6.01  It is conceptually possible for a market to consist of investors only. The investors would buy and sell from and to each other when the need arises to adjust their investment portfolios. However, today, markets would not be as liquid and efficient if the trading process were not facilitated by market participants who make it their business to intermediate and provide liquidity. Indeed, various markets would not exist without the intermediation of investment firms and banks that are willing to take either side of a trade on a regular and continuous basis. Thus, each marketplace divides into investors, the ‘buy-side’, and trading services providers, the ‘sell-side’. Investment managers, operating as agents for their clients as investors, are buy-side firms. Trades, therefore, are usually executed via a trading venue by a sell-side firm with or for a buy-side firm or with another sell-side firm Sell-side investment firms divide into firms whose regular business is to execute trades on the instruction and for the account of clients (brokers), and firms whose regular business is to execute trades for their own account with clients (dealers), which include investors as well as other sell-side firms.

(p. 236) 6.02  The MiFID II concept of ‘trading venue’ encompasses all facilities that on a regular and continuous basis offer multiple market participants, referred to as a ‘multilateral system’, the ability to seek to execute, directly or indirectly, a transaction in relation to a certain investment.¹ MiFID II divides the concept of ‘trading venue’ into three categories:

6.03  Accordingly, trading venues range from multilateral facilities offered by investment firms that do not operate subject to a formalized non-discretionary set of trading rules, OTFs, to multilateral facilities offered by authorized and regulated exchanges permitting the participant to organize (via a member of the exchange) the execution of a transaction in a formalized structure that operates with an organized trading system subject to sophisticated formal listing and trading rules that are policed by the exchange. The term ‘multilateral’ indicates that the trading venue (p. 237) in question facilitates simultaneous access by many market participants, allowing them to compete for transactions. Operation of an MTF or OTF is specified in the list of investment services and activities in Annex I, sub-para A of MiFID II and, accordingly, requires authorization by the competent regulator. RMs must be authorized and regulated by member states in accordance with Title III of MiFID II.

6.04  Originally, activities that are within the meaning of MiFID II’s definition of ‘MTF’ and ‘OTF’ were commonly referred to as ‘alternative trading systems’ (ATS), and although the investment firms that would offer ATS facilities would be regulated as a broker or a dealer, the ATS activity was not regulated as such. But ATS arrangements started to play an increasingly important role in the markets. For instance, in the equity markets (traditionally dominated by exchange-based trading), ATS facilitated the direct crossing of institutional orders outside the exchange, with the operator of the system acting as crossing agent, permitting the execution of relatively sizeable orders based on exchange prices, and thus limiting the market-impact cost normally associated with such orders. In the fixed-income markets (traditionally dominated by direct party-to-party, ‘over the counter’ (OTC) trading), ATS started to offer sophisticated multilateral trading facilities where dealers offer liquidity by giving binding quotes on a continuous basis subject to trading rules. Such ATS not only permitted the immediate execution of trades based on the available quotes, but by implication also have a price-discovery and price-forming function. MiFID I first brought some forms of ATS into the regulatory scope as ‘MTFs’, and MiFID II has increased the reach through the inclusion of OTFs.⁵ The operator of an RM, MTF, or OTF may not in its capacity as operator buy or sell on own account,⁶ and, subject to the trading rules, facilitates the execution of transactions between participants in the system as riskless principal, agent, or arranger.

6.05  The definition of ‘OTF’ is not meant to include ‘facilities where there is no genuine trade execution or arranging taking place in the system, such as bulletin (p. 238) boards used for advertising buying and selling interests, other entities aggregating or pooling potential buying or selling interests, electronic post-trade confirmation services, or portfolio compression’.⁷ Recital (9) of MiFIR confirms that OTFs complement RMs and MTFs, but continues to clarify that, unlike RMs and MTFs, which have non-discretionary rules for the execution of transactions, ‘the operator of an OTF should carry out order execution on a discretionary basis subject, where applicable, to the pre-transparency requirements and best execution obligations’, and consequently, that ‘conduct of business rules, best execution and client order handling obligations should apply to the transactions concluded on an OTF operated by an investment firm or a market operator’. Accordingly, Recital (9) observes that

Because an OTF constitutes a genuine trading platform, the platform operator should be a market neutral infrastructure provider to permit open access. Therefore, the investment firm or market operator operating the OTF should be subject to requirements in relation to non-discriminatory execution and neither the investment firm or market operator operating the OTF, nor any entity that is part of the same group or legal person as the investment firm or market operator should be allowed to execute client orders in an OTF against its proprietary capital.

6.06  Where the supplier of a trading facility does not strictly transact with market participants on own account but crosses an order internally with orders of other clients of the firm, or clients of the group to which the firm belongs, it can be challenging to distinguish between a bilateral trading facility that is essentially a gate into an order flow, and an MTF or OTF. Given the definitions, the outcome will depend on the nuances of the order processing rules and the extent to which the market participants can access the crossing system directly or are dependent on the discretion of the provider of the facility. This is of relevance, for instance, in so-called dark trading pools, which are arrangements that cross internal order-flow (p. 239) in equity instruments based on reference prices taken from representative price-forming markets. The reference to the pool being ‘dark’ indicates that the operators of the pools do not offer pre-trade price transparency.⁸ To assuage market participant concerns about potential abuse of conflicts of interest that arise if an agent acts for two principals, the dark pool will operate based on a formal crossing methodology, which moves the arrangement closer to the MiFID II definition of MTF. At its most basic level, this activity constitutes bilateral agency-crossing and it could be argued that, therefore, a dark pool operated by a single broker does not constitute an MTF because it is not truly multilateral. However, where several brokers combine their trade flow into a single pool, it will be difficult to maintain that the system does not permit multiple market participants to execute orders via a trading system that operates as an MTF based on formalized, that is, non- discretionary, trading rules, or otherwise as an OTF.

6.07  The term ‘bilateral’ indicates that the trading arrangement consists of only two parties, the sell-side liquidity provider and the buy-side market participant who wishes to execute a trade, which does not facilitate simultaneous competition for transactions. Together, the liquidity providers in a certain type of investment form the ‘market’ for that investment, and where an investment firm holds itself out ‘on the financial markets on a continuous basis as being willing to deal on own account by buying and selling financial instruments against that person’s proprietary capital at prices defined by that person’, such an investment firm is a ‘market maker’,⁹ and such bilateral trading arrangements are referred to as ‘OTC trading’. If the party that wishes to execute a trade requires an element of price competition, that party will have to contact several dealers independently to compare prices. Given that prices move continuously, there necessarily is only a very narrow window for price comparison, minutes at best, which has been a reason for the development of new ways of trading through MTFs and OTFs.

6.08  To capture trade flow that would be directed through a RM, MTF, or OTF, but for the fact that the investment firm fills the order from its own book, MiFID I introduced, and MiFID II continued, the concept of ‘systematic internalizer’ (SI). Recital (17) of MiFID II explains how SIs stack up against MTFs and OTFs in the regulatory spectrum of trading venues (emphasis added):

Systematic internalisers should be defined as investment firms which, on an organised, frequent, systematic and substantial basis, deal on own account when executing client orders outside a regulated market, an MTF or an OTF. In order to ensure the objective and effective application of that definition to investment firms, any bilateral trading carried out with clients should be relevant and criteria should be (p. 240) developed for the identification of investment firms required to register as systematic internalisers. While trading venues are facilities in which multiple third party buying and selling interests interact in the system, a systematic internaliser should not be allowed to bring together third party buying and selling interests in functionally the same way as a trading venue.

6.09  In non-MiFID parlour, an SI is a firm that makes a market outside an RM, MTF, or OTF. This could be as a dealer or as a broker, if that broker, as part of its brokerage business model, regularly fills client orders from its own book instead of executing on an RM, MTF, or OTF.¹⁰ MiFID I sought, and the MiFID II continues, to capture that liquidity flow under its trading venue framework. Article 4(1)(20) of MiFID II clarifies what is meant by ‘frequent and systematic basis’:

The frequent and systematic basis shall be measured by the number of OTC trades in the financial instrument carried out by the investment firm on own account when executing client orders. The substantial basis shall be measured either by the size of the OTC trading carried out by the investment firm in relation to the total trading of the investment firm in a specific financial instrument or by the size of the OTC trading carried out by the investment firm in relation to the total trading in the Union in a specific financial instrument. The definition of a systematic internaliser shall apply only where the pre-set limits for a frequent and systematic basis and for a substantial basis are both crossed or where an investment firm chooses to opt-in under the systematic internaliser regime.

6.10  Not every dealer or broker that offers liquidity in listed instruments is an SI. Those who do not meet the materiality and structural thresholds are not in scope, although may choose to opt in. To meet the threshold, the business must be material both in total EU market terms and relative to the liquidity provider’s other business and be carried on by specialist personnel or through an automated system dedicated to that business. The European Securities and Markets Authority (ESMA) publishes market data for comparison. The trading facility must be available to market participants on a regular or continuous basis and be carried on in accordance with certain non-discretionary trading rules and procedures. In other words, the facility must have the characteristics of a trading venue that offers continuity. To the extent that a trading facility is offered on an ad hoc and irregular basis only, and the transactions are executed outside the usual systems of the liquidity provider and may be characterized as above-standard market size, the activity is not treated as systematic internalization.

6.11  Recital (18) of MiFIR clarifies the rationale for the SI regime:

In order to ensure that trading carried out OTC does not jeopardise efficient price discovery or a transparent level playing field between means of trading, appropriate pre-trade transparency requirements should apply to investment firms dealing on own account in financial instruments OTC insofar as it is carried out in their (p. 241) capacity as systematic internalisers in relation to shares, depositary receipts, ETFs, certificates or other similar financial instruments for which there is a liquid market and bonds, structured finance products, emission allowances and derivatives which are traded on a trading venue and for which there is a liquid market.

Importantly, investment firms who execute trades via SI organized trading arrangements will fulfil their trading obligations under Articles 23 and 28 of MiFIR, discussed next.

6.12  In pursuit of its objective to move trading as much as possible onto organized platforms that are subject to transparency and reporting obligations, MiFIR introduced ‘trading obligations’. Recital (11) of MiFIR outlines the rationale:

In order to ensure more trading takes place on regulated trading venues and systematic internalisers, a trading obligation for shares admitted to trading on a regulated market or traded on a trading venue should be introduced for investment firms in this Regulation. That trading obligation requires investment firms to undertake all trades including trades dealt on own account and trades dealt when executing client orders on a regulated market, an MTF, a systematic internaliser or an equivalent third-country trading venue. However an exclusion from that trading obligation should be provided if there is a legitimate reason. Those legitimate reasons are where trades are non-systematic, ad-hoc, irregular and infrequent, or are technical trades such as give-up trades which do not contribute to the price discovery process. Such an exclusion from that trading obligation should not be used to circumvent the restrictions introduced on the use of the reference price waiver and the negotiated price waiver or to operate a broker crossing network or other crossing system.

6.13  Accordingly, Article 23 (Trading obligation for investment firms) of MiFIR requires each investment firm to ensure that trades in shares admitted to trading on an RM or that are traded on an MTF or OTF shall be executed on an RM, MTF or via an SI, or a third-country trading venue assessed as equivalent in accordance with Article 25(4)(a) of MiFID II,¹¹ as appropriate,¹² unless the characteristics of the (p. 242) trades ‘include that they: (a) are non-systematic, ad-hoc, irregular and infrequent; or (b) are carried out between eligible and/or professional counterparties and do not contribute to the price discovery process’. Article 28 (Obligation to trade on regulated markets, MTFs or OTFs) of MiFIR imposes a similar trading obligation in respect of OTC derivatives. ‘Financial counterparties’ within the meaning of Article 2(8) of EMIR and non-financial counterparties that meet the conditions of Article 10(1)(b) of EMIR¹³ must execute transactions that are (1) subject to the clearing obligation of Article 10(1)(b) of EMIR (but not intragroup transactions as defined in Article 3 of EMIR nor transactions covered by the transitional provisions in Article 89 of EMIR) and (2) have been designated as subject the trading obligation in accordance with Article 32 of MiFIR, on (a) regulated markets, (b) MTFs, (c) OTFs, or (d) equivalent third-country trading venues.

6.14  MiFID II has also introduced controls for algorithmic and high-frequency trading (HFT) techniques.¹⁴ As summarized in Recital (64) of MiFID II, investment firms that use these techniques must ensure that their strategies are properly tested, cannot be used in a way that would amount to market abuse under the Market Abuse Regulation (MAR),¹⁵ and do not create a disorderly market. Trading venues, on the other hand, must ensure that their systems are resilient and contain controls such as circuit breakers that are able to halt trading in the event of price turbulence.¹⁶

(p. 243) 6.15  MiFID II’s objective is to create ‘a comprehensive regulatory regime governing the execution of transactions in financial instruments irrespective of the trading methods used to conclude those transactions so as to ensure a high quality of execution of investor transactions and to uphold the integrity and overall efficiency of the financial system’.¹⁷ To enhance competition, MiFID II insists on open-access, requiring all trading venues, that is, RMs, MTFs, and OTFs, to offer access subject to ‘transparent and non-discriminatory rules’, although OTFs are permitted ‘to determine and restrict access based, inter alia, on the role and obligations which they have in relation to their clients’, based on specific parameters provided that these are ‘open and transparent’ and do ‘not involve discrimination by the platform operator’.¹⁸

6.16  Each trading venue or facility must have several qualities to be successful. It must provide liquidity, that is, the ability to trade without delay at relatively low cost and in relatively large sizes, price transparency, which means that the venue must make pre- and post-trade price information available in a timely and accurate manner, and it must offer a measure of completion assurance, that is, settlement risk must be manageable. Liquidity will be a function of both the general supply of and the demand for the investment in question, and the success of the trading venue in attracting a portion of that general order flow, which will depend on whether the trading venue is competitive relative to the facilities offered by other trading venues. One such competitive aspect is liquidity. Thus, liquidity begets liquidity. A start-up trading venue can only attract liquidity if it offers a service that is complementary to the services of existing trading venues, such as after-hours trading in the equity markets. Equally, trading venues that lose order flow to other venues and cannot reverse the trend will rapidly become moribund. Therefore, although the development of alternative trading venues (in particular in the equity (p. 244) markets, which have always had the benefit of a high level of concentration of supply and demand in the exchange-based venues) might result in fragmentation of liquidity, there is a logical limit on how far the market will fragment.

