The bar date over client asset claims in the UK investment bank special administration regime

1. Introduction

This article empirically investigates the use of the bar date over client asset claims under the UK investment bank special administration regime.¹ The bar date, a deadline, is routinely used in bankruptcy procedures to require creditors to file claims in a timely manner, because ‘dates matter in bankruptcy’,² and a claim lodged after is likely to be barred, as is literally suggested.³ The use of the bar date is key for the efficient management of the bankruptcy estate on the grounds that it can facilitate finality, otherwise, insolvency procedures may never be closed.⁴

For the UK legal community, the bar date is somewhat an alien term, since it is traditionally used by insolvency practitioners in the USA,⁵ and probably in the early 2000s, this phrase began to appear in some UK corporate insolvency cases.⁶ In the UK, by law, the equivalent of the bar date is formally called ‘the last date for proving’,⁷ which means that for creditors all claims must be submitted no later than a fixed point of time, but, by convention, British insolvency professionals tend to use the terms like deadline or cut-off date.⁸ It is, to a certain extent, only different linguistic preferences between the USA and the UK, but in substance, they are arguably the same.

Originally, the bar date applies to creditors’ debt claims, but it has recently been expanded to post-bankruptcy expenses⁹ and, further, to client proprietary claims in investment bank insolvencies.¹⁰ This article focuses on the use of the bar date over client asset claims (the bar date over client money claims will be investigated in a subsequent article) under the UK investment bank special administration regime, a tailored insolvency procedure for investment banks, since the bar date is designed as one of the key measures to pursue the objective of this special insolvency regime in expediting client asset returns.¹¹ The application of the bar date on pre-bankruptcy debts and post-bankruptcy expenses is equally important and could be investigated elsewhere.¹²

Although the bar date mechanism on client asset claims has been formally introduced into the UK investment bank special administration regime since 2011,¹³ many critical questions remain unanswered. The first and foremost question is whether the use of the bar date is really effective in achieving a speedy return of client assets, given that the British lawmakers view it as a critical tool in fulfilling the legislative goal when creating the investment bank special administration regime. The law on the books is different from, if not opposite to, the law in action;¹⁴ an evidence-based approach is necessary to test the effectiveness of this legal vehicle.¹⁵ Second, if the answer to the first question is negative, the subsequent question is why the application of the bar date is not as effective as envisaged by the lawmakers.

By investigating these two major questions, this article aims to propose evidence-backed policy recommendations so as to strengthen the UK investment bank special administration regime, in the belief that a client-friendly investment bank insolvency system will boost investor confidence and in return help develop the UK financial markets further.¹⁶

Bearing in mind these two questions, the rest of this article proceeds in four parts. Section 2 outlines the UK investment bank special administration regime in general and its adoption of the bar date in particular. Section 3 explains the methodology, clarifying how the empirical data are collected. Section 4 reports the findings and offers some analysis. The conclusions and some policy reform recommendations are placed at the end of this article.

2. The rise of the UK investment bank special administration regime and the development of the bar date

Before the 2007–2008 global financial crisis, in the UK, the bankruptcy of financial institutions including deposit-taking and investment banks was subject to the general insolvency law, as was practised in many other jurisdictions,¹⁷ namely there were no special bankruptcy procedures for them.¹⁸ To avoid further complexity, this article deliberately ignores the bankruptcy of a third type of financial institutions, insurance firms.¹⁹ Like many other jurisdictions, the UK was taught a lesson for the absence of a special insolvency system for failing financial institutions, since the general insolvency law is unable to accommodate the special issues emerging from financial institution insolvencies, like the maintenance of the national financial stability and the protection of depositors in a deposit-taking bank insolvency and of clients/investors in an investment bank failure.²⁰

In response to the challenges imposed by the 2007–2008 global financial crisis in general and by the collapse of the British bank, Northern Rock,²¹ in particular, the UK Parliament promulgated the Banking Act 2009 which is mainly tasked with establishing the bank resolution regime, a special regulator-centred insolvency procedure, to deal with failing deposit-taking banks.²² Meanwhile, fully aware of the deficiency of the British general insolvency law in handling the failure of investment banks, especially the failure of the giant investment bank, Lehman Brothers International (Europe),²³ the UK Parliament used the Banking Act 2009 to authorize the HM Treasury, the economic and finance ministry, to draft a secondary legislation for the bankruptcy of investment banks.²⁴ This led to the enactment of the Investment Bank Special Administration Regulations 2011,²⁵ which created the UK investment bank special administration regime and took effect on 8 February 2011.²⁶

Investment bank insolvencies are different from most general corporate insolvencies, since when an investment bank is bankrupt, most assets under the control of the company are not its own property, instead they are the property of the company’s clients. For example, in the 2008 administration of Lehman Brothers International (Europe), an investment bank as mentioned before, the company held some £48 billion in client assets, in contrast to the company’s own assets of only approximately £27 billion.²⁷ A typical investment bank is a securities broker holding various assets, like shares, bonds and derivatives, on behalf of and for its clients/customers (this article uses the terms client and customer interchangeably).²⁸

However, under the general insolvency law, protecting creditors is usually the first and foremost goal, and this inevitably leads to a situation where clients might be neglected.²⁹ In the highly publicized investment bank failure, the collapse of Lehman Brothers International (Europe), many clients were only able to see the return of their assets after one year,³⁰ and even a senior judge handling this case revealed that many clients in fact saw the return of their assets long after the creditors were paid.³¹ So, the key concern of an investment bank insolvency is how to return client assets quickly so as to alleviate disruption to clients.³² It is worth noting that the new investment bank special administration regime is largely based on the administration procedure under the UK general insolvency law, the Insolvency Act 1986.³³

Expediting the client asset return as the key objective

Under the Investment Bank Special Administration Regulations 2011 (hereinafter referred to as the Regulations 2011), when an investment bank is placed into a special administration proceeding, the special administrator is obliged to pursue three objectives: objective 1 is to return client assets as quickly as practicable, objective 2 is engaging with regulatory authorities and market infrastructure bodies, and objective 3 is essentially replicated from the general corporate administration procedure, which requires the special administrator either to rescue the failed investment bank as a going concern or to wind it up in the best interest of the creditors as an alternative.³⁴

Apparently, objectives 1 and 2 are quite unique, since the ordinary administration procedure under the Insolvency Act 1986 is not sophisticated enough to address these issues. Arguably, objective 1 of seeking a quick return of client assets is the key goal of establishing a special administration regime for investment banks, and this is the major purpose of having this special insolvency procedure.³⁵ Objective 2 is mainly on regulatory affairs, since the special administrator should work with the regulatory bodies, especially the Financial Conduct Authority (FCA) on regulatory compliance as well as with the Financial Services Compensation Scheme (FSCS) on investment insurance payments.³⁶

But there are at least two challenges in relation to these three objectives, which may derail this new regime entirely. First, although objective 1 states that the special administrator should seek a quick return of client assets, it remains silent on how quick the return of client assets should be; one day could be deemed to be quick, one month sounds okay, and one year could also be treated as being speedy if you are a generous commentator.³⁷ It can be argued that objective 1 might be too subjective to be meaningful in reality.

The second challenge is that no one objective by default takes precedence over others, which may considerably weaken the legislative intention of using the new law to expedite client asset returns.³⁸ That each one is important can lead to a situation where no one is important. In spite of the seeming hierarchy of the three objectives, the Regulations 2011 r. 10 (3) stipulates plainly that ‘the order in which the special administration objectives are listed in this regulation is not significant’.³⁹ To put it another way, objective 1 of seeking a quick client asset return is not automatically a priority. This is very puzzling.

Admittedly, although there is no statutory hierarchy elevating the client asset return as a top mission, the law leaves how to balance between these three objectives to the special administrator, authorizing that it is subject to the special administrator to prioritize any one objective as ‘the administrator thinks fit’.⁴⁰ In addition, as for the three objectives, the FCA, one of the UK tripartite financial regulators, ‘may direct the administrator to prioritise one or more special administration objectives’,⁴¹ which suggests that objective 1 has a chance to be given special attention; however, probably to further weaken client protection, the Regulations 2011 adds that if the Financial Conduct Authority gives a direction, the regulatory intervention must be exercised for the purpose of maintaining the stability of the UK financial systems and public confidence;⁴² a reasonable reading of this provision would lead to a gloomy conclusion that it may be highly unlikely for the FCA to direct the special administrator to single out objective 1 as the top priority.

Overall, although there are some ambiguities concerning the key legislative objective of expediting the client asset return, establishing the three objectives is still a step forward to reducing the disruption of investment bank failures to customers. Apart from pointing out what the special administrator should seek to achieve, the investment bank special administration regime also puts forward a favoured way of returning client assets, a client asset transfer.

The best option: client asset transfers

The most effective way of returning client assets would be through a transfer to a healthy third-party investment bank; this is usually achieved by way of a business sale. For UK lawmakers, it is somewhat of a rocky journey to recognize the effectiveness of this tool. In fact, many legislative/regulatory efforts have been made to pursue a transfer.

The earliest legislative step can be identified in the Banking Act 2009, which mentions the transfer vehicle. In authorizing the Treasury to draft the upcoming investment bank special administration regulations, section 233(3) of this Act addresses that ‘the Treasury shall have regard to the desirability of (a) identifying, protecting, and facilitating the return of client assets’.⁴³ Apparently, the Parliament was aware of the pressing issue of returning client assets under this scenario, which was why a special insolvency regime was needed in the first place. Immediately after this paragraph, the following one of the Act goes a step further and offers a more detailed guidance on how to conduct a client asset return⁴⁴:

A reference to returning client assets includes a reference to (a) transferring assets to another institution, and (b) returning or transferring assets equivalent to those which an institution undertook to hold for clients.

To a certain extent, two points can be generated from reading these two paragraphs. First, tasked with making the new special insolvency procedure for investment banks, the Treasury was instructed to pay particular attention to the return of client assets. Second, returning can be exercised in two ways, one of them returning them to clients themselves and the other carried out through a transfer. But, when the Regulations 2011 was finally promulgated, surprisingly, it deviated from the above provisions.

In the Regulations 2011, rather disappointingly, the transfer tool seemed to have been forgotten, since it was not even noted in the new legislation. It is more likely to be an error. Under the Regulations 2011 rule 10,⁴⁵ the administrator is given objective 1 ‘to ensure the return of client assets as soon as is reasonably practicable’. As for whether a return of client assets can be conducted by way of a transfer, the new law was silent. The Treasury was likely to either have inadvertently neglected the transfer mechanism or have simply abandoned this approach. It seems that the transfer tool was, at best, not preferred and, at worst, disregarded by the Treasury in the Regulations 2011, as it was very unlikely for the Treasury to have neglected this tool for the following reason.

The Treasury obviously knew the linguistic and practical distinction between the words return and transfer. In the Regulations 2011, the word ‘transfer’ was indeed used several times,⁴⁶ but not on the return of client assets; instead, it was on depositors’ accounts in the event that the investment bank also had the deposit-taking function and needed to return deposits to clients through a transfer to a third party.⁴⁷ Therefore, it can be safely speculated that at least the Treasury did not specifically highlight the use of a transfer as one way of returning client assets. Nevertheless, it is fine for lawmakers not to address this tool, but for practitioners who engage in everyday real cases, the usefulness of transfers cannot be neglected. This is extensively reflected in the review reports prepared by Mr Peter Bloxham, an eminent expert hired by the Treasury to review the Regulations 2011, in 2013.

