Burford Capital Ltd v London Stock Exchange Group Plc

The claimant (“Burford”), a company incorporated in Guernsey, is well known as a litigation funder. Its shares are listed on the Alternative Investment Market (“AIM”) maintained and operated by London Stock Exchange plc (“LSE”), a wholly owned subsidiary of the defendant (“the Stock Exchange”).

From about 2.00 pm on 6 August 2019 to the end of trading the following day, there was a sustained run on Burford’s shares that saw the share price fall very substantially. All times in this judgment are London time in August 2019, i.e. British Summer Time (GMT + 1).

Burford’s ordinary share price closed on 5 August 2019 and opened on 6 August 2019 at £13.81, valuing the entire issued share capital at c.£3 billion. On 6 August, it moved within a narrow range through the morning and into lunchtime, and at 1.30 pm stood at £13.92. Between 1.30 pm and a little after 2.00 pm the price eased down fractionally, then it dropped sharply, dipping below £11.00 a little after 3.30 pm before rallying somewhat (but in volatile fashion) to close that day at £11.21, a fall of 18.8% on the day.

The share price opened on 7 August 2019 at £11.30 and in early trading surged briefly to well above £12.00 before dropping back to, and hovering at, c.£12.00, where it stood at 8.53 am. After that, the price fell dramatically, touching as low as £4.00 by c.11.30 am, from where it rallied to £6.00 by c.12 noon. It then held relatively stable until after 12.30 pm before a steady, sustained fall back to £4.00 sometime after 3.00 pm, after which it rallied again to £6.00 or so as the close of trading approached.

The closing price on 7 August 2019 was £6.05, a fall of 46.5% on the day, leaving the share price 56.5% lower than it had been at 1.30 pm the previous day. Over the following week or so, the price strengthened back to approaching £9.00 after which it traded for the rest of 2019 generally in a range between c.£7.00 and c.£9.50. It fell steadily but substantially over Q1 2020 as the world suffered under Covid-19, and by the time of the hearing on 1 to 3 April 2020 it was at or below £4.00. (As I finalised this judgment on 14 May 2020 for handing down on 15 May 2020, the share price closed the day at c.£4.25.)

The share price movement for 6 and 7 August 2019 is shown by the following graphs produced by Prof Mitts, to whom I refer at some length below:

That 6 and 7 August 2019 were unusually busy days for market activity in relation to Burford’s shares may be shown by a number of measures. One of the simplest in the evidence before me is the daily number of mid-point price updates. Such an update will occur when the best bid or best offer changes, or both change, in such a way that the mid-point between current best bid and best offer moves. On 1, 2 and 5 August 2019, the three preceding trading days, there were, respectively, 1,961, 1,808 and 1,933 mid-point price updates. On 6 August, there were 8,544; on 7 August, 52,081.

The big news affecting trading in Burford’s shares on 6 and 7 August 2019 was that they were the subject of a short-selling attack (a ‘bear raid’) by Muddy Waters, a US investment and investment advisory business. The trial materials make reference to two different entities, Muddy Waters Capital LLC and Muddy Waters LLC, but for my purposes it is not necessary to distinguish between the two or consider which (if not both) would have any responsibility in law for what happened in relation to Burford’s shares.

At 1.30 pm on 6 August 2019, Muddy Waters tweeted as follows (“Tweet 1”):

“Muddy Waters is now in a blackout period until tomorrow 8 am London time [the AIM opening time] when we will announce a new short position on an accounting fiasco that’s potentially insolvent and possibly facing a liquidity crunch. Investors are bulled up about this company. We’re not.”

There was reason to suppose that the subject of Tweet 1, and thus the target of Muddy Waters’ announcement, might well be Burford, as in fact it was. Burford itself assumed that was so, and at 7.00 am on 7 August 2019 released via the LSE’s Regulatory News Service as RNS No.20851 a statement about what had happened the previous day. The statement lauded the company’s financial performance, health and transparency, and blamed the share price fall the previous afternoon on “a rumour of a potential “short attack” or “bear raid”, … where short sellers take on a short position in a company’s stock and then engage in claims about the company in an effort to alarm investors”. It went on to state that Burford was already “strongly suspicious” that actionable misconduct by way of market manipulation was responsible for the fall in the share price.

That Burford was indeed Muddy Waters’ target was confirmed definitively at 8.53 am on 7 August 2019 by a further Muddy Waters tweet (“Tweet 2”):

“Good morning, London. Apologies for the delay. Wanted to check in with counsel. These $BUR guys sure do have a guilty look to them, don’t they?”

Anyone familiar with the AIM in August 2019 would know that “$BUR” meant Burford (the listing code for Burford’s ordinary shares is ‘LON:BUR’); and Tweet 2 contained a banner-style link to a very negative 25-page opinion piece by Muddy Waters about the company. The link within Tweet 2 included text, with a heading “MW is Short Burford Capital Ltd. (BUR LN) …” and also the opening words of the opinion piece: “We are short BUR. For years, it was the ultimate “trust me” stock. …”. The opinion piece, if accessed via the link, continued, “Thanks to a light disclosure regime, the esoteric nature of its business, and unethical behavior by its largest shareholder, Invesco, [Burford] turned Enron-esque mark-to-model accounting into the biggest stock promotion on the AIM”, before claiming that detailed investment data “proves [sic.] that BUR has been egregiously misrepresenting its ROIC and IRRs, as well as the state of its overall business”. Burford was accused of manipulating ROIC and IRR metrics and actively misleading investors, a deception, as Muddy Waters called it, that it (MW) believed caused investors to give credence to Burford’s reported fair value gains (a major element, it was said, of Burford’s investment income, as reported by it). Burford was said to have risky liquidity, and to be “arguably insolvent” and “a perfect storm for an accounting fiasco”.

Tweet 2, at 8.53 am, appears to have been preceded, by a minute or so, by the appearance of a Burford tag on the Muddy Waters website, referred to in an online forum at uk.advn.com, Burford Capital Share Discussion Threads, on which a contributor posted at 8.52 am: “No content for the moment, but Muddy Waters has created Burford tag on their webpage”, with a link to the website.

Burford released RNS No.43901 at 12.45 pm on 8 August 2019 in response to the Muddy Waters short attack, opening with a claim that the Muddy Waters opinion piece “is false and misleading” and setting out detailed denials, explanations and counter-assertions from Burford by reference to Muddy Waters’ claims, point by point. Its concluding remarks included that “Short attacks such as this are a fundamental menace to an orderly market and to the value inherent in long-term investing companies such as Burford that are revolutionising industries. Burford is well-equipped to investigate and pursue market manipulators, and as stewards of investor capital, we are exploring doing so here, cognisant of the substantial losses our investors have suffered. Our early investigation already shows the hallmarks of market manipulation.”

Regulatory disclosures duly filed by Muddy Waters reveal its end-of-day short positions. Burford’s case for this Claim included an allegation that Muddy Waters did not comply with the applicable regulatory rules in that regard, but there was no real evidence for that, and the FCA has confirmed in correspondence that it was not so. At the close of trading on 5 August 2019, Muddy Waters was short 0.71% of Burford’s ordinary share capital (0.71% of Burford’s shares at the 5 August closing price being worth c.£21.5 million). It substantially exited that position over the period of the fall in the share price (as would be expected for the close-out phase of a short-selling raid): at the close on 6 August 2019, it was short 0.57%; by the close on 7 August 2019, that was down to 0.12%. It seems likely that Muddy Waters made a substantial profit (e.g. exiting at (say) £7.00 (on average) a short position priced at (say) £13.50 (on average) would mean profit of c.£8-8.5 million on the 0.59% closed out).

It has not been argued that Muddy Waters’ tweets, the opinion piece to which Tweet 2 linked, or its taking, announcing, and profiting from its short-sold position, is unlawful, or actionable in any way by Burford. Or at least, if those actions (individually or collectively) were or involved actionable conduct, that is not the subject of the present Claim. Burford has, however, contended that Muddy Waters’ visible activity implicates it in an alleged conspiracy unlawfully to manipulate the market on 6 and 7 August 2019; and the subject matter of the present claim is Burford’s contention that on those days there was unlawful market manipulation in relation to its shares.

It is obvious, and not in dispute, that, with or without any unlawful conduct on anyone’s part, Muddy Waters’ actions were calculated to generate a fall in Burford’s share price from which, having sold short, Muddy Waters could profit. However, Burford claims that the large extent of the fall in its share price was or may well have been caused or contributed to by spoofing and/or layering, a form of unlawful market manipulation and not (or not only) by genuine trading activity on the market.

The Stock Exchange and its regulator, the Financial Conduct Authority (“the FCA”), have both considered the evidence, in particular the analysis of trading data, put forward by Burford as suggesting that spoofing or layering occurred, including all of the evidence Burford submitted to the court for this Claim. Each has concluded, independently of the other, that Burford’s evidence does not support any claim that there may have been spoofing or layering; and each has satisfied itself, independently of the other and by reference to the full range of information available to it (which in each case is more extensive than the information available to Burford), that there is no reason to think that there was unlawful market manipulation in relation to Burford’s shares on 6 or 7 August 2019. Burford contests those conclusions and says it wishes to pursue further the contention that its share price was the subject of attack by unlawful means.

By this Claim, Burford seeks Norwich Pharmacal relief against the Stock Exchange. The Claim was issued in the Commercial Court as Claim No. CL-2019-000604 and transferred to the Financial List by an order I made at a case management hearing on 28 February 2020. The final hearing of the Claim had already been listed to commence on 1 April 2020, and that listing could be maintained after transfer. The final hearing was not then reserved to me, but as it happens it did come back before me.

That hearing was conducted remotely, via Skype for Business, but still as a hearing in open court (i.e. in public) in accordance with CPR PD51Y, during the UK’s national ‘lockdown’ in response to the Covid-19 pandemic that continues as I finalise this judgment. I pay tribute to the efforts of all involved in ensuring that the hearing ran efficiently, effectively and fairly, as it did, and to the clarity and quality of the arguments of Mr Dhillon QC for Burford, and of Mr Green QC and Mr de la Mare QC for the Stock Exchange, ably supported by Mr Dawid and Quinn Emanuel Urquhart & Sullivan (“Quinn Emanuel”), and Mr Adamson and Bryan Cave Leighton Paisner (“Bryan Cave”) respectively, a clarity and quality maintained in an exchange of additional submissions in writing after the hearing, at my request, to assist my consideration of one part of the evidence.