6.17  Broadly, price-forming processes divide into order-driven and quote-driven processes. Some trading systems combine elements of both. In an order-driven system, market participants submit buy-and-sell orders, which the trading venue then seeks to execute by crossing. Orders are collected in what is known as the ‘order book’, which is a reference to the traditional manner of order matching through open outcry on the exchange floor. Open outcry still exists in some commodities markets, but order-driven markets have otherwise automated the order book. Depending on liquidity, order matching may be continuous or interrupted, by way of a ‘call’ or ‘batch’. A precondition for order-driven trading is that the flow of buy-and-sell orders is reasonably continuous. Order-driven price forming can only work well if the subject matter is a liquid security or contract, that is, a financial asset that attracts regular supply and demand. Not surprisingly, therefore, order-driven trading methodologies are predominant in the equity markets, used both by the exchanges and by alternative trading systems.

6.18  Quote-driven markets rely on dealers to provide liquidity and make a market by continuously publishing quotes, that is, firm prices at which they are prepared to buy or sell, respectively, a certain quantity of the financial asset in which they are making a market. Trades are executed between a market participant and the dealer that is making the market. Dealers have a short holding horizon. Their business is to generate revenue intra-day by taking the buy (or ‘long’) side to one transaction and the sell (or ‘short’) side to a similar transaction, with the aim of earning from a spread. The spread is the difference between the offer price (the price at which the firm is prepared to sell) and the bid price (the price at which it is prepared to buy). Several considerations determine what spread the dealer will seek to achieve. First, it will wish to know what motivates the market participant to propose the trade, because the dealer will run its trading book on a market-neutral basis.¹⁹ It does not matter which way the markets move, that is, up or down. What matters is that they move. Profits can only be made from market volatility. The concern for the dealer is not as much which way the market is moving, but whether the trading book is tilted in the right direction given the expected movement. When considering entering into a trade, therefore, the dealer will be particularly concerned to know whether a proposed trade is driven by unique, non-public information that will soon change the price, or whether the market participant is trading for (p. 245) hedging or rebalancing purposes, that is, on a longer term view.²⁰ Typically, the dealer cannot discover the reasons for a trade and can only guess. As a rule, the larger and the more urgent the trade, the likelier it is that the market participant is informed, so that the dealer will demand a premium.

6.19  A second determinant for the spread that a dealer will demand is inventory risk. A dealer tries not to be net long or short in an investment for long during a day, and seeks no trading positions at the end of a trading day. Once it has taken a long or short position, literally every minute that passes before the opposing trade occurs is adding to the risk implied in the trading book. The spread, that is, the risk premium that the dealer will demand, will increase with the risk of the position. The quantum of risk depends on several factors,²¹ but in any event will increase directly in proportion to the relative size of the trade and the expected time that the position will remain on the trading book. These risks increase with trading volume and speed, and reduce as market liquidity increases, that is, spreads tighten in general if a market is liquid. Therefore, market-impact costs of a transaction increase with trading volume and speed, and decrease depending on market liquidity.

6.20  Price or trading transparency is a function of the level of publication of information about prices at which transactions at certain volumes may be executed, called ‘pre-trade transparency’, and information about prices at which transactions at certain volumes have been executed, called ‘post-trade transparency’. The more detailed, widespread, and immediate the information available about prices and volumes in a market, the greater the transparency of that market is perceived to be. Pre-trade transparency in an order-driven market is generally mixed. It depends on the level of public disclosure of the content of the order book, which will vary from market to market. Market participants limit the pricing risk by submitting limit orders to the market, that is, orders that have a maximum bid or minimum offer price (depending on whether they are selling or buying) as a condition for execution. Post-trade transparency in order-driven markets is generally good. Prices and trade volumes are usually widely available in near real-time. Pre-trade transparency in a trading venue that operates a quote-driven system must necessarily be immediate and complete to be useful. Unless the market making dealers publish their quotes, there cannot be a market. If the prices are merely (p. 246) indicative, the venue qualifies as a data vendor, rather than as a trading venue. One of the benefits to the participants in a quote-driven market is immediacy. Unlike market participants in an order-driven market, market participants in a quote-driven market will know the execution price with certainty, if they decide to trade. Post-trade transparency in quote-driven markets is generally mixed. Publication of prices and trade volumes is usually delayed, because the market making dealers resist immediate post-trade transparency, particularly in less liquid investments, on grounds that this changes the risk-return proposition.²²

6.21  In the absence of data consolidation, the development of alternative trading venues and fragmentation of liquidity may limit price transparency. This can adversely affect the quality of price forming, and therefore the efficiency of the markets. To address the concern relating to price-data consolidation in markets for listed instruments, MiFID I first introduced extensive pre- and post-trade price-publication requirements for operators of trading venues that maintain a market in listed instruments, and MiFID II and MiFIR expanded on that framework. Accordingly, RMs, MTFs, OTFs, and SIs must publish prices and volumes in a manner that permits consolidation of the data by market participants across trading venues. The publication requirements for RMs, MTFs, OTFs, and SIs cover bid and offer prices, the depth of the market, the price, the size, and the time of transactions, both prior to and following a trade in covered financial instruments.²³

6.22  When an investment firm executes a transaction, in legal terms, it can do so as agent for, or as principal with, the client:

1. (a)  Trading based on quote-driven price forming—OTC spot and derivative transactions and securities financing transactions—by its very nature concerns principal-to-principal business. Contracts are made between the market participant and the dealer as counterparties.

2. (p. 247) (b)  Trading based on order-driven price forming—on-exchange (RM or MTF) traded derivatives and spot transactions—by its very nature concerns agency business. The broker that has access to the relevant trading venue will make a contract for and on behalf of its client as principal so that the contract will be that of the client, although the broker will typically act as principal vis-à-vis the counterparty when executing the trade.

6.23  A broker is an agent and an agency relationship implies duties of skill and care as well as fiduciary duties.²⁴ Bowstead & Reynolds on Agency²⁵ address the purpose and scope of an agent’s fiduciary duty and submit that ‘the fact that an agent in the strictest sense of the word has a power to alter his principal’s legal position makes it appropriate and salutary to regard the fiduciary duty as a typical feature of the paradigm agency relationship’, and that ‘to do so will not mislead so long as two things are borne in mind’. The first matter noted by Bowstead & Reynolds on Agency is that the word ‘agent’ can be used in varying senses, and not all persons to whom the word is applied are agents in the full. That question does not arise if an investment firm accepts a client order as broker. The second matter is that the extent of an agent’s equitable duties may vary from situation to situation’.²⁶ The basic principle, at English law, appears to be that fiduciary duties commence and end with the relationship of agency, which may or may not coincide with any contract between the parties.²⁷

6.24  Fiduciary duties forbid an agent from acting under conflict and from profiting from the agency position. However, it will often be in the client’s interests as much as the investment firm’s that the conflict exists or the profiting takes place. A broker may offer the client to fill a trade as principal so that the broker will become the counterparty to the transaction rather than seek to fill the order with a third party through a trading venue. It may very well be in the client’s interest that the broker can fill an order as principal, for instance because the broker can offer a better price, can fill a large order immediately, can offer better confidentiality, or because the market does not offer the requisite liquidity. Indeed, pricing, liquidity, immediacy, and confidentiality benefits underpin the (p. 248) business models of systematic internalizers.²⁸ Nevertheless, the conflict is obvious and must be cleared by way of informed consent before any transaction can be made. Bowstead & Reynolds on Agency note that in the event of such a conflict

it becomes essential that the agent fully informs the principal of all relevant facts and then obtains consent to the conflict or profiting. Those duties of disclosure and consent, albeit arising only secondarily, are positive duties. In other words, the fiduciary duties are not outright prohibitions, but merely proscribe profiting and conflicts that have not been consented to by the principal. What constitutes a fully informed consent is a question of fact and ‘there is no precise formula which will determine all cases’.²⁹ Consent must be positively shown, but it can be inferred if the principal is plainly fully aware of all the facts and raises no objection. The burden of proving full disclosure of a conflict of interest and of obtaining consent lies on the agent. It is not sufficient for him merely to disclose that he has an interest or to make such statements as would put the principal on inquiry nor is it a defence to assert that had he asked for permission it would have been given. It is possible that the consent itself may be recalled if obtained by undue influence, duress or misrepresentation. Consent may be given in advance or retrospectively. In some circumstances it may not be necessary for the agent to give details of the amount of a commission being paid by a third party where the principal knows that the remuneration will be received and the commission is at a rate standard in the industry. It is not always necessary that the principal understand that without consent there would be a breach of fiduciary duty, but usually the principal should be aware that his consent is being sought.

6.25  Based on these principles, therefore, a broker may fill an order as principal subject to the following basic rules:³⁰

1. (a)  If the broker intends to fill an order as principal, the firm will need express authority to do so. In the absence of such authority, the dealing may be void at law and will be voidable ex debito justitiae at equity.

2. (b)  If the broker has obtained authority to fill the order as principal, the firm will be under a duty to act with perfect good faith, and make full disclosure of all the material circumstances, and of everything known to the firm respecting the subject-matter of the contract or transaction which would be likely to influence the conduct of the principal or the firm’s representative.

3. (c)  Where any question arises as to the validity of any such contract or transaction, or of any gift made by a principal to the agent, the burden of proving that no advantage was taken by the broker of its position, or of the confidence (p. 249) reposed in it, and that the transaction was fair, and entered into in perfect good faith and after full disclosure, lies upon the broker as the agent.

6.26  In other words, express authority, informed disclosure, and ‘fair-dealing’ will permit the broker to fill the order on its own account despite the conflict. A question is whether duties of disclosure and fair dealing arise while negotiating the terms of appointment of the broker. Bowstead & Reynolds on Agency suggests that the mandate is not uberrimae fidei, and that ‘until the agency is created the parties are at arm’s length and must look after their own interests, subject again to the law of undue influence and any statutory regulation that might exist’.³¹ In order-driven services, therefore, and all else being equal, fiduciary duties operate from the time the mandate has been given.

6.27  Some order-driven trading services present a different agency conflict because the broker is acting as agent to both clients on the transaction, known as ‘order matching’ or ‘crossing’.³² This is a typical structure for order-driven trading venues that operate a price-taking, rather than a price-forming, system. The market operator will collect buy-and-sell orders in accordance with the procedures of the trading venue and seek to match the orders on terms that are determined by reference to the trading venue’s mechanical, non-discretionary matching rules, which do not allow for market timing and take a price from an external, independent source, typically an RM. For each pair of matching orders, the market operator, or an affiliate, acts as broker on behalf of the parties that submitted those orders and will execute a trade as matching agent if orders match, in accordance with the rules of the trading system. This is a very common structure for alternative trading systems in the equity spot markets, which seek to provide a venue where market participants can execute sizeable trades at exchange prices, avoiding market impact. It is also the manner chosen to operate by brokers that offer equity-crossing facilities. This form of trading is also referred to as ‘guaranteed cross-trading’. As is true in relation to self-dealing, the broker as agent may not enter into a transaction in which duty to another principal may conflict with the broker’s duty to the principal, unless the principal, with full knowledge of all the material circumstances and of the nature and extent of the agent’s interest, consents;³³ the principles set out in para 6.25 above apply.

6.28  More often, the terms of the trading system prescribe that the broker interposes itself as counterparty to the market participants whose orders matched, so that two equivalent contracts result from the match. This trading process is often referred to as ‘matched principal trading’, ‘back-to-back trading’, or ‘riskless principal (p. 250) trading’. The benefits of matched principal trading are that market participants can remain anonymous, and that the broker as matching agent can deal with the administrative aspects of arranging settlement. Matched principal trading does not imply that the broker does not act as agent when it makes the contracts. It means that the broker acts as (matching) agent by self-dealing with each principal, resulting in two matching back-to-back contracts. The two contracts can exist, even though made by one agent acting for two parties, because the broker as matching agent becomes personally liable on both contracts—to the exclusion of the buyer as principal on one contract, and the seller as principal on the other—so that the seller has no rights and liabilities under the contract made by the matching agent with the buyer, and vice versa.³⁴ Accordingly, the broker is responsible and liable for settling the obligations arising under both contracts but is, subject to default risks, not putting capital at risk because the positions under the matching contracts cancel each other out.³⁵

6.29  Matched principal trading does not imply that the broker does not act as agent when self-dealing on a matched basis. The broker is subject to the fiduciary duties that an agent has to its principal. In conflict terms, the back-to-back or matched principal trading presents a double conflict: the broker has both a personal interest and a duty to another principal that may conflict with the broker’s duty to the principal on each side. These conflicts must be cleared in accordance with the principles set out in para 6.25 above. The broker’s discretionary authority as matching agent is limited and, consequently, the fiduciary duties are limited. The broker is expected to collect the buy-and-sell orders, ‘run the black box’ at some or several points during the trading day to match orders, and bring about contracts between the parties whose orders matched based on the independent reference price.

6.30  If the order-driven trading venue has a price-forming function, which would typically be a regulated exchange, buy-and-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. Each broker undertakes to access the trading venue and execute their client’s order by entering into a transaction with another broker for the account of the client on the best terms possible, subject to and in accordance with the rules of that trading venue.³⁶ The rules of trading venues that (p. 251) organize trading in this manner will ordinarily require that the brokers become personally liable for the contract made, to the exclusion of their principals, and that contracts are subsequently cleared in accordance with the operative CCP structure.³⁷ To that end, the contract made by the broker who acts as agent both for a buyer and for a seller incorporates terms that provide that the brokers are personally liable on that contract, and that their principals are excluded from enforcing rights arising under it. Thus, each broker need only be concerned with the other broker’s solvency, and not with the principal’s solvency, which facilitates the ease and speed of the trading process. Only the broker will have rights and liabilities on the contract made pursuant to the order of the client, who is excluded from that contract and consequently, cannot sue, nor be sued, nor otherwise intervene on it. The clients are excluded from the contracts made pursuant to their orders, because it would undermine certainty of execution, and thus market liquidity, to grant them a right of intervention.