Among many legislative reform recommendations, Mr Bloxham emphasized that the updated investment bank special administration regime should have a mechanism to facilitate a transfer, and more importantly, Mr Bloxham, boldly and correctly, called the Treasury to ‘amend relevant SAR (special administration regime) objective from “return” of client assets to a more explicit reference to “return or transfer”’.⁴⁸ Interestingly, the Treasury partially turned down this proposal on the grounds that the Banking Act 2009 had included transfers as one way of returning client assets; namely, return can broadly cover the transfer, as a result of which the wording of objective 1 of investment bank special administration did not need to be changed.⁴⁹

Anyway, although the wording of objective 1 remained intact, the Treasury did accept many recommendations on how to facilitate a transfer when amending the Regulations 2011 through the Investment Bank (Amendment of Definition) and Special Administration (Amendment) Regulations 2017 (hereinafter referred to as the Regulations 2017). Before outlining several key points in the amended Regulations 2011 made to seek a transfer, a technical nuance is worth noting, which shows the changed legislative attitude. In the text of the original Regulations 2011, the word transfer appears 15 times, none of them used for a transfer of client assets, as argued before; by contrast, this word is repeatedly used 112 times in the Regulations 2017, most of them presenting rules on how to make a client asset transfer more feasible.⁵⁰ This is the right direction and is backed by some important measures which are discussed below.

First, inspired by the American practice,⁵¹ rule 10B of the Regulations 2011 authorizes that a client asset transfer does not need the advance consent from any individual clients; a transfer can be operated by law; this means that any contractual restrictions to a scheduled transfer will be overridden.⁵² This powerful mechanism can be critical for a quick transfer, without which obtaining each client’s agreement prior to a proposed transfer can be time-consuming and costly.

Second, the client data protection rules can also be largely⁵³ relaxed if the special administrator wants to disclose to the transferee all information about the affected clients; before this, some administrators might be uncomfortable doing this without either client consent or regulatory approval. To ensure this rule is properly exercised, the amended Regulations 2011 requires the special administrator only to disclose the information ‘relevant to the transfer’.⁵⁴ Conditionally relaxing the data protection rules might significantly conduce to a speedy transfer.

Third, to overcome any obstacles hindering a proposed transfer, the updated Regulations 2011 reaffirms that the rights of set-off and netting, especially conferred to the investment bank, will not be affected by the transfer, which is to ensure a proper balance struck between clients and creditors.⁵⁵ Without this mechanism, the special administrator might be reluctant to pursue a transfer for fear that the general estate of the investment bank will be undermined, since many rights depend on either possession of certain client assets or the strict fulfilment of the mutuality principle.

In a word, after the Regulations 2017 took effect, a bundle of tools was available so as to pave the way for a viable transfer. These tools are definitely useful. But the question is whether transfers can be effectively achieved by these means. Apart from the recommended transfers in seeking a quick return of client assets, a bar date is highlighted by both the Regulations 2011 and the Regulations 2017 in order to expedite the process of returning clients assets. Of course, the bar date mechanism is the focus of this article, which will be comprehensively examined.

The creation and expansion of the bar date mechanism in the UK investment bank special administration regime

As noted before, the term ‘bar date’ emerged around the early 2000s in the UK, meaning a deadline to require creditors to submit claims before a fixed point of time. Later it was formally incorporated into the Regulations 2011 as a mechanism to facilitate the speedy return of client assets, and when the Regulations were amended in 2017, its coverage was further expanded, with its compelling powers further strengthened.⁵⁶ Historically speaking, the application of the bar date in the UK has experienced over three major phases.

The appearance of the bar date in UK corporate insolvencies before 2011

Arguably, the phrase ‘bar date’ first appeared in some schemes of arrangement,⁵⁷ and its earliest use can be tentatively identified in the 1998 case of Re Osiris Insurance Limited in which the scheme of arrangement was proposed and the bar date was set to urge scheme creditors to submit claims.⁵⁸ Before the year 2011, its use can be found in at least nine reported cases, seven of them on claims in schemes of arrangement, one on liquidation expenses and one about Company Voluntary Arrangement (CVA) claims.⁵⁹ Several major characteristics of the bar date in these cases can be found, and they have perhaps shaped what the bar date looks like in the future investment bank special administration regime.

First, court approval must be obtained before the imposition of the bar date over claims. This is partly because the law in these areas over whether a bar date can be used is not clear enough, as a result of which it seems safe to get a mandate from the court. As for schemes of arrangement, generally, Part 26 of the Companies Act 2006 is silent on whether a deadline can apply to scheme creditors’ claims; by the same token, in terms of liquidation expenses⁶⁰ and Company Voluntary Arrangement claims, a clearly worded statutory authorization also cannot be found.⁶¹ Court approval in the form of directions seems to be a safe option. Apparently, court scrutiny helps ensure fairness and equity, given the serious impacts of missing the bar date.⁶²

Second, to get the court’s blessing, in most cases, the company devised a comprehensive notification package in the hope of informing as many claim holders as possible. For instance, in Re Osiris Insurance Ltd, for scheme creditors already recorded on the company’s books, the company contacted them directly by sending them a copy of the scheme and its explanatory statement, and more importantly, for potential/unknown scheme creditors who did not appear on the company’s books and records, the company took ‘sufficient steps’ to notify them by advertising in six leading newspapers.⁶³ In other words, given the grave consequence of missing the bar date, all scheme creditors should be sufficiently informed. Sufficient notification to all affected parties seems to be a second major characteristic of using the bar date.⁶⁴

Third, the bar date is designed with a considerable degree of leniency/tolerance. In Re Osiris Insurance Ltd, for scheme creditors appearing on the company’s books, even if they did not submit claims before the expiry of the bar date, the company still counted their claims according to the company’s own records.⁶⁵ Namely, the bar date non-compliance of known scheme creditors would not affect their rights and subsequent distributions under the scheme. This is different from the rules in ordinary company insolvencies, under which if a known creditor fails to file a claim in writing before the deadline their claims will be in principle barred/extinguished.⁶⁶

Meanwhile, in Re Osiris Insurance Ltd, under the terms of the scheme, a late claim could still be considered and subject to an agreement between the scheme creditor and the company;⁶⁷ at best, it can be called leniency,⁶⁸ but at worst it can be said of being counterproductive. Since the use of the bar date is to achieve certainty and finality, the ambivalence here seems unjustifiable. In theory, a late claim cannot be admitted unless and until excusable neglect can be established, and in this regard British law surely has room for improvement.⁶⁹

Thankfully, years later in 2003, in Re Pan Atlantic Insurance Co Ltd, one condition of the bar date in the scheme of arrangement is noteworthy: a late claim cannot be accepted unless there is a good reason;⁷⁰ in 2008, the court overruled the use of the bar date over CVA creditors in Energy Holdings⁷¹ on the grounds that a late claim should be considered if there is no wilful default or a lack of diligence of the creditor, namely when designing the bar date, care must be taken on whether there is a legitimate reason for a late submission.

The bar date was originally used over debt claims, but a bold attempt was made in the Lehman Brothers International (Europe) administration⁷² to apply it to proprietary claims of investment bank clients. Although the Lehman scheme of arrangement itself⁷³ over client asset claims was later rejected by the court, interestingly using the bar date over proprietary claims has survived,⁷⁴ and years later it was formally adopted by the Regulations 2011.⁷⁵

Adoption of the bar date by the UK investment bank special administration regulations 2011

Perhaps because of the necessity of using a bar date for effective client asset management, as evidenced in the Lehman case, it was officially inserted into the UK investment bank special administration regime in 2011. It is noteworthy that the use of the bar date is mandated by the Regulations 2011 in parallel with the Investment Bank Special Administration (England and Wales) Rules 2011 (hereinafter referred to as the Rules 2011). Several key features of this mechanism can be identified.

First, the special administrator can exercise their own discretion over whether a bar date can be set, without the need to ask for advance court approval anymore.⁷⁶ To put it another way, the mandate of the special administrator in imposing a bar date can be straightforwardly derived from the statute, instead of from court authority. This may considerably improve efficiency, since seeking numerous court directions may artificially protract client asset management.⁷⁷ This move should be celebrated.

Second, the bar date only applies to client assets, not client money.⁷⁸ Excluding the bar date from client money is mainly because of the uncertainty on client money issues disputed in the ongoing litigation of Lehman Brothers International (Europe) when the Regulations 2011 was drafted; the HM Treasury, as the major drafter of the new regulations, did not want to see clashes with any pending judicial decisions.⁷⁹ Meanwhile, given the complexity of client money rules, the HM Treasury believed that the client money issues are better to remain being regulated by the Client Assets Sourcebook (CASS) made by the Financial Services Authority (later replaced by the FCA), rather than to be included in the Regulations.⁸⁰

Third, the investment bank special administration regime is silent on how long the period is between the commencement of the special administration procedure and the bar date, and this is left to the discretion of the special administrator.⁸¹ On the one hand, some can say that it is the flexibility offered to special administrators. On the other hand, certainty over this critical mechanism may be undermined. For better or for worse, the investment bank special administration regime allows a potential claim holder to request an extension of the bar date if extra time is needed to prepare a submission;⁸² a literal reading of this provision suggests that the extension may only apply to an individual claim, rather than generally to all client asset claims, which may not damage the usefulness of the bar date as a whole.⁸³ In addition, the regulator, the Financial Services Authority (later replaced by the FCA), may also request for an extension of the bar date over an individual client asset claim or over a class of claimants if there are justifiable difficulties in meeting the bar date.⁸⁴

Fourth, the bar date carries a specific notification requirement, which is aligned with what has been generated from case law as examined earlier. For known clients whose assets are recorded at the company’s books and records, the special administrator is obliged to inform them of the bar date if they have ‘a means of contacting’.⁸⁵ For unknown/potential clients,⁸⁶ the notice of the bar date should be gazetted, which means, in the UK context, publication at the journal London Gazette in England and Wales.⁸⁷ This is not enough. The Rules 2011⁸⁸ further authorizes the special administrator to advertise the bar date notice ‘in such other manner as the administrator thinks fit’.⁸⁹ On the face of it, the special administrator is given the right to consider advertisement issues, but essentially the special administrator is ‘obliged’ to advertise it to fulfil the sufficient notification requirement as analysed before. To avoid any doubt in this regard, the special administrator should advertise the notice of the bar date so as to ‘come(s) to the attention of as many of these persons who are eligible to submit a claim’.⁹⁰ This demands a high degree of diligence and care.⁹¹

Fifth, the bar date is accompanied by a similar degree of leniency or tolerance to late claims. Before examining the leniency of the bar date, it seems fair to appreciate the binding force of the bar date first. The key strength of the bar date seems to be that the claim submitted after the bar date, conventionally called a late claim, cannot disrupt ‘to those client assets that have already been returned’⁹² and that ‘the person to whom the assets have been returned acquires good title to them as against the late-claiming claimant’.⁹³ Such a restriction over late claims is constructively critical, especially given the tracing role of trust law offered to trust beneficiaries.⁹⁴ This means that if there is a late claim the assets that have already been returned will not be affected and that the late claim holder is barred from relying on the tracing principle of trust law to demand a recovery unless bad faith is found.⁹⁵ This clarification could tremendously contribute to legal certainty.⁹⁶ However, missing the bar date does not mean that the late claim will be categorically barred or extinguished.⁹⁷ This is where leniency arises.

