any time and without delay’, to distinguish assets and positions held for the account of one clearing member from the assets and (p. 130) positions held for the account of any other clearing member and from its own assets. Article 39(2) and (3) addresses two permitted models for segregation of ‘the assets and positions’ of a clearing member ‘from those held for the accounts of its clients (“omnibus client segregation”)’, and ‘the assets and positions held for the account of a client from those held for the account of other clients (“individual client segregation”)’. The clients must be given a choice.¹⁶³ Subparagraph (4) directs each clearing member to maintain internal records that ‘enable it to distinguish both in accounts held with the CCP and in its own accounts its assets and positions from the assets and positions held for the account of its clients at the CCP’. In the event that a clearing member defaults, its clients’ interest in their collateral assets and positions are protected by the operation of the portability mechanics of Article 48 (Default procedures) of EMIR. Subparagraph (4) prescribes that a CCP must ‘take all reasonable steps to ensure that it has the legal powers to liquidate the proprietary positions of the defaulting clearing member and to transfer or liquidate the clients’ positions of the defaulting clearing member’; subparagraphs (5) and (6) address the CCP’s portability responsibilities in relation to assets and positions subject to omnibus client segregation or individual client segregation.

3.88  Accordingly, the assets and positions can be removed from a clearing member’s insolvency. Recital (64) of EMIR observes in this context that (emphasis added):

CCPs should keep updated and easily identifiable records, in order to facilitate the transfer of the positions and assets of a defaulting clearing member’s clients to a solvent clearing member or, as the case may be, the orderly liquidation of the clients’ positions and the return of excess collateral to the clients. The requirements laid down in this Regulation on the segregation and portability of clients’ positions and assets should therefore prevail over any conflicting laws, regulations and administrative provisions of the Member States that prevent the parties from fulfilling them.

3.89  The United Kingdom has established a special market-default insolvency regime in Part VII (Financial Markets and Insolvency) of the Companies Act 1989. Part VII has been substantially revised after it became clear through the insolvency of Lehman Brothers that the then-prevailing default regime of Part VII did not take account of all material relationships. Part VII applies, inter alia, to ‘clearing member client contracts’, meaning a contract between a recognized CCP¹⁶⁴ and a clearing member, indirect client, or a client ‘which is recorded in the accounts of the recognized central counterparty as a position held for the account of a client, an indirect client or a group of clients or indirect clients’,¹⁶⁵ and to ‘qualifying collateral and transfer arrangements’ meaning, in summary, CCP margin (p. 131) collateral and default fund collateral transfers.¹⁶⁶ Section 159 aims to ensure that the proceedings of recognized CCPs precedence over insolvency procedures by confirming that the default rules of a recognized CCP, including the rules on which the recognized CCP relies to give effect to the transfer of positions or settlement in accordance with its default rules, or the transfer of a qualifying collateral or property arrangement ‘shall be regarded as to any extent invalid at law on the ground of inconsistency with the law relating to the distribution of the assets of a person on bankruptcy, winding up or sequestration, or in the administration of a company or other body or in the administration of an insolvent estate’, and that an insolvency administrator shall not exercise any powers in such a way as to prevent or interfere with the operation of the default rules.

3.90  Notwithstanding the protective mechanism of the CCP default rules and Part VII described above, and keeping the safeguarding requirement of Article 16(8) of MiFID II in mind, the question may still be asked to what extent the investor can claim a proprietary interest in the cleared contract or collateral received and held by the clearing member on an omnibus or individual client segregation basis. The rules of the trading venue and/or the CCP rules will stipulate that the investor is excluded from a contract made by a broker on exchange, and any cleared contract that is recorded as a position by the CCP. Although on such exclusion, the investor cannot sue or intervene otherwise, as each cleared contract is a contract the clearing member made as agent for the investor, the agency relationship may entitle the investor to proprietary remedies in respect of the contracts made pursuant to the mandate. Legal title to the property in the contract vests with the clearing member. Proprietary rights must therefore rest on the application of the principles of equity to the effect that at the time the contract was made the agent also became bound to deal with the contract so made as trustee for the benefit of the investor.

