The return of client money in investment bank special administrations: an empirical assessment

1. Introduction

This article empirically investigates the return of client money in UK investment bank special administrations.¹ As a matter of common sense, the return of client money is supposed to be subject to the same rules applicable to that of client assets. Considerable differences are found, however. For example, just to name but a few, a primary pooling event only applies to client money, not to client assets²; client assets are distributed following a court-approved redistribution plan, whereas the repayment of client money can be done without a court-sanctioned scheme; in the Financial Conduct Authority Client Assets Sourcebook (commonly known as CASS), client money and assets are regulated in different chapters, with Chapter 6 (CASS6) on client assets and Chapter 7 (CASS7) on client money. As is precisely addressed by an English judge, ‘a separate regime’ governs how client money is returned.³

Some may question whether the client money regime is worth investigating in the first place, given that in the UK investment banking sector there is only client money of around £100 billion in total held by investment banks, in contrast to some £10 trillion of client assets in the forms of various financial products.⁴ However, the significance of the return of client money cannot be judged purely by its relative sizes, for at least two reasons. First, in some investment bank special administrations, the firm may not hold client assets, but all firms in special administrations manage client money. In other words, challenges to the return of client money are more prevalent than those to client assets. Second, more important, like the return of client assets, the return of client money will equally affect investor confidence, failing to address which may lead to considerable damages to financial markets.⁵ This is especially true for the UK plc, since around ten percent of its national government annual income relies on the financial sectors.⁶ In addition, in many investment bank special administrations, it is not uncommon to see hundreds of millions of pounds of client money trapped, and clients deserve a speedy relief.⁷

To understand how client money is returned in investment bank special administrations,⁸ this article dissects all 19 investment bank special administrations, a census survey,⁹ entered into during the first-10-year implementation from 2011 to 2021 under the UK investment bank special administration regime, a novel insolvency procedure, investigating how this law works in practice and recommending policy reforms.¹⁰ In particular, this article focuses on three general questions. First, how is the primary pooling event conducted in reality? since both the statutes and regulatory rules do not offer specific guidance. Second, are client money shortfalls as widespread as anticipated? and more important, are the current measures in tackling client money shortfalls appropriate? What policy recommendation can be made to reform the practice for efficiency and cost-reduction? Third, since the bar date, a deadline urging clients to lodge their client money claims in a timely manner, is believed¹¹ to be a material tool¹² in seeking a speedy return of client money, this article raises the question over whether this mechanism does play its role in accelerating client money returns.¹³ Through answering these questions, this article attempts to offer a comprehensive picture over how client money return is conducted in UK investment bank special administrations.

To this end, the rest of the article proceeds in four parts. Part 2 briefly explains the legal frameworks over how a client money pooling event is to be carried out and the treatment of client money shortfalls, considering the major developmental phases of the client money bar date. Part 3 reports how the data were collected. Part 4 presents and analyses the findings. Part 5 concludes, proposing that whether to conduct a client money pooling event should be rethought, that client money records of failed firms could be assumed to be accurate by default,¹⁴ otherwise unintended costs and delays are inevitable, and that the client money bar date was designed with fatal weaknesses and should be radically strengthened.

2. The legal frameworks of the return of client money

As is widely known, the enactment of the Investment Bank Special Administration Regulations 2011¹⁵ (hereafter referred to as the Regulations 2011), which formally created the UK investment bank special administration regime, was a legislative response to the lamentable delays of returning both client assets and money encountered in the bankruptcy of Lehman Brothers International (Europe).¹⁶ It is worth noting that although the Regulations 2011 marks the establishment of the UK investment bank special administration regime, a combination of rules, statutory and regulatory, governs how an investment bank special administration procedure is handled. This is especially the case on the return of client money.

The client money primary pooling event and treatment of client money shortfalls

When the Regulations 2011 was promulgated, ironically, it probably only had three provisions related to client money, one of them positive and the other two negative. First, r. 10 of the Regulations 2011 states that one of the special administration objectives is to ‘ensure the return of client assets as soon as is reasonably practicable’,¹⁷ and in this situation client assets include client money,¹⁸ suggesting that seeking a quick return of client money is part of this statutory objective.¹⁹ Second, r. 11 of the Regulations 2011 allows a bar date to be set up for client asset claims, but it states clearly that this mechanism does not apply to client money.²⁰ Third, r. 12 of the Regulations 2011 emphasizes that any shortfall of client assets of a particular description should be borne by affected clients pro rata, but it adds a caveat that this redistribution rule is only for client assets, not for client money.²¹

Arguably, there were two considerations of why the Regulations 2011 did not touch too much on client money. First, if not foremost, when the Regulations 2011 was deliberated, the critical litigations over client money disputed in the Lehman case were ongoing,²² as a result of which HM Treasury, which was in charge of drafting this legislation,²³ did not want to rush to make client money rules which might clash with any upcoming judicial decisions.²⁴ Second, some rules on the return of client money already existed in the CASS, the regulatory rule book made by the FCA and mandated by the Financial Services and Markets Act 2000,²⁵ and probably it might not be too inconvenient to rely on CASS to solve these issues; needless to say, amending CASS would be far easier than obtaining Parliamentary blessing if some rules need to be updated. Over client money, apart from the aforementioned statutory objective in seeking its speedy return, two major points about the return of client money can be summarized from CASS.²⁶

First, under CASS,²⁷ a primary pooling event over client money occurs where the firm (the terms firm and investment bank are used interchangeably in this article) fails.²⁸ This means that if the firm enters into special administration, the primary pooling event is triggered, although various other incidents may also activate one.²⁹ The firm may hold client money in a general client bank account or a designated one³⁰; under normal and solvent circumstances, for a client whose money is placed in the firm’s general client account, she does not have a claim to the money deposited in any designated one, and vice versa.³¹ However, a turning point takes place when a primary pooling event happens. When it occurs, the client money placed in both the general client money bank account and any designated one, if there is one, is notionally pooled together, which means that the previous boundary between the general client bank account and any designated one is removed.³² All clients are equal and can make their client money claims to the newly created client money pool.³³

Why is the client money primary pooling event needed? Presumably, some may believe that, as was seen in the Lehman case,³⁴ when an investment bank is in trouble, the shortfall of client money usually emerges; for fairness, all client money is to be pooled together so as to distribute it to all clients in an equitable way; this leads to the next major CASS rule on the return of client money.

Second, when distributing client money after the failure of an investment bank, under CASS 7A.2.4, pari passu is to be observed, which means that ‘each client receives a sum which is rateable to the client money entitlement.’³⁵ This is the fundamental principle to handle client money deficits. Arguably, this arrangement may have failed to consider all three logical possibilities as for the (mis)match between client money entitlements/claims and the size of the notionally pooled client money. One scenario is as what is envisaged by CASS, which means that there is a client money deficit, and in this situation pari passu is the only realistic and fair option. The second scenario is that there is an exact match between what is claimed and what is deposited in all related client bank accounts; this means that the firm behaved properly pre-administration, and all clients can get back what they could claim; pari passu is unnecessary in such an event. The third one is probably a surprising occasion: when the firm fails and enters into special administration, it turns out that there is more client money than is claimed by all clients; if this happens, and it indeed did happen and will be reported later in this article, should pari passu still be observed? The CASS designers seem to have not anticipated this.

Thankfully, these gaps were filled by the Investment Bank (Amendment of Definition) and Special Administration (Amendment) Regulations 2017 (hereafter referred to as the Regulations 2017)³⁶ which authorizes special administrators to use a post-administration reconciliation to either move money from the firm’s house accounts to the client bank accounts in the event of a shortfall or send a surplus in the client bank accounts to the firm’s house ones if there is more client money than clients are entitled to.³⁷

After ascertaining the aforementioned distribution rules, the investment bank special administration regime offers the bar date mechanism in order to expedite the return of client money.

Statutory/regulatory rules on the bar date over client money claims

The rules of the bar date over client money claims are generally developed in three stages. The first stage is in advance of the Regulations 2011. Before the promulgation of the Regulations 2011, it was the general corporate insolvency law applicable to investment bank insolvencies, and in particular it is common to see the application of the traditional administration procedure, as was witnessed in the Lehman case.³⁸ Similar to dealing with creditor claims, a bar date, customarily called the last date for proving,³⁹ is necessary to seeking a speedy distribution of client money. Given that there was no specific provision in the general insolvency law authorizing administrators to set a bar date, they looked for a court order under the inherent authority conferred to the court by the Insolvency Act 1986.⁴⁰

This is what happened in the Lehman case, where the joint administrators, in 2011, applied for a court order permitting the use of the bar date over client money claims; in this case, under the bar date terms, if a client did not submit a client money claim before the bar date the joint administrators could not be held accountable after distributing client money on the basis of information available⁴¹; namely, if a client’s client money was not recorded in the company’s books and if the client did not submit a claim before the bar date, the joint special administrators could lawfully disregard the client money claims of an unresponsive client.⁴² It is worth reiterating that during this phase it was largely subject to the court deciding whether a client money bar date could be used and what were the consequences of missing it.

The second phase is after the Regulations 2011 took effect on 8 February 2012⁴³ and before the Regulations 2017 came into force on 7 April 2017.⁴⁴ Essentially, during this period, the rules on the client money bar date remained intact, and the only difference is that it was operated under the shadow of the client asset bar date explicitly authorized by the Regulations 2011.⁴⁵ This means that to use the client money bar date the special administrator had to seek an advance court order as before.⁴⁶

After the Regulations 2017, given that the client money uncertainties had been long settled judicially,⁴⁷ as was commented by Mr Peter Bloxham,⁴⁸ an eminent expert commissioned by HM Treasury to review the investment bank special administration regime around the year 2013, ‘there was no overriding policy reason for excluding client monies from the bar date procedure,’⁴⁹ the client money bar date was formally inserted into the Regulations 2017.⁵⁰ This is the third and current phase of the client money bar date rules.

Unlike in the Regulations 2011 which largely avoids rulemaking on client money,⁵¹ the Regulations 2017, which is to amend the former, has an extensive coverage on client money. One technical nuance is worth noting: the phrase ‘client money’ only appears twice in the Regulations 2011 but can be identified 84 times in the Regulations 2017.⁵² More detailed rules are made on how to return client money in investment bank special administrations. Before the Regulations 2017, as discussed before, it was at the discretion of the court determining whether to use a client money bar date and its terms. After that, the rules are more administrator-friendly.

First, the Regulations 2017 authorizes that ‘if the administrator thinks it necessary in order to expedite the return of client money, the administrator may by notice set a bar date for the submission of client money claims.’⁵³ This means that special administrators do not need to ask for a court order before setting a client money bar date, which is to help improve efficiency.⁵⁴ This also means that whether to set a bar date is subject to the special administrator’s own judgement.

Second, if a client money bar date is used, under the Regulations 2017,⁵⁵ the special administrator is required to make a distribution to a late client money claimant who has missed the bar date but submitted a claim before the return of client money. This can make the special administrator’s job more difficult, simply creating more uncertainty. To make things from bad to worse, third, the Regulations 2017⁵⁶ adds that when making client money distributions, the special administrator must reserve a portion of client money for those who have not made a client money submission before the bar date and have not received any distribution before. By common sense, if a client did not make a client money claim, how can the special administrator calculate how much should be reserved? Also, logically, if a client has already received a distribution before, this means that this client’s client money claim has already been fixed or agreed by the special administrator, why should the special administrator pay extra attention to this client again in a subsequent distribution; the special administrator could simply rely on the previously-agreed sum to make a dividend to this client in future distributions.