6.31  Trades in on-exchange traded derivatives are typically executed by admitted brokers pursuant to a mandate that provides that, in respect of every contract made by the broker in its own name on a trading venue, the broker and the client shall be deemed to have entered into an equivalent contract that in every respect is subject to the rules and procedures of that venue, which will include the exercise of any powers in respect of the CCP cleared contract by the broker, for example in the event of market disruption or a clearing-member default. The two derivative contracts are referred to as ‘back-to-back’ contracts, with the position of the broker being that of a riskless principal. However, there is a question of interpretation in relation to the nature of the equivalent contract that is deemed to have been made between the broker and the investor. Rather than creating an independent derivative contract, the ‘equivalent contract’ wording would appear to be creating a relationship of debtor and creditor between the client and the broker, under which the client and the broker undertake to pay amounts that are calculated by reference to the amounts due between the broker and the CCP in respect of the traded and cleared derivative contract. Principal and agent can be debtor and creditor in relation to payments made and received by the agent for the principal in connection with the performance of contracts made by that agent for that principal during the agency.³⁸ The resulting obligations are not to be treated as arising under a distinct or independent contract, but as obligations arising under the general agency relationship between the client and the broker by reference to the contract made by the broker.

6.32  MiFID II distinguishes between dealing on own account and orders executed on behalf of clients. ‘Dealing on own account’ is defined as ‘trading against proprietary capital resulting in the conclusion of transactions in one or more financial instruments’.³⁹. ‘Execution of orders on behalf of clients’ is defined as ‘acting to conclude agreements to buy or sell one or more financial instruments on behalf of clients’.⁴⁰ At first blush, these regulatory categorizations appear to align with the business of dealers, who are in the business of trading with clients on own account, and brokers, who are in the business of executing orders on the instruction and for the account of clients.

6.33  Nevertheless, MiFID II has muddied the waters by adding a second component to the definition of ‘execution of orders on behalf of a client’: ‘the conclusion of agreements to sell financial instruments issued by an investment firm or a credit institution at the moment of their issuance’.⁴¹ Prima facie, the activity of sell-side firms in the context of rights offerings in capital markets would not constitute an order from the client to whom the instrument is being sold.⁴² Recital (45) of MiFID II clarifies the rationale for the inclusion of the activity in the definition of ‘execution of orders on behalf of clients’:

Investment firms and credit institutions distributing financial instruments they issue themselves should be subject to this Directive when they provide investment advice to their clients. In order to eliminate uncertainty and strengthen investor protection, it is appropriate to provide for the application of this Directive when, in the primary market, investment firms and credit institutions distribute financial instruments issued by them without providing any advice. To that end, the definition of the service of execution of orders on behalf of clients should be extended.

6.34  MiFID II appears to be muddying the waters further by tagging certain activities that, on their face, would constitute ‘dealing on own account’ under MiFID II, as ‘execution of orders on behalf of a client’. Recital (103) of MiFID II Delegated Regulation (EU) 2017/565⁴³ stipulates rather bluntly that ‘dealing on own account with clients by an investment firm should be considered as the execution (p. 253) of client orders’, and therefore, is subject to the relevant requirements of MiFID II and the Delegated Regulation and, ‘in particular, those obligations in relation to best execution’. However, that blunt statement is nuanced by Recital (91) of MiFID II, which observes that the ‘best execution’ obligation is limited to circumstances ‘where a firm owes contractual or agency obligations to the client’.⁴⁴ As discussed in paragraphs 6.24 to 6.29 above, if an investment firm’s brokerage business fills an order on own account or is offering matched principal trading services, there is no question that the broker is still subject to agency duties, as stipulated by Recital (91). Accordingly, there is no question that the broker who fills an order on own account is executing an order on behalf of a client in regulatory terms. . The legal and regulatory capacity is clear and established at the time the broker enters into the mandate, and the fact that a subsequent trade is completed by the broker on own account does not change these legal and regulatory capacities.⁴⁵ The question is much more vexing in case of an investment firm’s dealer business, which offers to trade based on quotes. The ‘contractual or agency obligations’ reference in Recital (91) of MiFID II has been taken to mean that the relationship between the client and the dealer must have the character of supply of professional services, so that the firm can be said to be dealing ‘on behalf of’ a client, subject to ‘contractual or agency obligations’.⁴⁶

6.35  The European Commission, in response to questions from ESMA in the context of MiFID I,⁴⁷ has said that it considers self-dealing or matched principal trading by brokerage businesses to be examples of contracts that are made by the investment firm in a regulatory capacity of a firm that executes an order on behalf of (p. 254) the client, rather than in the capacity of a firm that deals on own account. The Commission further observed that, in its opinion, execution of a client order by entering a proprietary trade on a ‘request for a quote’ basis should not be treated as execution of an order on behalf of a client. The Commission nuanced that position as follows:⁴⁸

However, in some cases, proprietary trade will attract the best execution obligation. The application or otherwise of best execution will depend on whether the execution of the client’s order can be seen as truly done on behalf of the client. This is a question of fact in each case which ultimately depends on whether the client legitimately relies on the firm to protect his or her interest in relation to the pricing and other elements of the transaction such as speed or likelihood of execution and settlement that may be affected by the choices made by the firm when executing the order.

6.36  The question turns on the facts. The Commission helpfully outlined various relevant fact patterns for consideration:⁴⁹

1. (a)  Whether the firm approaches (instigates the transaction with) the client or the client instigates the transaction by making an approach to the firm.

2. (b)  Questions of market practice will help to determine whether it is legitimate for clients to rely on the firm. For example, in the wholesale OTC derivatives and bond markets buyers conventionally ‘shop around’ by approaching several dealers for a quote and in the circumstances, there is no expectation between the parties that the dealer chosen by the client will owe best execution.

3. (c)  The relative levels of transparency within the market will also be relevant. For markets where clients do not have ready access to prices while the investment firms do, the completion will be much more readily reached that the client is relying on the firm in relation to the pricing of the transaction.

4. (d)  The information provided by the firm about its services and the terms of any agreement between the client and investment firm will also be relevant but not determinative of the question. The Commission warns against the use of standard agreements that characterize the relationship in a manner that conflicts with economic reality.⁵⁰

6.37  The regulatory position, therefore, aligns in principle with the position at law. If the investment firm has assumed responsibility to execute the trade in the interest of the client—that is, the client is justifiably relying on the investment firm to use its professional skills to act in the interest of the client, rather than at arm’s length—best execution duties apply. In those circumstances, the investment firm acts as a broker rather than as dealer, notwithstanding the quote-driven basis for the trade. The brokerage relationship could arise as a matter of express or implied (p. 255) terms in a contract,⁵¹ or, more likely in an OTC quote-driven environment, as a duty of care at law.⁵²

6.38  Notwithstanding, article 64 of MiFID II Delegated Regulation (EU) 2017/565 (best execution criteria), sub-paragraph 4, arguably takes the regulatory obligation beyond the obligation at law. It provides that ‘[w]hen executing orders or taking decision to deal in OTC products including bespoke products, the investment firm shall check the fairness of the price proposed to the client, by gathering market data used in the estimation of the price of such product and, where possible, by comparing with similar or comparable products’. This might be characterized as an example of the client’s best interest duty, which introduces an element of good faith dealing in the relationship.⁵³ However, Article 24(1) of MiFID II operates only if an investment firm provides an investment service, that is, not when the firm engages in an investment activity.⁵⁴ Without prejudice to the obligation of Article 64(4), execution of a client order by entering a proprietary trade on a ‘request for a quote’ basis should not be treated as execution of an order on behalf of a client; it does not constitute the supply of an investment service.⁵⁵

6.39  The excess return that an investor might earn if certain transactions are carried out can be eliminated by the cost of the transaction. The transaction costs may be divided into direct and indirect costs. Direct costs include brokerage, exchange fees, taxes, settlement costs, and custody costs. Indirect costs include market-impact costs, opportunity costs, and the cost of delayed executions, including delayed or failed settlements. Market-impact costs are a function of the liquidity of the market, the size of the proposed transaction, and the speed with which it is executed. If the transaction is small relative to the size of the market, and the market is liquid, that is, there are plenty of buyers and sellers on any given day, the transaction will not move the market price much. However, if the transaction is large relative to the size of the market, a sudden spike in supply or demand might temporarily distort the price. Equally, investment firms who trade aggressively and seek to fill an order quickly (p. 256) should expect to incur higher market-impact costs. Missed-trade opportunity costs, on the other hand, may manifest if execution is delayed due to a failure to execute the trade in a timely manner, which could have been avoided if execution had been pursued more aggressively. The cost of delayed execution manifests itself if execution is delayed due to an inability, rather than a failure, to execute the trade in a timely manner, due to the size of the trade and market liquidity. The cost of delayed or failed settlements manifests itself as a replacement cost if a transaction does not settle on the day it is supposed to in accordance with its terms, due to operational errors, or counterparty inability, or unwillingness. The cost of replacement is equal to the difference between the original price for the investment and the price prevailing at the time the replacement transaction is executed, including transaction and settlement costs associated with that replacement transaction. Depending on market movement, the replacement transaction may of course turn out to be achieved at a more advantageous price than the original transaction which failed to settle.

6.40  The factors that contribute to the quality of execution are complex and require evaluation of several constantly changing variables such as the prevailing price, liquidity, and the potential market-impact costs. Therefore, the quality of effort of an investment firm that is engaged to execute a transaction for an investor is a key component of a successful investment cycle. The concept of best execution connotes the duty of the investment firm to ensure that the quality of the execution effort meets the appropriate standards of skill and care. Measuring the quality of the execution effort, however, is complicated by the fact that a transaction as such, or even just the knowledge in the market that a transaction is being contemplated, typically moves the most important market variables, that is, price and liquidity. Adams provides an elegant analogy to explain the conundrum of measuring best execution:⁵⁶

One of the corner stones for quantum physics (and some of its more bizarre offspring such as chaos theory) is the notion that you cannot know both the position and velocity of an atomic particle at the same time. The reason being that the act of observation (crudely, shining a light on the particle) changes either the position or the velocity of the particle. The act of observation changes the characteristics of what is being observed.

6.41  The point put forward is that the very act of executing, or seeking to execute, a transaction will often move the main market variables, thus making it difficult to assess a price at any time by comparing it to the price prevailing at that time in the same or some other trading venue, because the transaction that is being measured is reflected in the price of the other transaction, too. In other words, the price for the other transaction would very often have been different had the transaction that is being measured not been executed. Add the complexity of price and other data comparison in a market fragmented across different trading venues, and it (p. 257) follows that benchmark pricing to assess execution quality is to a large extent a futile effort if performed on a trade-by-trade basis. The duty to obtain best execution, therefore, is a duty to use skill and care in the design and implementation of execution processes, rather than a duty to obtain a quantifiable result.

6.42  The reality of trade execution has been recognized in the regulatory concept of best execution that operates pursuant to MiFID II and, prior to that, MiFID I, although the MiFID II framework is more detailed in terms of execution information, quality of execution monitoring, and venue selection procedures.⁵⁷ An investment firm is not required to achieve an execution result but is required to apply a certain level of skill and care in executing an order. More precisely, the firm is required ‘to take all reasonable steps to obtain the best possible result for its client taking into account price, costs, speed, likelihood of execution and settlement, size, nature, or any other consideration relevant to the execution of the order’.⁵⁸ In determining the relative weight of the best execution factors, a firm must consider the nature of the client, including the status as retail or professional, the nature of the client order, the characteristics of financial instruments that are the subject of that order, and the characteristics of the execution venues to which that order can be directed.⁵⁹ In respect of securities financing transactions executed for professional clients, Recital (99) of MiFID II Delegated Regulation 2017/565 notes firms may operate a differentiated order execution policy to obtain the best results:

In order to comply with the legal obligation of best execution, investment firms, when applying the criteria for best execution for professional clients, will typically not use the same execution venues for securities financing transactions (SFTs) and other transactions. This is because the SFTs are used as a source of funding subject to a commitment that the borrower will return equivalent securities on a future date and the terms of SFTs are typically defined bilaterally between the counterparties ahead of the execution. Therefore, the choice of execution venues for SFTs is more limited than in the case of other transactions, given that it depends on the particular terms defined in advance between the counterparties and on whether there is a specific demand on those execution venues for the financial instruments involved. As a result, the order execution policy established by investment firms should take into account the particular characteristics (p. 258) of SFTs and it should list separately execution venues used for SFTs. An investment firm should apply its execution policy to each client order that it executes with a view to obtaining the best possible result for the client in accordance with that policy.

6.43  Compliance with best execution duties, therefore, is a matter of the quality of the execution arrangements. Firms must institutionalize the way they will seek to comply with best-execution requirements by establishing effective execution arrangements that include an order execution policy that sets out the different venues where the investment firm executes its client orders and the factors affecting the choice of execution venue and the manner of execution.⁶⁰ It follows that a firm complies with its best-execution requirements in respect of an executed order if the execution arrangements and the execution policy meet the appropriate standards, and the order is executed in accordance with that execution policy and/or arrangements.⁶¹ The execution policy must include such venues as enable the firm to obtain on a consistent basis the best possible result for the execution of client orders, which means in practice that if an order is executed in a venue that is part of the list, the firm does not have to consider other venues. Ergo, ‘best execution’ is not a trade-by-trade test, but a test as to whether order execution is implemented on a consistent basis in accordance with effective execution arrangements.

6.44  Each firm must further provide ‘appropriate information’ to their clients on the firm’s execution policy, which information must include, if applicable, a disclosure that client orders may be executed outside an RM, MTF, or OTF. In practice, firms provide clients with a summary of their execution policy. MiFID requires that each firm obtain advance consent to the execution policy. That consent is based on the summary. In other words, although MiFID uses the term ‘consent’, as it is not conceivable that a firm would change the execution policy if one or more clients object, in practice, the consent requirement amounts to ‘no objection’ prior to supply of the investment service. A firm must be able to demonstrate to a client, on request, that orders have been executed in accordance with the firm’s execution policy.⁶²

6.45  The firm must review and monitor the effectiveness of its execution arrangements with a view to identifying and, where appropriate, remedying any deficiencies. Particularly, a firm must assess on a regular basis whether the execution venues included in the execution policy provide for the best possible result for the client or whether they need to make changes to their execution arrangements. The review must be an overall assessment of the effectiveness of the execution arrangements. Specifically, the firm should consider whether it could consistently obtain better execution results if it were to include additional or different execution venues or (p. 259) entities, assign a different relative importance to the best-execution factors, or modify any other aspects of its execution policy and/or arrangements.⁶³

6.46  Given that the primary purpose of the regulatory concept of best execution is to protect the interest of the investor as client of the regulated firm,⁶⁴ considerable weight ought to be given to the regulatory principles in determining the skill, care, and diligence, as is usual or necessary in, or for the ordinary or proper conduct of, the execution of transactions by the investment firm.⁶⁵ It would follow that, prima facie, an investment firm may be considered to have discharged its legal duty to provide best execution in relation to an executed transaction if it has established execution arrangements in accordance with its regulatory obligations, the execution arrangements meet the regulatory standards, and the order is executed in accordance with these arrangements. Again, the execution arrangements, prima facie, may be considered appropriate if the processes are designed in compliance with applicable regulatory standards.