For a late claim, there are two generous options, which are controversial but are client-protection measures preferred by the government.⁹⁸ On the first option, if the client’s assets can be identified on the investment bank’s own books and records but no submission is made before the bar date, the special administrator should contact the client individually and urge the latter to make a claim;⁹⁹ in this scenario, the contacted client is given 14 business days to reply.¹⁰⁰ The logical interpretation of this arrangement is that, if the client manages to submit the claim within the 14-business-day grace period, the claim is treated as a duly submitted one, and that, by implication, if the client still remains silent after the 14-business-day period is exhausted, no submission will be acknowledged. Unfortunately, such a literal reading is wrong. The bar date mechanism is more friendly to late claims than reasonably interpreted. The client asset distribution rules say that in this situation the special administrator still needs to calculate the amount of assets and distribute them to those clients whose assets are adequately recorded but who do not submit written claims, both before and after the bar date, at all.¹⁰¹ Such tolerance may simply encourage non-compliance.

The second clemency option is that, if a late client asset claim is made not only after the bar date but also after any distributions, be initial, second or third, have been made, this claim can still be fully met if the exact assets asserted by this late claimant have not been disposed of;¹⁰² this late claim may also be partially met if the asserted assets have been partially gone.¹⁰³ In principle, if a late client asset claim is finally submitted and all client assets have been distributed, this claim could only be converted to an unsecured claim and join the investment bank’s unsecured creditors to share the house estate pro rata.

In light of these two leniency options, if a client asset claim is submitted after the bar date, it might be premature to say that the claim will be barred or extinguished;¹⁰⁴ at best where some client assets are still available, a late claim can be satisfied, if not in full, and at worst the late claim will be relegated from a proprietary claim against the client asset estate to an unsecured one against the investment bank’s house estate. Why are late client asset claims treated with such indulgence? Arguably, the investment bank special administration regime is designed by the HM Treasury where most experts, like their US peers,¹⁰⁵ might specialize in financial, rather than bankruptcy, law, who may not be confident enough to understand that, in the context of insolvency, property rights can be curbed or restricted if proper procedures are followed,¹⁰⁶ as was clarified in Re MF Global UK Ltd.¹⁰⁷ Arguably, these two leniency options may considerably weaken the role of the bar date in accelerating the return of client assets.

Sixth, if a bar date is installed in an investment bank special administration, the client asset distribution plan must be approved by both the creditors’ committee and the court.¹⁰⁸ Seeking approval from the creditors’ committee may be easy, since usually there are only three–five committee members,¹⁰⁹ but seeking court approval can be lengthy and expensive, since court hearings will be involved. In contrast, where a bar date is not used, it seems that the investment bank special administration regime is silent over whether the distribution plan should be approved either by the creditors’ committee or the court or both. Meanwhile, very unfortunately, if the bar date is used, the beginning of returning client assets under the approved distribution plan should be at least three months after the bar date;¹¹⁰ this seems to be a design made to delay the return of client assets instead of expediting it.¹¹¹

In the eyes of British regulators and practitioners, the bar date seems to be useful, and this is why this mechanism was not only retained but also was further expanded and strengthened when the Regulations 2011 was updated in 2017 following an extensive review by the HM Treasury¹¹² and by the FCA.¹¹³

The bar date expanded and strengthened in the investment bank (amendment of definition) and special administration (Amendment) Regulations 2017

For many, it seems to be a regrettable vacuum for the bar date not to cover client money in the initial Regulations 2011.¹¹⁴ Meanwhile, the judicial uncertainties on the boundaries of the client money trust had been settled when the government launched a full review of the investment bank special administration regime in 2013 as required by law.¹¹⁵ Time was ripe for the government to expand the bar date mechanism to cover client money claims and to streamline it for efficiency.¹¹⁶ It should be aware that although in the UK financial markets as a whole investment banks only hold client money of around £100 billion, in contrast with client assets of some £10 trillion,¹¹⁷ client money rules seem to be more complicated than those on client assets.

First, the bar date started to apply to client money claims after the Regulations 2017 was enacted.¹¹⁸ Given the focus of this article, the bar date over client money is not to be examined here.

Second, a hard bar date can be added, subject to the discretion of the special administrator, to close the client asset estate finally under the Regulations 2017.¹¹⁹ This seems to be a significant step forward for both efficiency and finality. More specifically, a hard bar date can only be used after a bar date has already been set, whatever on client asset¹²⁰ or money¹²¹ claims. The previous bar date could be labelled as a ‘soft’ bar date by contrast. Unlike placing a (soft) bar date, in setting a hard bar date on client asset claims, the special administrator needs advance court approval;¹²² essentially, when deciding whether the hard bar date can be sanctioned, the court pays attention to whether the clients, known and potential, are sufficiently notified.¹²³ It remains to be investigated as to how often the hard bar date is applied in practice. This is also one of the empirical tasks of this study.

Again, leniency can be exercised even where the hard bar date on client asset claims is missed. Some may ask whether this hard bar date is still hard or whether it is hard enough. Under rule 12B (7) of the Regulations 2017, where a client asset claim is submitted after the hard bar date, the special administrator must meet this late claim if the disposal of the residual assets has not been arranged or if the arrangement for the disposal of the residual assets does not prevent the special administrator from meeting this claim.

The real strength of the hard bar date over client assets is that after the hard bar date, except for meeting some eligible late asset claims as noted above, the special administrator can dispose of all remaining client assets and transfer the proceeds to the investment bank’s general estate, namely the client asset estate will be closed, finally.¹²⁴ More importantly, unless client asset returns are carried out in bad faith or by deceit,¹²⁵ any late client asset claimants are prohibited from using the tracing role of trust law to demand a recovery.¹²⁶ All late client asset claims will not be extinguished but could be converted into unsecured claims against the investment bank’s house estate.¹²⁷

Third, under the Regulations 2017, the special administrator is somewhat relieved from the constraints of returning client assets where the bar date is used. As examined before, if a bar date is used, under the Regulations 2011, the special administrator can only return client assets subject to the court-approved client asset distribution plan.¹²⁸ This means that the special administrator is generally prevented from distributing client assets before the distribution plan is confirmed by the court. This seems a little harsh and mechanical, given that some client assets may be totally free from disputes.

The amendment made in the Regulations 2017 allows the special administrator to return client assets without court approval if the beneficial ownership can be manifestly identified and if ‘the assets are not held by the investment bank in a client omnibus account’.¹²⁹ This means that if client assets are adequately placed in segregated client asset accounts and the beneficiaries are correctly recorded, the special administrator can return these client assets straightforwardly, without asking for court advance approval or relying on a creditors’ committee agreed distribution plan. This is a step forward. But, realistically, given that most large investment banks use client asset omnibus accounts for trading efficiency,¹³⁰ the expectation of the real effect of this relief measure should be kept low.

Overall, this is how the bar date, soft and hard, is designed in the UK investment bank special administration regime up to date. Four major characters can be summarized. First, the use of a bar date should get advance court approval at the early stages, and later this power is devolved to insolvency practitioners. Second, the effect of a client asset bar date, like all other similar mechanisms, is subject to the fulfilment of sufficient notification, which means that affected parties should be notified by insolvency practitioners with diligence and care of a high degree. Third, the client asset bar date is made with a considerable degree of leniency towards late claims, since lawmakers aim to develop a special investment bank insolvency procedure in the best interests of clients; however, such leniency may significantly undermine the effectiveness of the bar date in achieving both efficiency and finality. Fourth, a client asset hard bar date can be ultimately imposed, subject to court approval, beyond which unclaimed client assets can be liquidated, with the proceeds transferred to the firm’s general estate. The chief objective of a client asset bar date is undisputed: it is expected to serve as a major means to expedite client asset returns. This article is testing its effectiveness.

It is worth noting that, apart from the development of the bar date mechanism in the UK investment bank special administration regime, it continues to be used over creditors’ claims in schemes of arrangement¹³¹ and Company Voluntary Arrangements (CVAs)¹³² and over liquidation/administration expenses,¹³³ and has been further expanded to debt claims in the newly established insolvency procedure, the Restructuring Plan,¹³⁴ but this is not what this article focuses on.

3. Methods in collecting the data for this study

Empirical legal research is rewarding but also challenging, since most legal scholars, with the author included, are not specifically trained for this.¹³⁵ This is an area that generally falls into the expertise of sociologists. Inevitably, methodological tribalism can be found.¹³⁶ However, without data from the real world, legal research, as is warned by Professor Elizabeth Warren, is ‘from nowhere to nowhere’.¹³⁷ Empirical legal research is an evidence-based approach, testing the gap between the law on the books and the law in action and pointing to the direction of how the law can be reformed.¹³⁸ Because of this firm belief,¹³⁹ the study collects the data from the sources as follows.

First, to identify how many investment bank special administration procedures were entered between 8 February 2011, the day on which the Regulations 2011 took effect,¹⁴⁰ and 7 February 2021, the day marking the end of the first decade of the implementation of this new regime, visiting the Gazette website at ‘thegazette.co.uk’ is the first step, since all investment bank special administrator appointments are legally required be advertised at the Gazette,¹⁴¹ the UK official journal publishing statutory notices.¹⁴² To be precise, the website ‘thegazette.co.uk’ is a UK-wide public information platform, since it archives public notices advertised on the London Gazette (covering England and Wales), the Belfast Gazette (for Northern Ireland) and the Edinburgh Gazette (for Scotland).¹⁴³ In fact, besides special administrator appointments, many other statutory notices in an investment bank special administration, including the notice of the initial meeting of creditors and clients and of the bar date, should also be published in this journal.¹⁴⁴

Therefore, the Gazette can give the most reliable data in terms of how many investment banks have been placed in the special administration procedures and who these companies are.¹⁴⁵ Although the two phrases ‘special administration’ and ‘investment bank’ were inserted into the search engine respectively under the Insolvency Notices section at the Gazette, and although all returned notices were manually checked, a mistake was still made: the investment bank, Avalon Investment Services Limited, which entered into the special administration in 2016, was not found; however, thanks to a report by the FCA made in 2016, which discusses this firm’s entry to the special administration,¹⁴⁶ this case was finally included.

Since there were only 19 investment bank special administrations in total identified, this study has the luxury of not sampling representative cases, as a result of which the census data can be examined.¹⁴⁷ Some may say that the number of 19 might be too small to be worthy of analysis, especially compared with the annual UK corporate insolvency number of around 27,000.¹⁴⁸ However, in the context of financial institution insolvencies, this number is considerably substantial, since during the same period the UK only handled three deposit-taking bank resolutions.¹⁴⁹

Second, after identifying these 19 investment bank special administrations, more detailed materials can be found at Companies House, a UK government agency for company registrations, whose website is https://www.gov.uk/government/organisations/companies-house. At the Companies House website’s ‘filing history’ section, the most valuable files, such as the special administrator’s proposals and progress reports, can be found and freely downloaded.¹⁵⁰ Companies House is the major data source for this study. But the shortcoming of Companies House files is that some documents might be very skeletal, as they only comprise of legally required content. The next resource is to further enrich data collection.