3.91  An investor’s proprietary claim to a contract made by a clearing member in its own name cannot be based on the assertion of an existing property right; the contract is created by the clearing member. Nor can it rest on the application of the doctrine of specific performance; at no time is the clearing member supposed to transfer the contract itself to the investor, only its proceeds. Neither can the claim rest on recognition of the title of the investor as a person who contributed to the purchase price, as settlement between investor and its broker and/or clearing member will be arranged on a delivery-versus-payment (DVP) basis.¹⁶⁷ In the absence of special circumstances, therefore, it is not likely that a constructive or resulting trust might be impressed on the contract made by the broker by operation of law.¹⁶⁸

(p. 132) 3.92  Accordingly, a proprietary remedy would have to be based on a consensual arrangement that gives rise to a trust. There ought to be no question as to the subject of that trust. A broker and the clearing member will be required to record that order and accurately track it through execution and clearance, including any novation to a CCP, so that, after execution and prior to settlement, all contracts made are clearly recorded as executed as agent on behalf of the investor.¹⁶⁹ It is perhaps tempting to suggest that the analogy between the concept of agency and the concept of the trust justifies a conclusion that an agent, who acquires legal title to property for the principal, always intends to do so as trustee, but this has never been the rule.¹⁷⁰ Mere agency, although a relevant factor in the construction of the brokerage terms, is not sufficient to establish intention to create equitable rights in the property acquired by the agent in its own name. Whether a common intention to create the relationship of trustee and beneficiary exists must turn on an interpretation of the entire trading and clearing relationships, objectively, against the background of all relevant circumstances, having regard to the overall business purpose of the transaction, not just the agency, and the reasonable expectations of the parties.¹⁷¹ The purpose of the trading and clearing service is for the broker and, subsequently, the clearing member to make a contract for the investor on a particular trading venue. The reasonable expectation of the investor is that the benefits of a cleared contract will accrue to the investor, which will typically regard an executed trade, as cleared, as its property. The clearing member, however, can only regard that contract as the property of the investor to the extent permitted by the rules and procedures of the CCP system concerned, which (p. 133) cannot be subverted by the agency relationship since, by their very nature, the terms of the broker and the clearing member’s services are subordinated to the rules and procedures of the relevant trading venue and CCP.

3.93  Based on these considerations, it could be argued that the common intention of the parties to a trading and/or clearing relationship, in general, might be construed as seeking to create equitable rights in respect of the contract made for the investor to the extent that the same would not subvert the purpose and intentions of the rules and procedures of the relevant trading venue and CCP system. Most importantly, any equitable rights should not interfere with netting or collateral and default arrangements. On these principles, extrication and insulation of assets and positions pursuant to the portability arrangements in a CCP’s default rules and Part VII of the Company’s Act 1998 in the event of clearing member insolvency, prima facie, would not appear to be inconsistent with the principle that only property owned absolutely by a company is part of its insolvency. Insolvency does not incapacitate a person from continuing to act as trustee in relation to trust property, or as agent.

3.94  The second stage of the investment cycle involves investment services and activities that concern the execution and settlement of transactions for the account of, or directly with, the investor. The services and activities result in obligations, whether contingent or not, either to make payments in a certain currency or to transfer financial assets of a certain type and description immediately or in the future, which denotes the third stage of the investment cycle. Securing payment or delivery obligations owed by an investor or counterparty pursuant to the execution of an investment service or activity, that is, taking proprietary interests in property of the debtor to support the satisfaction of the debt in case of default, can significantly reduce the capital requirements that apply to a creditor-bank or creditor-investment firm that are subject to prudential supervision pursuant to CRD IV framework, which is a material cost-of-business incentive for banks and investment firms alike. Taking proprietary security, also known as ‘collateralization’,¹⁷² therefore, is omnipresent in the investment cycle. It involves (p. 134) the use of account-based financial assets and money to secure the performance of a payment or securities delivery obligation arising pursuant to the relevant execution or settlement-oriented investment service or activity.

3.95  Five basic areas of collateralization may be discerned in the investment cycle:

1. a)  Bilateral OTC trades: a dealer makes OTC bilateral agreements with investors. These may be ‘cash’ or ‘spot’ trades, such as in fixed income securities or currencies, which are required to be settled within no more than so many business days after the trade date, the number of days depending on the trading venue. In that case, the dealer and the investor will typically assume credit risk for the settlement period. However, in case of longer term derivative or securities financing transactions,¹⁷³ the future payment and delivery obligations will be subjected to collateralization requirements. Derivatives and securities financing transactions usually require initial margin collateral and variation margin collateral determined at regular intervals until maturity or termination of the transaction, with completion settlement upon the earlier of maturity or termination.