The Claim concerns activity relating to Burford’s AIM-listed shares on 6 and 7 August 2019, directly on the AIM and on Turquoise, a Multilateral Trading Facility operated by Turquoise Global Holdings Ltd (“Turquoise Global”), which is majority owned by the Stock Exchange. In fact, the vast majority of the order book activity for Burford’s shares took place on the AIM itself, with only c.2.5% and 3.5% being through Turquoise on 6 August and 7 August respectively. The AIM and Turquoise are both regulated by the FCA.

Burford seeks an order that, for every buy and sell order event posted for Burford’s ordinary shares on 6 or 7 August 2019 (each a “Share Order Event” – there were c.360,000 in all), the Stock Exchange provide Burford with what I shall call the “Participant Identity” details, that is to say (in the terms of paragraph 1.2(a) of Burford’s draft Order):

“The names of the natural or legal persons, or third party entities (including but not limited to natural persons, brokers, banks and other financial institutions) which submitted the Share Order Event, including the names of the members or participants of the trading venue who submitted the Share Order Event to the trading venue, and the names of the client on whose behalf the member or participant of the trading venue submitted the Share Order Event to the trading venue.”;

together with a dozen specific pieces of information about the Share Order Event (“the Order Event Data”) listed in paragraphs 1.2(b) to 1.2(m) of the draft Order. However, Burford does not need the Order Event Data, as they are in the anonymised data Burford obtained at the time. Therefore, if the Claim were to succeed, it would be enough for Burford (so that meaningful identification evidence were provided and not just a long list of names) if the Participant Identity details were listed by reference to Share Order Event identifiers sufficient to enable Burford to match those details to the Order Event Data already in its hands. As to that, I understand it would be sufficient for Burford to be given a list of Share Order Events sorted by timestamp (to nanosecond accuracy) that gave for each Event (a) the Order ID to which it related and (b) the Participant Identity details. Burford did not pursue at trial a claim for additional information under paragraph 1.3 of the draft Order.

In opposition to the Claim, Mr Green QC for the Stock Exchange contended inter alia that it goes far beyond the proper scope of Norwich Pharmacal relief, citing in particular Ramilos Trading Ltd v Buyanovsky [2016] EWHC 3175 (Comm), [2016] 2 CLC 896, a decision of Flaux J (as he was then), at [28]-[62]. There, after a full review of the authorities, Flaux J held, at [46], that, “the [Norwich Pharmacal] jurisdiction cannot be used for wide-ranging discovery or the gathering of evidence, but is strictly confined to necessary information”, or again, at [62], that “the Norwich Pharmacal jurisdiction remains an exceptional jurisdiction within a narrow scope. The court will not permit the jurisdiction to be used for wide-ranging disclosure or gathering of evidence, as opposed to focused disclosure of necessary information … . … Furthermore, it is impermissible to use the jurisdiction as a fishing expedition to establish whether or not the claimant has a good arguable case … .”

It does not dispose of that contention to say, as Mr Dhillon QC did, that in the manner I indicated in paragraph 22 above, the information sought by Burford is only as to the identity of persons whose identities are unknown to Burford but known to the Stock Exchange, viz. those responsible, respectively, for each of the 360,000-odd Share Order Events. Whether a claim seeks only “focused disclosure of necessary information”, or whether it amounts to a fishing expedition, is a function, in combination, of what information a claimant seeks and why it seeks it. Burford does not say it has a strong arguable case, or any case at all, that every market participant posting at least one Share Order Event was guilty of wrongdoing. Indeed, its case is that it cannot distinguish between abusive market activity on the one hand and legitimate market activity on the other hand, so as to allow it to seek information only in relation to the former; and the Stock Exchange says this is a case where, in reality, Burford has no idea whether any unlawful activity may have occurred and so no case on the merits of any substance such as might support a Norwich Pharmacal claim.

As that illustrates, it is of the essence of the Norwich Pharmacal claim to identify not only what information is sought but also why it is sought. The latter is not just something that has to be identified in any event because of CPR 31.22 and CPR 32.12. That was recognised by Burford in framing its draft Order, stipulating as it does not only the information sought (paragraph 1 of the draft Order) but also the purposes for which Burford would be permitted to put it (paragraphs 3 and 4 of the Order). It is therefore not an oddity, as Mr Dhillon QC submitted, that in marshalling the Stock Exchange’s contentions, Mr Green QC considered first the relief sought and the Ramilos argument. It is not merely a question of the fine-tuning of the Order to be drawn up if the Norwich Pharmacal claim is made out; it is at the heart of the question whether relief should be granted at all.

The purposes for which Burford says it wants the information sought, and for which therefore it seeks to persuade the court that as a matter of justice it is necessary that the Stock Exchange provide that information, are these:

The latter ((ii) above) takes Burford nowhere. It is inchoate, open-ended and avowedly speculative (“if …”). It amounts to no more than a place-marker for the fact that the court may never say never, i.e. it might include a liberty to apply further if granting relief at all. The former ((i) above) is the substance of the Claim, but it is very generally worded. Why more precisely Burford says it wants the Participant Identity details from the Stock Exchange needs to be considered carefully by reference to its detailed explanations in the evidence filed in support of the Claim, and bearing in mind the starting point from which Burford accepts it is proceeding (paragraph 24 above).

In summary, then, Burford contends that:

The Stock Exchange contends, likewise in summary, that:

As may be taken to be implicit in those summaries, there is no dispute but that the Stock Exchange, although no allegation was made that it was itself a wrongdoer, was mixed up in, so as to facilitate, any unlawful market manipulation (if there was any), by operating through its subsidiaries the markets on which any such wrongdoing occurred. If the claim for Norwich Pharmacal relief otherwise runs, the Stock Exchange is a proper defendant with information it could make available to Burford if ordered to do so. Mr Green QC confirmed in oral argument that no point was taken that the Stock Exchange has been sued for Norwich Pharmacal relief rather than its market-operating subsidiaries, LSE and Turquoise Global.

Regrettably, in its determined insistence that its share price suffered at the hands of unlawful market manipulation on 6 and 7 August 2019, Burford made it part of its case in the Claim to launch a highly critical attack on the motivations and competence

The underlying principle was articulated in Norwich Pharmacal itself in these terms, per Lord Reid, [1974] AC 133 at 175B-C:

“… If through no fault of his own a person gets mixed up in the tortious acts of others so as to facilitate their wrong-doing he may incur no personal liability but he comes under a duty to assist the person who has been wronged by giving him full information and disclosing the identity of the wrongdoers. I do not think that it matters whether he became so mixed up by voluntary action on his part or because it was his duty to do what he did. It may be that if this causes him expense the person seeking the information ought to reimburse him. But justice requires that he should co-operate in righting the wrong if he unwittingly facilitated its perpetration”;

and Mr Dhillon QC also reminded me of Lord Kilbrandon’s statement of the basic idea, ibid at 203D-F:

“The plaintiff is demanding what he conceives to be his right, but that right insofar as it has patrimonial substance is not truly opposed to any interest of the defendants; he is demanding access to a court of law, in order that he may establish that third parties are unlawfully causing him damage. If he is successful, the defendants will not be the losers, except in so far as they may have been put to a little clerical trouble. If it be objected that their disclosures under pressure may discourage future customers, the answer is that they should be having no business with wrongdoers. Nor is their position easily distinguishable from that of the recipient of a subpoena, which, in total disregard of his probable loss of time and money, forces him to attend the court for the very same purpose as that for which discovery is ordered, namely, to assist a private citizen to justify a claim in law. The policy of the administration of justice demands this service from him.”

It was common ground that on a Norwich Pharmacal claim, the standard of proof as to the existence of wrongdoing is that of ‘good arguable case’, in the sense of a case that is more than just capable of serious argument, but is not necessarily a case the court considers to be well-founded, more probably than not. In that regard, both parties referred me to Ramilos, supra, in which Flaux J considered the issue at [12][23]. Thus, the existence of a good arguable case of wrongdoing is said to be a necessary requirement; once over that threshold, the stronger the case appears to be on the merits the more compelling the case may be for the court to grant relief, other things being equal.

It is now established at the highest level that it is not necessary for the Norwich Pharmacal claimant to intend to bring civil proceedings against any underlying wrongdoer whose identity a successful Norwich Pharmacal claim might reveal: see Ashworth Hospital Authority v MGN Ltd [2002] UKHL 29, [2002] 1 WLR 2033;

What is less clear, however, is whether nonetheless it must be the case that the alleged wrongdoing (if proved) would involve a private law (civil) cause of action vested in the Norwich Pharmacal claimant against the wrongdoer (the ‘cause of action issue’). The Stock Exchange says it must. Burford says it need not, so long as there is a legitimate purpose in seeking the disclosure sought that justice demands should be furthered by the provision of that disclosure by the Norwich Pharmacal defendant. There are dicta that favour Burford on that: e.g. per Popplewell J (as he was then) in Orb arl v Fiddler [2016] EWHC 361 (Comm) at [84], “The “wrong” may be a crime, tort, breach of contract, equitable wrong or contempt of court.” But I was not shown a case in which Norwich Pharmacal relief has been granted in respect of alleged wrongdoing that would not invest in the claimant a personal right of action against the alleged wrongdoer. Specifically tested on that, Mr Dhillon QC referred me to:

The focus of the cause of action issue, therefore, was upon the difference of view between the Court of Appeal in Financial Times Ltd v Interbrew SA [2002] EWCA Civ 274, [2002] EMLR 24 and the House of Lords in Ashworth Hospital, supra, on the possible involvement of Norwich Pharmacal relief in relation to the prosecution of crime. Mr Dhillon QC submitted that the view on that point of Lord Woolf CJ, with whom the other members of the House agreed, was part of the ratio in Ashworth Hospital so as to bind me come what may; on the other hand, Mr Green QC submitted that the view of Sedley LJ, with whom Longmore and Ward LJJ agreed, was part of the ratio in Interbrew so as to bind me, the contrary view of the House of Lords in Ashworth Hospital being, he argued, obiter.

In both Interbrew and Ashworth Hospital, relief was granted in respect of alleged wrongdoing that was in principle actionable in civil proceedings by the Norwich Pharmacal claimant against the alleged wrongdoer. The question raised in Interbrew was whether Norwich Pharmacal relief was available where the alleged wrongdoing was a breach of confidence actionable at the suit of the claimant (together with points on the confidentiality of journalists’ sources). The question raised in Ashworth Hospital (again alongside points on journalists’ sources) was whether the Norwich Pharmacal claim was defeated because it sought disclosure so that the claimant hospital authority could “discipline … the person responsible for the leak which in practice would mean that that person would be dismissed” (ibid. at [19]), rather than because it wanted to pursue against that person the civil proceedings that it could in principle have brought. Neither Norwich Pharmacal claim had anything to do with the prosecution of crime.