6.47  A firm will not escape best-execution requirements if it did not have the intention to act in the interest of the client because it is dealing on own account, but the client nevertheless reasonably relied on the firm doing so.⁶⁶ If a firm that deals on own account but, subject to the best-execution requirements, provides a quote to a client and that quote meets the investment firm’s best-execution requirements at the time the quote is provided, then the firm will still meet the best-execution requirements if it executes a transaction based on that quote after the client has accepted the quote, ‘provided that, taking into account the changing market conditions and the time elapsed between the offer and acceptance of the quote, the quote is not manifestly out of date’.⁶⁷ A firm may bring about or enter into transactions (p. 260) with eligible counterparties without being obliged to comply with best-execution requirements in respect of those transactions or in respect of any ancillary service directly related to those transactions.⁶⁸

6.48  The best-execution requirements apply mutatis mutandis to firms that ‘place orders’, rather than execute orders, with a broker for execution in the course of supplying portfolio management services, and to firms that receive and transmit orders without discretion, as if references to executing orders were references to those order-placing and transmission activities.⁶⁹ However, the MiFID best- execution provisions are not intended to require a firm that transmits or places orders with another firm to duplicate the efforts of the other firm. Instead, the firm that transmits or places the order should determine whether the executing firm is likely to execute the order in such a manner that the transmitting or placing firm will meet its own best-execution requirements. In that respect, it is relevant whether the executing firm must execute the order subject to MiFID best-execution requirements (which implies that this firm must treat the firm that places or transmits the order as a retail or professional client) and, if not, whether the executing firm will undertake separately to comply with the MiFID best-execution requirements, and to what extent it can demonstrate that it delivers a high level of execution quality for the type of orders that the other firm is likely to place with or transmit to it. If the executing firm is subject to MiFID best-execution requirements or has undertaken to comply with them, in general the firm that transmits or places orders to that executing firm will be able, to a large extent, to rely on the executing firm for purposes of compliance with its own best-execution requirements in respect of the execution of that order. That reliance, however, is without prejudice to the other best-execution requirements that apply to a firm that transmits or places orders, in particular the requirements to implement an appropriate execution policy and to monitor and review its effectiveness, including the execution quality actually delivered by such entities.⁷⁰ In addition, a firm should always consider whether it is reasonable to transmit or place orders with another (p. 261) firm for execution, and whether it is necessary to exercise some additional control over how its orders are executed. In a portfolio management context, that means that best-execution requirements of the investment manager differ considerably depending on whether the manager can rely on the sell-side firm to execute a transaction subject to MiFID’s best-execution requirements, or not.

6.49  A firm must implement procedures and arrangements which provide for the prompt, fair, and expeditious execution of client orders, relative to other client orders or the trading interests of the investment firm. The order-handling procedures must provide that client orders are carried out promptly and sequentially, unless, in view of the nature of the order or prevailing market conditions, this is impracticable, or the interests of the client require otherwise. Executed orders must be accurately recorded and promptly allocated. Brokerage firms are responsible for arranging the settlement of an executed order and must ensure that any client assets received in settlement of that executed order are promptly delivered to the account of the relevant client. A firm must not aggregate orders unless it is likely that such aggregation will not work overall to the disadvantage of any client whose order is to be aggregated. The firm’s intention to aggregate orders must be disclosed to each such client, together with a statement that the effect of aggregation might work to that client’s disadvantage. The firm may not aggregate orders unless it has established and implemented an effective order-allocation policy that provides for the fair allocation of such orders and transactions in sufficiently precise terms. When an aggregated order is partially executed, the firm must allocate the trade in accordance with its order-allocation policy. The firm must have procedures, which prevent detrimental allocation, in particular where the firm aggregates transactions for the firm’s own account with one or more client orders, in which case the firm must give priority to client trades over firm trades, unless the firm is able to demonstrate on reasonable grounds that without the combination it would not have been able to carry out the order on such advantageous terms, or at all. In the latter case, the transaction for own account may be allocated proportionally, in accordance with the general principle.⁷¹

6.50  After a trade has been executed and confirmed, cash and/or securities transfers will have to take place at some point, or several points, during the life cycle of the transaction. Performance of a cash or securities transfer obligation is referred to in the securities industry as ‘settlement’. To complete, spot, or ‘cash’, transactions require exchange of cash and securities, or currencies, within several business days (p. 262) after the trade date; the number of days depends on the trading venue. Futures, forwards, swaps, and options usually require collateral settlement on the first business day following the trade date, and then continuously on each following business day until expiry of the transaction, with completion settlement upon the earlier of maturity or termination. Securities financing transactions require an initial completion settlement in which securities and collateral are exchanged, continuous collateral settlements during the life cycle of the transaction, and final settlements upon the earlier of maturity or termination.

6.51  A party to a contract must perform exactly what it undertook to do, which is determined through construction of the contract. The parties to a securities transaction typically refer to ‘securities’ when executing the trade, but transfer book-entry securities to perform the resulting securities transfer obligation. It would be unconstructive to insist that references to ‘securities’ should be construed as references to original securities. That would ignore both the practical necessity to give efficacy to the transaction and the way securities transactions are settled in the financial markets. Trade confirmations and settlement instructions always provide details of securities accounts and of securities identification codes associated with the relevant book-entry securities.⁷² In the modern securities markets, unless there is express agreement to the contrary, the parties to a transaction in securities that exist in book-entry form must be presumed to have agreed that the seller shall transfer the property in book-entry securities, not in the original securities.

6.52  Under English law, a payment, that is, a valid tender in satisfaction of a payment obligation, is made by unconditionally offering legal tender (coins or notes), or otherwise by offering a performance in accordance with the contract.⁷³ That means that any value transferred to a creditor pursuant to a payment obligation in accordance with the terms of the transaction that created the obligation, constitutes a payment. In the absence of contractual terms specifying the way a payment must be made, the offering of legal tender will suffice. Payment by way of legal tender, however, is not a practical option in the modern securities markets, if only because the clearing and settlement systems operated by, or linked to, exchanges require that payments, regardless of the currency, be made by a funds transfer. It follows that references to ‘cash’, as exchanged between parties to a transaction in the financial markets, should be construed as references to ‘funds’, that is, ‘bank money’.⁷⁴ Not to do so would be to ignore what is in practice necessary to give (p. 263) efficacy to the transaction.⁷⁵ In addition, market participants normally provide account details when the transactions are confirmed, which should be construed as agreement to funds transfer as the payment method. It must be presumed, therefore, that large-value payment obligations that result from transactions in the securities markets ought to be considered validly paid if settled by a funds transfer, even where the terms of the transactions are silent as to the payment method.⁷⁶

6.53  Originally, a funds transfer obligation would have been settled by delivery of a cheque or other paper-based payment instrument, drawn on the payer’s cash account. The payee would have had to present the cheque to its bank, which would then proceed to collect the amount expressed to be payable by the cheque from the payer’s bank. Today, the organization of the post-trade infrastructure of the modern securities industry requires predominantly that payment obligations that arise pursuant to a transaction in the financial markets are discharged by electronic credit transfers. In addition, issuers of securities that are held as part of the portfolio of investments will need to make payments to the investor from time to time, for example dividend or interest payments. Just as in connection with payments made pursuant to transactions in many other areas of commerce, cheques and other paper-based payment instruments have been replaced as a method of payment by credit transfers that are initiated by the payer by means of a payment instruction to its bank, which will then proceed to execute that instruction by paying the payee’s bank. A role-reversal between the payer’s bank and the payee’s bank, so to speak, which has inspired some commentators to speak of a ‘push’ versus a ‘pull’ system. The term ‘funds transfer’ or ‘credit transfer’, in this context, refers to a payment made through a payment service offered by a bank, whereby a bank undertakes, pursuant to a payment instruction from an account holder, to use skill and care to arrange for a credit in the amount instructed to the cash account of a designated payee.

6.54  Consequently, transactions in the financial markets cannot be settled without the assistance of a firm that has access to the relevant payment and securities settlement systems,⁷⁷ and possesses the specialist skills, systems, and controls necessary to hold and (p. 264) administer the investment portfolio. Appointing an independent, specialist custodian is a practical necessity, particularly if the portfolio includes cross-border investments. Securities transactions often involve the transfer of funds against securities, or funds denominated in one currency against funds denominated in another currency. The usage and procedures of the post-trade infrastructures in the various markets typically insist on concurrence in the mutual transfers, an operative principle that is loosely described as ‘delivery versus payment’. It would be challenging to orchestrate the concurrent transfers and receipts of funds and securities for the investor, if the services are split between a firm that acts as custodian and a firm that acts as bank. The investor’s custodian will therefore usually also be the banker and establish a cash account relationship pursuant to the terms of the custody agreement. The custody agreement is an account agreement under which the custodian undertakes to maintain certain cash and securities accounts in the name of the investor, credit any cash or securities received for the investor to those accounts, and provide payment and securities settlement services. Further, the custodian will perform portfolio administration functions such as income collection and corporate actions.

6.55  If the investor has appointed an investment manager, the manager will have been given the authority by the investor, as agent, to instruct the custodian for and on behalf of the investor to settle transactions the investment manager has executed with a market counterparty, or has instructed a broker to execute, on behalf of the investor as part of the investment process. Communication protocols between the investor, or the investor’s investment manager, and the custodian about confirmed trades, settlements, and reconciliations are quite complex, but often largely automated and standardized, for example using the SWIFT messaging system or a custodian proprietary system. Nevertheless, these communications can be a source of errors causing failed trades in which case a dispute may arise between the broker or dealer who executed the trade, a potential investment manager, and the custodian as to responsibility for the error. In those circumstances, the question for an investment manager is to what extent it has undertaken to the investor to accept the custodian’s communication protocol. It should be kept in mind that the investor, not the investment manager, is the custodian’s client so that any limitation of liability that the custodian seeks to invoke in relation to errors occurring in the communication with the manager, the investor’s agent, should have been properly documented in the custody relationship, not in the custodian’s written communication protocol, as that may have been provided to the investment manager. The investment manager is not usually authorized to agree to any custody terms on behalf of the investor.

6.56  The time of settlement depends on the terms of the contract, which might vary considerably depending on the trading venue. A transaction that has been (p. 265) executed on an exchange that operates a central counterparty system will have detailed settlement procedures. Typically, settlement will take place on the second business day after the trading day, referred to as ‘T+2’, but shorter or longer cycles occur as well. At the other end of the spectrum there are bilateral transactions made with a dealer in OTC markets, which might not even have express terms and may rely on the usage, for example in the OTC fixed income market, to settle transactions on T+1.

6.57  There are several risks involved with settlement, the key risks being finality risk and principal risk. Finality risk is the risk of an adverse claim, so that the settlement is not final.⁷⁸ Principal risk is the risk that one party settles its obligation, but the other party does not. Principal risk is managed operationally through a process generally described as ‘delivery versus payment’ (DVP), which means that parties to the settlement seek to synchronize the settlement instructions.⁷⁹ The post-trade infrastructure will permit custodians, brokers, and dealers to access web-based ‘settlement engines’ offered by payment and settlement systems so that it is possible to verify whether matching settlement instructions have been given. However, that does not necessarily mean that the funds transfer and the book-entry securities transfer are fully synchronized. The funds transfer may occur on a net basis in batches, while the book-entry securities transfer may occur on a settlement-by-settlement basis.⁸⁰

6.58  The bilateral transactions made with a dealer in OTC markets⁸¹ or contracts made by a matching agent (not acting as riskless principal)⁸² will be settled (p. 266) directly between the dealer and the investor’s custodian, or between the matched investors’ custodians. To that end, after the contract is made, the investor, or the investment manager as agent for the investor, will provide settlement instructions to the custodian. The custodian will then seek to match the settlement instructions with the settlement intentions of the dealer or the other matched investor’s custodian. On the settlement date, the obligations will be settled via the relevant payment and settlement systems, either on the instigation of a national central securities depository (CSD), or on the instigation of an international CSD (ICSD) of choice, such as Euroclear SA, a settlement bank organized under the laws of Belgium and the prevailing ICSD for the European fixed-income markets. Figure 6.1 provides an overview of a direct DVP settlement using a CSD-operated linked payment and settlement system. Figure 6.2 provides an overview of a direct DVP settlement using an ICSD-operated integrated payment and settlement system.(p. 267)

• View full-sized figure

Figure 6.1  Basic OTC settlement infrastructure using CSD linked payment and settlement systems

• View full-sized figure

Figure 6.2  Basic OTC settlement infrastructure using ICSD operated integrated payment and settlement system

6.59  If the contract is made by a broker on a trading venue⁸³ or as matching agent acting as riskless principal,⁸⁴ it will have to be settled between the broker and the third party first, and funds and securities exchanged between the broker and the investor’s custodian subsequently. Basically, this entails two direct settlements. Thus, if the contract is made on an exchange that uses a central counterparty (CCP) system, settlement will take place between the CCP and the broker-clearing member first, and between the broker-clearing member and the investor’s custodian next. If the broker is not a clearing member, or if the broker is using a correspondent broker on a remote exchange, an additional settlement step may occur. Equally, if the contract is made as matching agent and riskless principal, the broker will have to settle with the seller first, as the rule is that ‘delivery of the stock’ drives the settlement process, and with the buyer next.