Third, it is the websites of the special administrators’ firms, where most special administrators use their own firms’ websites to publish case files to feed transparency. This is especially helpful if the special administrator’s firm is a large and prominent one. Several transparency best performers are KPMG (handling four cases), Smith & Williamson LLP (four cases), and pwc (three cases). Some documents that are not legally required to submit to Companies House are made publicly available at these websites. For example, in the special administration of MF Global UK Limited, KPMG, the joint special administrators’ firm, even placed the special administrators’ court hearing report on the website not only for clients to review but also for the public to access.¹⁵¹ This study takes advantage of the excellent disclosure by special administrators.

In addition, some newspaper reports are also collected. Along with some mainstream business-focused newspapers, like Financial Times, one boutique newspaper CityWire at citywire.com is particularly resourceful, as it not only reports the major news of the financial markets but also allows investors to post many insightful comments over failed investment banks placed in special administration procedures in the UK. Backed by these sources, this study confidently reports the findings.

4. The application of the bar date over client assets in the UK investment bank special administrations

The Regulations 2011 came into force on 8 February 2011, and months later, the Rules 2011 took effect on 30 June 2011.¹⁵² Technically speaking, both are procedural in nature,¹⁵³ since substantial issues remain subject to insolvency, property and trust laws, and in particular subject to the Client Asset Sourcebook (CASS) issued by the Financial Conduct Authority. After the legal framework had been established, the British courts were ready to accept investment bank special administration filings from 2011.

Overall implementation of the investment bank special administration regime between 2011 and 2021 in the UK

The first company applying for the special administration at the London High Court was MF Global UK Limited, a securities brokerage firm and a member of the London Stock Exchange, on 31 October 2011, with the three insolvency practitioners from the accounting firm KPMG jointly appointed as the special administrators.¹⁵⁴  Table 1 lists all 19 companies that entered into investment bank special administration between 2011 and 2021.¹⁵⁵ Of these 19 firms, 15 (79 per cent) were headquartered in London, suggesting the concentration of investment banking services at London as a world financial centre. No investment bank special administration took place in Scotland and Northern Ireland, justifying this study’s focus on England and Wales. Several major trends can be summarized.

Most investment banks in special administration are brokerage firms

The Banking Act 2009 defines the term investment bank very broadly,¹⁵⁶ which means that the investment bank special administration regime can cover a wide range of financial firms. Reportedly, in the UK, there are some 1,500 financial firms which could be subject to the special administration if they are in trouble.¹⁵⁷ In practice, most special administration users are brokers. As shown in Table 1, out of 19 firms in special administration are there 14 brokerage firms (74 per cent). Of course, all these firms entered into the special administration because they held either client money or client assets or both. In contrast, it is evident that an investment bank that does not hold client assets could alternatively apply for the ordinary administration procedure. For example, Seymour Pierce Limited was a broker and was the London Stock Exchange member, entering an ordinary administration procedure in 2013, rather than a special administration, as the firm mainly provided financial advice to its clients and did not hold client assets (including client money).¹⁵⁸

Intense regulatory actions before and when special administration was filed

Since all these firms are regulated by the Financial Conduct Authority (formerly the Financial Services Authority),¹⁵⁹ the regulator was heavily involved in the entry of these special administrations. The regulator’s involvement was conducted at two stages. At the first one, shortly before the special administration was filed, most firms were given regulatory notices/orders either restricting, suspending or terminating their regulated business activities. Table 2 shows that 14 out of all 19 investment banks (74 per cent) were imposed with regulatory actions in advance of special administration.¹⁶⁰

For example, on 9 January 2012, given the inadequate capital possessed by the investment bank, Pritchard Stockbrokers Ltd, the then Financial Services Authority issued a warning requiring the company to correct this; but unfortunately, the firm was unable to do so, and a month later, on 10 February 2012, the Financial Services Authority handed out a formal supervisory notice ‘preventing the company from carrying out its regulated activities except to close out transactions which had already been commenced’; following the regulator’s action, the London Stock Exchange suspended this firm’s membership on 16 February 2012, which essentially left the firm with no option but to seek a formal insolvency procedure.¹⁶¹

The diligence of the UK Financial Conduct Authority is highly recommendable, since its regulatory action played a key role in safeguarding client assets. For example, on 20 March 2017, because of a sudden departure of some key computer engineers who happened to be senior managers, the electronic trading platform of the investment bank Strand Capital Limited could not be operated as normal, and at the request of the newly appointed director, the Financial Conduct Authority intervened almost immediately, preventing the company from dealing with client money and assets and imposing that any payment in excess of £5,000 by the company was prohibited unless prior written consent of the Financial Conduct Authority was obtained.¹⁶²

For investment banks on which the regulator did not impose any regulatory action prior to special administration, usually their entry to special administration was because the distress of either the parent company or a sister company within the same company group brought the company’s operation to an abrupt end. For example, MF Global filed for special administration on 31 October 2011, mainly because its parent company, MF Global Holdings, was placed into a bankruptcy reorganization procedure, Chapter 11, in the USA on 31 October 2011.¹⁶³ The company itself did not breach any regulatory rules in the UK.

At the second stage, in most cases, the investment bank applied for special administration essentially under the direction of the Financial Conduct Authority, which was in line with rule 8 of the Regulations 2011,¹⁶⁴ and in the two cases of Beaufort and AFX, as shown in Table 1, it was the Financial Conduct Authority which directly acted as the applicant.¹⁶⁵ It is interesting to find that before the year 2014, written consent by the Financial Conduct Authority was routinely obtained before the court appointed the special administrator¹⁶⁶ and that after 2014 a subtle change was made: the FCA issued a letter indicating that it had no objection to the appointment of the special administrator,¹⁶⁷ which in substance was an implied consent before the special administrator appointment could go ahead.¹⁶⁸

It is also interesting to find that occasionally the Financial Conduct Authority directed a troubled investment bank to enter into an ordinary administration procedure rather than the investment bank special administration one, in spite of the existence of client money and assets held by the company. This happened in the administration of Greyfriars Asset Management LLP in 2018, where a pre-pack administration was achieved and the transfer of client assets and money to a new custodian was completed very quickly; this might be why the Financial Conduct Authority sanctioned the alternative insolvency procedure.¹⁶⁹

The investment bank special administration objective 1 has never been prioritized and is almost a dead letter

By law, as noted before, the investment bank special administration objective 1 of expediting client asset returns has two chances to be prioritized. One is that the special administrator can exercise their discretion, and the second is that the Financial Conduct Authority may direct the special administrator to do so. However, as listed in Table 3, in none of these 19 cases was objective 1 singled out for priority.

Interestingly, two ironic phenomena can be found. First, between 2011 and 2015, in the early years, it seems that most special administrators favoured the word ‘focus’ when addressing how the three objectives were simultaneously and equally prioritized: they ‘focused on all three’ objectives. After the year 2015, the preferred word is ‘pursue’: they pursued ‘all three’. The linguistic choices/preferences did not make any difference.

The second phenomenon is that in recent years it seems that the special administrator appeared to be ambivalent as to whether objective 1 should be given a priority. Is this a step forward? This is reflected in the latest three special administrations of Beaufort (in 2018), SVS (in 2019) and AFX (in 2019). In these three cases, the hesitation of the special administrators is palpable: when asserting that objective 1 was intended to be a priority, they immediately prevaricated by adding that objectives 2 and 3 were ‘concurrently’ pursued; if the three objectives are ‘concurrently’ pursued, it seems to be paradoxical to say that one of them is prioritized.

As for a regulatory direction from the Financial Conduct Authority, no direction was found in these 19 cases, although the Financial Conduct Authority sometimes boasted that it did have the power to do so.¹⁷⁰ It is true that the Financial Conduct Authority is given this power, but this power has to be accompanied with the condition, under which either the stability of the UK financial systems or public confidence to the UK financial markets is at risk.¹⁷¹ More trickily, this condition would be vague enough to befuddle regulatory staff. This study identified three cases where the special administrators explicitly stated that ‘in the absence of a direction from the FCA under Regulation 16, the Special Administration Objectives continue to be addressed in parallel’.¹⁷² This statement is generally fine but is not technically correct. As for whether to prioritize objective 1 or another one, the special administrator does not have to get a regulatory direction from the Financial Conduct Authority, since they can do it at their own discretion.

The application of the bar date on client assets in UK investment bank special administrations

The bar date seems to be a necessity, as the first investment bank special administration over MF Global immediately sees its application.¹⁷³ As noted before, after 2017, a hard bar date can be established to warn clients of the serious consequences in the absence of a timely response. Table 4 offers a glimpse of how the bar dates over various claims and of different categories are applied in these 19 special administrations.

The surprisingly low application rate of the bar date over client asset claims

Given the purported roles in accelerating client asset returns, the use of the bar date is assumed to be a routine exercise in every investment bank’s special administration. But it is not the case. As shown in Table 4 below, the bar date over client asset claims was only applied in eight out of all 19 cases (42 per cent). This means that, in over half of the investment bank special administrations, it was not deployed at all. This may considerably disappoint the proponents who assert the usefulness of this legal mechanism.

Admittedly, in six out of all 19 special administrations, the investment bank held client money only, without taking custody of client assets. If these six cases¹⁷⁴ are excluded, the client asset bar date application rate only climbs to a little over half at 62 per cent (8 out of 13). The next immediate puzzle is why it was not invoked in so many special administrations, given its asserted usefulness. By delving into all five remaining special administrations where there were client assets but the bar date was not installed, it is found that the quick, direct bulk client asset transfer took place in four of them (80 per cent). A direct bulk client asset transfer means that client assets are to be transferred to an alternative investment firm without going through a client asset claim process, which directly relies on the firm’s client asset records to calculate each client’s asset entitlement. To a large extent, it seems that the bar date is antithetical to a direct client asset bulk transfer. This assertion appears to be more convincing if an exceptionally atypical case, the special administration of WorldSpreads, can be statistically excluded. Disregarding the WorldSpreads case is mainly because of the meagre number of client asset claims in this case.

In WorldSpreads, there were only 11 clients having client asset claims worth £295,000.¹⁷⁵ The number of client asset claims in this case is a rare occurrence, because it is quite normal to see around 10,000 clients trapped in one investment bank special administration. Even in the special administration of WorldSpreads itself, there were at least 5,000 clients having client money claims (WorldSpreads had some 15,000 clients on record, but only around 5,000 of them were active).¹⁷⁶ Therefore, the WorldSpreads case is statistically negligible to counting the complexity of client asset returns. If the WorldSpreads case is disregarded, the real-world data suggest that the bar date over client assets seems to be 100 per cent (four out of four) unnecessary where a direct bulk client asset transfer takes place.

A direct bulk client asset transfer is perhaps the most desirable way of returning client assets, because it is exceptionally efficient in conducting client asset returns.¹⁷⁷ To measure the efficiency of client asset returns, this study calculates how long from the start of a special administration procedure it takes for clients to access their asset accounts again, and the methodology in counting this duration should be explained first. It is a tough job to identify an exact date when the client asset return takes place, since in many cases client assets are returned or transferred bundle after bundle spreading over several months and even over years. For simplicity and bias minimization, this study locates the date when the first bundle/tranche of client assets is returned or transferred as the overall client asset return date. Relying on this method, within these four special administrations resorting to a direct bulk client asset transfer and in the absence of a client asset bar date, as shown in Figure 1, on average, it only takes 19 days for clients to access their assets or accounts again. This looks very positive.