It will not be necessary to resolve the cause of action issue to determine this Claim. I shall therefore not lengthen this judgment by offering any analysis of or conclusion upon it, obiter, save to say that (a) it is not obvious to me why Sedley LJ was wrong in Interbrew to identify that the Norwich Pharmacal jurisdiction is rooted in the old equitable bill of discovery so as to require as its foundation an allegation of an actionable wrong in support of which that bill would have run, and (b) it is not necessary to disagree with that to get to Ashworth Hospital, since the issue there was whether there has to be an intention to bring proceedings upon the posited actionable wrong, not whether there has to be an allegation of such a wrong to give the court a basis, if it be a sufficiently strong allegation, to get involved.

The following concise summary of the conditions that must be satisfied before any question of granting relief arises is often cited, per Lightman J in Mitsui & Co Ltd v Nexen Petroleum UK Ltd [2005] EWHC 625 (Ch), [2005] 3 All ER 511 at [21]:

“(i)

a wrong must have been carried out, or arguably carried out, by an ultimate wrongdoer;

(ii)

there must be the need for an order to enable action to be brought against the ultimate wrongdoer; and

(iii)

the person against whom the order is sought must: (a) be mixed up in so as to have facilitated the wrongdoing; and (b) be able or likely to be able to provide the information necessary to enable the ultimate wrongdoer to be sued.”

It will be appreciated from what I have said, above, that in Lightman J’s first proposition, “arguably” should be “well arguably”, and that his second proposition and the final part of his third proposition (“necessary to enable …”) also require some qualification or explanation: firstly, they both use the language of bringing a civil suit against a wrongdoer, but it is now established that an intention to commence proceedings is not required; secondly, the supposed pre-condition of necessity “does not require the remedy to be one of last resort, but the need to order disclosure will be found to exist only if it is a “necessary and proportionate response in all the circumstances””, per Zacaroli J in Blue Power Group SARL et al. v ENI Norge AS et al. [2018] EWHC 3588 (Ch), at [17(ii)], derived from the Rugby Football Union case, supra, per Lord Kerr of Tonaghmore JSC at [16]. Thus, it is not correct to say that the jurisdiction is limited to cases of strict necessity. Rather, the question is whether, in

Indeed, in the Rugby Football Union case, Lord Kerr immediately went on, at [17], to emphasise that “The essential purpose of the remedy is to do justice”. Lord Kerr described the exercise of ascertaining what justice requires in the particular case as an “exercise of discretion by a careful and fair weighing of all relevant factors”, before listing the main factors identified as relevant in the authorities. I respectfully wonder whether it is right to characterise the balancing exercise to which Lord Kerr referred as an exercise of judicial discretion rather than one of judicial discernment, and whether therefore it is right (albeit now an orthodox description) to say that the Norwich Pharmacal jurisdiction involves (a) several pre-requisites, and then (b) a discretion as to whether relief should be granted.

I think there is room for the view that in truth there is:

In that analysis, of course the seeming strength of the claimant’s case that he has been wronged would be an important factor, but there would be no need to treat it as separately relevant to (i) the existence of a discretion and then (ii) a question whether a discretion should be exercised. The cases articulating a threshold merits requirement might be seen for what they really are, namely assessments that justice did not require that third parties mixed up in so as to facilitate the actions of others lend their aid in respect of a case of wrongdoing by or on behalf of those others that did not amount to a good arguable case. Or again, arguments over the scope of the information sought, or the purposes for which the claimant seeks an order, might be seen for what they are, namely matters going to whether, in the particular case, the justice in righting facilitated wrongs compels the facilitator to lend his aid to the claimant by providing that information.

Whether that be right or not – and both parties presented the argument in this case on the ‘multiple pre-requisites plus discretion’ model – Lord Kerr’s list of factors is plainly pertinent and important, both in point of logic and given its source:

In the present case, there are further factors, although they can be analysed as part of assessing the significance on the facts of one or more of Lord Kerr’s factors. Those further factors are:

Finally, as exemplified by R (Omar), supra, and the reason for its disapproval of R (Mohamed) (No.1), it is possible for a legislative scheme so to have occupied the territory that might otherwise be occupied by the Norwich Pharmacal jurisdiction as to make it inconsistent with that legislative scheme to grant relief under the common law doctrine. The regulatory scheme here, to which I adverted in the previous paragraph, was said by the Stock Exchange to give rise to a primary argument of that kind, although it was also relied on, in the alternative, as going to the discernment of what justice demands (or to discretion, if that be the preferred analysis).

Since this is a Part 8 Claim, there were not full statements of case. That does not diminish the importance, for the process to be fair to the defendant, of identifying with clarity the factual and legal basis asserted by the claimant for the relief it seeks. So far as concerns the wrongdoing alleged, Burford set out its case in the Claim Form, in paragraphs 4 to 6 of its “BRIEF DETAILS OF CLAIM”, as follows:

“4.

As set out in the evidence served with this Part 8 Claim, the fall in the price of Burford’s shares during the Relevant Period was caused or contributed to by trading patterns known as “layering and spoofing”. This involved very largenumbers of sell orders for Burford shares being submitted to and advertised viathe LSE’s trading platforms, without any genuine intention to trade, and which gave or were likely to give a false or misleading impression as to the market for Burford’s shares and thereby negatively influencing their price. [my emphasis]

5.

The persons responsible for such spoofing and layering (“Manipulators”) acted unlawfully and in breach of at least:

a.

section 12 of Regulation (EU) No 596/2014 (“ Market Abuse

Regulation”);

b.

section 90 of the Financial Services Act 2012;

c.

the tort of deceit; and/or

d.

section 2 of the Fraud Act 2006.

6.

Further it is to be inferred that Manipulators were acting in combination with each other and/or others with a financial interest in Burford’s share price falling during the Relevant Period, so as to commit the tort of conspiracy by unlawful means.”

In view of the language of paragraph 4 (“was caused or contributed to …”), at the hearing to which I referred in paragraph 19 above Burford confirmed (as recited in the order on that hearing) that it would seek to establish only that it has a good arguable case that the fall in its share price on 6 and 7 August 2019 was caused or contributed to by spoofing or layering. The Claim gives rise as regards alleged wrongdoing, therefore, to three questions:

The foundation for each of Burford’s allegations of wrongdoing is Article 15 of the EU Market Abuse Regulation (“MAR”), Regulation (EU) No.596/2014, by which “A person shall not engage in or attempt to engage in market manipulation”. Market manipulation is defined, for the purpose of MAR, by Article 12. The form of market manipulation relevant here is the sending of false market signals, as defined by Article 12(1)(a)(i), that is to say:

“(a)

entering into a transaction, placing an order to trade or any other behaviour which: (i) gives, or is likely to give, false or misleading signals as to the supply of, demand for, or price of, a financial instrument …”.

Mr Dhillon QC also referred me to Article 12(1)(b) (behaviour affecting or likely to affect price “which employs a fictitious device or any other form of deception or contrivance”), but that adds nothing in this case as the only form of deception suggested is the sending of false or misleading signals already caught by Article 12(1)(a)(i). He also referred me to Article 12(2)(c)(iii) which spells out that placing, cancelling or modifying orders on a trading venue by any available means of trading (including by electronic means such as algorithmic or high-frequency trading strategies), if it creates or is likely to create a false or misleading signal about supply, demand, or price, “in particular by entering orders to initiate or exacerbate a trend”, “shall, inter alia, be considered as market manipulation”.

Thus, the essential ingredient is the giving of (or likelihood of giving) false or misleading signals as to supply, demand or price. Contrary to a submission by Mr Dhillon QC, that cannot always be an entirely objective enquiry, as it depends upon the signal given out (or likely to be given out) by a particular activity or behaviour. Depending on what that signal is (an objective question, to be sure), falsity might involve some subjective enquiry. The point of substance, therefore, is whether, as Mr Dhillon contended, there is no subjective element in the supply, demand or price signalling involved in this type of case (or, therefore, in considering whether there was any false or misleading signal).

I do not accept that contention. A seller wishing to sell at £10 who offers to sell at £10 but, finding no takers at that price, withdraws his offer because he does not want to sell for less, and a seller who has no intention to sell at £10 but who offers to sell at £10 to initiate or exacerbate a price trend, then withdraws his offer, appear to the outside observer to have behaved identically. If their behaviour fell to be judged entirely by that appearance, they would have to be either both guilty or both innocent of a charge of market manipulation. But in reality, surely the former is innocent, the latter guilty, and that is because although the signals sent out were the same (e.g. their initially signalled intention to sell at £10), the truth or falsity of those signals turns on their actual intentions, which differed radically.

It is no defence for the guilty seller to say that he did not appreciate he was sending out a false signal, or in some other way that he did not intend to manipulate. In that sense, the manipulation does not have to be deliberate; and that is the relevant proposition to derive from Financial Conduct Authority v Da Vinci Invest Ltd [2015] EWHC 2401 (Ch), [2016] 3 All ER 547, per Snowden J at [104]-[108], decided under s.118 of FSMA 2000 implementing the EU Market Abuse Directive (Directive 2003/6/EC), which was the EU law predecessor to MAR. That is a different point, however, and the conclusion in that case, that there had been market abuse, required (and was justified by) the finding as to the actual purpose behind the relevant order activity (ibid, at [163]): “The purpose of the Traders was not in truth to sell or buy as indicated by the orders they placed. Instead, the Traders intended to cause a movement in the market price of the share … and to induce other market participants

Again, in my simple example, the time at which and market circumstances in which the initial offer to sell at £10 was placed might afford an argument that a seller looking to sell but unwilling to sell below £10 would have assessed that there was no sensible prospect an offer to sell at £10 might be matched and so would not have placed the offer. That might be an aspect of assessing the genuineness of the offer in fact placed, but it could not be a conclusion that foreclosed the inquiry. That MAR adopts that approach is confirmed by Part A of Annex I, with its list of ‘nonexhaustive indicators’ to be ‘taken into account’ when examining transactions or orders to trade for the purpose of applying Article 12(1)(a).

Article 13(2) of MAR provides that a ‘competent authority’ (the FCA for this jurisdiction) may establish ‘accepted market practices’, subject to taking account of various criteria the detail of which does not matter in this case. Then by Article 13(1), the prohibition in Article 15 does not apply to market manipulation falling within Article 12(1)(a) if the person responsible establishes that the manipulation was “carried out for legitimate reasons, and conform[ed] with an accepted market practice as established [under Article 13(2)]”. Thus, Article 15 does not render all market manipulation unlawful; by virtue of the ‘accepted market practice’ device of Article 13, and leaving aside questions of the burden of proof, MAR includes a concept of lawful market manipulation.