6.60  Each step in the process needs to adhere, as far as possible, to DVP protocols. That means that the broker-clearing member will often have to finance the first settlement, either by using its own book-entry securities or by using its own funds, to ensure that the first step in the settlement process takes place. Once the book-entry securities or funds are collected, the broker will have to settle with the investor via the custodian. That is not always simply a matter of passing on what has been collected, at least not in the case of a clearing member. Clearing members will collect funds and securities in omnibus accounts, which creates a settlement pool. Accordingly, incoming funds and book-entry securities are earmarked as settlements relating to one or more contracts made on behalf of clients, but the proceeds per se are not earmarked as belonging to any specific client. The proceeds are pooled, and the clients will be entitled to receive settlement out of the pool. On current practice, it is questionable whether the clients will have proprietary rights on a co-ownership basis in that settlement pool, because the settlement will have been funded by the clearing member as agent and it is therefore doubtful whether a common intention can be construed to the effect that the broker holds the proceeds on trust.⁸⁵ Principal and agent can be debtor and creditor in relation to payments made and received by the agent for the principal in connection with the performance of contracts made by that agent for that principal in the course of the agency.⁸⁶ If the broker is not a clearing member or if the broker is acting (p. 268) as matching agent on a riskless principal basis, the two settlements will often be synchronized so that the CSD or ICSD that arranges the settlements will release securities or cash if two pairs of matching settlement instructions are submitted via its systems.

6.61  Unlike securities financing transactions, spot contracts made with a dealer on an uncleared, OTC basis, that is, not via an MTF or OTF, are typically marginally documented. There may not be much more than a telephone conversation followed by a basic electronic confirmation that includes a description of the original securities purchased or sold, and settlement instructions. If the dealer defaults on the settlement, general principles of contract law will have to provide guidance as to the position. Much will depend on the answer to the question whether, expressly or implicitly, time was made ‘of the essence’. If the trade confirmation provides a specific date for settlement, it may probably be assumed, in the context of the sale and purchase of securities, that time will be construed to be of the essence.⁸⁷ If, somehow, a date has not been specified, the investor might have to rely on anticipatory breach and treat the contract as repudiated, so that it may be rescinded and damages claimed.

6.62  The terms and conditions on which a broker will execute transactions for an investor on a trading venue pursuant to an instruction from an investment manager may be equally ambiguous. Typically, the investment managers and the brokers exchange notices about terms that should apply to their dealings for their mutual clients, the investors, but express agreement is rarely reached. Trades are done under terms that are uncertain and that can lead to disputes in the event of a (p. 269) default. If the broker defaults on the second leg of the settlement, it will be difficult to construe a term that provides that time is of the essence. The broker will have sent a trade confirmation upon completion of the contract, but that confirmation relates to the settlement between the broker and the third party, for example the central counterparty. Those terms will not necessarily apply between a broker and the investor as its client, in the absence of clear terms. Therefore, if the broker fails to make the onward settlement from the settlement pool, the investment manager, acting as agent for the investor, may have to rely on anticipatory breach and treat the broker’s obligation to settle as repudiated, so that the mutual undertakings between the broker and the investor in respect of the transaction executed for the investor by the broker on the instruction of the investment manager may be rescinded and damages claimed.

6.63  Each funds transfer involves three operative stages.

1. (a)  First, the payer must send a payment order to his/her bank,⁸⁸ which must be validated and accepted by the bank in accordance with the terms and conditions of the account agreement. The instruction to make a payment is called a ‘payment order’ or ‘funds transfer order’.⁸⁹ The payer’s payment order initiates the funds transfer and will be referred to in this section as the ‘original payment order’ or the ‘OPO’.

2. (b)  Next, the payer’s bank (the paying bank) must execute the OPO by giving the payee’s bank (the receiving bank) a corresponding payment order, which in its turn must be validated and accepted by the payee’s bank, and, if it is, the amount must be paid by the payer’s bank to the payee’s bank. The payer’s bank’s subsequent payment order executes the OPO and will be referred to in this section as the ‘corresponding payment order’ or the ‘CPO’.

3. (c)  Finally, the payee’s bank, subject to validation of the CPO, must credit the payee, which results in discharge of the payer’s debt to the payee. Naturally, the payer and the payee may hold accounts with the same bank, in which case (p. 270) the CPO step is skipped and the payer’s bank executes the OPO by giving credit to the payee. Consequently, a funds transfer is constituted by a complex chain of transactions, rather than a single act. Each constitutive component and their internal coherence is discussed in turn below.

• View full-sized figure

Figure 6.3  Basic payment system transfer steps where payee bank is a different bank

6.64  Figure 6.3 shows the basic steps involved if the payer and the payee maintain bank accounts at different banks, which requires clearing through a central bank’s payment system.

6.65  A fund transfer starts with an instruction from the payer to its bank to place a certain amount at the disposal of a specified payee by means of a credit to the cash account of that payee. This payment order, the OPO, will specify the transfer amount, and the name and account number of the payee. Under common law principles, if the bank accepts the OPO, it agrees to use reasonable skill and (p. 271) care to ensure that the account of the payee at the payee’s bank is credited by an amount equal to the transfer amount specified in the OPO.⁹⁰ The payment services terms of an account agreement will normally provide that the bank will be deemed to have accepted the OPO unless it rejects the same promptly.⁹¹ Banks usually reserve the right to reject payment orders in their discretion, which will not normally be exercised unless the payer’s cash balance, or available overdraft facility, is insufficient to cover the amount of the requested payment, or in case a bank is confronted with operational difficulties that obstruct execution of the payment order.⁹² Unless the account terms provide otherwise, an OPO generally becomes irrevocable if the bank, having started execution of the OPO by giving a corresponding payment order, the CPO, to the payee’s bank cannot reverse the CPO anymore because the payee’s bank has given credit to the payee.⁹³

6.66  It is unlikely that an OPO and the resulting debit, as against the payer’s bank, may be successfully challenged by the payer on the basis that its intention to enter into the transaction was vitiated by some factor, as the source of the vitiation would ordinarily be the relationship with the payee that gave rise to the payer’s debt to the payee that it is seeking to discharge by way of the funds transfer. If the transaction between the payer and the payee is void or voidable, that is not a factor that is relevant in the account relationship between payer and paying bank except, perhaps, in special circumstances. The principle of abstraction implies that the reason or cause for the OPO is of no consequence in the banking relationship between the payer and the paying bank (as the order is accepted by the paying bank in the course of its ordinary banking business). The transaction between the payer and the payee is independent and separate from the transaction between the payer and its bank, which must be judged on its own merits.⁹⁴ In the event that the payer (p. 272) could nevertheless successfully claim vitiated intent as against the paying bank, the transaction and the resulting debit to the account can be avoided ab initio through rescission of the OPO.

6.67  Given that the account constitutes a contractual relationship, as a general rule those who are not privy to the banking relationship can assert no rights. Nevertheless, a third-party beneficiary might have an equitable interest in the balance and be able to base a claim on the fact that the giving of the OPO constituted a breach of trust or fiduciary duty. It is well established that a banker need not be unduly concerned as to the power of an authorized signatory that is acting in a fiduciary capacity to give valid payment instructions. For instance, in Gray v Johnston,⁹⁵ the beneficiaries of an estate sued the bank over the transfer of moneys from the estate account to the executor’s personal account. It was held that, for a claim to exist, among other things there must be proof that the bank is privy to the intention to misapply the funds.⁹⁶ Accordingly, a bank will not normally be liable, as constructive trustee, for carrying out a payment instruction that was given fraudulently by a trustee, a director, or an employee, other agent, or fiduciary who vis-à-vis the bank is authorized to give payment instructions in relation to a certain account, unless there is dishonest assistance.⁹⁷ The bank would need to be an accessory, that is, assist dishonestly,⁹⁸ to become liable, as constructive trustee, for the execution of a fraudulent payment instruction.

6.68  If the bank accepts the OPO, the payer must fund the execution cost and pay the bank the amount concerned.⁹⁹ Funding may occur in various ways, but it is customary in a banking relationship that the payer’s obligation to the bank is discharged by way of a debit to the payer’s account. Note that the mutual payment obligations that result from transactions between the bank and its customer, and by the bank for its customer, which are within the scope of the account relationship are netted continuously and instantaneously.¹⁰⁰ The acceptance by the bank of the OPO is the time at which the amount becomes due from the payer to the bank. It follows that the time at which the bank accepts the OPO is the time at which the debit to the account arises, not the time at which the bank completes its internal recordkeeping system process.

(p. 273) 6.69  The payer’s bank must generally execute a valid payment order from a customer promptly upon receipt.¹⁰¹ The number of the payee’s account enables the payer’s (paying) bank to identify the payee’s (receiving) bank. Where the payer and the payee hold accounts at the same bank, the original payment order can be executed by means of an in-house transfer. In that case, the bank performs the dual role of paying bank and receiving bank. In other cases, the paying bank will have to execute the original payment order by giving the receiving bank a corresponding payment order, the CPO, directly or indirectly, through a correspondent bank. If the receiving bank accepts the CPO, the paying bank owes the amount concerned to the receiving bank and must make a corresponding payment.

6.70  The giving and acceptance of the OPO, and the giving and acceptance of the CPO, are independent and separate transactions that give rise to independent and separate contracts between the party that gives the payment order, and the party that accepts it. Although the CPO is given pursuant to the OPO, the paying bank does not give the CPO as agent on behalf of the payer, but on its own behalf. Notwithstanding the customary reference in judicial authorities and academic writings to a bank as agent regardless of the type of payment service,¹⁰² it cannot be maintained that a bank that has agreed to provide payment services by funds transfers (rather than by payment or collection of cheques, which is of an entirely different character) has been given agency authority to carry out funds transfer services. Agency authority is the power that the law assigns to a person, the agent, by reason of the consensual legal relationship between the agent and another person, the principal, to effect the principal’s legal relations with third parties, that is, with persons other than the principal and the agent, including relations with such parties with regard to property.¹⁰³ There is no indication that it is the common intention of the bank and its customer as parties to an account agreement to create an agency relationship in connection with the execution of payment orders. Not everyone who agrees to carry out an instruction that involves making contracts benefiting the party that gave the instruction intends to make that contract as agent. Banks participate in payment systems as principal and do not intend to act pursuant to an agency relationship. Unlike the contracts made by brokers as principal on-exchange, a bank does not mean to give a CPO to a receiving bank as agent for the payer. The bank undertakes to the payer to make efforts to arrange for the account of the payee to be credited. The bank does not undertake to make efforts to bring about a relationship of any description between the payer and a third party. The affairs between (p. 274) a paying bank and a receiving bank, or in case of an in-house transfer, the payee, are not the affairs of the payer. Accordingly, there is no relationship between the receiving bank and the payer, or, with an in-house transfer, between the payee and the payer, at least not on account of the mandate of the paying bank, nor can the payer exercise any rights as principal in relation to any payment transaction executed by the payer’s bank pursuant to the OPO, or vice versa. The receiving bank cannot claim the amount due pursuant to the acceptance by it of the CPO, from the payer. Equally, the payer has no rights in respect of the transaction executed by the paying bank and may not intervene. To the extent that the account agreement gives rise to fiduciary duties between the payer and the bank as paying bank, this cannot be argued on grounds that the account agreement creates consensual agency, nor does any fiduciary relationship that may exist on other grounds justify the conclusion that there must be agency.

6.71  Payment by the paying bank of the amount due pursuant to the acceptance of CPO by the receiving bank can take place in one of two ways. It may be that the paying bank maintains an account at the receiving bank. Banks refer to cash accounts held in their name at other banks as ‘nostro’ accounts and to cash accounts opened for other banks as ‘loro’ accounts. Payment due to the receiving bank by the paying bank will be effected by means of a debit to the paying bank’s loro account, the nostro account from the paying bank’s perspective. In those circumstances, the relationship between the paying bank and the receiving bank is that of account holder and bank, respectively. The modern volume of payment flows, however, requires centralization of inter-bank payments. It would be impracticable, if not operationally impossible, for each bank to operate a myriad of nostro and loro accounts to facilitate payments by and to its customers. Accordingly, the banking industry has developed payment systems. These are formalized arrangements between a provider of centralized clearing and payment services and the member banks, often called ‘clearing banks’. The payment system enables paying banks to make payments to a receiving bank by using the services of the payment system. If the payee’s bank is not a clearing bank in the relevant system, for example because the receiving bank is a foreign bank, the paying bank may seek to make the payment to a correspondent-clearing bank of the payee’s bank. In that case, the payee’s bank must maintain a nostro account with that correspondent-clearing bank. The latter bank will transfer the payment from the paying bank to the payee’s bank by crediting the relevant account. The correspondent bank is then the receiving bank from the viewpoint of the payer’s bank, and then becomes the paying bank from the viewpoint of the payee’s bank, which is the receiving bank in that relationship.¹⁰⁴

(p. 275) 6.72  The procedures of the payment system require the clearing banks’ payment orders to be centralized in the system’s clearing house, usually by means of an electronic messaging system. The clearing house will process the payment orders and determine the amounts payable between the clearing banks.¹⁰⁵ The amounts are paid by in-house funds transfer carried out on the books of the central account provider. The collection and processing of payment orders is called ‘clearing’. The payment of the payment orders by the central account provider, on a net or gross basis, is called ‘settlement’ and the central account provider is called the ‘settlement agent’ or ‘settlement bank’. In most jurisdictions, the central bank will function as the central account provider and settlement agent in the domestic payment system. Each commercial bank will have an account with the central bank. In their turn, each central bank has an account with the Bank for International Settlement (BIS). Thus, the central banks normally act as settlement agents in relation to domestic payments, and BIS act as settlement agent in relation to international payments.

6.73  Just as regards the payment by the payer to the paying bank, it is difficult to imagine circumstances in which a paying bank that has sent a payment order will be able to claim successfully against the recipient of that order that its intention to give that order was somehow vitiated. If at all, the most likely scenario would be mistake. If the paying bank could nevertheless successfully claim vitiated intent as against the receiving bank, the transaction can be avoided ab initio through rescission of the transaction that resulted between the paying bank and the receiving bank pursuant to the acceptance by the receiving bank of the CPO. If the payment occurred in the nostro account of the paying bank with the receiving bank, the resulting debit to the account would be reversed as well. If the payment was made through a payment system, the paying bank might seek a proprietary right in the balance of the receiving bank’s account at the settlement agent. As the paying bank would be tracing a legal right, the balance must be a clean substitute, and may not be a mixture.¹⁰⁶ A payment system, whether operated on a net or a gross settlement (p. 276) basis, will make payments to the receiving bank’s omnibus account. Thus, unless the common law and equitable tracing rules merge, in normal circumstances it will be difficult for a paying bank to trace a payment made via a payment system. The balance in the recipient account would constitute a mixture and therefore would not be a traceable asset for purposes of the common law tracing rules.