Now it seems clearer why the bar date over client asset claims was not used in these cases. In these four direct bulk client asset transfers, if the bar date over client asset claims was used, the benefit of using direct bulk client asset transfers would have been totally ruined, since the average duration of 19 days between the commencement of the special administration and the occurrence of the bulk client asset transfer means that there is no room for the imposition of a bar date. If a bar date is proposed, client asset return can, by law, only be delivered at least three months after the bar date.¹⁷⁸ Namely, for a speedy client asset return, the bar date should be shunned, rather than be sought after.

Where there is a bar date, there is a disaster for client asset returns

To test how effective the use of the bar date is in accelerating client asset returns, this study uses the same method of identifying the duration between the entry of the special administration and the return/transfer of the first bundle, if applicable, of client assets, as explained before. The bar date was used in eight cases in which client assets needed to be returned. It should also be clarified that the special administration of Hartmann is excluded here, since paradoxically the majority of the client assets had been bulk transferred to a new broker long before the bar date was set; namely, the imposition of the bar date is largely irrelevant to the mission of accelerating client asset returns in this case.¹⁷⁹ As for the remaining seven cases, Figure 2 demonstrates the number of days spanning between these two key dates in each case.

For client asset claimants, the longest wait of 730 days happened in the two cases of City Equities¹⁸⁰ and Strand,¹⁸¹ that is, the clients waited for exactly two years after the commencement of the special administration to see the first tranche of client assets returned. The methodology of counting this duration has been explained before, but it is worth emphasizing that, after two years, only some clients saw the return of their assets, whereas many still waited after. The shortest wait is 193 days found in the special administration of Beaufort;¹⁸² even this most efficient client asset return took far longer than the average of 19 days in the four direct bulk client asset transfers without resorting to a bar date.

Of these seven special administrations using the bar date to ‘expedite’ client asset returns, on average, it took 476 days, around 16 months, to see the return of client assets. Between the figures of 476 and 19, the contrast reminds us that something was terribly wrong. Is the bar date a bane rather than a boon in facilitating a quick return of client assets? To a certain extent, the use of a bar date is partially responsible for the delay in returning client assets. Since if a bar date is used, first, the notification to all clients, known and unknown, takes time, and the Regulations 2011¹⁸³ rule 11(3) requires that when using a bar date, the special administrator should give a reasonable time for clients to calculate and submit asset claims. As for how long the reasonable time is, this is subject to the discretion of special administrators.

To probe how long on average a reasonable period is used by special administrators in setting a client asset bar date, this study collects both the special administration commencement date and the client asset bar date in each of these seven cases. As shown in Figure 3 below, the longest ‘reasonable’ time of 532 days was used in the special administration of Strand,¹⁸⁴ which means that clients were, unbelievably, given over 17 months to prepare and submit their client asset claims. This appears excessively generous. The shortest of the reasonable time of 99 days took place in the case of Beaufort.¹⁸⁵ Given that the Strand case appears to be considerably abnormal, with this case put aside, on average, in the remaining six cases, the clients were given 138 days to submit asset claims. Arguably, some clients may say that they are given sufficient time to make client asset claim submission preparation, but some may simply contest that this inadvertently causes delays instead.

For clients, the average wait of 138 days until the expiry of the bar date is not the end of the protraction. Under the Rules 2011,¹⁸⁶ where the bar date is used, client assets can only be distributed/returned at least three months after the bar date. Technically, if added with these three months (90 days), the simple calculation is that where the bar date is set, clients must wait at least for 228 days, around 7.6 months, to see the return of their assets. It is fair to say that the use of the bar date somewhat contributed to the delay in returning client assets, at least in the existing cases. Therefore, it can be comfortably inferred that where there is a bar date used, there is a disaster for clients seeking a quick return of client assets. The bar date looks to be a nuisance. Why was the bar date used in these cases?

For what purposes is the bar date set?

The orthodox answer is that the goal of using a bar date is to accelerate client asset returns, since the law says unequivocally that ‘if the administrator thinks it necessary in order to expedite the return of client assets, the administrator may set a bar date’.¹⁸⁷ The law on the books looks perfectly fine, but the law in practice gives a different, if not opposite, answer, because this purpose cannot be backed by the data.

In fact, the fundamental consideration of using the bar date is based on the assumption that when returning client assets, the company’s client asset records are not reliable, as a result of which each client asset claim must be verified one by one for accuracy.¹⁸⁸ Some special administrators do not hide this rationale. For instance, this is exactly what the special administrators in the Beaufort case justified their efforts in requesting client asset claim submissions and in assessing each of them individually on the grounds that the company’s books and records were not accurate; to make their conviction more persuasive, they provided a ‘concrete’ piece of evidence that ‘initial reconciliation work by the administrators had identified around 150 discrepancies on around 4,000 stock lines’.¹⁸⁹ The government is of the same view that a bar date can be used if ‘the records of the insolvent firm do not give a clear indication of what clients and third parties are owed’.¹⁹⁰ Is it really true?

By logic, before assessing the amount of client assets each client is entitled to, given the ‘unreliable’ records of the company, all clients must submit their client asset claims. If it is your right, prove it. In light of time constraints, definitely, a deadline called bar date can be legitimately justified, otherwise, as asserted before, confirming client asset claims will never be closed.¹⁹¹ This study questions the legitimacy of this fundamental assumption over the (un)reliability of the investment bank’s books and records by investigating whether it is true in the seven cases where there were client assets and where the client asset bar date was used.

In Table 5, the information generated from the documents, including progress reports and distribution plans, released by the special administrators in these seven cases, proves that the company’s books and records were generally, if not absolutely, accurate as for the proper match between the amounts of client assets held by the company and the amount of assets the clients were entitled to.¹⁹² It is categorically incorrect to accuse that the company’s books and records were not reliable.¹⁹³ More strikingly, because all assets belonging to the clients were there, it surprised no one that in these seven cases client assets were returned in full ultimately.

Critics may say that the company’s books and records are generally correct, but they are not 100 per cent precise, since there are still client asset discrepancies, as was identified in the Beaufort case. Discrepancies of client stock lines mean that the amount of client assets lying in the company’s client asset (either segregated or omnibus or both) accounts does not accurately match the amount of client assets claimed by the clients; to put it bluntly, discrepancies mean that there is likely to be a shortfall of client assets. The existence of any client asset shortfalls suggests that the clients could not get back their assets in full eventually. However, as clarified in Table 5, in all these seven cases, it was divulged that the clients got a return of their assets in their entirety finally, which overwhelmingly demonstrates that there were no substantial client asset shortfalls at all, thereby it is groundless to accuse the inaccuracy or unreliability of the company’s books and records on client assets. Granted, there were client asset shortfalls, but these shortfalls are either technical or negligible, and no substantial human-made shortfalls were identified.

For instance, in the Beaufort case, it is true that a discrepancy was found in 150 stock lines out of the total 4,000 ones held by the external custodians, but these discrepancies were due to technical causes and were easy to solve.¹⁹⁴ One common technical reason is an issue of time lags: A stock was sold by the investor and has been recorded on their own asset accounts, but the sale had not been completed at the stock exchange’s system before the broker entered into the special administration.¹⁹⁵ These discrepancies are routine and inevitable. Sometimes, the discrepancy can also go in the opposite way: there are a surplus of client assets because of unfinished client asset reconciliation, as is found in the Strand case.¹⁹⁶ As said before, these technical discrepancies are easy to be cleared. In the Beaufort case, following an investigation, it was found that client asset shortfalls only occurred in less than 30 stock lines;¹⁹⁷ given that Beaufort held some 4,400 stock lines on behalf of its clients, this means that only a tiny proportion (0.68 per cent) of client assets had problems, which generally was negligible.¹⁹⁸

It is also noteworthy that in the Beaufort case, of 17,600 clients holding assets, there were only 30 clients having an unsolved dispute with the company over the accuracy or legal status of their assets; by proportion, it is also a tiny minority (0.17 per cent) of clients raising substantial concerns.¹⁹⁹ The similar minor incident is also found in the case of SVS, where 250 (2.1 per cent) out of all 11,900 responding clients raised a dispute about their assets, but following a further investigation and consultation, the number of disputes was reduced to 90 (0.76 per cent).²⁰⁰ This is also the case in Reyker, in which 12 (0.15 per cent) out of the 7,832 responding clients had a dispute over the accuracy or legal status of their assets.²⁰¹ But for this tiny minority, the vast majority of the clients pay a hefty price by having all client asset accounts meticulously checked, which is too expensive. This is not justifiable and cannot be justified. Foreign experience suggests that when conducting a bulk client asset transfer a certain degree of record inaccuracies should be tolerated, otherwise quick and direct bulk transfers can never happen.²⁰² Mr Peter Bloxham, the top expert hired by the HM Treasury to review the investment bank special administration regime, also reminded this in 2013, but it seems that no one has listened.²⁰³ After all, the foundation of using a bar date does not exist in practice; it is time for a rethink.

In addition, based on the alleged unreliability of the company’s books and records, some special administrators advise that the imposition of the bar date is to ensure the safety of client asset returns, since it can prevent any potentially competing claims from disrupting the good title of distributed client assets; in a word, it is in the best interest of clients to have a bar date installed. This assumption is controversial, if not wrong, for two reasons. First, through checking the client asset sheets released in some special administrations,²⁰⁴ it seems very unlikely for competing claims to arise.²⁰⁵ On client asset sheets, usually, each client asset is recorded under the name of a specific client through a series of codes, like client identity, securities code, securities description, quantity (unit), and even whether the security is held by the company or by a third party (non-held assets).²⁰⁶ Arguably, unless fraud is committed,²⁰⁷ it is technically unlikely for client assets to be misrecorded. A chain of errors is needed to generate a wrong securities record.

Secondly, more importantly, this assumption of preventing competing claims cannot be supported by any evidence in these seven cases. It is frequently found that after investigation the special administrator reported that no competing claims were raised. In fact, in some cases, there were indeed some mistaken client asset claims, some of them frivolously submitted because they even could not tell the special administrator their securities codes, and some of them erroneously claiming an asset under asset title transfer (ATT) as a safe custody asset.²⁰⁸ Usually, these client asset claims would be quickly rejected.²⁰⁹ Ironically, the collected case materials suggest that what is witnessed in reality is quite the opposite: there are always some client assets unclaimed at the end of a special administration. Backed by the case materials, this study ventures to say that the existence of competing client asset claims is a remote possibility. Overclaimed? In theory, it is likely, but in practice, it has not been found. Hence, if returning client assets is based on this assumption, man-made delays are inevitable.

Apart from the existential justification of the bar date, as demonstrated in this section, the bar date has also proved to be an ineffective tool in urging clients to submit their claims in a timely manner. Before reporting this, attention should be paid to why a direct client asset bulk transfer, the most efficient way of returning client assets, was not attempted in the eight client-asset-bar-date cases.

Why was a direct client asset bulk transfer not attempted in all cases?