2. b)  On-exchange trade execution: a broker who executes spot or derivative transactions on-exchange will need to comply with the CCP collateralization requirements, and in turn, will seek collateral from the investor as its client.

3. c)  CCP clearing: each clearing member will need to provide initial margin collateral and, on completion of trade clearing cycles, variation margin collateral.

4. d)  Account-based funding of settlement services: a bank or custodian that provides payment and/or securities settlement services will need to access payment and settlement systems for the account of the investor but as principal participant,¹⁷⁴ and may therefore incur multi-day or even intra-day settlement risk, even if the settlement is carried out on a DVP basis. The account agreement will typically stipulate that the credit balances in the account serve as collateral for the account provider’s settlement exposure.¹⁷⁵

5. e)  Account-based funding or arranging of securities financing transactions: an account-providing firm may offer to fund or arrange securities financing transactions, which are to be secured by the client’s assets credited to cash and securities accounts. This type of service is often part of a ‘prime brokerage’ service. The prime broker will provide cash or securities financing and in return, will seek a general lien over the accounts and the right to use the client assets credited to the securities accounts. An investment firm that uses custodied assets under a securities financing transaction—regardless of whether it is party to a securities financing transaction, acting as an agent, or if it concerns a (p. 135) tripartite transaction between a borrower a client and the investment firm—must adopt specific arrangements to ensure that the borrower provides appropriate collateral and that the firm monitors the continued appropriateness of such collateral and takes the necessary steps to maintain the balance with the value of the client safe custody assets.¹⁷⁶ The investment firm’s internal records must identify the clients that authorized the use of their custody assets, the securities financing transactions that have been effected for their account, and the assets used so as to enable the correct allocation of any loss.¹⁷⁷

3.96  In each case, some form of collateralization of financial assets or money will be deployed to secure the exposure. Two principle techniques exist: ‘classic’, security interest-based collateralization, and outright transfer-based collateralization. A security interest in account-based assets seeks to encumber the interest arising as a matter of the balance standing to the credit of the account from time to time. Money standing to the credit of a bank account is a simple debt of the bank as account provider. Book-entry securities, or indeed, client money credited to an account provided other than as a banker, are a more complex hybrid of in personam rights against the account provider and equitable co-ownership interests in the pool of client assets held by the account provider.

3.97  English law, strictly speaking, only knows four forms of consensual security interests: the pledge, the contractual lien, the mortgage, and the equitable charge. The pledge and the lien concern tangible assets, including tangible instruments of title. Both are possession-based rights, although the latter only in relation to goods that have been delivered to the creditor for some service purpose, such as repairs. Also, the contractual right of lien, unlike the right of pledge, does not include a right of sale on default; only a right to detain the goods until the service debt has been paid, so a right to sell must be stipulated contractually. A mortgage is a transfer of ownership of tangibles as well as intangibles, by way of security, and is non-possessory. Equitable mortgages permit the transfer of future assets. An equitable charge, which may be fixed or floating, involves neither possession nor transfer of title, but instead, operates as an encumbrance on an asset, which can be appropriated and sold to discharge indebtedness.¹⁷⁸ Accordingly, liens and pledges technically cannot operate on intangible account-based assets. Curiously, financial services transaction documentation, in particular account documentation, stubbornly uses terms such as ‘general lien’ or pledge, but what is meant is a charge. Briggs J had to decide on the meaning of the word ‘general lien’ in the context of LBIE master custody account documentation (MCA), that used the (p. 136) term purportedly to create a security interest in book-entry securities, that is, intangible property, in the LBIE Lien case:¹⁷⁹

In my judgment the answer to Issue 1(a) is that the MCA did not, either as between LBIE and its intended street clients, still less between LBIE and [another group company,] LBF, create a general lien in the strict sense. It seems to me highly improbable and therefore most unreasonable to attribute to commercial parties an intention to create security rights of a type incapable of applying to the overwhelming bulk of the property likely to be held as custodian by LBIE under its standard form MCA. Whether the phrase ‘general lien’ was intended to displace any risk that the security right might be specific rather than general, or merely a hangover from the days when most securities were in certificated rather than de-materialised form, I do not need to decide. It is obvious that the security right was not intended to be transaction specific, since it was expressed to extend over ‘all … Property’ held under the MCA, and to constitute security for all liabilities and obligations of the client.