6.74  A third party, such as the payer or the beneficiary of the payer, might seek to assert a right against the receiving bank in relation to the amount paid to it by the paying bank, based on the receiving bank having received trust property disposed of by the paying bank, or the payer, in breach of trust, knowingly and to its benefit. If the receiving bank receives the payment in the normal conduct of its payment services, a claimant will not succeed easily. From Millett J’s dictum in Agip (Africa) Ltd v Jackson, banks and their clients may take comfort that a receiving bank will not be regarded as benefiting from a receipt unless the amount was credited to an account that showed a debit balance, or the bank acted unconscionably.¹⁰⁷ That it will be difficult to hold a receiving bank liable for the receipt based on lack of bona fides follows from Macmillan Inc v Bishopsgate Investment Trust plc (no 3),¹⁰⁸ which permits a bank to assume that a receipt is bona fide unless there are facts that ‘made it imperative for [the bank] to seek an explanation because in the absence of an explanation it was obvious that the transaction was probably improper’.¹⁰⁹ In other words, similar to a paying bank that receives a payment by a debit to an account pursuant to the acceptance and execution of a payment order from a customer, a receiving bank that accepts a payment for credit to a payee’s account in the ordinary conduct of its banking business will not be considered to be acting unconscionably.¹¹⁰ Accordingly, a problem would normally only arise if the claimant managed to establish a claim¹¹¹ before the receiving bank makes the payment to the payee,¹¹² and that is unlikely in practice.

6.75  Once the receiving bank has accepted the CPO, it will have to credit the payee. Here, too, it is often suggested that the payee’s bank acts as the payee’s collection (p. 277) agent,¹¹³ but that position cannot withstand scrutiny in the context of funds transfers. The acceptance of the CPO by the receiving bank is a function of the relationship between the paying bank and the receiving bank, not of the account agreement between the payee and the receiving bank. Often the bank is not instructed by the payee to accept the payment.¹¹⁴ Only after the receiving bank has accepted the payment order, which is a decision that it makes independently from the account terms, without prejudice to its duty not to reject the payment unless it has good reasons to do so, does it become obligated to the payee to give a credit. There is no indication that parties to an account agreement share common agreement to grant agency authority to the receiving bank. Under the account terms, the receiving bank undertakes to credit the payee, should it accept a payment order from a third party, whether in-house, another account holder, or a paying bank. A bank, as receiving bank, does not undertake to bring about a relationship of any description between the account holder as payee and a third party in connection with a payment order given to it by that third party. The affairs between the receiving bank and the third party are not the affairs of the payee. Accordingly, there is no relationship between the third party with whom the receiving bank is dealing and the payee. The payee has no rights in respect of the contract made between the receiving bank and the third party pursuant to the payment order that the receiving bank accepted, and the payee may not intervene. To the extent that the account agreement gives rise to fiduciary duties between the account holder as payee and the bank as receiving bank, those duties do not imply that the account agreement creates consensual agency.

6.76  When the receiving bank gives credit to the payee, the funds transfer is complete. The question must be asked, at which point the receiving bank may be considered to have credited the payee. Three moments in the cycle appear relevant in this respect: the time at which the receiving bank accepts the CPO, the time at which it receives payment from the paying bank, and the time at which it completes its internal recordkeeping processes so that the funds are de facto available to the payee for use.¹¹⁵ The decision in Momm v Barclays Bank International Ltd suggests that credit is given at the time the bank accepts the payee as creditor for the transferred amount, not at the time the bank actually completes its internal recordkeeping (p. 278) process and makes the funds available.¹¹⁶ The bank in question, Barclays, had received a payment order from one account holder, Herstatt, another bank, to credit the account of another account holder, Momm. Barclays initiated the in-house transfer notwithstanding the fact that the balance of Herstatt’s account did not cover the payment order, but then became aware of Herstatt’s insolvency. Barclays subsequently decided to reverse the credit process. It was held that Barclays was not entitled to refuse to give Momm credit as it had made a decision to accept Momm as creditor for the amount, which acceptance had manifested itself through the initiation of the internal recordkeeping process.

6.77  Thus, at English law, subject to the provisions of the account agreement, the bank’s decision to accept a CPO, or, in the case the payer and the payee have the same bank, the OPO, and credit a payee may be sufficient to render the bank liable to the payee for the amount in question.¹¹⁷ Momm v Barclays concerned an in-house transfer, but nothing suggests that the principle that the decision to give a credit commits the receiving bank to the payee for the relevant amount should not apply in the event of a CPO that is accepted through a payment system. The corollary of that approach is that the paying bank has executed the OPO, and discharged its duties in that respect, at the time that the receiving bank accepts the payment for credit, not necessarily at the time that the funds are made available to the payee.¹¹⁸ It also follows that the credit to the payee’s account does not depend on receipt of payment by the receiving bank. This conforms to clearing and settlement practices. It is not always possible to match a CPO with a receipt before giving credit because payment systems often operate on a net settlement basis. Subsequent reconciliation procedures may show that the payment failed, in which case the receiving bank will reverse the credit, which it will typically have reserved the right to do in the account agreement.

6.78  In determining whether to give a credit in relation to a CPO,¹¹⁹ the receiving bank must use care and diligence so as to accept or to reject the paying bank’s payment order, promptly.¹²⁰ If the receiving bank fails to accept the CPO and to credit the payee when it should have, it may be liable for breach of contract and the payee may sue for money had and received.¹²¹

(p. 279) 6.79  Finality of the credit by the receiving bank to the payee is a function of the ability of the receiving bank to reverse that credit. In principle, unless contrary terms apply, the receiving bank’s decision to accept a payment order that names the payee as beneficiary and to credit the payee is sufficient to render the bank liable to make that payment. The named payee will have become the bank’s creditor for the amount in question.¹²² Unless a fundamental mistake has been made regarding the identity of the payee or the amount payable, it is not conceivable that the receiving bank will be able to claim successfully that its intention to accept the payment order and to give a credit was somehow vitiated. Just as the paying bank against the payer (or vice versa), the receiving bank cannot avail itself of any remedies, claims, or defences against the payee that are available to the payer. The root of the payee’s entitlement against the receiving bank is the account agreement. The principle of abstraction implies that the cause for the OPO, or any CPO, is of no consequence in the banking relationship between the receiving bank and the payee, as the order is accepted by the bank in the course of its ordinary banking business. The terms of the account agreement might specify the time at which credit is given, or any reversal rights that the receiving bank may have. In particular, it is customary for the terms of the account agreement to stipulate that the bank may reverse any credit if it cannot collect the corresponding payment.

6.80  At each completed transaction in the funds transfer cycle, it may technically be argued that payment occurred in relation to the payment obligation the payer is seeking to pay. The judicial authorities are not conclusive. There is an interesting difference between the cases that had to determine what constitutes payment by the paying bank for purposes of discharging that bank’s duty to execute the payment order, and the cases that had to determine what constitutes payment by the payer for purposes of discharging the payer’s debt to the payee.¹²³ The former suggest that the paying bank has performed its duty to the payer to make a payment to the payee’s bank when that (receiving) bank gives credit to the payee.¹²⁴ The latter, however, suggest that the payer has performed its duty to the payee to make a payment when the payee’s (receiving) bank not only has credited but also has unconditionally made the funds available to the payee.¹²⁵ It could therefore be considered to be unlikely that a uniform rule will be developed in this respect.¹²⁶

(p. 280) 6.81  It should be noted, though, that the cases suggesting that payment occurs between payer and payee only when funds are unconditionally available to the payee were concerned with charter contracts that required payment ‘in cash’ and stipulated that time was of the essence. Accordingly, it was held that nothing short of unconditional use would be sufficient. It is suggested that this rule might not constitute proper law or even apply in circumstances where the terms of the relevant transaction between payer and payee, whether express or implied, provide that a funds transfer is a valid means of payment. In those circumstances, if there are no express terms relating to the timing of the availability of funds, the better position is that the risk relating to completion of the funds transfer cycle passes to the payee, so that payment occurs, at the time that the payee’s bank or its correspondent bank is put in a position where it ought to credit the payee.¹²⁷ The law of funds transfers should take into account that the choice of bank by the payee is a relevant factor in allocating the risk of failure or delay during the funds transfer cycle. In particular, the risk of insolvency of the payee’s bank (or its correspondent bank) after completion of the payment by a paying bank, that is, at the time the funds transfer cycle has moved to the payee’s leg of the structure, should be assumed by the payee. However, the cycle cannot be said to have passed to the payee’s side of the funds transfer chain properly for this purpose, until the payee’s (receiving) bank has received payment from the paying bank or its correspondent bank. It would not appear justified to discharge the payer on grounds that the payee’s bank has credited the payee in relation to a payment order in anticipation of payment. That is a matter for the banking relationship between the payee and its bank, not the transaction between payer and payee. Therefore, discharge of the payer where the receiving bank has credited it is subject to the condition subsequent that the receiving bank has unconditionally received payment in respect of the paying bank’s payment order.¹²⁸

6.82  If the transferring and the receiving parties each hold intermediate securities accounts at separate custodians, the book-entry securities transfer involves the following operational stages.

1. (a)  The transferor initiates the book-entry securities transfer by sending a transfer order, called a ‘settlement instruction’, to its custodian, which must be validated and accepted by the custodian in accordance with the terms and conditions of the securities account agreement. Differently from a funds transfer, the validation cannot take place unless the transferee sends a collection order, known as a ‘matching settlement instruction’, to its custodian so that the transferor’s custodian and the transferee’s custodian can match and compare settlement instructions. The transferor’s and the transferee’s settlement instructions will be referred to in this section as the ‘transferor’s settlement instruction’, or ‘TorSI’, and ‘transferee’s settlement instructions’, or TeeSI’, respectively.

2. (b)  If the transferee’s custodian confirms that it will accept settlement based on a matching TorSI and TeeSI, both custodians will give matching settlement instructions to the relevant central securities depository (CSD) after which the CSD will carry out an in-house transfer by way of book-entries. The custodians’ subsequent settlement instructions to the CSD will be referred to as the ‘CSD settlement instructions’, or ‘CSD SIs’, respectively, in this section.

• View full-sized figure

Figure 6.4  Securities settlement system infrastructure

6.83  Figure 6.4 shows the structure of a book-entry securities transfer involving two custodians, a sub-custodian, and a securities settlement system.

(p. 282) 6.84  A book-entry securities transfer starts with an instruction from the transferor to its custodian, the TorSI, to place a certain quantity of book-entry securities at the disposal of a specified transferee. If it is accepted that an account holder takes an absolute, vested, and indefeasible share in the undivided capital and interest of the trust property as equitable tenant in common,¹²⁹ it follows, in accordance with the rule in Saunders v Vautier,¹³⁰ that the TorSI constitutes a demand from the transferor, as account holder and beneficiary, to the custodian as bare trustee to divide the trust property partially and deal with the property as instructed.¹³¹ The rationale of the rule in Saunders v Vautier is that the beneficiary of a bare trust is entitled to direct the trustee to deal with the trust property as the beneficiary wishes. The application of the rule is subject to any rights the trustee may have against the assets. An absolute entitlement to a pro rata share of undivided fungible trust property generally entitles the beneficiary to demand a transfer of a proportionate quantity of the fungible mass.¹³² Division must necessarily take place at the time at which property is transferred out of the pool in execution of the TorSI. At that time, the pool is partially divided and the transferor’s share is reduced. A debit to the account does not necessarily constitute the division because the debit does not necessarily coincide with the transfer out of the pool. Thus, a TorSI, like an OPO, is a request by the account holder to the account provider that, if accepted, results in a mandate to provide a transfer service in accordance with the terms of the account agreement.

6.85  It is unlikely that the TorSI and the resulting division may be challenged successfully by the transferor as against the transferring custodian on the basis that the intention to give the TorSI was vitiated by some factor. The principle of abstraction applies to a TorSI, so the fact that the transferor made a mistake in respect of the underlying securities settlement obligation the transferor is seeking to discharge, or the fact that the transferor was unduly influenced by the transferee in assuming the securities settlement obligation or giving the TorSI, is not a factor (p. 283) that is relevant between transferor and its custodian as the transferring account provider except, perhaps, in special circumstances.¹³³

6.86  A third party may have a right in the balance of the transferor’s securities account as beneficiary and seek to claim against the transferor’s custodian based on the fact that the giving of the TorSI constituted a breach of fiduciary duty. Given the functional and purposive similarities between a TorSI and an OPO—both are requests under a contractual relationship, the securities or cash account, that result in a mandate to provide electronic book-entry securities or funds credit transfer service in accordance with the provisions of the account agreement—it appears right and appropriate to apply judicial precedent that relates to funds transfers by analogy. Accordingly, a custodian or CSD should not be unduly concerned about the power of an authorized signatory that is acting in a fiduciary capacity to give valid transfer instructions.¹³⁴ Only if the actions or inaction of the custodian or CSD amounts to dishonest assistance should there be liability as a constructive trustee.¹³⁵

6.87  The transferor’s custodian as trustee will have a duty to use reasonable skill and care in accepting and carrying out the TorSI, having regard in particular to the special knowledge and experience that may be expected of a professional account provider.¹³⁶ The settlement processes require that both the transferor and the transferee must give matching settlement instructions to their custodians. The TorSi and the TeeSI are usually given upon confirmation by the broker-dealer of the market transaction in question.¹³⁷ The transferor’s custodian will start the execution of the TorSI by giving the transferee’s custodian a notice of settlement, which will be rejected if the notice cannot be matched with a corresponding TeeSI. Once the transfer has been cleared by the receiving custodian, both the transferring and the receiving custodian will give matching settlement instructions to the CSD, the CSD SIs. Usually, the CSD offers services that are comparable to the services offered by a clearing house in connection with payment systems. The settlement notice by the transferor’ and acceptance by the transferee’s custodian will be carried out as part of a pre-settlement matching process that is initiated by the matching of the TorSI and the TeeSI and that will produce the CSD SIs. Upon matching the CSD SIs, the CSD will generate an in-house transfer by way of book-entries. The relationship between the transferor’s custodian and the transferee’s custodian, therefore, is different from the relationship between a paying bank and a receiving (p. 284) bank. The transferor’s custodian does not give an instruction to the transferee’s custodian and no transaction arises between them.