Presumably, what the Regulations 2017 hesitates to clarify is whether the firm’s own records on client money can be relied on when calculating each customer’s client money entitlement.⁵⁷ Arguably, the overwhelming policy-makers’ attitude towards firms’ client money (including client assets) records is suspicion.⁵⁸ This article will use evidence later to prove how this attitude is unfounded and misleading, causing tremendous delays and costs to the detriment of clients and damaging investor confidence in this country.

Third, if the bar date was used but there is still unclaimed client money left, the Regulations 2017⁵⁹ permits the special administrator to set a hard bar date prodding unresponsive clients to make a final client money submission. If having failed to observe the hard bar date, the client is ultimately barred from claiming a share in the remaining client money pool, instead this client, real or imagined, is allowed to make an unsecured claim against the firm’s general estate afterwards.⁶⁰ After the hard bar date, any remaining client money will be transferred to the firm’s general estate, with the client money pool finally closed, which is to avoid wasteful delays.⁶¹ However, unlike the bar date, the hard bar date should get an advance court order before being laid out.⁶² When assessing whether to give an order approving the use of a hard bar date, the court will generally consider: (1) whether ‘the special administrator has taken all reasonable measures to identify and contact persons who may be entitled to’ client money, and (2) whether ‘there is no reasonable prospect’ for the special administrator to receive further client money claims.⁶³

All looks very well, and a seemingly sophisticated arrangement is designed to safeguard client money claims. But the question is that in reality whether the intended aims can be effectively achieved. And more important, what costs have been paid in fulfilling these objectives. These will be reported in the findings section.

3. The methods of collecting the data

As noted before, given that there is not a huge number of investment banks entering into special administration, this article has the luxury of doing a census survey. First, to identify which firm was placed into special administration during the studied period, the Gazette website was mainly relied upon, since many official special administration notices, including those of special administrator appointments,⁶⁴ initial meetings of creditors and clients⁶⁵ and of the bar date,⁶⁶ are legally required to be publicized there. Given that the Gazette website offers information advertised on all three publications, the London Gazette, the Belfast Gazette and the Edinburgh Gazette,⁶⁷ technically speaking, investment bank special administrations taking place across the UK could be found there. On the Gazette website, there is a special column dedicated to insolvency notices, and the phrases ‘investment bank’, ‘special administration’ and ‘special administrator’ were typed into the text search box, which generated a great number of notices on relevant cases. Of course, in many occasions, one case had many notices of different types advertised there, and each was manually checked.

It is worth noting that although the Regulations 2011 took effect on 8 February 2012,⁶⁸ this study counts the starting point of the first 10-year implementation of this special administration regime from 30 June 2011 instead, because the latter is the time when the Investment Bank Special Administration (England and Wales) Rules 2011 (hereafter referred to as the Rules 2011) became effective.⁶⁹ Without the support of proper procedural rules, this special administration regime could not operate effectively. Some may question the relevance of the related procedural rules in Scotland and Northern Ireland. As will be reported later, thankfully, all investment bank special administrations in the studied period took place in England. Therefore, the doubt of whether this study is incomplete in its jurisdictional coverage could be cleared. By this measure, this study includes all investment bank special administrations commencing between 30 June 2011 and 29 June 2021; the commencement of a special administration procedure is marked at the appointment of the special administrator(s).⁷⁰

Second, after identifying the firms that entered into special administration during the studied period, substantial information was obtained from the files uploaded by these firms to the Companies House website. Two types of files are critical to this study: one is the special administrator proposals, which routinely have one section on client money, and the second one being regular 6-month progress reports which inform the stakeholders as to the development of the case. On the Companies House website,⁷¹ after clicking on the icon ‘Find company information’, simply type the company name into the search box, which will return a specific webpage for the company. Wonderful treasures lie in the folder called ‘Filing history’ where the special administrator proposals and progress reports were faithfully uploaded, thanks to Companies House.

Third, although the bulk of the data were generated from Companies House materials, some supplemental, but more valuable, data were collected from many special administrator firms’ websites where the special administrators made them digitally available to feed transparency. Many large firms use their firms’ websites as a major venue for information exchanges with clients. Some documents, such as witness statements presented by special administrators to courts when seeking a special order, are not available on the Companies House website but are disclosed on the special administrator firm’s website, which tremendously enriched the quality of the data collected by this study.

Based on the three major sources above, this study collates them into a comprehensive data sheet comprising of: (1) the name of the company placed into investment bank special administration, (2) the date of the special administrator appointment, (3) the time of the client money primary pooling event, (4) the number of clients holding client money claims, (5) the total amount of client money claims, (6) the amount of client money in the client money pool, (7) the percentage of client money return, (8) the client money shortfall causes, if applicable, (9) the client money bar date, if established, (10) the client money bar date compliance, (11) the consequences of client money bar date non-compliance, (12) the number of client money distributions, (13) the client money hard bar date, and (14) the ways of closing the client money pool.

However, the data collected have certain flaws.⁷² First, there were some conflicting figures on the same subject; therefore, this study had to pick up a less erroneous one. For example, in the special administration of LQD opened in 2015, according to the special administrators’ proposals, the balance of the client money was £1,455,131.40,⁷³ but this figure became £1,407,615.70 in the February 2017 bi-annual progress report.⁷⁴ Faced with such conflicting information, this study discarded the old one and chose the new, although it should be emphasized that this mismatch might be largely caused by accounting adjustments and that the margin of errors seemed to be tolerable.

Second, some figures are actually an estimate. For example, shortly after entering into special administration, the investment bank European Pensions Management Limited successfully sold its holding of both client assets and money to a buyer, but after all available files were carefully examined, the figure of how much the client money was in this case could not be found; to fill the gap, this study obtained a figure of estimated client assets worth some £650 m reported in the administrators’ proposals⁷⁵ and the figure of an overall value of both client assets and money at some £1bn disclosed by one special administrator,⁷⁶ as a result of which the sum of the client money was estimated at £350 m. It must be clarified that such an estimate is only used in exceptional circumstances, and that most figures were directly obtained from the documents released by special administrators.

Third, some figures could not be obtained either through reading documents publicly available or by an estimate based on relevant data, as a result of which the data collection of this study is inherently incomplete.⁷⁷ For example, the special administration of Solo Capital Partners LLP was commenced in 2016; partly because of the ongoing criminal investigation, the majority of useful information, such as the amount of client money claims and the return rate of the client money, has not been disclosed by the special administrator; at the time of writing, this case has not been closed⁷⁸; it looks very unlikely for expected information to be released in the near future. Facing the lack of key information, this study will address this to remind the reader when summarizing the pattern or trend of each subject.

Backed by these data, this article confidently reports the findings.

4. The findings on the return of client money in UK investment bank special administrations between 2011 and 2021

Between 30 June 2011 and 29 June 2021, as shown in Table 1, there were 19 investment banks entering into special administration under the new regime, all of them taking place in England. Unsurprisingly, given London’s eminent position as one of the global financial hubs, 15 out of all 19 firms are headquartered in the capital. This also convinces the selection of the jurisdiction of England and Wales for this study, since Scotland, Northern Ireland and even Wales are simply irrelevant.

Of these 19 investment bank special administrations,⁷⁹ as presented in Table 2, all firms held client money, with six firms holding client money only. In the majority of the cases, the firm looked after both client assets and client money. Why did some firms only have client money in their custody? This depends on the business model of each firm.⁸⁰ A good example is Alpari (UK) Limited, a Financial Conduct Authority (hereafter referred to as FCA) regulated firm; unlike traditional brokerage firms usually holding both client assets and client money, the firm only held client money because it exclusively provided online financial services to customers,⁸¹ and the firm fell within the broadly defined category of investment banks and was subject to special administration when in trouble.⁸²

As shown in Table 2, the amount of client money in the firm’s custody varies considerably. The largest amount of client money of £724,390,000 is found in the special administration of MF Global, the first company using the new regime.⁸³ At the other end of the spectrum is the special administration of City Equities where client money only amounts to £59,090.⁸⁴ The average sum of client money held by a failed firm is £58,205,289, a huge figure justifying the importance of this study.

The client money listed in Table 2 is only what was identified when the special administrators were appointed. The final amount of client money available for clients is subject to further reconciliations. For example, in the special administration of MF Global, when the special administrators were appointed on 31 October 2011, the firm’s records showed that the client money balance was £724,390,000⁸⁵; but following the subsequent reconciliations, the total of client money was adjusted to £1,081,468,000 several months later.⁸⁶ To handle client money, for special administrators, the imminent problem would be whether to pool client money together.

The client money primary pool events

As for a client money primary pooling event, three specific issues arise: first, whether to pool client money? Second, when to pool? And third, how to pool?

As shown in Table 3, the primary pooling event occurred in 16 (84.21 per cent) out of all 19 investment bank special administrations; in the remaining three (15.79 per cent) cases client money was not pooled, that is, the client money pooling event did not take place at all. As noted earlier, under CASS 7A.2.2, the client money pooling event is supposed to be automatically triggered when the firm enters into special administration.⁸⁷ Why did it not happen in these three cases? The reason is very simple: the primary pooling event was unnecessary since there were no client money shortfalls identified. The assumption of pooling client money together is that there is a shortfall and that creating a client money pool is to ensure the subsequent pari passu distribution among all clients. Given that there is enough client money in client bank accounts, pari passu does not have its foundation, and so does the primary pooling event which seeks to materialize the former.

A typical case of not pooling client money is the special administration of City Equities; after being appointed, the special administrators inspected the firm’s client money accounts and found no shortfalls, addressing unequivocally that ‘should there be a shortfall this will trigger a pooling event’, which implies that a primary pooling event cannot be justified in the absence of a client money shortfall.⁸⁸ It is also worth adding that in this case the special administrators finally identified a client money surplus of £829.52 instead.⁸⁹

In the special administration of European Pensions, a client money primary pooling event was exactly what the special administrator wanted to avoid and had successfully escaped from, because there was no client money deficit and because a primary pooling event would be a ‘much more complex and costly process’ impeding the proposed bulk transfer of all client money accounts to an acquiring party.⁹⁰

This also raises a further question: since a further nine cases,⁹¹ as recorded in Table 3, also see no client money shortfalls, why did the special administrators pool client money together? The real reasons will be reported later. A client money pooling event can be very destructive. Now, attention is drawn to the second issue over when to pool client money.

Except for the three special administrations in which client money was not pooled, as listed in Table 3, the primary pooling event was triggered by the appointment of the special administrator in 12 (75 per cent) out of 16 special administrations, suggesting that the primary pooling event took place simultaneously when the special administrator was appointed, which is in line with what CASS mandates.⁹² But in the remaining four (25 per cent) special administrations the primary pooling event occurred, usually days, prior to the firm formally entering into special administration, which was actually prompted by a regulatory notice issued by the FCA (or its predecessor the Financial Services Authority, commonly referred to as the FSA)⁹³; indeed, under CASS 7A.2.2, the FCA is given such a power forcing the firm to take the client money pooling action so as to protect customers.⁹⁴

Whatever a special administration-triggered primary pooling event or a regulator-imposed one, the common issue is that some action should be taken over client money. As for how special administrators conduct pooling client money, unfortunately, CASS, it seems, does not give any detailed instructions.⁹⁵ Special administrators, arguably, have to improvise.