He further concluded that the terms of MCA ‘clearly permitted LBF a right to substitute or withdraw excess property from the ambit of the charge (pending crystallization) so that, pursuant to the analysis of the House of Lords in National Westminster Bank plc v Spectrum Plus Ltd,[¹⁸⁰] it could not be a fixed charge’.¹⁸¹

3.98  Equally, the mortgage as an instrument that requires transfer of title is a less practical instrument in relation to account-based assets as the account is in constant use. Accordingly, in practice, as demonstrated in the LBIE Lien case, security-interest based collateralization takes the form of a charge over the relevant account or accounts and is typically provided to the account provider, rather than a third party, to secure obligations arising under account-based settlement services and account-based transaction financing mentioned under d) and e) in paragraph 3.95 above. Separately, a broker will normally seek to take a charge over any cash or securities received pursuant to the settlement of a transaction executed for a client, prior to the client funding the execution of the trade.

3.99  The second form of collateralization is outright transfer based, known as a title transfer collateral arrangement, or ‘TTCA’. TTCAs rely on the fungibility of financial assets and money, and creates a contractual netting arrangement that is based on mutual obligations to transfer—not by way of security, but by way of unencumbered ownership—financial assets or money to the collateral taker who will be under return equivalent financial assets or money when the secured obligation has been performed. The contractual netting provisions constitute netting by novation and are structured as rolling netting or close-out netting. Rolling netting aims at extinguishing debts arising pursuant to a specific transaction or set of transactions, replacing mutual settlement obligations with a single settlement obligation, thus reducing the number of settlements required to complete (p. 137) a transaction. Close-out netting extends to all outstanding obligations that have arisen or may arise under the contractual arrangement and, on the occurrence of an enforcement event—typically an act of insolvency or a material and continuing default—aims to accelerate or terminate all transactions, determine their current value, and calculate a net sum equal to the balance of account due between the parties. The only sum due is that which is due from the party that owes the net balance. All securities financing transactions documented under standard market terms rely on and constitute TTCA-based transactions. Bilateral OTC derivative transactions documented under standard market terms, such as the ISDA Master Agreement, also rely on novation netting, but do not qualify as TTCA unless they are liquidity swaps,¹⁸² or, if not, include what is known as ‘credit support’. Credit support in respect of transactions documented under an English law-governed ISDA Master Agreement is structured as a TTCA that is documented as a separate transaction and, therefore, will be swept up into the calculation of the balance of account between the parties if the close-out provision becomes operative.

3.100  CCP terms and conditions usually rely on rolling novation netting, supported with initial and variation margin supplied by the clearing member by way of a TTCA, so that close-out netting operates in the event of a clearing member default event occurring, particularly insolvency. Investors who acquire an on-exchange traded derivative or an OTC derivative that is cleared through a CCP system will be required to provide collateral to the clearing member, who will in turn provide collateral to the CCP, although on a net basis.

3.101  TTCAs transfer full and unencumbered ownership and, therefore, by their very nature, permit the collateral taker to use the collateral for subsequent transactions, leaving the collateral provider with a mere in personam right for return of equivalent assets, albeit subject to the netting provisions. The use of the TTCA technique should not cause significant risks to the collateral provider in the event of a securities financing transaction that is subject to a bespoke and regular variation margin collateral calculation that ensures the net balance is as close to nil as possible, taking any ‘haircuts’ into account.¹⁸³ The haircut reflects the volatility of the market price of the collateral so as to protect the collateral taker against ‘under collateralization’. Naturally, this could lead to ‘over collateralization’. The regular reset of the collateral value through the variation margin collateral mechanism should guard against a swing too far off the mark either way. The risk of the use of the TTCA technique is less transparent where the collateral is taken from a client’s securities account by an account provider subject to the terms of the account (p. 138) agreement. This is typical for prime brokerage arrangements, which stipulate that the balance in the securities account serves as collateral for any present and future claims the prime broker may have against the client in connection with the investment services and activities for which the client engages the prime broker. Unless the terms of the account agreement require the prime broker as account provider to document the use of the assets in the account by way of a separate securities financing transaction that collateralizes the prime broker’s obligation to return equivalent securities, the client may be looking at a meaningfully over-collateralized position, which represents a net in personam exposure to the prime broker.