There is no suggestion in this case that Article 13(1) might be in play. It is relevant, however, because Mr Dhillon QC argued that it proved or supported the submission I rejected in paragraphs 50 to 53 above. Because Article 13(1) refers to the putative wrongdoer’s purposes or motives, so the argument went, they cannot be relevant under Article 12. I do not agree. The correct logic is this: Article 13(1) does not qualify the definition of market manipulation, rather it provides circumstances in which market manipulation within that definition is not prohibited; those circumstances are defined inter alia by reference to a manipulator’s purpose or motive; but a putative wrongdoer has no need to call upon Article 13(1) unless manipulation be established first; that requires (so far as presently material) proof of false signalling to the market; what it means to say that there was false signalling to the market depends on the signal.

Finally, for this review of the legal framework, Mr Dhillon QC referred me to Commission Delegated Regulation (EU) No.2016/522, supplementing MAR, one purpose of which (see recital (5)) is to supplement and clarify Annex I of MAR. It confirms (see recital (10)) that its relevant subject matter, ‘orders to trade’, “encompasses all types of order, including initial orders, modifications, updates and cancellations, irrespective of whether they have been executed …”. Section 1 of Annex II to the Delegated Regulation sets out lists of particular practices, one list for each of paragraphs (a) to (g) of Part A of Annex I of MAR. The effect of Article 4(1), then, is that each listed practice is to be regarded as ‘specifying’, i.e. being a specific instance of, the ‘indicator’ listed in the corresponding paragraph of MAR Annex I.

The main point of substance arising from the Delegated Regulation is only that it confirms and reinforces what I said about Annex I of MAR in paragraph 53 above. It serves also, in the detail, to support the logic of paragraphs 50 to 55 above. Most pertinently, the practice listed at paragraph 5(e) of Annex II to the Delegated Regulation is specified as an instance of the indicator at paragraph (e) of Part A of MAR Annex I, as to which:

Paragraph 6(a) of Annex II to the Delegated Regulation specifies, for the indicator at paragraph (f) of Part A of MAR Annex I, the practice of “Entering … orders that are withdrawn before execution, thus having the effect, or which are likely to have the effect, of giving a misleading impression that there is demand for or supply of a financial instrument [etc.] — usually known as ‘placing orders with no intention of executing them’.” That does not, as it may be Mr Dhillon QC sought to suggest, equate the withdrawing of an order prior to execution with a lack of intention to trade when the order was placed. The middle step is important – there has to have been an effect, or a likelihood of creating the effect, of a misleading supply or demand signal. That will not be true if the reason for the withdrawal is not that the order was not genuine when placed.

It might have been helpful for Burford to particularise more fully in the Claim Form the case it alleged to the effect that its shares were or may have been subject to spoofing or layering. As it is, the Claim Form just cross-referred to the evidence served with it for the case that was to be advanced. Thus, the case to be advanced was articulated, in the first place, through the argumentative witness statement of Mr East of Quinn Emanuel (“East 1”), which contained the following allegations:

Prof Joshua Mitts is an Associate Professor of Law at Columbia University and principal of M Analytics LLC, a financial economics consultancy. His PhD, from Columbia, is in Finance & Economics and his relevant expertise for present purposes is in the statistical analysis of trading data (‘data analytics’) to identify signs or possible signs of improper activity. His academic work has included teaching courses at Columbia on securities regulation, data science and contracts. His courses cover economic theory and quantitative methods of valuation, asset pricing, investments, and data analytics, as well as the economics of securities fraud, market manipulation and insider trading. He has been consulted in matters involving allegations of securities fraud and market manipulation, and his research and expert opinions have been presented to the U.S. Securities and Exchange Commission and other regulatory bodies.

Prof Mitts’ opinions are central to Burford’s contention that there was or may have been spoofing or layering (see paragraph 59(ii)(a) above). A number of other witnesses provided evidence, in particular because the Stock Exchange’s evidence in response to the Claim raised matters beyond any remit that might be given to Prof Mitts. It remains the position, though, that there is no case at all unless, upon the basis of Prof Mitts’ work, and taking account of the direct response to it submitted by the Stock Exchange, there are, as East 1 put it, strong grounds to consider that there was spoofing or layering on 6 and/or 7 August 2019. The major part of what follows, therefore, is my consideration of that hypothesis.

Prof Mitts appears to have been engaged by Burford immediately after the Muddy Waters short attack. The precise terms of his brief were not disclosed. Mitts 1 was not prepared as an expert report conforming to CPR Part 35, but permission to rely on it as expert evidence was given under CPR 35.5(1) at the case management hearing in February. The report says that it “summarises the findings of an investigation into the manipulation of the share price of Burford Capital’s common stock on 6 and 7 August 2019”, which is question-begging, and claims of Prof Mitts’ data analysis that it “shows that market manipulation in the form of spoofing and layering caused an artificial decline in Burford’s share price.”

Prof Mitts worked on the order book data provided by the Stock Exchange to Burford. That data set is, in general, anonymised, i.e. the identity of the market participant placing the order is generally not given, but it is otherwise detailed and complete for the c.360,000 Share Order Events. The exception to that anonymity in the data set is that Executable Quotes (“EQs”) and the market-makers providing liquidity by placing them can be identified from the order book data and a publicly available key. Each new share order placed will have a unique Order ID; and the anonymised order book data gave, for each Share Order Event, the Order ID to which it related.

The focus of Prof Mitts’ analysis and opinions is sell-side orders, i.e. orders placed to sell at a given price if there are buyers with whom the market can make a match so as to execute a trade. Any descriptive examples I give will therefore focus on the selling

If a seller with an unmatched order wishes to change his offered price or transaction size (or both), whether up or down, he can do so either by posting an amendment, which will be a further Share Order Event showing the same Order ID as the original order, or by cancelling that order and placing a fresh order at the new price and/or size. The latter is equivalent to the former, functionally, legally and economically; but where the seller is not identified in the anonymised data, that data will not reveal the latter for what it is in substance, i.e. an amendment rather than an outright cancellation.

Building on terminology used by Prof Mitts in his reply report (“Mitts 2”), three illustrative sequences can be imagined:

For present purposes, Sequences 1 and 2 are the same, and in substance they do not involve the cancellation of Order X, only its re-pricing to US$9 a share. Sequence 3 is different. It does involve, in substance, the cancellation of Order X. Put another way, all three Sequences involve a withdrawal of supply, to the tune of 100 shares at US$10 a share; but Sequences 1 and 2 involve, in addition, the creation of replacement supply to the tune of 100 shares at US$9 a share, whereas with Sequence 3 the withdrawal of supply is absolute. If it is not known that Order Y was created by the Order X seller by way, in substance, of a price amendment, Order Y is liable to be omitted from any analysis of Order X, and its cancellation is liable to be misidentified as an instance of Sequence 3.

It is self-evidently material to any consideration of the implications of cancelling Order X, upon the genuineness of the seller’s intention to trade when placing it, to know that the placing and cancelling of Order X were part of Sequence 2 rather than constituting Sequence 3. That is not to say that when placing Order X, the seller in Sequence 3 necessarily lacked or the seller in Sequence 2 necessarily had a genuine intention to trade, but it is to say that they are, and obviously so, materially different sequences, giving rise to different analyses in any further interrogation of the data or of the seller if testing the genuineness of the apparent trading intent.

I should also note, having discussed those illustrations, that in these days of high frequency trading by algorithmic trading programmes, there could be huge numbers of Share Order Events that can be shown to have involved one or other of those Sequences but that will not have involved individual human decisions, Event by

By Mitts 1, Prof Mitts opined that the order book data provided to Burford by the Stock Exchange gives “strong evidence of spoofing and layering, i.e. the creation and cancellation of a large volume of limit orders which distorts the supply and demand for a security”. The creation and cancellation of a large volume of limit orders may or may not involve any seller in deceptive behaviour (false signalling) – I take it that by ‘distorts’ Prof Mitts means ‘misrepresents’ and not merely ‘reshapes’, otherwise this judgment could stop here. But then to spell it out more fully, Prof Mitts’ definition is (or should be) “… spoofing and layering, i.e. the creation and cancellation of a large volume of limit orders placed by sellers not intending, when placing those orders, totrade on the offered terms, which distorts the supply and demand for a security”.

There is an inherent weakness in any claim that a case can be built upon Prof Mitts’ analysis, namely that there is no true control for comparison. Burford’s share price suffered a dramatic and very substantial decline on 7 August 2019, following Tweet 2, on the back of a significant initial fall on 6 August 2019, after Tweet 1. To posit that unlawful conduct was involved is to posit that the Tweets (including under Tweet 2 the vitriolic attack on Burford in the Muddy Waters opinion piece) did not in fact do, and/or would not in fact have done, that much damage to the share price on their own. All that data analysis of the sort undertaken by Prof Mitts can ever claim to do is identify quantitatively measurable features or patterns in the data. The conclusions to draw from those features or patterns are for the court, and not a matter for Prof Mitts.

Of course, as Mr Dhillon QC emphasised in his submissions, I do not have all the evidence (nor does Burford, hence this Claim) that would be available at the trial of a claim against one or more market participants (or their principals or clients) alleging wrongdoing on their part. The merits question is not whether Burford has proved that there was spoofing or layering, or that it caused some or all of the fall in the share price, but how strongly arguable it is that spoofing or layering occurred, or was causative.

Before I turn to the detail of Prof Mitts’ analysis, I should indicate how he used the terms spoofing and layering in his reports in this case because the precise use of those terms can vary.

I would understand that spoofing is often used generically to mean the placing of limit orders to a large volume on one side of the limit order book when not intending to trade on that side of the book. A lopsided limit order volume signals excess supply or demand, as the case may be; and that has the capacity to move the price. The lack of intention to trade on the side of the book where the spoof orders were posted (which defines them as spoofs) means that the excess supply or demand signal was a false signal, and if the price does indeed move away in response, it has been tricked (spoofed) into doing so. This seems to me consistent with the use of the term by Snowden J in FCA v Da Vinci Invest, supra.

Prof Mitts however used the term more narrowly, or specifically, to mean “placing a high volume of orders at a price equal to or better than the national best-bid-best-

In relation to layering as a term, I understand that some use it interchangeably with spoofing, while others use it as a more specific term for one or more particular species of spoofing. Obviously what matters for any consideration of Prof Mitts’ reports is how they used the term; and Mitts 1 explained that for this case Prof Mitts meant by layering something similar to but different from spoofing (as he used that term), namely and specifically the placing of a high volume of orders “deeper in the order book, i.e., at prices above the best offer [for sell-side layering] or below the best bid [for buy-side layering].”