Generally speaking, there are two steps taken by special administrators to pool client money. The first one is that once appointed, the special administrators contact the bank(s) holding client money to freeze the accounts immediately. For example, in the special administration of WorldSpreads, after taking office, the special administrators informed 18 banks holding 70 client money accounts for the firm ‘with a request to place the accounts on stop with immediate effect’.⁹⁶ But the trouble here is that if the bank is from a foreign jurisdiction other than the USA, Canada, Australia, or Hong Kong,⁹⁷ obtaining its cooperation can be very challenging or costly.⁹⁸ In the aforementioned WorldSpreads case, the special administrators were a little lucky, since some foreign banks only demanded ‘further proof of the special administrators’ authority to issue instructions in relation to these accounts, which was provided in the form of notarised and apostilled documents’.⁹⁹ This inevitably generates more costs and leads to more delays. In the worst scenario, foreign banks simply refuse to cooperate. This happened in the special administration of AFX, where the firm operated a client money account with Swissquote Bank, which is based in Switzerland; the Swiss bank straightforwardly declined the special administrators’ requests for the details of the account, let alone to help freeze it, and lawyers had to be hired to tackle this conundrum.¹⁰⁰ Further costs and additional delays.

The second step of pooling client money is to move client money deposited in various bank accounts into one for the convenience of forthcoming distribution. For instance, in the special administration of LQD, the firm had 10 client money accounts with Santander UK plc and five with Barclays Bank plc, with a total of client money of some ‘US$2.195 million across these accounts held in various currencies’; after the cooperating action of these two banks, the special administrators later transferred all client money into a single client bank account opened at Lloyds Bank plc.¹⁰¹

Moving client money from a financial institution, notably a bank, into the full control of special administrators seems relatively easy.¹⁰² The challenge here is that if client money is held by a non-bank business entity, repatriating such money can be considerably problematic.¹⁰³ Two specific problems can arise. First, like what happened in the AFX case in which the majority of its £7,665,711.47 client money was in the custody of its parent company AFX Capital Markets Limited based in Cyprus and in which no client money was recouped from there due to the latter’s insolvency, the third company holding client money is in financial trouble itself, which means that demanding the return of client money to special administrators is predictably gloomy.¹⁰⁴ This is also witnessed in the special administration of WorldSpreads where some client money was held by a broker which was also in a corporate bankruptcy procedure.¹⁰⁵

Second, the third-party entity holding client money may either dispute the status of client money¹⁰⁶ or may refuse to recognize the trust law protection offered to client money. The latter scenario is especially acute if the third party is from a jurisdiction where trust law is traditionally not used or is not well observed for client money protection.¹⁰⁷ Admittedly, the picture here is mixed, since it is equally found that special administrators smoothly repatriated client money held in affiliated companies in cases such as the special administration of MF Global.¹⁰⁸ Thankfully, in most cases (14 out of 19), client money was deposited in domestic banks, and special administrators found no difficulty in collecting it.¹⁰⁹

To a certain extent, the primary pooling event in response to a regulatory notice seems to be incomplete if compared with that straightforwardly carried out by special administrators, because in the former situation the regulatory requirement for a primary pooling event only leads to the firm’s client bank accounts being frozen, with no action of transferring client money from various accounts into one as often practised by special administrators. For example, it was 9 March 2012 when the special administrators were appointed for the firm Pritchard, whereas it was 10 February 2012, 27 days earlier, when the firm’s client bank accounts were halted because of the supervisory notice issued by the FSA, but the firm did not take any further actions on these frozen accounts¹¹⁰; as anticipated, after the commencement of the special administration, later, supported by the FSA, the special administrators were able to transfer all unfrozen client money to a bank account set up by themselves.¹¹¹

As noted before, the assumption of pooling client money together is that there are client money shortfalls; now attention is moved to these troubles, real and imagined, witnessed in these special administrations.

Client money shortfalls and their causes

When a failed firm is placed into special administration, the existence of client money deficits is assumed, which is to justify the use of the primary pooling event. However, the data presented in Table 4 give a largely different picture.

Client money shortfalls not as prevalent as assumed

With the two special administrations of Solo and AFX excluded, since the outcome of client money distribution in these two cases was still pending at the end of data collection, it is found that the ultimate client money return rate in the remaining 17 special administrations fluctuates significantly, ranging from the full repayment at 100 per cent seen in three cases, the special administrations of Maple, Avalon, and European Pensions, to the lowest figure at 8.17 per cent identified in the special administration of LQD. On average, clients as a whole received 73.95 per cent of their cash back from failed investment firms, suggesting around a quarter of client money losses.

However, this average figure should be read with caution. A closer look reveals that out of these 17 cases the client money shortfall was not originally found in 13 of them, which means that in the majority (76.47 per cent) of special administrations no client money deficit was caused by the firm itself. The failed firm behaved very well in safeguarding client money. Why did clients still suffer approximately a quarter of losses? In fact, the eventual client money shortfalls were mainly generated by two sources, one of which was because of the firm allegedly having failed to fulfil the client money safeguarding role and the second of which was due to special administrators charging distribution costs.

As demonstrated in Table 4, in terms of the average client money shortfall at around one quarter, around one-third of such losses (8.41 per cent) was because of internal issues within the firm, and two-thirds (17.83 per cent) were external and were to pay distribution costs most of which are special administrator fees. However, the figure 8.41 per cent of firm-caused client money losses could be read from a slightly different angle, since it might have unintentionally exaggerated the fault of the firms. This is because many high firm-caused client money loss percentages occurred in smaller firms, as recorded in Table 4. From a more substantial perspective, of these 17 surveyed firms, as shown in Table 4, the average amount of client money claims is £86,386,893, which is made against the average size of client money pools of £85,985,352, indicating that the substantial firm-caused client money deficit rate only stands at 0.46 per cent. The FCA might be happy to see this figure, since it suggests that the UK investment banks as a whole are adequately regulated/supervised, especially on safeguarding client money.¹¹² Irrespective of the statistical methods, the clear conclusion regarding client money losses is that the more substantial client money trouble is from the special administrators rather than from the firm.

That 13 out of 17 special administrations did not see any original client money shortfall indicates the propriety of these 13 firms in safeguarding client money. But a closer check finds that more firms did the right thing in looking after client money, since the presence of client money shortfalls cannot always be linked with the firm’s impropriety or misbehaviours.

Causes of client money deficits

Admittedly, it seems fair to say that if a client money shortfall is identified, some wrongdoings can be generally assumed. For example, shortly before being placed into special administration, Pritchard was not able to maintain ‘a minimum of qualifying capital to capital resource requirement by the FSA’, one consequence of which is that when the client money shortfall was found, the firm did not have enough cash in reserve to fill the gap; although this regulatory breach did not directly contribute to the shortfall, it at least failed to cushion this deficit as required by regulation.¹¹³ In the worst scenario, client money shortfalls are simply because of fraud perpetrated by the firm, which can be found in WorldSpreads where the firm deliberately concealed the shortfalls and provided misleading financial statements to the auditors.¹¹⁴

However, on the side of the firm, a client money shortfall may also be the consequence of two other technical factors: a delayed reconciliation or the controversial open position valuation method. For client money, all firms are required to conduct daily reconciliations so as to match what clients are entitled to with what is available in client bank accounts.¹¹⁵ But given the disruption of entering into special administration, a delayed reconciliation is inevitable.¹¹⁶ This is also why the updated Regulations 2011 requires that ‘immediately after being appointed as the (special) administrator, the (special) administrator must carry out a client money reconciliation.’¹¹⁷

In the special administration of Fyshe, for instance, the special administrators initially found a client money deficit of £70,000,¹¹⁸ but after a post-administration reconciliation, it was realized that a larger sum of the client money shortfall at £991,664 actually arose because some clients had failed to top up their transaction accounts, namely the firm did nothing wrong; following a debt collection effort against the concerned clients, there was no client money shortfall at all; unsurprisingly, in this case all client money was returned only after the deduction of the distribution costs.¹¹⁹ In other words, before a post-administration client money reconciliation, it is arguably too early to assert that there is a client money shortfall.

Also, it should be noted that the outcome of a post-administration reconciliation can go in the opposite way: client money shortfalls are first spotted, but client money surpluses are identified after a reconciliation. This happens in at least two out of these 13 non-original-client-money-shortfall cases.¹²⁰ Take the special administration of Strand as an example. When the firm entered into special administration on 17 May 2017, the special administrators warned that a shortfall of client money was likely, but it was not possible to ‘estimate the quantum of such a shortfall’, given the early stages of this case.¹²¹ But later it is divulged in a progress report that the sum of client money deposited in two bank accounts, following the post-administration reconciliations, was £12,570,091.51 in contrast to the client money claims at £12,525,301.55, implying a surplus of £44,789.96.¹²² Client money surpluses are most commonly generated because of interest-bearing bank accounts¹²³ or foreign currency exchange windfalls.¹²⁴

Similarly, the open position valuation method clarified in Re MF Global UK Ltd (In Administration)¹²⁵ can also artificially give rise to either a client money shortfall or a surplus.¹²⁶ In the special administration of MF Global, because the client open positions were valued at the prices when the client money primary pooling event occurred rather than at the prices when these open positions were actually closed out, this valuation method artificially produced a client money shortfall of £36.65 million (US$59.1m).¹²⁷ Whatever a delayed reconciliation or the open position valuation method, the firm is innocent in the face of a prima facie client money shortfall or an eventual one.

In addition, in some cases, client money shortfalls are purely caused by market volatility, and the firm did nothing inappropriate.¹²⁸ This is what happened in the special administration of LQD. On 15 January 2015, the Swiss National Bank suddenly removed the informal peg between the Swiss Franc and the Euro; as a match principal broker, LQD was unable to process the trades at the requests of its clients at the price of 1.540 per point, since ‘banks stopped offering a market for 17 minutes;’ when the market was resumed, the price of those positions collapsed to 0.82 per point, leading to the heavy losses to both the firm and the clients, as a result of which a client money shortfall of US$742,995.69 was found; the firm immediately informed the regulator to seek solutions.¹²⁹

If the aforementioned three factors are taken into account, apart from the 13 investment banks which did not see any original client money shortfalls, one further firm, LQD, could be exonerated. This means that 14 (82.35 per cent) out of all 17 firms did not perform improperly in looking after client money. But this summary should be understood in its context, since two pending cases, the special administrations of Solo and AFX, are excluded because client money distribution did not begin at the end of this study’s data collection. However, in terms of whether there are client money shortfalls, at least the AFX case should be included in the statistics, otherwise it might be an unforgivable bias. The Solo case can still be safely excluded, given that an international criminal investigation over its founder and its businesses continued when this study’s data collection was closed.¹³⁰ By the standards adopted in this study, AFX cannot be said to be innocent/competent in safeguarding client money, since it overly relied on its unreliable parent company AFX Capital Markets Limited,¹³¹ which was based in a foreign jurisdiction, to hold client money, and when things went wrong, it was unable to repatriate the bulk of the client money back to the UK, resulting in a catastrophic client money deficit.¹³²

Therefore, with AFX included, it is more appropriate to conclude that at least in 14 (77.78 per cent) out of all 18 special administrations the firm did not perform as inadequately in looking after client money as assumed by lawmakers. Whatever the original statistical figure and the adjusted one, it can be safely summarized that assuming the existence of client money shortfalls in all investment bank special administrations lacks supporting evidence, and it is time to rethink the policy designs.