3.102  In view of the fact that the TTCA technique may result in over-collateralization, MiFID II has introduced limits on the use of TTCAs, in particular in relation to retail clients. Recital (52) MiFID II observes:

The requirements concerning the protection of client assets are a crucial tool for the protection of clients in the provision of services and activities. Those requirements can be excluded when full ownership of funds and financial instrument is transferred to an investment firm to cover any present or future, actual or contingent or prospective obligations. That broad possibility may create uncertainty and jeopardise the effectiveness of the requirements concerning the safeguard of client assets. Thus, at least when retail client assets are involved, it is appropriate to limit the possibility of investment firms to conclude title transfer financial collateral arrangements as defined under [the Collateral Directive],¹⁸⁴ for the purpose of securing or otherwise covering their obligations.

3.103  Accordingly, Article 16(10) of MiFID II provides that an investment firm may not ‘conclude title transfer financial collateral arrangements with retail clients for the purpose of securing or covering present or future, actual or contingent or prospective obligations of clients’. This does not mean that the investment firm may not execute securities financing for and on behalf of the client. Recital (51) of MiFID II notes that the rule against retail client TTCAs for general present and future business transactions does not ‘prevent a firm from doing business in its name but on behalf of the investor, where that is required by the very nature of the transaction and the investor is in agreement, for example stock lending’. Ergo, an investment firm can still execute bespoke securities finance transactions for a client, and presumably also with a client.¹⁸⁵

3.104  Given the increased risk of loss in the event of default of the collateral taker that emanates from over-collateralized TTCAs, MiFID II also seeks to curb the use of TTCAs for professional clients outside bespoke securities financing transactions. Recital (6) of MiFID II Delegated Directive 2017/593 notes that there is ‘a risk (p. 139) that without further guidance investment firms could use TTCA more often than reasonably justified when dealing with non-retail clients, undermining the overall regime put in place to protect client assets’. Investment firms are therefore required to ‘consider the appropriateness of title transfer collateral arrangements used with non-retail clients by means of the relationship between the client’s obligations to the firm and the client assets subject to TTCA’ and should only use TTCAs ‘if they demonstrate the appropriateness of TTCA in relation to that client and disclose the risks involved as well as the effect of the TTCA on his assets’. Accordingly, Article 6(2) of MiFID II Delegated Directive 2017/593 requires an investment firm in determining whether a TTCA is appropriate to consider whether the link between the secured obligation and the collateral is sufficiently robust, including the likelihood of default,¹⁸⁶ whether there is a risk of material over-collateralization,¹⁸⁷ and whether the TTCA uses assets from a client pool that commingles assets from clients that owe the investment firm very different secured obligations.¹⁸⁸

3.105  Recital (7) confirms that the obligation to demonstrate ‘a robust link between collateral transferred under a TTCA and client’s liability’ does not interfere with an investment firm’s ability to ‘require a sufficient collateral and where appropriate, to do so by a TTCA’. The requirement of Article 6(2), therefore, ought not to conflict with the collateralization obligations arising under securities financing transactions or indeed CCP collateralization requirements arising under, for example, EMIR. However, it will restrict the scope for use of the custody portfolio for a firm that offers prime brokerage services. A prime broker offers a bundle of services linked to the opening of a cash and securities account. The Financial Conduct Authority’s (FCA) Handbook of Rules and Guidance defines ‘prime brokerage services’ as follows in the Glossary:

A package of services provided under a prime brokerage agreement which gives a prime brokerage firm a right to use safe custody assets for its own account and which comprises each of the following: (a) custody or arranging safeguarding and administration of assets; (b) clearing services; and (c) financing, the provision of which includes one or more of the following: (i) capital introduction; (ii) margin financing; (iii) stock lending; (iv) stock borrowing; (v) entering into repurchase or reverse repurchase transactions; and which, in addition, may comprise consolidated reporting and other operational support.

3.106  In summary, a prime broker will act as broker, dealer, and custodian, and will provide securities financing and clearing services for on-exchange and OTC trading. The prime broker combines traditional custody services with traditional broker-dealer services, and couples the custodian’s ‘general lien’, discussed in the section E2, with broad ‘use’ of the custodied assets by way of outright transfer. In those circumstances, as noted in para 3.101 above, the investor may be exposed to very substantial over-collateralization. A broad discretion to over-collateralize (p. 140) is subject to the regulatory obligation of Article 6(2) of MiFID II Delegated Directive 2017/593 and the investment firm will have to demonstrate that the arrangement is appropriate. In addition, in using that discretion and designing its services, the prime broker will have to consider its obligation under Article 24 of MiFID II to act fairly in the investor’s best interest.¹⁸⁹ The FCA has made specific reporting rules for prime brokers in CASS 9 to ensure that the way the custodied assets are used is transparent to the investor.