6.88  If the transferor’s custodian maintains an account at the transferee’s custodian, the transfer may occur by way of an in-house transfer across the books of the transferee’s custodian involving the account of the transferee. In practice, even in an in-house transfer, the transferee’s custodian will still generate matching CSD SIs, taking both sides of the transfer, acting de facto as the correspondent custodian, or sub-custodian, for the transferor’s custodian, and arranging a transfer via the CSD. The reason is that the transferee’s custodian will wish to take the benefit of the CSD’s operational infrastructure and minimize the risk of shortfalls and discrepancies. Almost all transfers will ultimately settle pursuant to matching CSD SIs across the books of the relevant CSD.

6.89  The position concerning a challenge by the transferor’s custodian or a third party as against the transferor’s custodian on the basis that the intention to give the CSD SI was vitiated by some factor is similar to the position of the transferor in the relationship with the transferor’s custodian discussed in para 6.85 above.

6.90  Upon completion of the CSD book-entry transfer, the pool of assets held by the transferee’s custodian will have swollen. It must be assumed under the terms of the account agreements that the additional assets will have accrued to the pool as trust property.¹³⁸ If the transfer is received pursuant to a TeeSI, the custodian may be presumed to have accepted the additional assets into the pool for the benefit of the relevant account holder so that its share in the pool will increase proportionately at the time of receipt. If this were different, shortfalls or surpluses in the fund will occur as a matter of operational certainty. It follows that a subsequent manifestation of assent by the custodian to the transferee/account holder is not required to create proprietary rights for the transferee in relation to the receipt. Equally, it must be concluded that the receiving custodian’s assent (as manifested by the settlement instruction to the CSD and the subsequent acceptance of the transfer) constitutes an assumption of contractual liability, too, so that liability arises as soon as the transfer in question is completed.

6.91  If the transferee’s custodian gives credit by mistake, and the transferee is a bona fide recipient without notice, the transferee as account holder may acquire an increased share in the pool. The erroneous notice of receipt must be distinguished from notices given pursuant to contractual settlement arrangements. In that case, the transferee’s custodian undertakes to credit securities to the transferee’s account on the day that the transfer is due under the transaction between the transferor (p. 285) and the transferee, subject to the right to reverse the credit if the transfer is not completed by the CSD within a reasonable period.¹³⁹

6.92  It is perhaps tempting to argue, analogous to the cases on funds transfers, that the transferor has performed its duty to the transferee to transfer property in book-entry securities when the transferee’s custodian, as its account provider, not only has given credit but also has unconditionally made the book-entry securities available to the transferee, that is, has completed its internal recordkeeping process.¹⁴⁰ However, as has been argued in this respect in relation to funds transfers,¹⁴¹ where it follows from the terms of the relevant transaction that the parties agreed that the transferor would transfer book-entry securities, unless there are express terms relating to the timing of the availability of the book-entry securities, the risk of completion of the transfer should pass to the transferee at the time that its custodian or that custodian’s sub-custodian is put in a position where it ought to give credit to the transferee. The law of book-entry securities transfers should consider that the choice of custodian by the transferee is a relevant factor in allocating the risk of failure or delay during the book-entry securities transfer cycle. Particularly, it seems that the risk of insolvency of a transferee’s custodian or its sub-custodian after completion of the book-entry securities transfer at the level of the CSD should be assumed by the transferee.

6.93  Liquidity is a condition precedent for the successful existence of a market in financial assets. However, it is not just a function of certainty of execution based on the levels and continuum of supply and demand. Liquidity is also very much determined by certainty of settlement ex post execution, and certainty of title transfer ex post settlement. A prospective buyer of securities might be able to execute a trade without delay or noticeable market impact, but unless the buyer can be reasonably certain that the seller will transfer property of the description and quantity that the buyer purchased at such time as may be expected under the terms of the transaction, the buyer will not be in a position to sell that property onwards until it has actually been received. Equally, if the buyer cannot be reasonably certain that the property is free of adverse claims upon receipt, it cannot sell that (p. 286) property onwards free of all claims. The likelihood that settlement may or may not occur, or that transferred property is subject to adverse claims, may be referred to as ‘settlement risk’ and ‘transfer-finality risk’, respectively. Both risks have an adverse effect on liquidity if they exceed a threshold level. The market infrastructure, thus, is designed to minimize settlement risks, where possible.

6.94  Settlement finality connotes the degree to which a transfer of an agreed value to a creditor in satisfaction of a transfer obligation may be challenged after completion.¹⁴² The law should not seek to protect a recipient of a transfer if that recipient has acted unconscionably or if the transfer would enrich the recipient unjustifiably. The degree of finality of a transfer, therefore, should be benchmarked against a transfer to a bona fide recipient without notice who has given value. A payment has a high degree of finality if a good-faith transferee for value may irrevocably retain the transferred value free of adverse proprietary claims.

6.95  It is well established that a funds transfer does not constitute assignment of a claim, nor novation.¹⁴³ A payment order is not an instrument or act of disposition but an instruction to the payer’s bank which, if accepted, results in a mandate to provide a payment service in accordance with the provisions of the account agreement.¹⁴⁴ Unlike the offering of legal tender or delivery of a bearer instrument, no property in an asset in specie, tangible or intangible, passes from the payer to the payee as a result of a funds transfer. Rather, such a transfer comprises a chain of (p. 287) payments: the payer pays the paying bank pursuant to the original payment order (OPO), which pays the receiving bank pursuant to the corresponding payment order (CPO), which pays the payee pursuant to both the acceptance of the CPO and the terms of the account agreement between the payee and its bank. Each time, the payment is made pursuant to an independent transaction, so that, at the end of the cycle, the payee becomes the receiving bank’s creditor for the amount in question and the payer is discharged.¹⁴⁵ The final payment is a function of a chain of causally linked but legally independent transactions whereby one party accepts a payment instruction from another party to make a payment to a third party.¹⁴⁶

6.96  Lord Millett in Foskett v McKeown observes in respect of the nature of a funds transfer that no money passes from the paying bank to the receiving bank, but that there is ‘simply a series of debits and credits which are causally and transactionally linked’.¹⁴⁷ It is probably not the debits and credits that are somehow linked, but the contracts for services made first between the payer and the paying bank, next between the paying bank and the receiving bank, and last between the receiving bank and the payee. The debits and credits are the consequential by-products of the method of performing the payment obligations but are not themselves the reason for the next stage in the funds transfer cycle. It is acceptance of the payment order that is the reason for the next step.¹⁴⁸ Each party that accepts a payment order at some point during the cycle could also agree, although not practically, to accept or make payment otherwise, which then would not result in a debit or a credit, but still result in payment. The credit to an account creates a new interest or increases an existing interest, in the hands of the account holder.

6.97  Unlike the credit of funds to a cash account maintained by a bank, which only creates a right in personam against the bank and no proprietary rights in funds received by the bank for credit to the transferee’s account, the credit of book-entry securities to a securities account maintained by a transferee’s custodian purports to create proprietary interests in the original securities or book-entry securities received by the custodian for credit to the transferee’s account. That, however, does not mean that somehow, interests in that property are assigned or otherwise pass from the transferor to the transferee. The reduction of client asset pools through (p. 288) the custody chain to the CSD following the execution of the transferor’s settlement instruction (TorSI) does not constitute a transfer of property interests from the transferor to the transferee. It has been argued that an in-house transfer at the level of the CSD, that is, a debit to the transferor’s custodian’s CSD account and a corresponding credit to the transferee’s custodian’s CSD account, amounts to assignment or novation.¹⁴⁹ Matching settlement instructions by the custodians to the CSD (CSD SIs) do not constitute an assignment, nor can the CSD SIs be said to constitute a novation agreement between the transferring custodian and the receiving custodian, in either case, because there is no transaction between the custodians.¹⁵⁰ There are matching but legally distinct and independent settlement instructions to the CSD to reduce one share and simultaneously increase another, which amount to a simultaneous division and accrual of trust property. No property passes from the transferor to the transferee. A TorSI is a request to partially divide the pool held by the transferor’s custodian and transfer the divided part, as book-entry securities, to the transferee’s custodian for the benefit of the transferee. The property in book-entry securities that is acquired by the transferee on completion of the transfer is created de novo. The analysis is not different when the TorSI is carried out in-house if the transferor and the transferee have accounts at the same custodian.

6.98  In each transaction in a transfer cycle, an adverse claim may originate from the transferor, based on a complaint that its intention to transfer the value was vitiated by some factor so that the transfer is void or voidable.¹⁵¹ An adverse claim might also originate from a third party, based on a complaint that the transferor misappropriated the funds or securities used for the transfer because the assets either were acquired pursuant to a void or avoided transfer, or, generally, used in breach of a fiduciary duty.

6.99  A valid and effective transfer of legal title to personal property requires that the transferor intends to pass property of a certain description, in a certain quantity, (p. 289) to a certain recipient. If the transferor by mistake has a false belief about any of these elements at the time the transfer is made, for example the settlement instruction identifies the wrong person, or the wrong amount, there is a ‘fundamental’ mistake. The mistake must go to the essential identity of the transferee, the property, or the quantity, not to false beliefs the transferor holds about certain qualities of the transferee, the property, or the quantity,¹⁵² which is a matter of misrepresentation, rather than mistake. A fundamental mistake negates the transferor’s intention to pass property and thus renders the transfer void. No title passes pursuant to a void transfer and the transferor may bring an action for conversion against the initial transferee, including in the recipient’s bankruptcy or liquidation, or anyone to whom the asset is subsequently transferred, but not against a bona fide purchaser of negotiable instruments for value.¹⁵³

6.100  If the transferor cannot follow the original property into the hands of any transferee or subsequent transferee and effect specific restitution because the original property cannot be identified anymore,¹⁵⁴ the transferor will have to rely on equity to give effect to the existing property right in some manner by preventing the recipient of the traceable substitute value from denying the rights of the original transferor, or from taking a benefit from that value while knowingly acting contrary to the rights of another.¹⁵⁵ In Chase Manhattan plc v Israel-British Bank (London),¹⁵⁶ the plaintiff bank sought to establish a proprietary interest in a balance held by the defendant, who had become the subject of insolvency proceedings, on grounds of a fundamental mistake, after having made the same payment twice. Goulding J held that the defendant could not have retained the money in good conscience, and hence, that a constructive trust arose at the time of receipt. The reasoning in Chase Manhattan was revisited by Lord Browne-Wilkinson in Westdeutsche Landesbank Girozentrale v Islington London Borough Council,¹⁵⁷ who noted that, in his view, Goulding J’s analysis that a proprietary right arises immediately on the bank’s receipt of mistaken payment was not supportable, but (p. 290) that the case was probably rightly decided, because the bank became aware of the mistake only two days after the transfer. At that point, the bank became bound as trustee under a resulting trust.¹⁵⁸ Thus, subsequent transferees, other than bona fide purchasers for value of the legal title, will be bound by that trust so that the transferred property or its traceable substitute may be reclaimed, even in the event that the holder is the subject of insolvency proceedings.¹⁵⁹ On the merits of the decision in Westdeutsche Landesbank, it would appear that the effective operation of resulting trusts in respect of mistaken payments of money will be limited to circumstances in which the recipient realized immediately upon transfer that the mistake was made, because until the recipient’s conscience is affected, the money can be used without limitation. Subsequent transferees will not be bound by a trust that does not exist at the time of the transfer.

6.101  If the intention to transfer property is formed under the influence of fraud, coercion, or misrepresentation, that intention is regarded as vitiated and the resulting transfer voidable. A voidable transfer is usually effective to pass property, but the party making the transfer will have a right at law or in equity to rescind it.¹⁶⁰ Rotherham summarizes the position as follows:¹⁶¹

The common law has traditionally taken a stricter approach to rescission than has equity in two respects. First, equity provides for rescission in response to a wider range of vitiating factors than does the common law. Thus, while at law rescission is available only for those misrepresentations that were fraudulently induced, equity equally allows for rescission in cases of innocent misrepresentations. In addition, equity provides for rescission of contracts vitiated by a common mistake that is ‘fundamental’—a criterion that is understood to set a lower threshold than that demanded at law. Similarly, in equity, a plaintiff may set a contract aside on the grounds of undue influence or unconscionable bargain. Secondly, equity takes a more flexible approach in determining when it is possible to unwind a partially performed transaction. At common law, the requirement that restitutio in integrum must be possible is interpreted to mean that rescission will not be permitted if judicial assistance is required to restore the status quo. For this reason, rescission will be denied if plaintiffs have received any benefit under the contract in question. In contrast, equity allows for rescission on terms that ensure that the status quo can be substantially restored.

6.102  The principle that operates upon rescission is that the transaction is retrospectively treated as void ab initio, so that the parties’ outstanding executory obligations are cancelled and any transferred property revests in the original transferor.¹⁶² The equitable rights that flow from the right of rescission are deemed to vest in the (p. 291) transferor at the time of the transfer.¹⁶³ Except in the case of a bona fide purchase for value, third parties who acquire property from the transferee take subject to any equities and are bound by the original transferor’s right of rescission. Upon effective rescission, the recipient holds the property on a bare trust for the transferor. There is some uncertainty as to the nature of the trust that arises upon rescission. The prevailing view appears to be that it is a resulting trust.¹⁶⁴ The trust operates prospectively, so that the recipient is not liable as trustee for dealings with the transferred property prior to the rescission,¹⁶⁵ but that is without prejudice to the fact that the recipient takes subject to the original transferor’s equitable rights.

6.103  As noted above, a funds or book-entry securities transfer does not constitute an assignment of a claim, nor novation. It is intrinsic to the structure of a funds or book-entry securities transfer that no original property passes hands. Therefore, no person can identify the original property in the hands of any transferee or subsequent transferee and effect specific restitution. A claimant will have to identify a substitute for the original property in accordance with the applicable tracing rules.