Turning to the major cause of client money shortfalls, that is, distribution costs, this provokes considerable resentment from clients; in some cases, the tension between clients and special administrators over such costs can escalate into fierce confrontation as was witnessed in the Beaufort case.¹³³ The so-called distribution cost comprises mainly (not entirely) of the fees paid to special administrators themselves and to hired solicitors; needless to say, the bulk of these costs go to special administrators. A typical example is the special administration of Strand, in which until 16 November 2021 the special administrator’s charge over distributing client money was £688,244.87, along with the legal costs paid to the solicitors accumulated to £193,845.82,¹³⁴ suggesting that 78 per cent of the distribution costs are special administrator fees. At the core of client resentment/anger over such fees is whether these costs can be justified.

How to justify distribution costs as a major cause of client money shortfalls?

Before examining the statutory/regulatory mandate for such costs, it seems appropriate to probe how special administrators justify these costs by themselves in the first place. Generally speaking, there are two alarm bells ringing to convince why distribution costs are inevitable and worthwhile.

Special administrators’ own justification over distribution costs

The first bell is that there is a client money shortfall, real or imagined, and for fairness between clients a redistribution process managed by special administrators is necessary. This is to legitimize two further specific actions: first, as noted before, all client money accounts are to be frozen, followed by transferring all client money into one bank account opened by special administrators themselves; second, all client money accounts should be checked and verified through a client money claim process individually by special administrators to ensure the accuracy of each so as to prevent mistaken redistributions.

But the question is that, as reported earlier, 13 (76.47 per cent) out of 17 investment banks in special administration saw no client money shortfalls, therefore, it seems to too weak to support special administrators using client money shortfalls as a reason to mandate the subsequent actions. One may ask that in these 13 non-original-client-money-shortfall cases whether the special administrators raised this alarm for self-authorization. To test this, all written documents in these 13 cases were carefully checked.

As recorded in Table 5, although in these 13 cases where no client money deficit occurred, four types of reactions of special administrators can be identified. The first type is seen in the special administrations of City Equities, Maple, and SVS, where the special administrators declared straightforwardly that there was no client money shortfall, which was in line with what were reflected in the firm’s client money records. The second type is noticeable in the special administrations of Avalon, European Pensions and Beaufort, in which the special administrators were, deliberately or not, silent over whether there was a client money shortfall but conducted a client money verification process meticulously as if there had been a considerable deficit. The third type can be summarized from the four special administrations of MF Global, Alpari, Strand, and Reyker, where the special administrators avoided stating explicitly whether there were client money shortfalls or not, but the tone of their expressions over client money issues implied that there were shortfalls.

For example, in the special administrators’ proposals of MF Global, no words like client money shortfall, deficiency, or deficit can be found throughout the documents, but the real message can be observed/guessed from two points presented by the special administrators. Point one is that for the time being only 82 per cent of client money had been recovered, and the rest was held by the parent company which happened to also be in an insolvency procedure.¹³⁵ Therefore, it was uncertain whether client money could be fully recovered. It sounded precarious.¹³⁶ Point two is that the special administrators warned that the final client money distribution depended on a cohort of factors, one of which was whether the client money deposited at the parent company could be thoroughly recouped.¹³⁷ Any reasonable customer would infer that a client money shortfall was very likely, which necessitated both the primary pooling event and the planned redistribution carried out by the special administrators.¹³⁸

The fourth type is the actions found in the three special administrations of Boston, Hume and Fyshe, where the special administrators stated, contrary to what eventually turned out, that a client money shortfall was found. In the special administration of Boston, the clients were initially reminded that based on the special administrator’s then assumption ‘there would be a significant (client money) shortfall to clients’.¹³⁹ However, in this case, it soon transpired that the amount of client money claims was US$900,344¹⁴⁰ exactly matching the client money of US$900,344 safely lying in the firm’s ‘client funds account’,¹⁴¹ with no shortfall occurring at all. Unfortunately, in all subsequent progress reports released in the special administration of Boston, no single word was used to inform the stakeholders that the initial assertion of a client money shortfall was a premature observation.

In the special administrations of Hume and Fyshe, the so-called client money shortfalls of £30,949.71 and £70,000, respectively, are arguably mislabelled. In the special administration of Hume, the special administrators identified a client money deficit of £30,949.71, but classifying it as a client money shortfall can be technically controversial, since this gap was because one client had failed to deposit enough cash to balance her/his own individual client money account before the primary pooling event, which is supposed to be very common, given the dynamic of cash traffics between the firm and its clients¹⁴²; in this situation, the most appropriate response would be to chase the client debtor rather than to declare a general client money shortfall¹⁴³; indeed, the special administrator did pursue the client debtor and later got a partial recovery¹⁴⁴; solving this individual client’s cash relationship with the firm should be a significant part of client money reconciliations; jumping to the conclusion that a client money shortfall has been found seems to be problematic. The £30,949.71 client money deficit in the Hume case should also be read in its context. The firm’s client money pool amounted to £6,666,250.88,¹⁴⁵ with a deficiency of £30,949.71 only, which can be translated into a meagre deficiency rate of 0.46 per cent. However, for this trivial flaw, the clients as a whole had to bear the distribution costs of £1,272,301.21.¹⁴⁶

It should also be pointed out that in the Hume special administration, shortly after the post-administration reconciliation, the special administrators were contacted by the firm’s overseas share-dealing partner, with the latter transferring a client money surplus of £129,149.50 to the special administrators.¹⁴⁷ This means that after offsetting the aforementioned deficiency of £30,949.71, there was actually a net client money surplus of £98,199.79 in the Hume special administration.¹⁴⁸ But, unfortunately, in all subsequent progress reports, the original statement that there was a client money deficit remained.¹⁴⁹

In the special administration of Fyshe, the story of a client money shortfall was almost identical as what happened in the special administration of Hume. The Fyshe special administrators rightly estimated a client money deficiency of £866,000, but given that there was a sum of £992,619 owed to the firm by some clients,¹⁵⁰ it seems too early to rush to the judgement that there was a client money deficit; again, this is a natural result of client money reconciliations.¹⁵¹

Although these four types of behaviours vary significantly, surely, the first type is very recommendable, since the special administrators stated exactly what was really recorded in the firm’s books and client bank accounts. However, even this type of reactions can be a problem. If the special administrators knew that there was no client money shortfall, which, to a large extent, suggested that the firm’s client money records were correct, why the clients were asked to go through an expensive client money claim process to verify each client’s client money claim again; simply relying on the reliable client money records to return client money saves time and saves costs. It is puzzling that they did not do this.

The second alarm bell raised by special administrators is the potential of competing client money claims, namely all client money accounts should be individually checked in a separate claim process by special administrators to avoid any hidden competing claims, without which client money distributions may be reversed.¹⁵² This also sounds very serious, and this assertion is found at least in the special administrations of SVS and Reyker. In the special administration of SVS, for example, although the special administrators clarified that there was, following a post-administration reconciliation, no material client money deficit, the clients were reminded that the firm’s client money records could not be relied upon to directly return client money because there were ‘risks of claims from other clients or third parties who may argue that the company’s records were not correct,’ therefore, a client money claim process should be carried out to ‘offer protection against claims like that’.¹⁵³

The threat exerted by the special administrators in the SVS case could be problematic on at least two grounds. First, the existence of competing client money claims was perhaps initially disputed in the Lehman case, where essentially it was a disagreement over whether some professional investors could be offered client money protection under CASS rules.¹⁵⁴ If a professional client has cash in the investment firm under a title transfer collateral arrangement, CASS states clearly that such cash is treated as the property of the firm and is not subject to client money protection.¹⁵⁵ In fact, in the SVS case, around 530 professional clients were adequately classified and recorded on the firm’s books, which means that it was very unlikely for any competing client money claims which might dispute their client status arising, given the correct demarcation between retail and professional clients in this case.¹⁵⁶ What has been clearly recorded in the firm’s client money books could not back the special administrators’ assertion over the existence of potential competing client money claims.¹⁵⁷ Second, it soon transpired in this case that before the bulk client money account transfer to a third party there was a client money surplus of £21,500 found,¹⁵⁸ suggesting that there were no competing client money claims at all. This alarm proves to be premature.

In the special administration of Reyker, a similar warning was seen.¹⁵⁹ But no competing client money claims came forward as asserted by the special administrators; on the contrary, after the deadline set for submitting client money claims, there were only 7,832 clients stepping forward by lodging their client money claims, whereas there were 3,294 clients who remained silent.¹⁶⁰ Therefore, the tricky reality is that there is always a lack of eligible client money claims rather than a flood of unmerited competing ones. If the special administrators’ assertation in both the special administrations of SVS and Reyker over the spectre of competing client money claims is correct, a bulk transfer of client money accounts without the use of a client money claim process, the most effective way in returning client money, is impossible in all investment bank special administrations.¹⁶¹ In reality, quite the opposite can be seen, such as the client money bulk transfers used in the special administrations of Avalon¹⁶² and European Pensions,¹⁶³ without having to go through the lengthy and costly client money claim process. In a word, for special administrators, using the alleged culprits of client money shortfalls and competing claims cannot adequately justify distribution costs of returning client money.

Legal justification over distribution costs

Attention is now turned to legal mandates over client money distribution costs. The direct legal backing would be CASS, section 7.17.1 of which authorizes that ‘in the event of failure of the firm, costs relating to the distribution of client money may have to be borne by the trust.’¹⁶⁴ The word ‘may’ in this paragraph can be literally understood as a possibility.¹⁶⁵ This means that these costs may, or may not, have to be borne by the trust.¹⁶⁶ But CASS does not offer any clues as for in what situations these costs may not be paid out of the trust. It is unknown whether CASS deliberately retains such vagueness, although it is understood that this cost is normally borne by the trust.¹⁶⁷ Given that CASS is made by the FCA, a regulatory body, it seems appropriate to trace the higher authority of this mandate further upwards.

Although the FCA is empowered to produce CASS by the Financial Services and Markets Act 2000,¹⁶⁸ allowing distribution costs to be drawn from the trust cannot find its origin from the Act itself, since the latter only generally permits the FCA to ‘make provision which results in that clients’ money being held on trust’. ¹⁶⁹ Specifying that distribution costs may be paid from the trust is due to the FCA’s own initiative, at least at the first impression.¹⁷⁰ In fact, the FCA’s confidence in this regard is backed by common law instead of by the Act. The first precedent where British judges allow the costs of distributing trust assets to be borne by the trust itself is re Berkeley Applegate¹⁷¹ adjudicated in 1988. But re Berkeley Applegate should be properly understood, and two key points are worth addressing.