3.107  The enforceability of the collateral rights purportedly created in respect of collateral in the form of bank funds or book-entry securities had in the past been clouded in considerable uncertainty about the legal character of account-based assets and the way traditional consensual security interests and netting techniques operate on account based- assets and obligations to transfer account-based assets. Much of this uncertainty was reduced with the introduction of the Collateral Directive,¹⁹⁰ which is implemented in the United Kingdom through the Financial Collateral Arrangements (No 2) Regulations 2003 (FCAR).¹⁹¹ The Collateral Directive clarifies, with respect to ‘financial collateral’ within the meaning of the Collateral Directive, the formalities that apply to the creation of collateral rights;¹⁹² abolishes legal technical obstacles to enforcement¹⁹³ and the right to use;¹⁹⁴ confirms that TTCAs and the netting arrangements are valid;¹⁹⁵ and disapplies certain adverse insolvency provisions.¹⁹⁶ The Collateral Directive has created a harmonized EU legal framework for the receipt and enforcement of financial collateral and facilitates the cross-border use of financial collateral by abolishing many formal requirements such as the need to register the collateral, simplifying enforcement rules, clarifying that the law of the place of the account provider, also known as the ‘Place of the Relevant Intermediary Approach’ (PRIMA) rule, enhancing legal certainty as to which law applies to book-entry securities collateral in cross-border situations.¹⁹⁷

(p. 141) 3.108  The Collateral Directive operates on TTCAs¹⁹⁸ and consensual security interests¹⁹⁹ relating to ‘financial collateral’, which means ‘cash, financial instruments or credit claims’.²⁰⁰ ‘Cash’ is defined as ‘money credited to an account in any currency, or similar claims for the repayment of money, such as money market deposits’.²⁰¹ Confusingly—in view of the fact that MiFID II defines the term ‘financial instruments’ more broadly to include derivatives—the term ‘financial instruments’ is defined narrowly as securities, including securities that embed a derivative, as follows:²⁰²

[S]hares in companies and other securities equivalent to shares in companies and bonds and other forms of debt instruments if these are negotiable on the capital market, and any other securities which are normally dealt in and which give the right to acquire any such shares, bonds or other securities by subscription, purchase or exchange or which give rise to a cash settlement (excluding instruments of payment), including units in collective investment undertakings, money market instruments and claims relating to or rights in or in respect of any of the foregoing.

It must concern a collateral transaction between a collateral taker and a collateral provider who must be, broadly speaking, two financial institutions or a financial institution and a professional client,²⁰³ and it must seek to secure a ‘relevant financial obligation’,²⁰⁴ that is, a payment or securities settlement obligation.²⁰⁵

(p. 142) 3.109  Although the implementation of the Collateral Directive has done much to reduce legal uncertainty around the use of account-based assets as collateral and to simplify enforcement, naturally, given the operational complexity of the financial system, certain questions remain. The notion of ‘control’ or ‘possession’ provided for in the Collateral Directive can give rise to interpretation issues when held against the background of domestic law concepts of security. Article 1(5) Subparagraph 5 provides that the Collateral Directive ‘applies to financial collateral once it has been provided and if that provision can be evidenced in writing’. Article 2(2) clarifies that references in the Collateral Directive to financial collateral being ‘provided’, or to the ‘provision’ of financial collateral, are to be interpreted as a reference to the financial collateral ‘being delivered, transferred, held, registered or otherwise designated so as to be in the possession or under the control of the collateral taker or of a person acting on the collateral taker’s behalf’ (emphasis added). Importantly, Article 2(2) further clarifies that variation margin collateral arrangements do not fatally undermine the requirement of control or possession: ‘any right of substitution, right to withdraw excess financial collateral in favour of the collateral provider or, in the case of credit claims, right to collect the proceeds thereof until further notice, shall not prejudice the financial collateral having been provided to the collateral taker’.

3.110  The Recitals explain the balancing act that underpins the notion of control. Recital (9) notes that the Collateral Directive aims to ‘limit the administrative burdens’ for parties using financial collateral arrangements and therefore, that:

Recital (10) raises the concern that, notwithstanding the aim to reduce the administrative burden, the Collateral Directive must

Ergo, the notion of control, defined as some form of dispossession of the collateral provider in relation to the collateral, is driven by the need to balance the protection of the parties, and third parties, from the risk of fraud.