Prof Mitts illustrated again by his imagined £9.99 NBBO, this time positing a sellside manipulator, worried that sell orders at £9.98 might trade, placing instead a high volume of orders at £10.01, seeking to exploit the fact that putting up a wall of sellside orders can push the price down, even though it is erected above the best offer, if it is nonetheless close to it, either directly (by inducing market-makers to adjust their quotes downward) or indirectly (by suppressing (genuine) demand because of the appearance of a limit on how far the price can recover). Prof Mitts went on to explain that, in that definition of layering, “for a layering strategy to succeed, the manipulator must place and cancel these orders at prices which maintain a given distance from the best offer (or bid).”

So as Prof Mitts used the terminology in this case, spoofing and layering both involve the repeated placement, cancellation and replacement of limit orders in a high volume, at incrementally reducing prices. Going back to Sequences 1 to 3 and paragraphs 64 to 68 above, Prof Mitts was not looking for Sequence 3, he was looking for Sequences 1 and 2, at or below best offer (for spoofing, as he used that term), or above but close to best offer (for layering, as he used that term).

Although Sequences 1 and 2 are, as I have said, functionally, legally and economically equivalent, Prof Mitts’ ability to find them was not. Sequence 1 should be visible in the anonymised data, because the Order ID stays the same; Sequence 2 will only be visible in the anonymised data if generated by EQs; but the sell-side orders, order cancellations and new orders generated by EQs that might be seen as instances of Sequence 2 cannot be spoofing or layering, because by definition they will have come from positionally neutral liquidity providers and will have been matched by paired buy-side activity (each EQ being a matched sell-side/buy-side order pair equidistant from the NBBO). A sophisticated manipulator, aware of the expected behaviour of market-makers with EQs, might seek to factor that expected behaviour in to the detailed design of his manipulation strategy, but that is a different point – the Share Order Events generated from EQs cannot themselves be spoofing or layering.

Then finally, to be clear, and applying what I said in paragraphs 50 to 51 above, it is not the case that to find a pattern of repeated placement, cancellation and replacement is to find spoofing or layering, because the question remains one of intention to trade.

Before getting into the detail, therefore, I am clear that Prof Mitts is wrong to opine that “repeatedly placing and cancelling orders in a very short time is strong evidence of manipulative intent”, or that “a wave of abnormal order cancellations [i.e. sell-side cancellations in statistically atypical volumes] at or above the best offer … indicates intentional manipulation of Burford’s share price”. Mitts 1 set out the logic for those claims, but it immediately betrays their error. The logic set out is that there is no economic justification for a short seller to place a large volume of sell orders above the best offer. The opinion that logic might justify, a more limited opinion than that expressed by Prof Mitts, is that a short seller repeatedly placing and cancelling orders in a very short time, above the best offer, is strong evidence of manipulative intent.

That creates a significant problem for Burford’s contention that Prof Mitts’ data analysis provides a strong basis for suggesting that its shares were the subject of manipulative behaviour. Prof Mitts has no idea whether those who would be accused of such behaviour upon the basis of his opinions as quoted above were short selling, but unless they were, they could not properly be so accused.

Finally, before turning to the individual elements of Prof Mitts’ data analysis, I note that the only case put forward by Mitts 1, and therefore by Burford as its claim, is a case founded upon sell-side cancellations at or above the best offer. Thus, in Prof Mitts’ use of the terms spoofing and layering (paragraphs 75-76 above), the case involves only an allegation of layering.

The first element in Prof Mitts’ analysis is a large increase and aggregate volume of sell-side order cancellations following each of the Muddy Waters tweets. For these purposes, Prof Mitts counted as cancellations both Share Order Events identifying themselves as such and non-execution price or size updates (amendments), the latter then also counting as new orders on the amended terms for the purpose of Prof Mitts’ analysis of the data. In other words, he treated Sequences 1, 2 and 3 as equivalent for the purpose of counting and drawing conclusions from sell-side order cancellations.

Prof Mitts found that on 6 August 2019, whereas the aggregate share volume of sellside order cancellations up to 1.30 pm was 2.2 million shares, it was 12.85 million shares from 1.30 pm to the end of trading. On 7 August 2019, the aggregate share volume of sell-side order cancellations was 91.77 million shares. Where sell-side orders are being placed, or cancelled, by positionally neutral market-makers (whether registered market-makers posting EQs, or other participants posting matching sellside and buy-side orders not identifiable as EQs), that cannot be manipulation and should be removed from the equation.

The Stock Exchange’s responsive evidence came principally from Mr Liam Smith, Head of Market Supervision, who I am satisfied is qualified to respond in an expert manner to Prof Mitts’ analysis and who, with his team and with the full information available to the Stock Exchange (and thus, in particular, without the anonymity affecting the data used by Prof Mitts), reconstructed and reviewed the entire order book for 6 and 7 August 2019. Mr Smith confirmed that of Prof Mitts’ 12.85 million and 91.77 million aggregate sell-side order cancellation volumes, 7.2 million and 43.5 million respectively came from EQs amending to a new price.

Mitts 1 stated that stripping out sell-side order cancellations that are matched by buyside order cancellations because they are the activity of positionally neutral market makers “adjusts for the possibility that the observed cancellation rates are simply driven by changes in the price, for such changes should lead to symmetric cancellation rates on both sides of the order book”. No explanation for or science behind that claim was provided. It is another example (as in paragraph 81 above) of imprecision or over-statement from Prof Mitts, as is clear from the paragraph of Mitts 1 that makes the claim. Prof Mitts there explained that the symmetric cancellations on both sides of the order book that he was saying should be generated by price movements are the order cancellations of positionally neutral market makers triggered by such movements. For an order cancellation volume imbalance to be of any moment would require some basis, not provided by Prof Mitts, for a view that, absent manipulative activity, sell-side and buy-side order cancellation volume should balance out even when the share price in question is falling (or rising) dramatically.

During Mr Dhillon QC’s opening submissions, I raised with him my concern that Prof Mitts focused on the number of individual trading minutes during which there was a sell-side order cancellation volume imbalance, which is the point to which I turn next, but Mr Smith’s evidence in response, to which Prof Mitts provided no answer in Mitts 2, was that there were similar numbers of trading minutes during which there was a buy-side cancellation volume imbalance (see paragraph 91 below). In a written Note provided in response, Mr Dhillon submitted that what mattered was that the cancellation volumes involved in the sell-side imbalance minutes exceeded those involved in the buy-side imbalance minutes. This was said to matter because if minute-wise sell-side and buy-side order cancellation volume imbalances were generated by chance out of higher overall trading volumes, they ought to even themselves out. That suggestion about the chance effect of higher trading volumes was not founded upon any of the expert evidence from either Prof Mitts or Mr Smith, although it has a certain intuitive logic to it (if intuition is to be trusted here). But the conclusion submitted in the Note does not follow, since there was no evidence explaining why one should not find, with a dramatically and rapidly falling price, an order cancellation volume imbalance (sell-side volume larger, perhaps much larger, than buy-side volume). The point of interest raised by the supplementary Note but not answered by it was not the pattern that random chance and high trading volumes might generate, but why, absent manipulative behaviour, a large, dramatically sharp share price movement should not be accompanied by an order cancellation volume imbalance in favour of the side away from which the price is rapidly moving (i.e. a sell-side cancellation imbalance if the price is falling; a buy-side cancellation imbalance if it is rising).

Whereas Mr Dhillon QC submitted that Mr Smith “offers no explanation as to why net cancellations on the sell-side should so far exceed those on the buy-side”, in fact (a) Mr Smith’s evidence was clear that everything he saw in the patterns identified and described by Prof Mitts is what he would expect to see, in the absence of market manipulation, when a stock is subject to a major sell-off, and there is no basis for any suggestion that he failed to notice the cancellation volume imbalance, and (b) this is Burford’s claim, and the significant omission is that Prof Mitts offered no explanation for why net cancellations on the sell-side should not substantially exceed those on the buy-side (by aggregate share volume) when the share price is plummeting.

In that regard, Mr Dhillon QC’s supplementary Note also referred to Prof Mitts’ regression analysis to find what he (Prof Mitts) called abnormal sell-side (or buy-side) order cancellations. I discuss that analysis below and am unable to accept that it generates a good arguable case of layering. It does not plug the gap in Burford’s case that I have just been considering.

Returning, then, to Mitts 1, having acknowledged the need to net out buy-side order cancellations, the analysis focused on trading minutes and, specifically, what Prof Mitts labelled ‘spikes’ in net sell-side order cancellations, i.e. trading minutes during which the sell-side order cancellation volume exceeded the buy-side volume by more than 10,000 shares. The claim was that these are significant in themselves and are not plausibly to be interpreted without positing market manipulation. That however cannot be right, as there was a similar pattern of ‘spikes’ on the buy-side, and no suggestion that they are markers for buy-side manipulation. I accept Mr Smith’s opinion that reliance on a large number of sell-side ‘spikes’ after each Muddy Waters tweet as some evidence of sell-side manipulation is fundamentally flawed. As Mr Dhillon QC accepted, Prof Mitts gave no answer to that opinion in Mitts 2, his report replying to Mr Smith’s evidence. (There is then the point on order cancellation imbalance volumes – the combined height of the sell-side spikes exceeds that of the buy-side spikes – but that is a separate aspect.)

Mr Smith’s further view was that since the order counts (new, amendment and cancellation) remained equivalent on both sides of the book throughout and the order to trade ratios remained consistent (in fact, Mr Smith found the order to trade ratio on 6 and 7 August 2019, at 9.22 and 9.86, to be slightly lower than on the prior trading day, 5 August 2019 at 10.65), there is no basis for concluding that the sell-side order cancellations observed by Prof Mitts were materially abnormal (i.e. other than might be expected, without market abuse, for a dramatically falling stock). In my judgment, Prof Mitts’ reports do not provide any clear or substantial basis for supposing that Mr Smith might be wrong about that.

Prof Mitts suggested that sell-side order cancellations cannot be explained by genuine market activity in reaction to the Muddy Waters tweets (and then in reaction to that reaction, etc.). His logic was that if the market takes the view that Burford’s stock is overvalued, market participants “would immediately place–but not cancel–sell orders at or below the best offer to maximise their profits before the share price [fell] further. A wave of panicked selling by investors should not induce the sort of extreme spike in cancellations observed in the data” (original emphasis). That is overly simplistic, and obviously so, as it fails to have regard to the wide range of lawful execution strategies to maximise profits (or minimise losses) that might be adopted in the face of a rapidly moving price target.