First, in this case, the firm’s house estate only amounted to up to £80,000, in contrast to the client money trust totalling around £11,400,000, whereas the distribution cost of the client money trust had already risen to £644,500, suggesting that it was practically impossible to use the firm’s house estate to cover this cost.¹⁷² This is a key contextual fact. Second, the court ruling should be read in its entirety; after considering the facts and all related jurisdictions, Edward Nugee J. delivered the verdict that ‘the liquidator was entitled to be paid his proper expenses and remuneration out of the trust assets if the company’s assets were insufficient’.¹⁷³ Arguably, re Berkeley Applegate seems to have only answered half of the whole question.

If the firm does not have sufficient assets, the cost of distributing client money held on trust by the firm should be paid from the trust itself; however, no answer has been given in relation to whether such a cost should be shouldered by the firm if the firm has sufficient assets.¹⁷⁴ But, in a following court order of re Berkeley Applegate, Peter Gibson J. from the same court went a step further, addressing that the cost of distributing the client money trust should be paid from the trust itself even if there are free assets in the firm’s house estate,¹⁷⁵ which seems in conflict with the previous ruling given by Edward Nugee J. In practice, as is seen in many subsequent cases,¹⁷⁶ it looks to be routine for insolvency practitioners to draw from the client money trust to cover the distribution costs irrespective of whether the firm has enough assets or not.¹⁷⁷ This means that the ruling issued by Peter Gibson J. prevails.

This article is, however, of the view that if the firm has any free assets, the client money distribution costs should be paid from the firm’s house estate first and that unless and until the firm’s house estate is exhausted, the client money trust starts to bear these costs.¹⁷⁸ This view is also consistent with what happened when the firm was solvent, as was clearly stated by a respondent in re Berkeley Applegate that:¹⁷⁹

Had the company not been in a creditors’ voluntary winding up but in a members’ voluntary winding up and solvent, there would be no difficulty in treating all expenses properly incurred in the winding up as including the remuneration and expenses incurred by the liquidator in the administration of the trust and that they would have been involved in winding up the affairs of the company.

What should also not be forgotten is that before insolvency it was the firm bearing the costs of managing/administering the client money trust; shifting these costs to clients after the firm became insolvent is not persuasive enough. But the key question is how the special administrators in the surveyed cases relied upon both CASS and the re Berkeley Applegate principle when deciding whether to use client money to pay the distribution costs. A mixed picture far beyond what the rules say.

Was the client money trust always used to pay distribution costs?

In summarizing what happened over whether distribution costs were imposed on client money estates, the special administrations of Solo and AFX should be excluded from the statistics for the same reasons noted before. Of the 17 cases, it is found that the special administrator charged distribution costs on the client money estate in 14 cases (82.35 per cent), as shown in Table 6.

This means that the special administrators imposed distribution costs on the client money estate in the majority (82.35 per cent) of the special administrations. Given that it is allowed by the rules, it seems counterintuitive for special administrators not to exercise this right. The piles of cash in client money bank accounts also look irresistible, especially compared with what is available in the firm’s house accounts. For example, when Reyker started its special administration in 2019, there was a sum of £57,229,599 found in the client money bank accounts, in contrast to a meagre amount of cash of £52,000 in the firm’s own house accounts.¹⁸⁰

Making most of the client money estate seems also to be a necessity, since, as detailed in Table 6, all these 14 firms in special administration were financially insolvent, which means that it would be quite challenging, if not impossible, to resort to the firm’s house estate to foot any bills of distributing client money. In particular, 11 (78.57 per cent) out of these 14 firms were overwhelmingly insolvent, as recorded in Table 6, because the unsecured creditors ultimately recouped either zero or almost zero,¹⁸¹ suggesting that it was hopeless to look at the house estate for relief.

A closer examination surprisingly divulges that in many special administrations the special administrators actually overstepped their charging authority by using the client money estate to pay the costs incurred in the interests of the firm’s house estate. Since the Regulations 2011 lists the three objectives for special administrators to pursue, the expected practice is that special administrators split the costs between these objectives: the costs of fulfilling objective 1 on the return of client money as well as client assets are paid from the client money/asset estate, with those for objectives 2 and 3 in liaising with regulators and market infrastructure authorities and administering the firm’s general estate drawn from the firm’s house estate.¹⁸² This is the rule, although it is contentiously client-unfriendly. In reality, even this rule is not well complied with.

Granted, this rule is not relaxed in all cases. Of these 14 special administrations, six (42.86 per cent) special administrations saw that the special administrators rightfully divided the overall costs between the three estates, that is, the client asset estate, if applicable,¹⁸³ the client money one and the firm’s house estate.¹⁸⁴ By contrast, in another five special administrations,¹⁸⁵ the special administrators either prevaricated over whether the overall costs had been allocated between the three estates or simply gave no explanation at all.

For instance, in the special administration of Beaufort, no exact information as for how the overall costs were split between the estates could be found; the special administrator only gave a figure of the overall costs at some £59,016,000, which ‘includes costs relating to Objective 1 (ie the return of Client Portfolios) and costs associated with our other objectives including the realisation of House Assets’, without giving any further information as to how to allocate it between different estates.¹⁸⁶ It can be safely speculated that in the special administration of Beaufort, part of the client money estate was used to pay for the costs incurred in pursuit of objectives 2 and 3.¹⁸⁷ This is more likely to be the same in the remaining four cases of this group.¹⁸⁸

Using client money to pay for the firm’s house estate cost is legally debatable. However, a more controversial step could be identified in the sixth case, the special administration of Strand, in which, probably because there was only £140,270.05 realized from the firm’s house estate, the special administrators expansively interpreted that ‘costs incurred in respect of objective (sic) 2&3 for the benefit of objective 1 can be chargeable to the client estate’.¹⁸⁹ If this interpretation can be widely applied, it may derail the crucial principle of allocating costs between each estate in investment bank insolvencies.

The difficulty of the special administrators getting properly paid is understandable in the special administration of Strand, since the objective 2 costs alone had already accumulated to £527,465.55 at the end of November 2018, and obviously the £140,270.05 house estate was too low to cover these costs.¹⁹⁰ But the question is that the pendulum here swings too far at the expense of clients.

A further two special administrations seem to be innovative in allocating costs, which appears to be more contentious. These are the special administrations of Alpari and Boston; surprisingly, the special administrators divided the overall costs, either literally¹⁹¹ or essentially,¹⁹² between the client money estate and the firm’s house one pro rata.¹⁹³ Such a step cannot be encouraged, since usually the size of the client money estate would be far larger than the firm’s; allocating the overall costs pari passu may lead to the costs disproportionately levied on clients. Admittedly, the special administrators of Alpari seem to have obtained the legal mandate to do so, since they got an advance approval from the creditors’ committee, which comprised both creditor and client representatives,¹⁹⁴ before dividing the costs between the two estates pro rata.¹⁹⁵

By contrast, in the special administration of Boston, it seems that adopting the pari passu principle in dividing the overall costs between the two estates was purely at the initiative of the special administrators themselves.¹⁹⁶ The result in the special administration of Boston may surprise no one, since each client had to pay the distribution cost of £45,465.32 in order to get their money back, a figure more likely hitting the historic record.¹⁹⁷

Overall, it seems that in eight (57.14 per cent) out of 14 studied special administrations client money was used to pay the costs of managing the firm’s general estate. When the Regulations 2011 was made, the policy-makers emphasized that clients cannot be protected at the expense of creditors,¹⁹⁸ but in reality quite the opposite can be found: clients are being dragged into paying the costs which are supposed to be paid by creditors.¹⁹⁹ This study can prove that the widespread accusation that client assets (including client money) are used to pay for the liquidation of investment firms is largely true.²⁰⁰ Striking a balance between client and creditor in investment bank insolvencies deserves to be rethought.

Distribution costs not charged on the client money trust

However, the picture is not entirely negative, especially from the perspective of clients. In three (17.65 per cent) out of 17 studied special administrations the special administrators, as shown in Table 6, did not charge distribution costs on the client money estate, using the firm’s house estate to pay instead. One of these three cases is the special administration of Maple, in which the special administrators decided not to impose distribution costs on the client money estate so as to ‘avoid the relatively limited CMP (client money pool) being impaired or exhausted by the Joint Administrators’ remuneration and expenses.’²⁰¹ In fact, it would be awkward for the Maple special administrators to levy the distribution costs on the client money estate because the unsecured creditors got fully repayment; this would be the only case where the firm was finally proved solvent after realizing all company assets. If the re Berkeley Applegate principle had been mechanically followed, the consequence would be very absurd, meaning that the clients only got a partial client money return but the creditors were fully paid. This case also raises a red flag over the judgement given by Peter Gibson J. in re Berkeley Applegate, which imposes distribution costs on the client money trust disregarding whether the firm is insolvent or not.

In the remaining two cases, although the firm was insolvent, the special administrators still went ahead by not charging distribution costs on the client money estate; to prevent any unwanted challenges, the Avalon special administrator obtained an approval from the unsecured creditors to do so.²⁰² But in the European Pensions case, the special administrators took this decision by themselves.²⁰³

In the special administration of European Pensions, the special administrators were confident in not charging distribution costs over the client money estate mainly because the sales of the client money/asset accounts generated an income of £1,320,000 for the firm’s house estate, and it would be contradictory to, on the one hand, hand over the benefit from selling the client estate to the firm’s general one, but to, on the other hand, impose the cost of such a sale to the client estate itself; probably because of this justification, there was no challenge over this arrangement from the unsecured creditors.²⁰⁴ In a word, the re Berkeley Applegate principle must be updated to reflect what is practised in reality.

Three lessons could be learned regarding client money distribution costs. First, distribution costs are often broadly interpreted to encroach on the client money estate. The initial argument of protecting creditors from paying for the return of client assets (including client money) is largely groundless.²⁰⁵ The reality is that clients are often sacrificed in favour of creditors. Second, the re Berkeley Applegate principle should be treated as an optional choice, not as a mandatory arrangement, and it should be set aside when the firm is solvent.²⁰⁶ Third, any benefit generated from the client money estate should be used to pay for the costs over dealing with this estate in the first place, otherwise practical oddity can ensue.

The evidence presented above suggests that the client money shortfalls due to distribution costs could not be equitably justified in most cases and that CASS and the Re Berkeley Applegate principle simply opened the floodgates, leading to widespread misapplication of client money to the detriment of clients. Seeking a higher rate of client money returns is a significant goal of the client money regime, but a speedy return of client money is also a major task, especially for the investment bank special administration regime. To accelerate client money returns, a bar date as a major concrete legal mechanism was created to facilitate this goal. But the question is whether it is really useful in practice.

The application of the bar date in pursuing a speedy client money return

To test whether the use of the bar date has contributed to a quicker return of client money is a complicated business. Three things should be sorted out before a meaningful result can be presented.

Three parameters in understanding the usefulness of the client money bar date

First, how long did the client money return take in each case? This demands identifying how many days elapse between the return of client money and the commencement of the special administration. Locating the exact date when the client money return took place is complex, since in most cases client money was returned tranche by tranche²⁰⁷; for consistency with the previous article in calculating the efficiency of client asset returns,²⁰⁸ the date of the first client money distribution is deemed to be the date of the overall client money return.