3.111  Briggs J had occasion to consider the notion of control of the Collateral Directive, as implemented in the United Kingdom through the FCAR, in the LBIE Lien case.²⁰⁶(p. 143) As discussed in para 3.97 above, Briggs J decided that the term ‘general lien’ in the context of LBIE MCA, coupled with the right to substitute collateral or withdraw excess collateral, was to be interpreted to create a floating charge of the client assets. He then had to consider whether the floating charge was saved by the FCAR from any of the forms of invalidity or challenge that are excluded thereby. The question centred on the notion of control. Conform his familiar methodology—given his many important LBIE decisions—the judge made several very helpful observations.

3.112  First, he noted that the FCAR add little to this debate, because of the use of almost identical language to that found in the Collateral Directive, save in one significant respect. In Regulation 3, the expression ‘security interest’ is defined by reference to five familiar categories, the fourth of which, at subpara (d) is a charge created as a floating charge where the financial collateral charged is delivered, transferred, held, registered or otherwise designated so as to be in the possession or under the control of the collateral-taker or a person acting on its behalf; any right of the collateral-provider to substitute equivalent financial collateral or withdraw excess financial collateral shall not prevent the financial collateral being in the possession or under the control of the collateral-taker. Briggs J observes:²⁰⁷

It is evident, at least at first blush, that HM Treasury did not in framing the regulations think that the requirement to show possession or control in the collateral-taker necessarily precluded floating charges from qualifying as relevant security interests, even though it is now well settled that it is precisely the absence of possession or control of the subject matter of an English charge by the chargee that makes it likely to be characterised as a floating rather than a fixed charge. Nonetheless the extensive debate about that question among lawyers and bankers was only resolved some years after the framing of the Regulations, in the Spectrum Plus case, in June 2005.

3.113  Next, he addressed the extempore judgment of Vos J in Re F2G Realisations Ltd, Gray v GTP Group Limited (Gray).²⁰⁸ Gray concerned a floating charge over a £100,000 balance standing to the credit of cash account in the name of the collateral taker who was not also the account provider. The plaintiff claimed the floating charge to constitute a security financial collateral arrangement under the FCARs so that it would not fail for want of registration. Vos J held that the claim under the FCARs failed because the terms of the charge were insufficient to confer control on the collateral taker even though the subject matter of the charge was held by the collateral taker in a bank account in its name. Briggs J observes (emphasis added):²⁰⁹

While acknowledging that the phrase ‘possession or control’ had to be given an independent meaning, [Vos J] held that possession had no sensible application to intangibles, and that there was an absence of control in the collateral taker because, pending crystallisation, the collateral provider had the legal right to require the collateral taker to deal with the deposited money at its direction. The absence of what he called (p. 144) legal or negative control in the collateral taker (due to it having no right to prevent the removal of the charged money from the account by the collateral provider) was in his view fatal to the claim under the FCARs, even though the collateral taker had what he called administrative control, in the sense that, on a day to day basis, it was able to deal with the money as the relevant account holder.

Vos J accepted that his conclusion ‘left little room for any floating charge (as understood in English legal terms) to qualify for protection under the FCARs, despite the express reference to it as a potentially qualifying species of security interest in Regulation 3’.²¹⁰ Briggs J considered, in light of academic and professional writing,²¹¹ that the common theme was that ‘although Vos J had recognized the need to identify an independent meaning for the phrase possession or control, he had then disregarded possession as a sufficient means of satisfying the test because of a narrowly English law view that possession was irrelevant to intangibles’.²¹²

3.114  The publication of the FMLC report on Gray²¹³ was followed by an amendment to the FCARs, which inserted a new Regulation 3.2 that confirmed that ‘possession’ of financial collateral in the form of cash or financial instruments

Briggs J observes that the amendment did little to clarify the situation:²¹⁴

Its relevance is only that in discharging its duty to understand and apply the [Collateral] Directive within the context of British law, HM Treasury acceded only to an immaterial and largely theoretical extent to the FMLC’s invitation to do something by way of ameliorating the restrictive effect of Vos J’s analysis.