Unlike the four non-client-asset-bar-date special administrations which resorted to a direct bulk client asset transfer,²¹⁰ the eight client-asset-bar-date special administrations also, surprisingly, used the transfer tool ultimately achieving the mission of returning client assets, as elaborated in Table 6 above. Unbelievably, the findings of this article can clarify that the debate over whether the wording of objective 1 of special administration should be changed from ‘return’ to ‘return or transfer’ could be further updated, since in reality no client assets can either be electronically or physically returned to clients; return is a wrong word and has never been, and cannot be, practised in reality.²¹¹

Generally speaking, to handle client assets, special administrators only have two options: to transfer them to an alternative investment bank either as arranged by themselves or as instructed by clients, or to liquidate them in the event that client assets have not been claimed and are essentially abandoned by their beneficiary owners. A typical example is the special administration of City Equities, where after transferring the vast majority of client assets to alternative firms at the instruction of the clients individually, the special administrators liquidated/sold the remaining client assets belonging to 40 clients who had not made a claim, simply to cover the costs.²¹² A transfer is a transfer, and liquidating is essentially also a transfer: after being sold to the buyer, the client assets are in substance transferred to another custodian chosen by the buyer. Client assets by nature cannot be held by clients, since almost all trading platforms only accept investment banks as intermediaries.

Some may ask whether client assets in the form of paper certificates can be physically returned to clients. The answer is no. As witnessed in the special administration of Beaufort, 380 client securities were physical certificates, and the special administrators had to work with the alternative brokers for a transfer.²¹³ The paper certificates cannot be returned to clients/investors directly, for the same reason noted above: trading platforms, including stock exchanges, do not interact with investors directly, and investors have to trade through an intermediary.²¹⁴

Moving to the central question of this section as to why a direct client asset bulk transfer was not pursued in these eight cases. The reason is simple. It is because special administrators did not believe in the accuracy of the firm’s client asset records. The assumption is that the firm’s client asset records are messy and unreliable so that a client asset verification process should be used before client assets can be transferred. As evidenced in the previous section, this assumption is wrong. Also, very significantly, a direct client asset bulk transfer without a client asset claim process is not in the interest of special administrators, since prolonging the client asset transferring procedure will increase the fees charged by special administrators.²¹⁵

As shown in Table 6 above, in these eight cases, the client asset transfers were generally conducted in three ways. The first is piecemeal transfers by which client assets were moved to alternative firms at the choice of client by client; this is the most destructive and costly way of transferring client assets and was astonishingly seen in four (50 per cent) out of all eight cases.²¹⁶ Did the special administrators in these four cases attempt to find a single transferee in view of a bulk transfer? There is no evidence suggesting the existence of such an endeavour.

The second way seems to be less destructive, since multi-bulk transfers are used, which is found in the case of Reyker; in this case, since a solicited buyer finally withdrew its offer, the special administrators had to find five alternative firms to take over the client assets.²¹⁷ Given the difficulty in finding a single willing transferee, multi-bulk transfers themselves are understandable and are not the problem; the problem is the disruptive insertion of a separate client asset claim process; in the Reyker case, since there was ‘no material shortfall in client assets’, why did the special administrators establish a client asset claim process asking all clients to submit claims for ‘custody assets online’?²¹⁸ Admittedly, before the completion of client asset reconciliations, it might have been reasonable for the Reyker special administrators to cast doubt on the accuracy or reliability of the firm’s client asset records. But after having known the accuracy of the records following the reconciliations, the Reyker special administrators still insisted on using a wasteful and unnecessary client asset claim process;²¹⁹ this can be very controversial.

The third way is a single bulk transfer, which is supposed to be the least worst among these eight cases and was used in two cases, the special administrations of Hume and SVS. As for the case of Hume, it is not exaggeration to say that the single bulk transfer is actually a fake one, since all client assets were held by a nominee company XCAP Nominees Limited which was a wholly controlled subsidiary of the firm, and since what was really transferred was the shares of this nominee company, not the client assets, to the buyer; from the perspective of the clients, both before and after the ‘transfer’, their assets were continuously held by the same company called XCAP Nominees Limited;²²⁰ what affected all Hume clients was that they were forced to pay the transfer costs, £4,992.63 per client asset account, a huge figure.²²¹ In fact, the real contentious issue in the Hume case is not the fake transfer, instead, it is the inserted client asset claim process, which was also used in the SVS case.

The SVS special administrators successfully found a buyer who was willing to take over all client assets; but, like in many similar cases, although ‘the reconciliation was completed on 18 September 2019 and identified no material discrepancies between the company’s records and the custody assets and client money actually held by it’, a wasteful and unnecessary client asset claim process was again used.²²²

In a word, two points can be summarized from these eight cases. Point one is that in half of these cases, special administrators simply did not attempt to find a single transferee, since it was not in their financial interest to do so; point two is that, in remaining cases, whatever a single- and multi-bulk transfer, the benefit of bulk transfers was overwhelmingly undermined by the interruption of a client asset claim process, in spite of the fact that the firm’s client asset records were proven correct. Granted, if a client asset claim process is really warranted, the question is why this process could not be expedited by the use of the widely acclaimed bar date.

Why is the bar date made toothless?

The bar date, as is literally composed, is supposed to bar/extinguish any claims submitted after the fixed line of time. However, as analysed before, it is designed with a considerable degree of leniency. In practice, the intended leniency indeed results in some level of non-compliance. As shown in Table 7 above, in the seven special administrations where the bar date was installed, non-compliance can be widely found.

The highest client asset bar date compliance rate of 100 per cent is found in the Strand case, since all 76 clients were approached and submitted the asset claims on time; apparently, the small number of clients is perhaps the key reason for the maximum response rate here.²²³ The lowest compliance rate is seen in the SVS case, where only 11,900 (55.48 per cent) out of all 21,451 notified clients had made a timely client asset submission;²²⁴ this means that astonishingly almost half of the notified clients, knowingly or unknowingly, ignored the bar date call. On average, at the end of the bar date, only 76.73 per cent of notified clients make a timely client asset submission, suggesting a non-compliance rate of around one quarter.²²⁵ Why was the bar date not treated seriously? There are perhaps three contributing factors.

First, it is intrinsic leniency resulting in non-compliance. Two examples can help explain. One is the Reyker case, where the client asset bar date was set on 7 April 2020, before which only 7,832 (70.39 per cent) of the total of 11,126 clients had made a client asset claim; to dutifully meet the statutory requirement clarified in the Rules 2011,²²⁶ 2 months later, on 26 June 2022, the special administrators sent a second round of client asset claim invitations to known clients urging them to make a claim within 14 business days; the additional grace period ended on 21 July 2022, with 1,514 new submissions received, pushing the submission rate up to 84.00 per cent.²²⁷

On the face of it, it is to kill two birds with one stone: more clients were persuaded to make submissions, and the statutory rules were faithfully followed. However, to a large extent, a hidden cost may have been forgotten: the leniency given here is simply an unintended device undermining the deterrence force of the bar date itself. In principle, as addressed before, unless and until a late claim is made under excusable neglect, blindly allowing any late claims without asking questions is not something rationally recommendable.²²⁸ This is also against the latest principles enshrined in some court precedents.²²⁹

The second example is more self-contradictory than the first one. It is the Beaufort case, in which the client asset bar date was scheduled on 8 June 2018, with 9,210 client asset claims duly received at its expiry; however, after the bar date, the special administrators continued to make ‘considerable efforts to engage with clients who have failed to submit claims’; interestingly, it is unknown whether the special administrators in this case relied on rule 143 of the Rules 2011 in nudging late claimants, as was quoted in the Reyker case above; these ‘considerable efforts’ did bear fruits one month later; before 25 July 2018, 1,227 new client asset claims were received; to weaken the effect of the original bar date further, for unclaimed client assets, the special administrators would return them according to Beaufort’s records.²³⁰

The question here is that if client assets can be identified in the company’s records, why the clients are required to make client asset submissions in the first place. Simply replying on the company’s records to return client assets saves costs and makes client asset return quicker, which is also the initial intention of the UK government.²³¹ This study has already reported that the company’s books and records are generally reliable and correct,²³² hence, it is difficult to understand the usefulness of the bar date package as a whole.

Second, from the perspective of many small-claim clients, it may not be financially worthwhile to make a claim because of the expensive fees charged by special administrators. Presumably, for many small-claim clients, giving up is the best option, which results in low submission rates, as reported in Table 7 above. It is fair to say that although the average bar date compliance rate only amounts to some three quarters, the claims submitted before the bar date represent a larger proportion by value. For instance, in the Beaufort case, in response to the bar date calls, 9,210 client asset claims duly arrived, representing 65.79 per cent of clients by number but 92.30 per cent in value of client assets, which suggests that most unresponsive clients had smaller amounts of assets.²³³

The case documents released by the special administrators of MF Global can give a vivid account of how small claims these silent clients hold. In MF Global, after the bar date, there were only 234 claims received, with 162 clients not responding; so the special administrators contacted them again, and unfortunately, only one client holding assets worth £1,000 stepped forward, with the remaining silent clients remaining silent; however, by value, these 162 clients in total only had client assets worth £67,000, suggesting each client only had the average assets worth £413.58.²³⁴ They are small-claim clients and are reluctant to participate at all.

In fact, as for expensive fees charged by special administrators, the MF Global case is not very frightening, since the special administrators here charged the clients at the rate of 6.74 per cent of the value of client assets returned.²³⁵ Small claims in value mean small amounts of costs imposed. But in the majority of these seven cases, at least for small claim clients, the fees charged by special administrators may be high enough to deter them from coming forward. Table 8 above lists how much special administrators charged over client asset returns.

As shown in Table 8, only in the MF Global case, the special administrators charged over client asset returns by a percentage of the returned asset in value, but in the majority of the cases, including Hume, Strand, Beaufort, SVS, and Reyker, the special administrators charged a fixed amount of fees either per client or per client account. Given that some clients may have more than one asset account,²³⁶ they would bear more costs if the fees are levied per account. The amount of fixed fees can be significantly high. The highest fixed fee is seen in the SVS case, in which each client was required to pay £12,000 before getting their assets back,²³⁷ and the lowest fee per client at £2,250 is found in the Strand case.²³⁸

Admittedly, it appears too embarrassing to charge a client whose assets are only worth £100.00 a fixed fee of £10,000, as witnessed in the Beaufort case. Hence, in many cases, probably to lessen the guilt, the special administrators offer some fee concessions to alleviate the injustice caused. For instance, in the Beaufort case, the special administrators ‘generously’ designed a compromise package, by which for the clients whose assets are worth less than £10,000 the fee would be capped at the value of the client assets,²³⁹ which means, for example, if a client had £500 worth assets, the fixed fee would be capped at £500; in real terms, the clients having less than £10,000 worth assets would get nothing from cooperating with special administrators.

Some concession packages are less ‘generous’. In the case of Hume, a fixed charge of £4,992.63 per client asset account is waived only if ‘the client asset account has a £nil value’, which suggests that, for example, if a client had assets worth £100.00, a fixed fee of £4,992.63 should be paid before getting back the trapped assets.²⁴⁰ For these small-claim clients, why bother to submit a claim whatever before or after the bar date?