For example, in the special administration of Strand, the special administrators made two interim client money distributions in addition to the third and final one between April 2019 and March 2021,²⁰⁹ and for simplicity and for defence, the time of the first client money distribution is treated as the date of the client money return. Of course, occasionally, in some cases, there was only a first and final client money distribution, as was seen in the special administration of LQD.²¹⁰

In trying to identify the date of the client money return, two practical difficulties arose as anticipated. The first difficulty is that the exact date of the first client money distribution was not disclosed in some cases. For example, in the special administration of Strand, the special administrators only generally stated that they had ‘paid a first interim distribution of all client money claims of 65 pence in the pound at the beginning of April 2019’, without giving the specific date of this distribution.²¹¹ Therefore, the first distribution date is conservatively estimated as 1 April 2019, the earliest time. To estimate in this way is to ensure this study’s conclusion as for the effectiveness of the bar date is more robust, given that this study is to question the efficacy of this mechanism. This difficulty was encountered and has been reasonably solved in the special administrations of Fyshe,²¹² Hartmann,²¹³ Maple,²¹⁴ Strand,²¹⁵ Beaufort²¹⁶ and Reyker.²¹⁷

The second difficulty is that surprisingly at the end of this study’s data collection, there were still two cases, the special administrations of Hume²¹⁸ and AFX,²¹⁹ in which the client money distribution had not been made. Again, the distribution date is conservatively assumed to be the end of the latest progress report period; this study takes a step back by accepting that the bar date was used at its maximum in facilitating a speedy return of client money so as to enhance the vigorousness of the attempted criticism.

The second thing key for this test is that achieving a speedy client money return is relative, requiring benchmarks to compare with. To judge the usefulness of the bar date, this study compares the average speed (the length of time between the return of client money and the commencement of the special administration) of client money return in the cases where the bar date was installed with those without using the bar date. In addition, given the well-known Lehman case, which led to the establishment of the investment bank special administration regime, this study also compares what has been achieved under the bar date mechanism after the Regulations 2011 with what was realized in the Lehman case.

The third thing is to declare the incompleteness of the data collection, which is common in empirical studies.²²⁰ Although many efforts were made to locate when the client money return took place in the special administration of City Equities, this study was even unable to estimate, since the special administrators only vaguely stated that some client money had been returned before 10 April 2014,²²¹ which renders an estimate too wild to be credible. It is time to give up. The same data are also marked as missing in the special administration of Solo due to its ongoing criminal investigation.²²²

With these three parameters elaborated, the findings are presented in Table 7.

Did the bar date really facilitate a quicker return of client money?

With the special administrations of City Equities and Solo excluded, it can be seen that the bar date was widely used both before and after the Regulations 2017. In particular, it was deployed in 12 (70.59 per cent) out of 17 special administrations, with only five cases (29.41 per cent) not resorting to this mechanism. Before the Regulations 2017 took effect, special administrators had to apply for a court order for imposing a bar date, as was seen in the seven cases listed in Table 7. Unsurprisingly, following the Regulations 2017, it was special administrators themselves setting up the bar date on their own initiative. It is also interesting to know that, in some special administrations which began before the Regulations 2017 came into force, the special administrators still petitioned to the court for an order when the bar date was needed post-Regulations-2017, as was witnessed in the special administrations of Pritchard²²³ and LQD.²²⁴ In doing so, perhaps it is because the special administrators believed that they got their mandate from the Regulations 2011 rather than the Regulations 2017, thereby they followed the old regime and relied on the original law for a court order to legitimize the use of the bar date.

Before comparing the efficiency of client money returns in the cases of these two categories, that is, the cases where the bar date was installed and those where no bar date was applied, the speed of the client money return in the Lehman case should be recalled. By the standards adopted by this study, the first Lehman interim client money distribution on 23 April 2013 is deemed to be the client money return date,²²⁵ and given the Lehman administration started on 15 September 2008,²²⁶ this means that it took 1,680 days for the clients to see the return of their cash. It is also worth noting that a client money bar date was sanctioned by the court in the Lehman case.²²⁷ Based on the data collected in Table 7 as well as the information of the Lehman case, Fig. 1 shows the contrasts.

Some may celebrate the achievement made by the new investment bank special administration regime. Look at the progress: pre-Regulations-2011, the Lehman case used 1,680 days to manage the return of client money under a bar date; under the Regulations 2011, the time was reduced to 817.92 days, more than halving the waiting time in the best interests of clients. However, the 1,680-day waiting time in the Lehman case should be more closely examined. The Lehman clients suffered considerable delays in getting their money back for various reasons. But one of the key causes is the litigations disputing whether the client money pool includes client money that has not been segregated and is still trapped in the firm’s house accounts²²⁸; these litigations lasted from 1 May 2009 to 29 February 2012²²⁹; if this 1,034-day litigation-caused suspension is taken out of the calculation, the net waiting time in the Lehman case should be equitably updated to 646 days. Accordingly, Fig. 1 could be rightly amended to Fig. 2 as follows.

Now, any sense of complacency suddenly disappears. The new investment bank special administration regime, which was mainly designed to expedite the return of client money (including client assets), only damaged this chief goal by protracting the frustrating wait.²³⁰ In the Lehman case, the clients waited for 646 days, but under the new special administration regime, the clients on average prolonged their waiting agony to 817.92 days. This is in fact not the worst.

The worst contradiction is that the clients, after the Regulations 2011, on average, got their money back in 71.6 days in the cases where the bar date was not installed, in contrast to the average 817.92-day waiting time in the cases facilitated by a client money bar date, as shown in Fig. 2. The use of the bar date was made for the purpose of accelerating client money return. But in practice it is not exaggeration to say that where there is a bar date applied there is a disaster to the speed of client money returns instead. What is wrong?

Granted, of the 12 special administrations using the client money bar date the waiting time varies sharply, ranging from the longest of 2,919 days²³¹ to the shortest of 102 days.²³² But the question is that even the shortest waiting time of 102 days is longer than the average of 71.6 days seen in the non-bar-date cases, suggesting that the bar date might be considerably counterproductive. Why was the client money return delivered more quickly when the bar date was not used?

Why was client money returned more quickly without the use of the bar date?

There is no fixed answer, unfortunately. Of these five non-bar-date special administrations, no magic wand can be identified. In the special administrations of Fyshe and Hartmann, a client money claim process was conducted, but it seems that a relative efficiency was achieved, because it only took 103 days and 140 days, respectively, to see the return of client money, as shown in Table 7. The special administrators may have found it easier in the special administration of Maple, since there were only 20 professional clients involved, and given such a small number, a quick client money return seemed not to be a surprise. Within this group, what can be truly recommended is the remaining two special administrations of Avalon and European Pensions which achieved the speediest client money returns, 2 days in the former case and 23 days in the latter. If there is a magic wand, this is what was used in the last two special administrations: bulk client money transfers.

In the special administration of Avalon, a buyer was secured shortly after the commencement of the special administration, with all client assets and money accounts successfully transferred to the buyer within 2 days.²³³ This case can also reveal that the Regulations 2011 is an obstacle rather than an advantage in facilitating a speedy client money return. For a quick wholesale transfer of client money accounts, the application of a client money pooling event under the Regulations 2011 would be undesirable. To avoid the client money primary pooling event which is mandated by the Regulations 2011, the special administrators of Avalon had to apply for a regulatory waiver from the FCA so as to make a speedy client money return a reality.²³⁴ What should not be forgotten is the possibility of potential competing client money claims purported in the special administrations of SVS and Reyker; if such an assertion in these two cases is correct, a quick client money account bulk transfer can never take place, since a bulk transfer is fundamentally incompatible with a client money primary pooling event in addition to a subsequent client money claim process, which can be wastefully lengthy and costly.

In the special administration of European Pensions, the same regulatory exemption was sought and obtained before the special administrators were able to carry out the bulk client money account transfers to the acquiring parties ‘in order to ensure there was no primary pooling event of client money’.²³⁵ To some extent, a primary pooling event is a curse rather than a blessing for a speedy client money return. This article is, it is worth repeating, of the view that a primary pooling event is unnecessary if there is no client money shortfall found when a firm enters into special administration, as argued before. The rules, both statutory and regulatory, may be updated to reflect this.

Why was the bar date unable to facilitate a speedy client money return?

Now, attention is turned to why the application of the bar date did not accelerate the client money returns in the 12 cases reported earlier. The first reason would be that most client bar dates were set up too late. Figure 3 records how many days had passed between the entry of the special administration and the installation of the bar date in each of these 12 cases. On average, the special administrators established the bar date 817.92 days after the firm commenced the special administration (coincidentally the same as the average duration between entry and client money return). Admittedly, four abnormal special administrations of MF Global, Pritchard, Alpari, and Beaufort might have distorted this summary, since in these four cases the client bar date was set up long after the first client money distribution had been made.

In the special administration of Pritchard, the first client money distribution was made on 13 November 2014, but the client money bar date was established at 12 April 2019, around 5 years after.²³⁶ A similar oddity can also be found in the special administrations of MF Global, Alpari, and Beaufort.²³⁷ If these four cases are excluded, the average period between the bar date and the commencement of the special administrations in the remaining eight cases could be updated to 628.13 days, still a long delay.

Why did it take so long for the client money bar date to be set up? Many causes. Before the Regulations 2017, given that setting up a client money bar date should be approved by the court, it seems understandable that it might have unintentionally protracted this process. It might be true. With the four abnormal cases excluded, it can be found that before the Regulations 2017 took effect, it on average took 908.5 days to see the installation of the client money bar date in the five cases.²³⁸ By contrast, post-Regulations-2017, this gap was reduced to 446 days in the four cases,²³⁹ a significant improvement, suggesting that the Regulations 2017 played its role in facilitating efficiency.

However, that the average 446 days were used to wait for the arrival of the bar date cannot be said to be a success. Unlike many non-asset corporate bankruptcies in which a debt claim process is unnecessary and is routinely skipped and in which a deadline called the last date for proving is accordingly unwarranted, in an investment bank special administration, the pool of client money is always there, although there might be a shortfall; this means that if a client money claim process is used, the deadline called the bar date should be installed as soon as possible after the special administration procedure starts. The average delay of 446 days means that over 1 year is used to prepare, leading to considerable sufferings to innocent clients.

Admittedly, on client money claims, it is wrong to say that special administrators did nothing immediately after the commencement of special administration. In fact, it is almost regular to see that special administrators initiate a client money claim process soon after being appointed.²⁴⁰ But the problem is that no deadline of proving client money claims was used simultaneously, as clearly evidenced in the four abnormal cases²⁴¹; this is counterintuitive; without imposing a deadline immediately, how can efficiency and finality be achieved?

Before the Regulations 2017, special administrators might have the excuse of having to seek a court approval before setting up a bar date; but after that, for example, in the Reyker case, almost 9 months²⁴² after the special administration procedure started, the special administrators then decided on 10 July 2020 that the bar date at 7 August 2020 was to be installed for client money claims²⁴³; they could have done it almost immediately after being appointed, but it was delayed for 9 months.²⁴⁴

The second reason would be over-relaxation of the client money bar date by special administrators, mainly due to a conflict of interests.²⁴⁵ The client money bar date is designed to discipline clients to make a timely client money submission. As is suggested by its name, the bar date means that the client money claim will be barred from a distribution if the claim is lodged after the deadline without good reason. However, in reality, as will be revealed below, the client money bar date is considerably weakened, because it is too generously enforced in different ways both before and after the Regulations 2017.