Furthermore, it forgets that Prof Mitts counted amendments of existing orders as cancellations. A genuine seller amending his offered price down (whether by Sequence 1 or by Sequence 2) is, in substance, doing exactly what Prof Mitts says he would expect, viz. placing an order at a new (lower) price to try to trade, because the market has moved, or is moving, away from him, or he anticipates that it will do so. Yet that seller is treated as manipulative by Prof Mitts’ stated conclusion.

That Prof Mitts’ analysis, or more strictly the conclusion Mitts 1 claims can be drawn from the analysis, is fundamentally flawed in that way is demonstrated by Tables 2 and 3 in Mitts 1:

Prof Mitts, it turns out, was wrong to claim that the instances he listed in Tables 2 and 3 were or might be cancellations by market participants that had no genuine intention to sell (the claim made, in terms, in Mitts 1). Mr Smith’s response demonstrated, with respect, that Prof Mitts had no basis for making that claim. He should not have offered any stronger claim for Tables 2 and 3 than that they listed a number of substantial sell-side orders placed and cancelled within a very short timescale, and he could say nothing one way or the other as to whether the cancelled orders may not have been genuine.

Furthermore, echoing paragraph 81 above, if Prof Mitts had taken care in formulating his basic propositions, it should more naturally have led him to qualify his later claims, e.g. to explain, as regards Tables 2 and 3, that they could only be evidence of possibly manipulative behaviour if the sellers in question were selling short. Mr Smith was able to confirm that 32 of the 38 examples across Tables 2 and 3 were generated by a single seller exiting a pre-existing long position and sell-side cancellations by positionally neutral liquidity providers matched by simultaneous buy-side cancellations.

The other 6 examples, Mr Smith confirmed, were generated by best-execution algorithms. In relation to execution algorithms, as Mr Smith explained: “… it is quite typical for legitimate execution algorithms (programmed to execute orders) to be set so as … not to “cross the spread” and aggressively execute against an outstanding buy order. This dynamic … is often confused with the appearance of spoofing by thosewithout an operational knowledge of trade execution” (my emphasis). In reply, Mitts 2 again misunderstood the point and did not answer it, making instead:

Finally, on Tables 2 and 3, some 20% and 31% respectively of the orders in question actually traded and were thus examples of successful amendments to chase down and find demand that Prof Mitts accepted (in Mitts 2) will not have been manipulative, though it may be Prof Mitts could not see that in the anonymised data. Furthermore, those execution rates are markedly higher than overall rates: on the order to trade ratios confirmed by Mr Smith (paragraph 92 above), overall only c.11% and c.10% of

Overall, it is surprising that Mitts 2 did not simply acknowledge, as it should have, that while Prof Mitts could not know this when writing Mitts 1, Tables 2 and 3 can now be seen to have provided no evidence whatever that spoofing or layering occurred. Whereas they were claimed by Mitts 1 to list exemplars of sell-side cancellations suggesting that the orders cancelled were not genuine when placed, in fact there was and is no basis for that claim.

The same is true for further tables – Tables 4 and 5 – that Mitts 1 presented as exemplars, this time exemplars of ‘spoofing and layering loops’. Again, the focus is on (alleged) sell-side spoofing. A sell-side spoofing and layering loop, as Prof Mitts explained it, would be where:

In that definition, the spoof orders amounting to unlawful market manipulation are the original sell-side orders (step (i) above). They are spoof orders not (or not without more) because of their effect – the downward revision of best bid they induce – but because the purpose behind placing them was to induce that effect rather than to trade at the offered price. The manipulator’s buy-side order that executes (step (ii) above) is not a spoof order, to the contrary the whole point of the loop is to get that order to trade. Placing that buy-side order is not itself spoofing, it is the means by which the manipulator seeks to profit from the spoofing at step (i).

Prof Mitts noted that on 6 August 2019, prior to Tweet 1, there were on average, per second, 120 ‘sell-side shares’ created but also cancelled between successive buy-side order executions, of which 100 (83.3%) were “non-marketable sell orders created deep in the limit order book”. I understand that these ‘sell-side shares’ measure share volume, not number of orders. Thus, what Prof Mitts means by the creation in a second of X ‘sell-side shares’ is that within that second there were sell-side orders for (in aggregate) X shares.

On 6 August 2019, after Tweet 1, the 100/120 ratio was 1,064/1,259 (84.5%); on 7 August 2019, it was 2,658/3,528 (75.3%). Prof Mitts said that those leaps in volumes “are extremely unlikely to have occurred by random chance”. He did not make anything, one way or the other, of the ratio, i.e. whether it matters that it was the same on 6 August 2019 before and after Tweet 1, but c.10% lower on 7 August 2019. I do not understand the comment about random chance to be contentious. On the Stock Exchange’s case, the obvious cause of the massive jump in volumes, not being random chance, was the Tweets and the massive share price slide that followed them.

Prof Mitts then claimed to ‘illustrate’ spoofing and layering loops by finding instances where there was a price drop from one buy-side execution to the next and a large volume of ‘sell-side shares’ created and also cancelled between the two. But immediately there is again the problem of the chicken and the egg, if the suggestion is that the sellers responsible might have placed the sell-side orders that were cancelled without genuinely intending them to trade. The measurable phenomenon itself is consistent with genuine trading intent, and it is in truth simply speculative, on the data available to Prof Mitts, to say anything about the trading intention of the sellers responsible.

Rather than giving that opinion, Mitts 1 proceeded to present, as Tables 4 and 5, data available to Prof Mitts for his illustrations, describing them (as if that data established this) as examples of spoofing and layering loops. But from the full data set available to the Stock Exchange, Mr Smith was able to show convincingly that none of the Share Order Events picked out by Prof Mitts is suspicious.

That brings me to the regression analysis to which I referred in paragraph 90 above. Mitts 1 claimed to show that the “correlation between the number of orders created and the number of orders cancelled” on the sell-side (by number I think Prof Mitts in fact meant share volume) shows that “the order cancellations observed immediately following the posting of the Muddy Waters tweet[s] are statistically abnormal, and not merely expected due to an increase in new sell-side orders” (original emphasis).

Subject to considering the detail underlying that claim, it seems to me a problem for it that the order to trade ratio remained consistent (see paragraph 92 above); and if the suggestion is that this analysis gives evidence of market manipulation it suffers from the same unanswerable problem that order amendments (whether Sequence 1 or Sequence 2) by sellers trying to sell out of long positions as the share price plummeted are included as cancellations and (for all Prof Mitts is able to say) could be the entire phenomenon, so the data is consistent with an absence of manipulative behaviour.

Does the detail overcome those problems, or otherwise justify the claim made? In my judgment, the answer is no, for the reasons set out in paragraphs 110 to 118 below.

Prof Mitts explained that he banded the Share Order Events as follows: at or below best offer, up to 0.25% above best offer, then 0.25-0.33% above best offer, 0.33-0.5%, 0.5-1%, 1-3%, 3-5% and 5-10%. He said that he used that banded distribution to “estimate the ordinary relationship between the creation and cancellation of sell-side orders at those bands over an unaffected period (1 July to 31 July 2019)”. He did the same for the buy-side – he did not spell this out, but I assume his buy-side price banding was: at or above best bid, up to 0.25% below best bid, then 0.25-0.33% below, etc.

Prof Mitts used a linear regression model applied to trading data for July 2019 to derive ‘expected’ sell-side cancellation volumes for each pricing band by “volume of

Mitts 1 included dramatic-looking graphs of cumulative atypical sell-side order cancellation volumes derived in that way (Prof Mitts’ Figures 7 and 8). That cumulative volume appears to be insignificant until after Tweet 1 at 1.30 pm on 6 August 2019, in fact until the share price starts to drop rapidly after 2.00 pm that day, after which it builds through the rest of that trading day, then again throughout the day on 7 August 2019 (both before and after Tweet 2).

Finally, for each trading minute and each pricing band, Prof Mitts subtracted his atypical buy-side order cancellation volume from his atypical sell-side order cancellation volume to give a net atypical order cancellation volume. Mitts 1 then included cumulative volume graphs for that net measure for 6 and 7 August (Prof Mitts’ Figures 9 and 10). They show a significant cumulative net sell-side volume building through the morning and into the afternoon on 6 August 2019 until when the share price began to drop rapidly, at which point the cumulative net sell-side volume initially drops as the price drops but then from c.2.45 pm or so builds rapidly again until just after 3.00 pm, falls back sharply again very briefly, then builds again until the close of trading. On 7 August 2019, a cumulative net sell-side volume builds and builds all through the day.

It is evident visually from Figures 9 and 10, but not quantified in Mitts 1, that the very substantial majority of that build-up of cumulative net sell-side volume is in the first pricing band above best offer (i.e. above, but not more than 0.25% above, best offer). I add for completeness that Prof Mitts gave overall totals for this net sell-side imbalance of just over 1 million shares on 6 August 2019 and c.10.4 million shares on 7 August 2019. It is not clear to me how those figures fit against Prof Mitts’ Figures 9 and 10, as Figure 9 shows the cumulative net sell-side imbalance building to what looks like c.3.75 million by the end of 6 August, and Figure 10 shows the cumulative net sell-side imbalance for 7 August hitting 10 million by what looks like c.10:15 am and then heading ‘off the chart’, as indeed Mr Dhillon QC put it, because Figure 10 is shown with a vertical axis cut off at the 10 million mark. Also, I should note that whilst these totals are large numbers, c.1 million and c.10.4 million, there were in total, as mentioned above, some 12.85 million and 91.77 million sell-side cancellations respectively, in comparison to which whether Prof Mitts’ atypical cancellation volume imbalance totals have any real statistical significance is not addressed.

So far, so good, but the question is ‘so what?’. Prof Mitts opined that his net sell-side atypical cancellation volume “is inconsistent with ordinary market trading”. That conclusion was not asserted by reference to any empirical assessment, given the lack

Put another way, all that this quantitative analysis tells the court is that on 6 and 7 August 2019, as the Burford share price suffered so dramatically: order cancellation volume exceeded a statistical measure of expected cancellation volume (given the order volume); it did so to a much greater extent on the sell-side than on the buy-side; and how that phenomenon was distributed by reference to the pricing bands used by Prof Mitts. Prof Mitts’ argument was that “This evidence strengthens the inference that Burford’s share price was subject to market manipulation [(a)] in the form of spoofing and layering [(b)] because the volume of abnormal sell-side order cancellations deeper in the order book far exceeded the volume of abnormal buy-side order cancellations” (‘(a)’ and ‘(b)’ added for reference).