Although for most clients, if they have paid the fixed fees prior to the return of client assets, they could get compensation from the Financial Services Compensation Scheme (FSCS), many still are unwilling to come forward. To have a glimpse of how often small-claim clients may be reluctant to participate, this study incidentally identified three sheets of client asset claims in the Beaufort case. According to these files, Beaufort had 230 client asset accounts, 46 (20.00 per cent) of them having value of less than £1,000.00 and 24 (10.43 per cent) of them less than £100.00; another sheet on submitted client claims over alleged client assets comprises of 332 received claims, 27 (8.13 per cent) of them having value below £1,000.00 and 5 (1.51 per cent) of them below $100.00;²⁴¹ a simple mathematical calculation suggests that for clients having assets of less than £1,000.00 they are 59.33 per cent likely to give up, and that for those having assets of less than £100.00, the likelihood of walking away rises to 85.52 per cent.

For practical consideration, it seems that it is time to treat small-claim clients in a more sensible way, otherwise, the delays caused are to serve no one’s interests.

Third, the bar date non-compliance is also probably because many clients are from outside the UK, with logistics and language barriers hindering communication efforts of special administrators, given that the UK financial markets are considerably globalized.²⁴² In many cases, some foreign investors are trapped, which may pose a significant challenge to the notification campaigns of special administrators. For example, in the Beaufort case, some 1,740 (6 per cent) out of 29,000 clients are from abroad, implying difficulties in reaching all clients.²⁴³ In the SVS case, many clients are from as far as China.²⁴⁴ Although no clear evidence is found, this study ventures to speculate that it may be particularly challenging for foreign clients to comply with the bar date requirement.

After all, the bar date helps special administrators in managing the client asset estate to some extent, but its effectiveness is considerably weakened by its design and by the presence of many small-claim clients; and arguably the bar date in the UK investment bank special administration regime may not deserve its name. Responding to bar date calls is essential for protecting client interests, but if the bar date is finally missed, what are the eventual consequences?

How are unclaimed client assets treated long after the bar date?

To investigate the consequences of missing the bar date finally, this study delves into the progress reports uploaded to Companies House by the special administrators in these seven cases, and surprisingly finds that the ultimately unclaimed client assets are exceedingly small by proportion and by value.

The largest amount of client assets of £400,000.00 remaining unclaimed long after the bar date, as shown in Table 9, is witnessed in the Beaufort case, where 76 clients did not come forward eventually.²⁴⁵ The lowest is found in the SVS case, in which the unclaimed assets (including client money) were only valued at £34.12 and US$275.29 and were owned by 39 clients.²⁴⁶ Whatever the amounts of client assets, the striking point is that there were always some client assets left unclaimed, which defies the key legislative assumption that there would be insufficient assets to meet client asset claims when the investment bank special administration regime was made in 2011.

To recall what the law says, two key points are worth repeating. First, if a late client asset claim arises, whenever before or after the execution of the court-approved distribution plan, this claim should be fully met if the claimed assets are still there and intact.²⁴⁷ Second, if the assets requested by the late claimant are partially available, two solutions are prepared: the unmet balance of the late client asset claim is automatically treated as an unsecured claim against the company’s general estate, and the clients who have received the assets claimed by the late claimant have good title over the returned assets.²⁴⁸ These two points are essentially based on two perceptions.

Perception one is that there are more likely to be client asset shortfalls, with more client asset claims chasing fewer assets. But in reality, as was reported, this perception cannot be backed by the evidence collected in this study. Perception two is that, because of fewer client assets left, there would be two or more clients claiming ownership over the same assets, which is to justify why the tracing role of trust law should be disabled to maintain certainty.²⁴⁹ Again, this perception cannot find supporting evidence. Generally speaking, no competing client asset claims are found.

To some extent, these two perceptions are derived from one vital assumption: the management of client assets by the failed investment bank prior to the special administration procedure is chaotic, and the investment bank’s books and records on client assets are messy and are not trustworthy.²⁵⁰ This study has offered evidence proving this legislative assumption is misguided. The reality is quite the opposite: there are always, ironically, client assets remaining unclaimed at the end of special administrations, in spite of the meagre amounts of them.

Although the Regulations 2011 and the Rules 2011 do not anticipate the presence of unclaimed client assets long after the bar date, this gap is later filled by Regulations 2017, which authorizes special administrators to set a hard bar date so as to close the client asset estate by liquidating any remaining assets and transferring the proceeds to the company’s house estate.²⁵¹ Therefore, the treatment of unclaimed client assets depends on whether this happens before or after the Regulations 2017 took effect.

Before 2017, despite the legislative vacuum, special administrators improvised. Of these seven cases, four of them (MF Global, City Equities, Hume and Strand) were commenced before the Regulations 2017. In these four cases, the unclaimed client assets were treated in three unique ways.

The boldest and quickest measure is found in the special administration of City Equities, where 3 months after the expiry of the bar date, the special administrators straightforwardly liquidated/sold the unclaimed client assets to cover the costs, which essentially means that the proceeds were used to pay the special administrators themselves; to do this, the special administrators relied on rule 196 of the Rules 2011, which authorizes special administrators to draw remuneration from returning client assets, and for safety, the special administrators obtained an advance permission from the Financial Conduct Authority and claimed to seek ratification from the court afterwards.²⁵² Given that the proceeds were even not enough to cover the shared costs, the special administrators did not attempt to explain two further difficult questions: how to face a late client asset claim if it emerges at a later stage, and whether the proceeds should be transferred to the company’s house estate if there is a surplus after meeting the shared fees charged by the special administrators.

The second way is more cautious and is found in MF Global, where the special administrators treaded the uncharted waters more carefully. The bar date in this case was set on 29 February 2012, but the special administrators did not take immediate action on unclaimed client assets.²⁵³ Instead, they waited for almost two years; in January 2014, perhaps because of running out of patience or perhaps because of anticipating that the affected clients would never step forward, the special administrators liquidated/sold the unclaimed client assets for £67,000.00 to cover the costs, namely to pay the special administrators themselves as was done in the aforementioned City Equities case; also, like in the City Equities case, since there was no surplus after meeting the fixed fees charged by the special administrators, no proceeds were transferred to the company’s general estate; however, equally bold in a different sense, the special administrators stated that in the event that the late asset claimants emerged they would not ‘have a claim in the general estate’ with respect to their liquidated clients assets.²⁵⁴ This means that due to 2 years’ silence, the late client asset claims are extinguished, which is pragmatically reasonable but is controversial by law. Luckily, it seems that there were no legal challenges followed. If the MF Global special administrators were not pragmatic, the long waiting game might have been going forever, which is exactly found in the cases of Hume and Strand.

Both Strand and Hume adopted the long/perpetual waiting strategy. Since Hume is more representative, here its details are dissected for analysis. In this case, the bar date was scheduled on 30 June 2015, after which 114 client asset accounts were unclaimed;²⁵⁵ the special administrators seemed to be considerably tolerant and did not take any drastic action. Over one year later, in September 2016, progress was made since the number of unclaimed client asset accounts dropped to 101,²⁵⁶ but the special administrators might have run out of patience at this point, threatening that if no further responses were received they considered ‘making an application to the court to liquidate (unclaimed) client assets to meet the costs’.²⁵⁷ The threat, it appeared, did have an effect, leading to the arrival of 13 late client asset submissions during the following six months.²⁵⁸ It is unknown whether the court was really approached, but what is known is that there was no court order found.

From September 2018, more than three years after the bar date, probably turned away informally by the court, the special administrators updated the threatening tactic by warning that they were working with the solicitors to apply to the Financial Conduct Authority to deal with unclaimed client assets, and such a threat was repeated six times during the following three-year period between September 2018 and September 2021.²⁵⁹ The renewed threat somewhat worked, because the number of unclaimed client asset accounts fell to 56 at the end of September 2021.²⁶⁰ Interestingly, in spite of asserting to apply for approval from the Financial Conduct Authority to handle unclaimed assets, no action was found; presumably, there might have been a humiliating rejection from the regulator.

In March 2022, unbelievably, seven years after the bar date, still facing 55 unclaimed client asset accounts, the special administrators softened, realistically or reluctantly, their warning tone, emphasizing that they themselves were currently progressing this matter so as to reach a conclusion in this regard, without referring to any official orders sought from either the court or the regulator.²⁶¹ On 7 October 2022, the special administrators were still trapped in the unclaimed client asset quagmire, unable to bring the 55 unclaimed client asset accounts to an end;²⁶² it may never come to a conclusion. The similar story also happened in the Strand case, where the bar date was set on 31 October 2018, and the latest progress ending on 16 May 2022 states that the special administrators were still waiting for the responses from the claimants over two client asset accounts.²⁶³

What happened in these four cases exposes at least three flaws of the investment bank special administration regime. The first flaw has already been addressed earlier: in designing the client asset bar date mechanism, the legislative hypothesis is that there would always be client asset shortfalls after the bar date, but the reality cannot be aligned with this assumption. The second flaw is that under the original statutory provisions, if there are unclaimed client assets, the only option for special administrators is to wait, and to wait permanently; it should not be forgotten that one vital justification of resorting to the bar date is to bring finality, since the client asset management cannot go endlessly; therefore, in principle, giving special administrators only one option to wait is contradictory to what the bar date is created for in the first place.²⁶⁴

The third flaw is the legislative failure of not paying attention to small-claim clients. As reported in Table 8, the fees charged by special administrators allocated to each client asset claimant are disproportionally high if the assets demanded by an individual client are low in value. The tricky dilemma is that if the small-claim client gives up for fear of high fees or of an unworthwhile fight, whether special administrators can still charge fees on the abandoned assets. If the client asset is returned to the client, it seems to be legitimate for the special administrator to recoup the cost, but if the client, explicitly or implicitly, discards the assets, it can be legally controversial for the fees to go ahead. In the City Equities case, it was purported by the special administrators that permission from the Financial Conduct Authority had been obtained before liquidating unclaimed client assets to cover the fees, but it seems that the regulator was not given such a power to offer any approval like this; in fact, the only legally viable choice is to seek an ad hoc order from the court, which can give a direction as generally authorised under common law.²⁶⁵

In MF Global, it is probably wrong for the special administrators to sell the unclaimed client assets without getting any extra authorization from an authority; moreover, it is considerably audacious for them to state that the late client asset claimants are deprived of the right to make an unsecured claim against the company’s house estate late. Granted, from the point of view of efficiency, the special administrators in both City Equities and MF Global cases are pragmatic, since in the absence of what was improvised, they would share the same lengthy and endless waits as witnessed in the Hume and Strand cases. Probably aware of this challenge, the hard bar date was inserted into the investment bank special administration regime when the Regulations 2017 was enacted. The question is whether the hard bar date can effectively fill the gap.

Applying the hard bar date to close the client asset estate

Before the Regulations 2017 was made, it was in unchartered waters for special administrators to handle post-bar-date unclaimed client assets, as a result of which confusion should not be a surprise. But after the year 2017, a new tool of the hard bar date is available in the arsenal of special administrators and is supposed to empower them to manage client asset estates more efficiently.