3.115  Briggs J then proceeded to formulate his own conclusion, finding that he did not regard himself ‘bound by judicial comity to follow the decision of Vos J unless sure that it is wrong. The matter has been argued so much more fully before me, (p. 145) and by reference to so much more admissible travaux preparatoires, and subsequent academic and expert comment, that it is in my view both legitimate and necessary to address the question afresh’.²¹⁵ As he did in respect of the client safeguarding provisions of MiFID I and CASS 7 in the context of the LBIE client money case, LBIE v CRC,²¹⁶ he emphasized the need for a purposive interpretation of the Collateral Directive and FCAR:²¹⁷

That interpretation must be purposive, in the sense that it is responsive to such purpose as can be identified from the [Collateral] Directive as a whole and, in the present context, primarily from the second half of Recital 10. That explains that the balance between market efficiency and creditor protection is to be struck by limiting the scope of the [Collateral] Directive to financial collateral arrangements which provide for some form of dispossession, by which Recital 10 means ‘provision of the financial capital’ provided that the provision can itself be evidenced in writing or in a durable medium. Article 2.2 comprehensively defines what the Directive, and Recital 10 in particular, means by ‘provision’. Article 1.5 implements the requirement for evidence in writing, both in relation to the provision of financial collateral, and in relation to the financial collateral arrangement itself.

In Briggs J’s opinion, Article 2.2 of the Collateral Directive does not intend to limit ‘possession’ to ‘exclude any application to intangibles’. True to fact, he notes that ‘[i]ntangibles are, and were by the time the [Collateral] Directive was being prepared, the very stuff of modern financial collateral’, and concludes that, therefore, there is ‘some force in the criticism of Vos J’s analysis that control (whether legal or administrative, negative or positive) lies at the heart of the concept of possession or control’.²¹⁸ He concludes:²¹⁹

It follows that I broadly agree with Vos J’s analysis of the possession or control requirement, save for his apparent inclination to think that possession is of minimal relevance where the financial collateral takes the form of intangibles. In my judgment what needs to be shown (in order to bring a particular collateral arrangement within the protection of the FCARs), is that the terms upon which it is ‘provided’ (Article 2.2) or ‘delivered, transferred, held, registered or otherwise designated’ (Regulation 3) are such that there is shown to be sufficient possession or control in the hands of the collateral taker for it to be proper to describe the collateral provider as having been ‘dispossessed’ (Recital 10). There will be cases in which the collateral is sufficiently clearly in the possession of the collateral taker that no further investigation of its rights of control is necessary. In other cases, such as the case before Vos J, it will be necessary to analyse the degree of control thereby conferred on the collateral taker. There may be some cases, in particular where there is no delivery, transfer or holding to or by the collateral taker, but merely some form of designation, where the collateral remains wholly in the possession of the collateral provider, (p. 146) but on terms which give a legal right to the taker to ensure that it is dealt with in accordance with its directions.

3.116  All of this means in practice that if there is some minimum balance level beyond which the collateral provider cannot withdraw, the floating charge will be enforceable as an FCAR arrangement. LBIE’s MCA contained a proviso in Clause 9 that ‘[t]he Custodian shall have no obligation to deliver the Property of the Client where the Custodian believes that there may be insufficient Property in the Custody Account to cover any exposure that the Custodian has to the Client’, which Briggs J considered to be a ‘written explanation of the precise extent of the custodian’s right to retain the client’s property’.²²⁰ The ‘Client’ referred to in Clause 9 was Lehman Brother Finance SA, an affiliate, ‘LBF’. The floating charge created under the MCA, however, purported to secure liabilities not just of LBF as the Client, but also of other affiliates. Briggs J considering the complete body of secured obligations, not just the liabilities of LBF that were referred to in Clause 9, found that the floating charged failed because:²²¹

the distribution of the parties’ respective rights to it, pending crystallisation, are insufficient to satisfy the possession or control requirement. This is because LBF retained, pending crystallisation, uncontrolled rights of recall and disposal of the property held in custody to a substantially greater extent than rights of substitution or withdrawal of excess, treating ‘excess’ as referable to the whole of its liabilities for satisfaction of which the security existed. In short, leaving aside its debts to LBIE, LBF could do what it liked with the property, regardless of its liabilities to LBIE’s affiliates.

A cautionary tale for those who draft security arrangements, indeed. Considering the above analysis, it is no surprise that the result in Gray, though not the reasoning, was correct in Briggs J’s view. Floating charges of account balances that are accessible without sufficient restriction pending crystallization will not be sheltered under FCAR.²²²