In Prof Mitts’ use of the terms spoofing and layering (paragraphs 75-76 above), spoofing does not involve sell-side order cancellations above best offer. Therefore, on Prof Mitts’ terms, he should have referred only to layering in (a). His contention, then, at (b), was that there being a larger than expected volume of sell-side order cancellations above best offer to a greater extent than there was a larger than expected volume of buy-side order cancellations below best bid supports a notion that the share price was being attacked by layering. Prof Mitts’ argument was that:

There is an obvious and basic flaw in that. The logic for Prof Mitts’ premise (paragraph 117(i)-117(ii) above) was that a surge in sell-side supply manifests as a surge in buy-side executions, and a surge in buy-side executions ought to lead to (net) buy-side cancellations by market makers because they will lower their buy-side orders. But the positionally neutral market maker posts matching sell-side and buyside orders; if one is adjusted, so is the other, and there is no net buy-side (or sellside) cancellation; and if it is the market maker’s own buy-side order that trades, its matched sell-side order will be cancelled so, if anything, a surge in sell-side supply manifesting in buy-side executions would mean that market makers contributed to the net sell-side cancellation imbalance.

To be clear, I do not mean by that to find or to suggest that ordinary market making activity by positionally neutral providers of liquidity explains the large overall net sell-side atypical cancellation imbalance highlighted by Prof Mitts. But that brings me back to the basic point. Correcting for positionally neutral market-making activity in calculating that imbalance, or finding that it does not explain it, is not the same thing as finding that there is a serious likelihood that the imbalance resulted from manipulative behaviour. It means only that it resulted from the activity of parties who were not positionally neutral. I cannot say there is a seriously arguable case that the activity in question was manipulative as claimed by Burford.

What has happened in this case is, in my judgment, clear. The anonymity (in general) of the data on which Prof Mitts has worked caused Burford to think of itself as a possible Norwich Pharmacal claimant. But, as Mr Smith put it, such anonymity “is also precisely why public analysis and allegations of market manipulation can be flawed. That is primarily and simply because public participants do not know the trading positional data associated with the orders they observe. This leads frequently to erroneous accusations of manipulation, when in fact the orders are genuine – as is the case here. Errors such as failures to account for algorithm behaviour; lack of understanding of execution algorithm strategies; or a general lack of knowledge related to order or trade types are commonly made by market participants alleging spoofing and layering.”

Mr Green QC went so far as to submit that Prof Mitts was “out of his depth” in this case precisely because his relevant experience has been with data that, so far as material, is not anonymised, so that he is looking for, and at, patterns in the stock market activity of parties whose motivations for their activity he can sensibly assist a court or regulator to assess. Without necessarily endorsing Mr Green’s turn of phrase, the point he makes is well made in substance. In my judgment, it is a fair and accurate characterisation of Mitts 1 to say this: it contained data analysis that, so far as it goes, may well be right; however, it made claims for what that data analysis shows that are obviously unjustified; in particular, it repeatedly (a) described with care, when read closely, what exactly the data analysis involved, but (b) asserted conclusions that cannot properly be drawn from that data analysis if attention is paid to that initial description.

Burford founds its case that there was or may have been market manipulation on statistics, and its undoing is the truth in the old adage that there are ‘lies, damned lies, and statistics’. That truth is not that one cannot trust carefully defined, accurately calculated statistical propositions, but that any truth illuminated by a correctly calculated statistical proposition is strictly confined by a careful definition of what has been calculated.

Burford’s and Prof Mitts’ inability to see fully what went on, because the data available to them is incomplete and in general anonymised, may mean they feel they cannot independently rule out all possibility of market manipulation, i.e. independently of what they are told by the Stock Exchange and/or the FCA. But that is as far as any of Prof Mitts’ analysis, carefully considered, takes Burford. Carefully considering Prof Mitts’ reports to see if they really provide any serious support for a market manipulation claim is not to attempt to conduct a trial of such a claim without disclosure and cross-examination when this is ‘only’ a Norwich Pharmacal application. It is merely to insist that it is not enough, to say that there is serious cause

I find that it does not, and that is to say that the data analysis simply does not support the claims Prof Mitts made, and therefore Burford made to found this Claim, that there is strong evidence of spoofing or layering in this case. In truth, there is no substantial evidence of that at all; Prof Mitts’ data analysis leaves Burford, and the court, in the dark as to whether spoofing or layering might really have occurred; the case that it did is speculative. It has taken me some time to explain why that is so because of the complexity of the subject matter, not because, once that complexity has been taken on board, it takes more than a careful first look at the claims made to find them wanting.

Burford’s unwillingness to rest content with the Stock Exchange’s and the FCA’s conclusions or reassurances is not a substitute for a seriously arguable case that there has been wrongdoing, and is not a sufficient basis upon which to consider granting Norwich Pharmacal relief. Prof Mitts said in Mitts 2 that “… it is impossible to determine with certainty whether an order is manipulative based on the publicly available data that I have reviewed. The determination as to whether a given order is likely to be manipulative requires a careful examination of the overall course of conduct of a given trader, whether there are indications that the given trader is attempting to execute the order, and so forth. This can only be done by evaluating the deanyonymised evidence.” (emphasis added). I agree, unless Prof Mitts intended by the words I emphasised to suggest that a provisional conclusion that orders were or may have been manipulative can be reached without the careful examination and evaluation to which he refers.

In my judgment, the fully true statement is that it is impossible to determine from the data Prof Mitts reviewed whether any Share Order Events were or may have been manipulative; and any claim that there was market manipulation on 6 or 7 August 2019 by reference to Prof Mitts’ data analysis is speculative.

In Mitts 2, picking up on some observations in Mr Smith’s evidence about possible indicators of market manipulation, Prof Mitts reported finding in the order data examples of “order amendments which did not fully execute, were not executable quotes, remained within 10 levels of the best offer, and were posted “close to (but not at) the best price level” but “amend[ed] the price away [as] the best offer move[d] towards the order”” (the internal quotations being Prof Mitts quoting Mr Smith). Prof Mitts purported to agree with Mr Smith that if that pattern were found, “there is a high likelihood that the order flow was manipulative”, but Mr Smith did not say that, and listed numerous examples from the data of the pattern just described (Mitts 2, Appendix B). In addition, Prof Mitts provided a long list of examples where sell-side orders were amended down towards the falling price as best offer moved away from the order (Mitts 2, Appendix C), noting that that is an amendment pattern that can be created by manipulative behaviour.

Across Appendices B and C there were 621 Order ID examples: Appendix B listed 40 Order IDs for 6 August 2019, 129 for 7 August 2019; Appendix C had 80 and 372 respectively.

This represented a new case, if the suggestion was that although Mitts 1 did not provide a good arguable case of spoofing or layering, the court should find such a case on the basis of these Appendices in Mitts 2. That indeed was Mr Dhillon QC’s submission, in reply, if, contrary to his primary argument but as I have concluded, Mitts 1 gave Burford no proper basis for this Claim. The Stock Exchange was not entitled to reply to Mitts 2 and there was no directly responsive evidence from Mr Smith.

That led to some exchanges in argument over what might have happened if the Stock

As I indicated earlier in this judgment, with (I acknowledge) the benefit of hindsight, it would have been better if Burford had provided a more detailed pleading setting out its allegation to the effect that spoofing or layering may have occurred. Having not done so, but rather pleaded that its case was the case in the evidence it served with the Claim Form, and that case in turn being whatever case could be made out of Mitts 1, that is the only case it is fair to the Stock Exchange to consider. That case has failed.

I could not in any event find on the basis of Mitts 2, Appendices B and C, that there was a good arguable case of spoofing or layering. The claims made for them by Prof Mitts suffer the same error as various of the claims he made in Mitts 1. What he has reported finding are examples of patterns in the order data that sellers, if behaving manipulatively in certain ways, could have created. However, whether the sellers involved were behaving manipulatively is a matter of speculation on the anonymised data. In truth, neither Prof Mitts nor Burford has any idea one way or the other whether, with access to de-anonymised data for the Appendix B or Appendix C trades, any seriously arguable allegation of spoofing or layering would emerge. The track record of Prof Mitts’ best candidates to examine (Mitts 1, Tables 2, 3, 4 and 5) is unpromising for Burford. To found a case upon these Appendices is, to my mind, to attempt to use this Norwich Pharmacal claim as a fishing expedition to establish whether or not Burford has a seriously arguable case in precisely the manner deprecated by Flaux J in Ramilos (see paragraph 23 above).

In my judgment, Mitts 2, Appendices B and C, are at best for Burford lists of Order IDs that Prof Mitts might say could usefully be looked at carefully to double-check that nothing improper occurred, but neither he nor Burford has any basis for proposing that anything improper may in fact have occurred. That does not come close to justifying intervention by the court under the Norwich Pharmacal jurisdiction.

The key causation claim by Prof Mitts was that an association he can measure in the data, between very high intra-minute net sell-side cancellation rates and particularly sharp intra-minute falls in the share price, is to be explained by saying that the latter resulted from the former, rather than that the former was triggered by the latter, or that there was no causal link in either direction between the two. Prof Mitts put it this way, in Mitts 2: “There is compelling evidence that order flow unaccompanied by trading caused the Burford share price decline. Non-execution order flow induced a downward revision in quotes …, with no shares changing hands. Only thereafter did traders sell at these lower, quoted prices. These sales often occurred at a price plateau, i.e., the share price remained unchanged after the sales. At times, sale transactions were followed by price increases.”

Care is needed over what is meant here by any statement of cause and effect. Taking Prof Mitts’ statement at face value, it is consistent with a lack of demand triggering sell-side cancellations, as sellers either lowered their offers (Sequence 1 or Sequence 2, each counting in the data as a cancellation) or simply withdrew (being unwilling to go lower), in turn resulting in the next traded price showing a sharper drop (from the immediately prior traded price) than might otherwise have been the case. But none of that would involve any manipulation such as is alleged by Burford – false signalling to the market – and the cause of the sharp intra-minute price drop would simply be the acute lack of demand, the high net sell-side cancellation rate being itself caused thereby. Nor would manipulation be required to explain some element of apparent temporary ‘bounce-back’ immediately thereafter.

The association observed by Prof Mitts (paragraph 134 above) is therefore uninformative as to whether manipulation occurred. I can see that if there were strong evidence of manipulative Share Order Events, that association might mean it could not be ruled out that, if in due course manipulation were finally established, it might also be shown that some individual sellers’ traded prices were lower than they would otherwise have been, due to the manipulation. I am not sure that would amount to a good arguable case of causative manipulation, let alone of manipulation causing Burford any loss. But that is all hypothetical, as in my judgment there is no strong evidence of manipulation in the first place.