6.104  Lord Millett restated the concept of tracing in Foskett v McKeown:¹⁶⁶

6.105  Tracing, as an identification process, can perhaps be compared to the concept of remoteness in a claim for compensatory damages. On a strict causative approach, an asset may be a substitute for the original property, but tracing rules may limit the scope of the recovery. The body of tracing rules that applies is (still) dependent on whether the original property was owned at common law or in equity. The main difference is that the common law tracing rules limit the category of traceable products to ‘clean substitutes’. That means that the substitute property should not be, nor be the result of, a mixture with other assets, even if (p. 292) the balance has swollen with profits from various transactions.¹⁶⁷ Thus, in the scenario where the receiving bank opens a new account for the payee and credits the transferred amount to that account, resulting in a fresh claim of the payee against the bank for the credited amount, the balance would constitute a traceable asset for purposes of the common law tracing rules.¹⁶⁸ In the scenario where a debit balance results after the credit to the account, no traceable product remains.¹⁶⁹ If the receiving bank credits the payment to an existing account, thereby increasing a credit balance, that balance constitutes a mixture, and tracing will only be available in equity.¹⁷⁰

6.106  To be able to trace the product of a transfer of property in equity it is a prerequisite that the claimant had an equitable interest in the original property, or that the person who transferred the property away had a fiduciary relationship to the claimant.¹⁷¹ Book-entry securities create equitable co-ownership interests, but funds do not. This means that it will be difficult for the payer, or any other party in the funds transfer cycle who makes a payment, to trace the proceeds of a funds transfer based on the fact that its intention to transfer property was vitiated. Prima facie,¹⁷² it would be tracing a legal interest at law.¹⁷³ Where the payment is instructed by an agent on behalf of the payer, the situation may be different.¹⁷⁴ In circumstances where an investment manager instructs the investor’s custodian to (p. 293) make payments pursuant to a transaction that exceeded the manager’s authority, it may be a breach of a fiduciary duty. However, where the investment manager simply instructed in error it may be a mere breach of duty of care, but not necessarily a breach of a fiduciary duty. An effective adverse proprietary claim on the account of the payee would have to originate from a third-party beneficiary and be based on the fact that the payer acted in some fiduciary capacity and transferred the funds used for the transfer in breach of that fiduciary duty. The payer could be bound as a trustee under an express trust, or a trust imposed by law if it has taken funds pursuant to an earlier transfer that was rescinded, or (if it knew this) that was void on grounds of a fundamental mistake, in which case a beneficiary might seek to employ the equitable tracing rules in respect of the funds transferred by the party that is making the payment.

6.107  Equitable tracing claims fail if the traceable product ceased to exist, for example where the bank account into which the money is traced has gone overdrawn.¹⁷⁵ However, this does not mean that, if the money had been paid into that bank account and subsequently transferred to another account, it cannot be traced into the next account.¹⁷⁶ On balance, the authorities appear to permit equitable tracing of transferred funds into an account, even if that account is frequently used for receipts and payments by the (recipient) account holder. The funds can be said to be mixed in the sense that the value derived from the credit to the account is mixed with the total value represented by the balance of the account, which belongs to the account holder. It becomes a question of allocation between the account holder and the party claiming a proprietary interest in the balance on account of the offensive funds transfer. The long-standing rule is that, if the mixture comprises value attributable both to the equitable claimant and to the (constructive) trustee who is breaching its duty by mixing the money, then it is presumed that any value which is dissipated from the mixture is attributable to the trustee.¹⁷⁷ If there are multiple equitable claimants, priority must be established and shortfalls must be allocated.¹⁷⁸

6.108  Certain conclusions might be drawn in respect of void or voidable transfers, and in respect of transfers in breach of trust. First, assuming the original property can be identified, to assert a proprietary right in the product of the transfer, the original transferor (or the beneficiary, as the case may be) will have to comply with the tracing rules, as the original property will have ceased to exist. The original transferor of funds will need to apply the common law tracing rules and, if the funds have not been credited to a separate account, will not be able to identify the product of the funds transfer as it will have been credited to an account that is generally (p. 294) used to make and receive payments, so that the proprietary right is lost through mixture. On the other hand, a transferor of book-entry securities, or, in case of a funds transfer, the beneficiary of an express, resulting, or constructive trust may trace funds into a mixture, in accordance with equitable tracing rules, and can thus identify the product of the funds transfer and establish an equitable right.

6.109  Second, the transferee of a funds transfer will be taking a legal estate on the balance, so that any equities would be defeated if it is a bona fide recipient for value without notice.¹⁷⁹ This means that, if the defrauded original transferor can somehow identify a clean substitute in accordance with common law tracing rules, the claim will succeed.¹⁸⁰ In that respect, funds transfers would appear to occupy a position that is different from transfers of physical cash, which are recognized as negotiable and thus subject to an exception to the principle that is expressed by the maxim nemo dat quod non habet, so that bona fide acquisition for value repairs title defects. The exception has grown out of economic necessity. As Best J observed in Wookey v Pole:¹⁸¹ ‘By the use of money the interchange of all other property is most readily accomplished. To fit it for its purpose the stamp denotes its value and possession alone must decide to whom it belongs.’¹⁸² The rule ought not to be different for funds transfers. In economic life, a funds transfer has the same purpose as cash, that is, coins and notes. The degree of finality afforded to each transferee in the funds transfer cycle ought to be similar to that bestowed by the law on bona fide recipients for value of cash and, indeed, other negotiable instruments. Without prejudice to obligations arsing as a matter of financial crime regulations, there should be no obligation on the recipient to research the title of the transferor or the source of the funds unless there are facts that ‘made it imperative for [the recipient] to seek an explanation because in the absence of an explanation it was obvious that the transaction was probably improper’.¹⁸³ For funds to perform the intended economic function they must be liquid, that is, be a value that, once received in good faith and for value, may be retained and transferred without delay and without concern. JH Sommer writes in this context:¹⁸⁴

Commercial law has an essential, usually invisible, role in providing liquidity. … What if Lord Mansfield had decided Miller v Race[¹⁸⁵] the other way? In such a case, (p. 295) payments would never be final and could be reversed if the payor had not had good title to funds. If a payment could be reversed, it is scarcely liquid. When it came time for the recipient to make a payment, it would not know if it could rely on the fund it had received. … Such a transfer would scarcely be liquid, ie fully usable without delay.

6.110  As equity protects the bona fide transferee of the legal estate, if for value without notice, but not the transferee of an equitable interest, the hybrid character of the rights of a recipient of book-entry securities leads to a curious situation. A bona fide transferee for value can defeat any equities in respect of the rights in personam on the balance, but not equities in respect of the equitable co-ownership rights. Further, just as with the rather inconsistent implication of the application of traditional principles to a funds transfer, where bearer instruments are concerned, the transferee could have defeated even a defective title had it acquired the original securities, as bearer securities are negotiable instruments. Most commentators take the view that registered securities and bearer securities in electronic form have not yet acquired, or cannot acquire, negotiable status.¹⁸⁶ These objections ignore the function of securities in commerce. Negotiability enables a transferee to rely on the physical embodiment of the original promise to the bearer without further need to research the title, but the original promise is embodied in the paper so that it can be transferred without much ado. Negotiability is the natural consequence of the commercial use of the underlying debt, not of the fact (alone) that the debt is embodied in paper. In the interest of liquidity in the markets, the law on book-entry securities transfers should recognize the need for protection of an innocent transferee for value.

6.111  An important source of finality risk in connection with a fund or book-entry securities transfer is insolvency of the transferor, or indeed, the transferor’s bank or custodian. Under section 127(1) of the Insolvency Act 1986, on avoidance of property dispositions, in ‘a winding up by the court, any disposition of the company’s property … made after the commencement of the winding up is, (p. 296) unless the court otherwise orders, void’. For the purpose of section 127, the winding-up is deemed to commence on the presentation of the petition on which the winding-up order was made,¹⁸⁷ unless the company was already in voluntary liquidation at the time of the petition, in which case the winding-up is deemed to commence on the passing of the resolution for voluntary winding-up.¹⁸⁸ The problem is that the section does not distinguish between preferential payments,¹⁸⁹ payments at undervalue,¹⁹⁰ and bona fide payments. Any payment order that is given by a company at or after a winding-up petition has been made in respect of that company constitutes a disposition for purposes of section 127 of the Insolvency Act 1986,¹⁹¹ rendering the payment order void. Technically, that company cannot even receive money into its bank account. If the account is in credit, receipt constitutes conversion of a debt owed by the debtor to a debt owed by the bank, which is a disposition.¹⁹² If the account is overdrawn, payment into the overdrawn account results in a disposition of the receipts to the bank, which is (p. 297) caught by section 127.¹⁹³ Consequently, any payments or transfers instructed as of the moment the winding-up petition is made could be void, effectively paralysing the company’s business,¹⁹⁴ which is a considerable source of payment-finality risk for innocent recipients.

6.112  Section 127 provides that a post-petition disposition not sanctioned by the court is void, but unlike sections 238 and 239, does not provide a method for recovery. This has to be determined by general law.¹⁹⁵ If the original property can still be identified, it simply means that the property may be followed and recovered.¹⁹⁶ In other circumstances, the law appears to remain somewhat unsettled.

6.113  If the payment order is given before commencement of the insolvency proceedings, but is executed after that time, rather than challenging the payment order itself, a liquidator may seek to challenge the resulting netting, that is, the validity of the corresponding debit to the account. This will be a difficult argument if the payment order stands, which must be assumed if the payment order had been accepted before the making of the winding-up petition. First, netting in current account is not legally distinct from the acceptance of the payment order. Netting occurs automatically if a payment order is accepted and therefore cannot be void or avoided separately from the transaction that caused it.¹⁹⁷ Even if that were not true and the netting were treated as legally distinct, assuming that the payment order is valid and enforceable, in any event the insolvent corporate payer’s payment obligation would be eligible for insolvency set-off.¹⁹⁸ If it is accepted (p. 298) that acceptance of the payment order denotes the moment at which the payer’s payment obligation becomes due, and that at that time, simultaneously, the payment obligation is netted in current account, it should not generally matter for the application of section 127 whether the bank has started execution or not. The transaction ceases to be executory at the time the netting is complete, which is the time at which the payment order is accepted. That position is without prejudice to the right of the payee, or a liquidator, to revoke a payment order in accordance with the terms of the account agreement.

6.114  If a winding-up petition has been made against the payee, it can receive money into its bank account because, if the account is in credit, receipt constitutes a conversion of a debt owed by the debtor, to a debt owed by the bank, which is not a disposition. If the account is overdrawn, however, the payment of proceeds into the overdrawn account results in a disposition of the receipts to the bank, which is caught by section 127.

6.115  A liquidator could potentially reject a transfer and seek to terminate the transaction between the transferor and the transferee under section 178 of the Insolvency Act 1986. Under section 178 a liquidator may ‘disclaim’ any ‘onerous property’, which means, in particular, any unprofitable contract, as well as ‘any other property of the company which is unsaleable or not readily saleable or is such that it may give rise to a liability to pay money or perform any other onerous act’.¹⁹⁹ A disclaimer ‘operates so as to determine, as from the date of the disclaimer, the rights, interests and liabilities of the company in or in respect of the property disclaimed’, but ‘does not, except so far as is necessary for the purpose of releasing the company from any liability, affect the rights or liabilities of any other person’.²⁰⁰ If a person incurs a loss or damage as a consequence of the operation of the disclaimer, s/he ‘is deemed a creditor of the company to the extent of the loss or damage and accordingly may prove for the loss or damage in the winding up’.²⁰¹ The power may affect executory contracts, not executed contracts. But if the liquidator can effectively disclaim a payment, s/he could potentially restore a contract to executory status by rejecting both the transfer out and the transfer in under section 127, and subsequently disclaim the transaction. This potentially puts ‘out of the money’ transactions at risk. The liquidator cannot disturb accrued rights and liabilities.²⁰²

6.116  The provisions of the Settlement Finality Directive²⁰³ substantially mitigate the risk of adverse claims due to insolvency of a bank, investment firm, custodian, CSD, or CCP that form or participate in a payment or securities settlement system. The Directive seeks to ensure that transfer orders from participants²⁰⁴ that enter into eligible systems continue to be settled and become final, regardless whether the sending participant has become insolvent or transfer orders have been revoked in the meantime. Accordingly, funds and book-entry securities transfers that are carried out through designated systems enjoy the benefit of finality protection as certain insolvency displacement rules are set aside in relation to participants in such systems. The Settlement Finality Directive aims to insulate designated payment systems,²⁰⁵ securities settlement systems,²⁰⁶ and CCP systems,²⁰⁷ from the application of certain adverse insolvency provisions in the event that insolvency proceedings are commenced in respect of a member of the system. To enjoy the benefits, a system must be designated by the Member State, the law of which is applicable.²⁰⁸ A system is eligible for designation as a ‘system’ and enjoys finality protection if it is subject to a ‘formal arrangement’ that is expressed to be governed by the law of a Member State and that sets out ‘common rules and standardized arrangements for the execution of transfer orders[²⁰⁹] between the participants’.²¹⁰ Participants must be banks, investment firms, public authorities, and publicly guaranteed undertakings, or ‘any undertaking’ whose head office is outside the EEA and ‘whose functions correspond to’ those of the banks or investment firms, which participate in a designated system and are (p. 300) responsible for discharging the financial obligations arising from transfer orders within that system.²¹¹

6.117  The key protections of the Settlement Finality Directive are set out in arts 3 and 5. These provisions seek to insulate the giving and the execution of ‘transfer orders’, as defined by the Directive, from an intervening insolvency. Article 3 provides in its first section that a transfer order shall be legally enforceable, even if the participant becomes insolvent after the order became irrevocable but before the in-house transfer is completed, provided that the order is received by the system before the insolvency proceedings commenced, or, where the system was unaware, provided that the process was completed on the day that the insolvency proceedings commenced. Article 3(2) provides that certain insolvency rules that allow for the setting-aside of transactions executed before the commencement of the insolvency proceedings shall not apply. In addition, Article 5 provides that transfer orders are irrevocable from the moment so defined by the system, that is, irrespective of the provisions of the applicable insolvency rules. Accordingly, the implementation of the Settlement Finality Directive alleviates much of the uncertainty around the effect of insolvency displacement rules in the event that a participant in a domestic payment system executes a payment order by instructing the payment system to carry out a funds transfer. The effect of section 127, and also sections 178, 238, and 239, of the Insolvency Act 1986²¹² has been much curtailed.