Before the Regulations 2017, there were seven client money bar date cases (MF Global, Pritchard, WorldSpreads, Alpari, LQD, Boston and Hume). These seven cases were generally conducted in three different manners. The first manner is that the client money bar date should not be applied in the first place due to the small number of clients affected and is only seen in one case, the special administration of Boston; in this case, there were only two clients whose cash was subject to client money protection, but the problem is that, given this single digit of clients, how the special administrators’ actions in going through the lengthy, costly court process and getting a court client money bar date approval could be economically justified; they could simply solicit the consent from these two clients and got the job done immediately; unsurprisingly, in this case, on average, each client had to pay the distribution costs of £49,669.50 to get their money back.²⁴⁶ Of course, in the strictly legal sense, rather ironically, it is perfectly fine for the Boston special administrators to do this.

Under the second manner, the court-approved client money bar date was too loosely observed at best and was essentially ignored at worst by special administrators, which happened in two cases, the special administrations of MF Global and Hume. Under the court order authorizing the use of the client money bar date by the MF Global special administrators, three points were unequivocally clarified: first, the special administrators were permitted to set up a client money bar date; second, any client who did not lodge a client money proof by the bar date was not entitled to a share in the proposed client money distribution; third, for a client who did not submit a client money claim but whose client money balance was recorded on the firm’s books, the special administrators would calculate this client’s client money entitlement according to the firm’s records.²⁴⁷ The MF Global special administrators chose 19 July 2013 as the client money bar date.²⁴⁸ In principle, after the bar date, any late claim would not be accepted, and for non-responsive clients, their client money entitlements would be decided solely on the basis of the firm’s records. This aimed to achieve both speed and finality of the client money return. But this was not the case.

Rather than disregarding any late claim as mandated by the bar date, the MF Global special administrators acted proactively by soliciting late claims long after the bar date expired, which looks confusing. For example, the special administrators stated in a report released early in 2014 that the unresponsive clients were ‘chased by email, post and phone’ to submit client money proofs (long after the bar date), which reduced the number of clients indifferent to the bar date by 200 to 817, with the remaining clients being chased continuously.²⁴⁹ In a subsequent report late in the same year, it is stated that they contacted the clients holding a client money entitlement of over £100 but having failed to lodge a client money proof by the bar date in writing and ‘contacted at least three times by email, telephone or advert all clients with an entitlement over £1,000’.²⁵⁰ After the bar date, any late client money claim was supposed to be declined straightforwardly, but what the special administrators did was exactly opposite to what was clearly authorized in the client money bar date court order. Admittedly, for a non-responsive client, her client money entitlement can be decided according to the firm’s records.

But it seems that the MF Global special administrators were legally allowed to do so. In line with what is practised in normal corporate insolvencies on creditor claims, which gives insolvency practitioners discretion in handling late claims, ²⁵¹ the MF Global client money bar date court order did have a clause permitting that ‘the administrators are not obliged to deal with client money proofs lodged after the last date for proving but may do so if they think fit’.²⁵² However, this article is of the view that this cannot be used as a defence on two grounds.

Ground one is that this discretion clause can apply where a late claim was initiated and lodged by a client after the bar date; proactively demanding the submission of a late client money claim did not suit the factual condition of this clause; to put it in a different way, the special administrators could exercise their discretion when approached by a late client money claimant, but were not given the discretion to solicit clients to make a late client money submission. Ground two is that even if the special administrators were given such a discretion, using the discretion by chasing late client money claimants is against the spirit of the client money bar date mechanism itself, since the installation of the bar date was for the purpose of seeking efficiency and finality. A similar story also happened in the special administration of Hume. The client money bar date was pinpointed at 2 October 2015 in this case,²⁵³ but almost 7 years later in March 2022, the Hume special administrators still generously asked unresponsive clients to lodge client money proofs.²⁵⁴

Under the third manner, which was adopted in the remaining four pre-Regulations-2017 cases (Pritchard, WorldSpreads, Alpari and LQD), it looks good since the client money bar date was adequately adhered to, but the real motivation of special administrators is to do so is, to a large extent, to mitigate their own fees charging risks. In the special administration of Pritchard, after the client money bar date expired on 12 April 2019, the special administrators announced that ‘any client who had not submitted a client money claim by the bar date has lost any right to receive a distribution from the client money pool’, with no hesitation to leave 2,326 unresponsive clients behind.²⁵⁵

But, the inconvenient fact is that shortly after the bar date what was left in the client money pool only amounted to £1,161,180, which shrank from the original size of £23,849,631 over the years partly because of the previous distributions and partly because of footing the distribution costs²⁵⁶; the special administrators had lost interests in soliciting unresponsive clients to make late client claims, since they were more likely to get unpaid; this is because they could only draw their fees and relevant expenses from the client money pool²⁵⁷; given that the client money pool has been reduced to a minimum, it is not financially worth doing more work by contacting silent clients any more.

An almost identical story happened in the special administrations of WorldSpreads and LQD, where the client money pool fell to £1,315 from £7,295,252.54²⁵⁸ and to £15,242.17 from £1,642,568.78,²⁵⁹ respectively, shortly after the bar date ended. This dampened the appetite of the special administrators to relax the bar date set up by themselves. The special administrators of Alpari also stopped accepting late client money claims after the bar date, largely because the creditors’ committee had put a cap of £10,500,000 on the fees²⁶⁰ they could earn and because at the expiry of the bar date the special administrators’ commission had already reached £9,833,824.87.²⁶¹ Needless to say that in the special administration of Alpari, after the bar date, the client money pool dropped to £1,488,339.41 from £97,478,692.23,²⁶² which was perhaps also a major factor motivating the special administrators to strictly enforce the bar date.

Some may ask why so many clients did not lodge their client money claims by the bar date. For example, in the Pritchard special administration 2,326 out of all 11,211 clients remained silent to the bar date call.²⁶³ In fact, the proportion of unresponsive clients by number looks huge, but they only represent a tiny minority by value. In the Pritchard special administration, the indifferent clients counted as 20.75 per cent by number but only represented some 3 per cent by value.²⁶⁴ Most of them held a small claim, which might be why they essentially had abandoned their client money claims by staying unengaged.²⁶⁵ To get a glimpse of how small some client money claims may be, this study happened to have obtained a list of client money claimants in the Pritchard special administration.

Pritchard had 11,211 clients, most of them individuals; for individuals, the special administrators were obliged not to disclose their names and addresses,²⁶⁶ therefore there was only a list of corporate clients holding client money claims made publicly available.²⁶⁷ Under this list, there were 485 corporate client money accounts (one client may have more than one client money account), 210 (43.30 per cent) of them having the balance of £1 or less; the highest balance at £365,835.56 can be seen in one account.²⁶⁸ By common sense, compared with corporate clients, individual ones may hold smaller client money claims. This means that potentially a higher proportion of individual clients may have a negligible amount of money in their client money accounts. This could well explain why 20.75 per cent of clients by number stayed in silence to the client money bar date call; it is simply not worthwhile to step forward.²⁶⁹

Meanwhile, small-claim clients might be deterred from coming forward for fear of disproportionate distribution costs. For example, in the special administration of Hume, each client was asked to pay a fixed fee of £3,172.82 to have their client money balance returned²⁷⁰; it seems unreasonable for small-claim clients to lodge a claim. Although in most cases these costs would be ultimately born by the FSCS,²⁷¹ understandably there is no incentive for small-claim clients to get involved.

Three defects of the making of the client money bar date

Before the Regulations 2017, generally speaking, the role of the bar date was arguably undermined by two defects. The first defect is the inconsistency over how to handle unresponsive clients whose cash balances are recorded on the firm’s books. On this question, three different approaches could be found, which led to considerable confusion. Approach one is that the court orders special administrators to only accept timely-lodged client money claims but says nothing on how to deal with unresponsive clients whose cash balances are clearly shown on the firm’s records, as is seen in the special administration of LQD.²⁷² One may guess that this implies that only duly-submitted claims can share in the distribution and that two groups of clients can be lawfully excluded, one of them the clients who do not make a submission before the bar date and whose client money balances could not be verified from the firm’s records and second of them the clients who do not lodge a proof but whose cash balances are recorded in the firm’s books. Such ambivalence should be removed.

Approach two is more lenient in favour of some clients; under this approach, the court instructs special administrators to make distributions to duly-filed client money claimants as well as to those who do not make a claim but whose balances are clearly recorded, which was used in the special administration of MF Global.²⁷³ Approach three could be the harshest but is a white-and-black formula feeding certainty; under this approach, the client money distributions can only be made to duly-filed client money claimants, and for non-responsive clients no distribution will be made even if the firm’s records show the cash balances under their names; this approach can be found in the special administration of Alpari.²⁷⁴

Technically speaking, approach two would be preferred, since it strikes an adequate balance between efficiency and fairness. But the problem is that even if approach two is used, it can still be derailed by the second defect, which sees the lack of willingness of special administrators to vigorously enforce it. These two defects seem to be forgivable since a trial-and-error learning curve was inevitable during the first several years of implementing the special administration regime. After the year 2017, given that many detailed rules on the client money bar date have been clarified in the Regulations 2017, the use of the bar date is supposed to see an improvement. In fact, the new application of the client money bar date simply went from bad to worse.

There were five post-Regulations-2017 special administrations using the client money bar date (Strand, Beaufort, SVS, AFX and Reyker); on the use of the client money bar date, two old defects remained, and one new, perhaps more damaging, defect arose, which was ironically somewhat created by the Regulations 2017 itself. The old defect is that the Regulations 2017 says nothing on whether the firm’s records can be used to determine the client money entitlements of clients who do not lodge a proof by the bar date. Of the aforementioned five special administrations, no information was obtained from the special administration of Strand; based on the remaining four cases, the special administrators explicitly stated, in the three special administrations of Beaufort,²⁷⁵ SVS²⁷⁶ and Reyker,²⁷⁷ that for nonresponsive clients their entitlements could be counted on the basis of the firm’s records; by contrast, in the special administration of AFX, although the special administrators did not disclose their principles in writing, their actions indicated that they did not rely on the firm’s records to include silent clients.²⁷⁸

Also, it is worth noting that in spite of the disclosed commitment the Reyker special administrators did not practise their written principles and still kept on requesting silent clients to make client money submissions long after the bar date.²⁷⁹ Therefore, on this old defect, if the special administration of Reyker is regrouped with the AFX case, it can be said that in half of the new cases the firm’s records were used to calculate the client money entitlements of unresponsive clients and in another half they were not. Inconsistency persisted stubbornly.

Defect number two is the inconsistency of special administrators towards the use of the client money bar date. This trouble remained in place. In the Beaufort case, the client money bar date was adequately enforced, mainly because the creditors’ committee imposed a fees cap of £55.1 million and because shortly after the client money bar date the fees claimed by the special administrators had already exceeded the ceiling²⁸⁰; the Beaufort special administrators could not benefit from relaxing the bar date, which was perhaps why they did not urge unresponsive clients to lodge client money claims after the bar date.