There are three cases, Beaufort, SVS and Reyker, which were commenced after the Regulations 2017 came into force. However, the hard bar date was only imposed in the SVS case.²⁶⁶ In SVS, the company entered into the special administration on 5 August 2019, and the (soft) client asset bar date was set on 10 January 2020; as reported in Table 7 above, after the selected date, there were still some 9,551 (44.52 per cent) client asset accounts remaining unclaimed.²⁶⁷ The waiting game was played as predicted and tested the special administrators’ endurance. Over one year after the (soft) client asset bar date, the special administrators realized the significance of applying the brakes, submitting a client asset hard bar date application to the court on 15 March 2021, with the court approving the client asset hard bar date of 30 July 2021.²⁶⁸

Unfortunately, the SVS special administrators did not fully disclose how many clients were affected by the client asset hard bar date, since they only generally stated that they had exercised their powers under the hard bar date provisions (covering both client assets and money) and had transferred unclaimed client money and the proceeds of liquidated/sold client assets (totalling £34.12 and US$275.29) belonging to 39 clients to the company’s house estate;²⁶⁹ given that in an early progress report the special administrators mentioned that there were 38 non-engaging clients holding client money claims of ‘£0.19 and US$275.29’,²⁷⁰ presumably after the client asset hard bar date the special administrators only sold one client’s assets for £33.93. How could this action be economically justified?

It is also puzzling to read the statement of the SVS special administrators over the proceeds of £33.93 from selling the unclaimed client assets transferred to the company’s house estate, since in this case each client was required to pay the shared cost of £12,000 prior to seeing the return of their assets.²⁷¹ The sheer value of this single client’s assets was too small to cover the shared special administrators’ fees, thereby there should be nothing transferred to the house estate in the same way that the special administrators did in the cases of City Equities and MF Global. The likely explanation is that the special administrators exercised their ‘absolute’ discretion by waiving the shared fees under the approved distribution plan.²⁷² In spite of the meagre value of £33.93 involved in this case, it raises two significant questions.

First, under the approved distribution plan of SVS, the special administrators were indeed given the discretion to charge individual clients or not over the return of their client assets.²⁷³ But, since the overall fees charged by the special administrators are determined by their working/charging hours in this case,²⁷⁴ this means that the special administrators simply avoided being embarrassed for charging disproportional fees on small-claim but unresponsive clients at the expense of cooperative/engaging ones. The diligent clients are arguably punished for their engagement, which is not something that could be encouraged. Second, although this single client was entitled to make an unsecured claim against the company’s house estate under the distribution plan,²⁷⁵ the transferred proceeds of liquidating the unclaimed client assets still constitute a kind of subsidy for the benefits of the creditors in this case. Since most investment banks in special administration are insolvent, this subsidy is more likely to become a reality.

Moreover, rather disappointingly, the client asset hard bar date in the SVS case is also not hard enough to close the client asset estate, since after the end of the hard bar date on 30 July 2021, the special administrators surprisingly continued to ‘deal with the custody assets and client money belonging to the remaining 2 clients who have failed to meaningfully engage with’ them.²⁷⁶ This might be a step too far in indulging late client asset claimants. In the case of Reyker, the hard bar date was not applied, but this case does remind the necessity of reforming the hard bar date mechanism itself.

In Reyker, at the expiry of the client asset (soft) bar date on 7 April 2020, of 11,126 client asset accounts, 7,832 (70 per cent) client asset claims were received, and after a new round of notification efforts, up until 7 October 2020, the client asset submission rate rose to 84 per cent (some 9,345 clients).²⁷⁷ The special administrators did not disclose whether more late client asset claims arrived long after the (soft) bar date. But the subsequent progress reports prepared by the special administrators suggest that they had no difficulty in locating client assets and identifying clients.

One apparent reason is that the client assets as a whole were accurately listed on the company’s 1,089-page client asset sheet.²⁷⁸ It is fascinating to scan this 1,089-page file, since it painstakingly lists all the details of each client asset account, including client identity codes, client securities codes, the amounts of securities held for each client, and even whether the assets are in the physical or electronic form; more strikingly, under one column on whether there is a shortfall, the number ‘0’ is used throughout all the 1,089 pages, suggesting the accuracy of the company’s client asset recording.²⁷⁹ In light of this client asset sheet faithfully recording the assets belonging to 11,138 clients,²⁸⁰ the common sense question is why the clients were asked to submit asset claims, which is factually unnecessary.

In this case, by calculation, there should be some 1,781 clients who did not submit a client asset claim before 7 October 2020, and as stated before, since the special administrators did not announce any troubles in identifying all client asset claimants, presumably they relied on the company’s records in returning client assets by transferring them to alternative brokers.²⁸¹ The Reyker special administrators perhaps deliberately avoided stating the reliability of the company’s client asset records. Of course, this is not the key point in the Reyker case. The real question raised in the Reyker case is the conditions of applying for a client asset hard bar date.

Whatever the late client asset claims submitted by the clients themselves long after the (soft) bar date or straightforwardly acknowledged by the special administrators through the company’s records, the Reyker special administrators implicitly assumed that all client asset claims had been received, but the considerable delays in returning client assets still happened. According to the latest progress report released on 5 May 2022, more than two years after the (soft) client asset bar date, there were still 1 per cent of client assets stuck in the hands of the special administrators, and these client assets were unable to be returned not because the affected clients did not make a submission, but because they either had not proposed a third broker to accept their assets or had not cooperated with the special administrators to pay shared costs.²⁸² In one word, according to the Reyker special administrators, it was the inaction of clients leading to the delay of returning client assets.

Given the protracted delays of returning some client assets for over two years, did Reyker consider applying for a client asset hard bar date to finish the job? Although the special administrators did not manifestly state whether such a proposal was deliberated, by law, the special administrators were unable to do this. It was also likely for the Reyker special administrators to approach both the Financial Conduct Authority and the court behind closed doors, but the proposal for setting up a hard bar date was quietly declined on the basis that there is no statutory mandate to do so.

The key trouble is that in this situation the condition of using a client asset hard bar date to end the client asset estate was not met. Under the Regulations 2017, the condition of installing a client asset hard bar date is that some clients fail to make client asset claims, not that they do not cooperate after client asset claim submissions by way of not nominating a third broker to accept client asset transfers or not paying the shared costs.²⁸³ In both ways, the client’s indifference causes avoidable delays, but the Regulations 2017 did not anticipate a wider range of apathetic client behaviours leading to protracted client asset returns, and time is ripe to think of how to fill this new gap.

In the third case, Beaufort, the (soft) client asset bar date was pinpointed on 08 June 2018, and as listed in Table 7, there were only 65.70 per cent of the clients who had submitted the client asset claims on time. As noted, facing the unresponsive clients, the Beaufort special administrators mainly used the company’s client asset records in returning the assets.²⁸⁴ However, over one year later, in August 2019, there were still 76 client securities whose beneficiary owners (clients) stubbornly stayed silent; the special administrators asserted that these clients had ‘remained unresponsive to request for the return of a claimant options form’, but did not clarify whether these clients had voluntarily submitted a client asset claim or not in the first place.²⁸⁵ Ingeniously, probably in anticipation of an uphill legal battle to get a client asset hard bar date approved by the court, the special administrators made an amendment of the court-approved distribution plan, which authorized themselves to liquidate the unclaimed client assets to cover the costs.²⁸⁶

The 76 unclaimed client securities were sold for £400,000.00.²⁸⁷ In Beaufort, as reported in Table 8, the special administrators charged a fixed fee of £10,000 per client account,²⁸⁸ but the special administrators did not say how many client accounts these 76 unclaimed client securities were from. Given that most unresponsive clients are small-claim holders, it can be safely speculated that there were likely to be around 76 client accounts. On average, the value from realizing the assets from each client account only amounted to some £5,263.16, a sum not enough to cover the allocated cost of £10,000.00. No surplus can be transferred to the company’s house estate.

The Beaufort special administrators did not choose the hard bar date route to close the client asset estate but achieved the same result. They innovatively got the mandate from the distribution plan which was first confirmed by the creditor committee and later approved by the court; it is a shortcut, effectively circumventing a separate and costly court approval procedure if a client asset hard bar date is used.

Overall, the client asset hard bar date mechanism seems to be largely useless. Moreover, it unexpectedly raises more questions. First, if the client asset hard bar date is applied, it does not appear to be justifiable for the proceeds from realizing/liquidating the residual client asset estate to be transferred to the company’s house estate; it constitutes a wealth relocation from the pockets of clients to those of creditors, as were shown in the SVS case. Ideally, such proceeds should go to the Insolvency Service’s unclaimed asset pools instead. Second, as was evidenced in SVS, the availability of the client asset hard bar date is not treated seriously by insolvency practitioners, albeit only in this individual case.

Third, when the Regulations 2017, and in particular its rule 12(B), was made, to use the client asset hard bar date, the legislative assumption is that the client asset hard bar date only binds the client who has failed to make a client asset submission, but not the client who has submitted, or is assumed to have submitted, a claim but refuses to collaborate afterwards in various ways, which also causes unnecessary client asset return delays or deadlocks. A wider range of causes should be considered before deciding whether a client asset hard bar date can be justified.

However, even though the law is updated as recommended here, the general fragility, it seems, of the client asset hard bar date cannot be substantially changed. Granted, if there is a choice between two evils, the pragmatic approach used in the Reyker case might be more recommendable, because at least it ultimately closes the client asset estate and is conducted in a more cost-effective way. The harsh reality is that both the (soft) bar date and the hard bar date on client assets seem weak in nature and are weakly implemented.

5. Conclusions

During the first 10-year implementation of the UK investment bank special administration regime, the use of the bar date in seeking a speedy return of client assets in investment bank special administrations was largely a failure. The bar date mechanism itself was designed with structural flaws, and the application of the bar date in real cases simply made things from bad to worse, since the bar date was largely powerless in forcing clients to submit client asset claims in a timely manner. The UK lawmakers’ hesitation in imposing a strong, deterrent client asset bar date is mainly because of the concern of not depriving investors’ proprietary rights lightly.²⁸⁹ But in the context of insolvency law, this understanding is problematic and is perhaps wrong.²⁹⁰ With due process, even property rights can be curbed in insolvency, and UK lawmakers could be advised to be more confident in establishing decisive measures so as to strike a balance between property rights and efficiency.²⁹¹

More fundamentally, the findings of this study, astonishingly, reveal that the foundation of using the bar date does not exist at all, since the legislative assumption that the investment bank client asset records are not reliable cannot be backed by the real-world data collected by this project.²⁹²

For a speedy return of client assets, the only feasible and realistic option is to facilitate a transfer of client assets to a third healthy investment bank.²⁹³ And this study supports Peter Bloxham’s argument that the term ‘client asset return’ should be updated to ‘client asset transfer’;²⁹⁴ return is largely a wrong word.²⁹⁵ This is because, in practice, it is found that an individual investment bank special administration may be conducted in different ways, but for client assets the overall destination is to be transferred to a third party either through a bulk transfer, which is arranged by the special administrator or by the investment bank itself during or shortly before special administration, or through a series of piecemeal transfers at the request of client by client, with the former cheap and efficient and the latter expensive and time-consuming.²⁹⁶

This study calls for legislative reform. When an investment bank fails, client assets should be transferred, as a default arrangement, to a third investment bank in accordance with the investment bank’s client asset books and records, unless fraud is committed by the firm.²⁹⁷ Further legislative reforms should focus on how to facilitate a transfer, since there is still more work to be done on transfers in this jurisdiction.²⁹⁸

An early version of this article was presented at INSOL International Academic Colloquium held in London in June 2022, and the author wishes to thank the conference participants for helpful comments; this article has been significantly improved following the constructive feedback given by an anonymous reviewer, which is greatly appreciated. All errors remain the author’s responsibility.

Footnotes