In the special administration of SVS, in accordance with the client money bar date terms,²⁸¹ the special administrators did not chase unresponsive clients after the expiry of the (extended) client money bar date and straightforwardly used the firm’s records to calculate the entitlement of the silent clients²⁸²; why did the special administrators do this? The likely answer is that in this case the FSCS, rather exceptionally, agreed to pay the client money distribution costs directly to the special administrators, and for unresponsive clients the special administrators were happy to use their names on the firm’s records to claim fees from the FSCS²⁸³; if some clients chose to stay away, it did not bother the special administrators very much, since these clients were deemed to have submitted a claim.

It is also worth noting that in this case the original client money bar date which was set on 10 January 2020 was not strictly followed by the special administrators; it was extended to 7 May 2020 to allow further submissions²⁸⁴; by doing so, on the one hand, perhaps the special administrators needed to offer the FSCS a reasonable response rate of client money claims so as to be qualified for the direct payment from the latter, and on the other hand the extension did prolong the special administrators’ working hours and hence push up the scale of the overall fees.

Contrary to what happened in the special administrations of Beaufort and SVS, the client money bar date in the special administration of AFX²⁸⁵ and Reyker²⁸⁶ was virtually ignored, in spite of the fact that the Reyker special administrators clearly claimed to calculate unresponsive clients’ client money entitlements directly on the basis of the firm’s records after the expiry of the client money bar date.²⁸⁷ Meanwhile, it seems that the Regulations 2017 unexpectedly offered a tool to special administrators who can further relax the client money bar date, which gave rise to the new defect.

Defect number three is that the enactment of the Regulations 2017 counterproductively weakened the already flimsily enforced client money bar date by allowing late client money claims to join proposed distributions. The Regulations 2017 supplements the old Regulations 2011 by adding a sub-paragraph, which becomes r. 12A(4) of the Regulations 2011, demanding special administrators to include a late claim in the proposed client money distribution if it is submitted after the bar date but before the proposed distribution.²⁸⁸

The Regulations 2011 r. 12A(4) is apparently a step backwards in two ways. First, it sends a wrong message to the client/investor community that missing the client money bar date may not have a dire consequence. Second, it probably unintentionally removes the discretion often sanctioned by the court in the past allowing special administrators to either accept or decline late client money claims as they think fit; under the new rules, special administrators must accept late client money claims, with the old discretion unconditionally removed.

This new defect was immediately materialized in the special administrations of AFX and Reyker. In the AFX case, almost 3 years after the expiry of the client money bar date, the special administrators stated:²⁸⁹

Under regulation 12A(4) of the Special Administration Regulations a person who submits a claim after the bar date but before the return of client money after that date must, as far as is reasonably practicable, be included in the distribution of client money. Therefore, we are still accepting claims notwithstanding the expiry of the “bar date.”

The client money bar date in the special administration of Reyker was initially set up as 7 April 2020, and the special administrators unequivocally stated that for non-responsive clients their client money entitlements would be calculated on the basis of the firm’s records, emphasizing that what nonresponsive clients lost was not their substantial client money entitlements but their right to dispute the accuracy of the firm’s client money records.²⁹⁰ This was correct. Before the 7 April 2020 client money bar date, 7,832 (70.39 per cent) out of all 11,126 clients lodged their proofs.²⁹¹

At this point, what the Reyker special administrators were expected to do was to verify what had been submitted by the 7,832 responding clients and to directly calculate the client money entitlements for the remaining 3,294 silent clients according to the firm’s books prior to proposing a distribution. It is also worth clarifying that Reyker’s client money records were checked before and that ‘no material discrepancies have been identified and the company’s cash records and client statement are, overall, very good’.²⁹² proving their reliability. However, for non-responsive clients, the special administrators did not go to the firm’s records as scheduled.

Instead, in early July 2020, the Reyker special administrators wrote to all silent clients again and gave them a new deadline of 21 July 2020 to submit client money claims.²⁹³ To offer the silent clients a new deadline, the special administrators quoted r. 143 of the Investment Bank Special Administration (England and Wales) Rules 2011,²⁹⁴ saying that it was a statutory requirement to give unresponsive clients ‘an additional 14 business days from receipt of the notice to submit a claim’.²⁹⁵ This can be a problem, since r. 143 applies to client asset claims, not on client money ones.²⁹⁶

The leniency of the special administrators did yield a seemingly positive result. After the deadline extension, the overall claim submission (including both client asset and money claims) rate rose from 70.39 per cent to 84 per cent.²⁹⁷ The devil is always in the details. The tricky detail is that the special administrators did not elaborate how much these 13.61 per cent of late claims by number were represented by value. Presumably, a disproportionately tiny percentage of client money claims by value were submitted after the first postponed deadline, given that there is usually a significant number of client money accounts having small and negligible balances. Throughout the process of returning client money in the Reyker case, relying on the firm’s records to calculate unresponsive clients’ client money entitlements was largely forgotten.

Overall, it seems to be an irony that the new investment bank special administration regime has been proven to be ineffective in facilitating a quick return of client money, especially with the use of the client money bar date. Before the promulgation of the Regulations 2011, in the Lehman case, the return of client money took 646 days under the client money bar date; but supported by the brand-new special administration procedure under the Regulations 2011, this waiting time was, instead, prolonged to 817.92 days under a similar client money bar date. To a large extent, the new special administration procedure is a failure rather than a success. The most striking contrast is that under the new regime, without the use of the client money bar date, clients only waited for 71.6 days before seeing the return of their money.

On the client money bar date, at least two lessons can be learned. First, if no client money shortfall is found, which indicates the accuracy and reliability of the firm’s client money records and can be found in most cases, the client money claim process should not be used in the first place; special administrators can simply rely on the firm’s records in returning client money; putting aside the correct client money records of the firm, resorting to a client money claim process is wasteful and costly; if a client money claim process cannot be justified, undoubtedly, the use of the client money bar date does not have its foundation at all.

Second, in the rare event that the firm’s client money is genuinely messy and incorrect, for instance, when a fraud is committed, if a separate client money claim process is warranted to determine the client money entitlements of each client, the client money bar date is surely necessary, but the law should be updated to clarify two points: late client money claims will not be considered, and special administrators should be obliged to use the existing client money records to calculate the entitlement of unresponsive clients following the expiry of the bar date, otherwise, the client money bar date would be entirely useless, as was seen in the reported cases.

5. Conclusion

As for the first-10-year implementation of the new investment bank special administration regime on the return of client money in the UK, several key points are worth summarizing.

First, as anticipated, the client money pooling event was regularly triggered and took place in 84.21 per cent of the special administrations, which was largely unnecessary and counterproductive to seeking a speedy return of client money. The implied assumption of using the client money pooling event is that there is usually a client money shortfall when a failed firm collapses into special administration. But this study finds that no client money deficit was found in 76.47 per cent of special administrations, and more strikingly, on average, the overall substantial client money deficit rate only stands at 0.46 per cent, indicating that the majority of the failed investment firms behaved properly in safeguarding client money. The assumption that there are widespread client money shortfalls cannot be backed by the real-world data, as a result of which using this unfounded assumption to justify the use of the client money pooling event only leads to disruption and delays.²⁹⁸

Moreover, in achieving the quickest return of client money by way of a bulk client money transfer, it is exactly the client money pooling event that should be avoided and was indeed strategically avoided in three successful special administrations which expedited the client money returns most effectively.²⁹⁹ Practise suggests that unless and until there is a client money shortfall, which happened in the minority of the existing cases, the client money pooling event should be avoided.

Second, the real threat to the client money estate is the distribution costs imposed by special administrators; the data collected by this study suggest that in non-original-client-money-shortfall special administrations, due to an acute conflict of interests, most special administrators used the existence of client money shortfalls to justify a separate client money claim process, which was wastefully costly and lengthy. This article can also confidently argue that imposing distribution costs on the client money estate cannot be rightly justified,³⁰⁰ and that the re Berkeley Applegate principle should be updated to reflect what has been adequately practised in reality: special administrators do not charge distribution costs on the client money trust when the firm is solvent.³⁰¹ Meanwhile, during the making of the Regulations 2011, lawmakers were mainly concerned that clients should not be protected at the expense of creditors.³⁰² But, in practice, quite the opposite can be found: clients are unfairly dragged to pay the costs which are supposed to be paid by creditors, in around 57.14 per cent of the existing special administrations. It is time to unequivocally prioritize the interests of clients.

Third, for a speedy client money return, the only practical and effective way is to return client money through a transfer.³⁰³ As reported, at least in 77.78 per cent of the existing investment bank special administrations, the firm behaved adequately in safeguarding client money, suggesting the accuracy of the firm’s client money records.³⁰⁴ Relying on the firm’s records for a bulk transfer is safe, and it is economically irrational to go through a separate client money claim process by ignoring the accurate records.³⁰⁵ Moreover, what should not be forgotten is that the client book comprising client money is a valuable asset, as a result of which it was sold at a considerable price in many cases; with the sale of the client book, one benefit is that it can generate an income to improve the return to creditors, but more critically, its by-product, if not its major advantage, is to achieve a quick return of client money.³⁰⁶

This article proposes that in the event of no willing acquiring party emerging, the FSCS should proactively offer monetary incentives up to 17.83 per cent of client money to a third party that is able to take over the client book.³⁰⁷ By doing so, it can kill two birds with one stone. On the one hand, a client money bulk transfer can be achieved³⁰⁸; on the other, it may considerably reduce the compensation paid by the FSCS. This article can safely speculate that probably up to 5 per cent of client money as a financial incentive would be enough to persuade a third eligible party to come forward. Without offering a financial incentive to a buyer, the FSCS has to pay more as insurance compensation to cover the so-called distribution costs as examined in this article, which stands at 17.83 per cent, an astonishingly high percentage.

Fourth, the use of the client money bar date has been proved almost useless in facilitating a quick return of client money.³⁰⁹ In exceptional circumstances, such as fraud, if a client money claim process is really warranted, the bar date can be used but should be strictly defined.³¹⁰ In line with British convention on the last date for proving on creditor claims in corporate insolvencies,³¹¹ if having missed the bar date, the client should be prohibited from joining the proposed client money distribution, unless an excusable neglect can be proven.³¹² The current practice of overly indulging late client money claims simply led to endless delays, contrary to the fundamental objective of the investment bank special administration regime in expediting the return of client money. The empirical data also suggest that whether the client money bar can be vigorously enforced depends on whether it is in the interest of special administrators, and this unhelpful discretion could be removed so as to enhance both certainty and efficiency.

Footnotes

Acknowledgements

This article was presented at the 5^th International & Comparative Insolvency Law Symposium held at Royal Holloway University of London in April 2024 and at the INSOL International Academic Colloquium taking place at San Diego, California, USA, in May 2024, and the author wishes to thank the conference participants for helpful comments. The author also wishes to sincerely thank the anonymous reviewer who gave generous support which led to the publication of this article. All errors remain the author’s responsibility.

Funding

No funding is involved.

Conflict of interest. None declared.