The main other points arising out of the evidence concern market depth and Millennium Pentagon, and I can take them quite briefly.

Mr Smith’s view was that if spoofing and layering of the kind, and scale, proposed by Burford’s claim had been experienced, it would have needed to involve spoof orders making up over 50% of the top ten price levels. Prof Mitts’ view, in Mitts 2, was that manipulation involving greatly lesser market depth can be effective. I do not think those two views are inconsistent with each other; and Mr Smith’s view certainly does not justify suggestions that were made that the Stock Exchange does not understand

Millennium Pentagon is the Stock Exchange’s proprietary, automated market abuse surveillance system. It includes specific algorithms, also proprietary to the Stock Exchange, for detecting spoofing and layering. Mr Smith gave evidence too of the human resources deployed, on a daily basis, upon the monitoring of activity on its markets, assisted by Pentagon, specifically to detect and investigate possible signs of abuse. Mr Smith expressed himself confident that spoofing or layering as alleged by Burford would have been detected, which led to unrealistic submissions by Burford, and evidence adduced by it from Mr Wim Nelen (an expert with experience in the regulation of financial markets), that because the Stock Exchange had not provided extensive technical detail on the programming of Pentagon, the court should doubt its adequacy.

This Claim created no demand for the Stock Exchange to open up to Burford, or to the court, the confidential detail of its market abuse detection systems. Mr Smith’s evidence, together with that of Mr Denzil Jenkins, Group Head of Compliance at the Stock Exchange, provide general reassurance that the Stock Exchange is well aware of the types of manipulation discussed by Prof Mitts and that Pentagon, and relevant human resources, are targeted at it and are reasonably expected to spot it if it occurs.

Ultimately, however, this part of the evidence is neutral on this trial. If without reference to it Burford had made out a good arguable case that there was unlawful market manipulation, then I do not think the general evidence about the Stock Exchange’s detection systems would have done enough to negative that. Burford having not made out any such case, however, it cannot sensibly ask the court to say that there is any significance in the fact that the Stock Exchange did not, in real time, detect any abuse.

Having mentioned Mr Nelen in passing, I would add that he was asked, or at any rate chose, to pass comment as if he were the court on the credibility and weight of Prof Mitts’ and Mr Smith’s evidence, both generally and including as to whether Mr Smith (or the Stock Exchange) should be giving the court full technical detail on the workings of Pentagon. That was most unwelcome evidence, and it is not to Mr Nelen’s credit that he found nothing to criticise in Prof Mitts’ opinions.

Despite the vigour with which, from the off and throughout, Burford has pursued a desire to say that its share price was the victim of unlawful market manipulation on 6 and 7 August 2019, there is no good arguable case on the evidence that any such unlawful conduct occurred.

The truth or falsity, and if false the actionability (if any), of the very negative claims about Burford that Muddy Waters published as part of what is likely to have been, for it, a successful short-selling attack or ‘bear raid’ on the Burford share price, has not fallen for consideration on this Claim. What has fallen for consideration is what strength of case there is for a proposition that the Share Order Event activity in

In my judgment, there is no strong case for supposing that spoofing or layering occurred. On a careful assessment of the evidence put forward by Burford, the case that spoofing or layering occurred, or may have occurred, is speculative. That case was argued earnestly, but I am not persuaded it is right to regard it as capable of serious argument. There is a gulf between the conclusions that Burford’s case has to advance by reference to Prof Mitts’ data analysis and the conclusions that analysis, carefully considered, is capable of supporting.

In those circumstances, this Claim, pursuing Norwich Pharmacal relief against the Stock Exchange, fails and will be dismissed.

The legislative scheme that in R (Omar), supra, was held so to occupy the territory that it would be inconsistent for the court to exercise the Norwich Pharmacal jurisdiction was the Crime (International Co-operation) Act 2003. This again illustrates the point I made at the outset that the purpose for which information is sought, by the Norwich Pharmacal claimant, is of the essence of the claim. The claimant in R (Omar) sought Norwich Pharmacal relief so as to obtain evidence for use in foreign criminal proceedings. But by the 2003 Act, Parliament had created a statutory scheme for precisely that purpose, i.e. the obtaining of evidence by someone in the position of the claimants in that case for use in foreign criminal proceedings. That scheme was different in several respects from the common law jurisdiction under Norwich Pharmacal and (as accurately summarised in the headnote at [2014] QB 113A-B): “those differences were so substantial that Parliament was to be taken, in enacting the 2003 Act, to have created an exclusive procedure, not a parallel one; that, therefore, where the statutory regime was in play, the Norwich Pharmacal remedy did not run”. That ratio was then applied to the Evidence (Proceedings in Other Jurisdictions) Act 1975 in Ramilos, supra, for the conclusion that the Norwich Pharmacal jurisdiction was ousted as regards the obtaining from a third party of evidence for use in foreign civil proceedings.

There is nothing in play here akin to the statutory scheme under the Crime (International Co-operation) Act 2003 or the Evidence (Proceedings in Other Jurisdictions) Act 1975. The regulatory scheme under MAR accords a critical role to national regulators – the FCA in this jurisdiction. Mr de la Mare QC built upon that the following submission: “there is no case … where Norwich Pharmacal relief has been sought against a regulator or a … regulated body that is specifically charged with monitoring for and investigating the very wrongdoing in question. Still less is there such a case in circumstances where those bodies have looked at the matter in depth, using all the materials available to them in their special capacity, and declined to exercise their powers on the basis that there is no reasonable suspicion. The basic problem is that a Norwich Pharmacal application is, however you look at it, in those circumstances a collateral attack on the expert regulatory judgments of [those bodies]”.

That is an important argument, when considering the balancing or competing factors to assess whether justice requires that the Stock Exchange (being the ‘regulated body’

Suppose an investor in AIM-listed shares were able to say, unlike Burford, by reference to evidence available to it, that (a) a particular sequence of trades, if a single party was responsible for it, was very probably market manipulation that caused the investor loss, (b) a single party very probably was responsible for it, and (c) the investor wished only to know the identity of that party, so it could issue proceedings against it to pursue a good arguable cause of action in deceit. There is nothing in the MAR generally, or the role of the FCA under it, or in the regulatory obligations imposed on the Stock Exchange because of it as a market operator regulated by the FCA, that seeks to affect such an investor’s entitlement to pursue that claim before this court, assuming the wrongdoer, once identified, could indeed be sued here. One incident under English law of such an entitlement is the Norwich Pharmacal cause of action.

The MAR and the regulatory regime under it does not purport to legislate in the territory of such an investor’s rights (if any) under national law to obtain information to enable it to pursue such a private law claim. I do not accept that there is an argument here for the Stock Exchange, as regards the legislative intent of the European Parliament and the Council of the EU when enacting MAR, that is similar to the argument as to UK Parliamentary intent that prevailed in R (Omar).

Indeed, the counter-example could be extended to a case where, again unlike the present case, the Stock Exchange (and/or the FCA) concurred in the investor’s assessment that it had probably been defrauded by means of market manipulation, but the Stock Exchange was understandably unwilling without a court order to divulge the identity information the investor needed to pursue its rights. If the MAR regulatory scheme ousted the Norwich Pharmacal jurisdiction, there would be no basis for the making of the court order needed.

Mr de la Mare QC submitted that, “it is quite impossible to think that this is a situation in which the parallel application of the Norwich Pharmacal jurisdiction is going to do anything other than cause unacceptable and unintended friction with the statutory regime. Indeed, it is really hard to think of a better case than the present that illustrates the sorts of friction that will be created.” But again, on analysis, any strength that argument has derives from the fact that the Stock Exchange and the FCA, with access to trading data that is not anonymised, and more besides that is not available to Burford, both say that upon a full and careful review they see no evidence of unlawful conduct. An argument that the Norwich Pharmacal jurisdiction is statutorily ousted, absent an express legislative provision to that effect, has to work without reference to such a fact-specific assessment. Thus, this line of defence does not in my judgment go to the existence of the jurisdiction but whether it should be exercised in this case.

Putting the point another way, I disagree with Mr de la Mare QC’s submission that R (Omar) and Ramilos were decided “in the particular context of exclusion by reference to parallel regimes of evidence provision for foreign courts, but … based uponreasoning of wider principle.” (emphasis added). They were cases of statutory ouster because, for evidence sought for foreign criminal or civil proceedings (respectively), the legislature had occupied the very territory that would otherwise be occupied by the common law remedy, viz. the Norwich Pharmacal cause of action, and had done so in a way that was inconsistent with the notion that the common law remedy was intended to persist. Thus, the particular context of parallel regimes of evidence provision was of the essence of the decisions that the common law regime did not survive. Other Factors

I turn finally to the factors weighing in favour of, or against, the grant of relief, had there been a good arguable case of wrongdoing, or (as I have proposed might be a simpler analysis) that go to whether justice would have required of the Stock Exchange that it provide the assistance sought in respect of that good arguable case because it was mixed up in so as to have facilitated the wrongdoing if wrongdoing did occur. The list in paragraph 44 above, taken from Lord Kerr’s judgment in the Rugby Football Union case, serves as a sufficient template; and I shall take in the additional elements raised by this case (paragraph 45 above) within that template, not as independent points.

Merits

The strength of the case that there was wrongdoing here I have already dealt with, concluding against Burford that there is an insufficient showing to warrant relief. The consideration of Lord Kerr’s other factors assumes, to the contrary, that I had found there to be a good arguable case. Had that been the finding, on any view it would not have been by any substantial margin.

Vindication

The public interest in allowing Burford as claimant to vindicate its legal rights would weigh in favour of granting relief, if the alleged wrongdoing would vest in Burford a personal, civil cause of action against the wrongdoer(s). There is no reason not to take at face value Burford’s avowed intention to pursue any such cause of action by ordinary civil proceedings in the English court, if it had a serious claim to bring. The stronger the merits (beyond having passed the good arguable case threshold), the weightier that public interest. The Stock Exchange submitted, however, that Burford would not have any cause of action against the alleged wrongdoers upon the wrongdoing alleged.

Burford contended that the market manipulation alleged would give rise to one or more of the following causes of action:

Taking the possible conspiracy claim first, there is, apart from one point, no evidence to support the notion that there was any unlawful combination. If it be supposed that there were manipulative Share Order Events, it could not be thought impossible that there might have been some coordination. But that is as far as it goes, meaning the conspiracy case is entirely speculative, along these lines: if, which cannot be entirely ruled out, there was coordination, then it might have amounted to an unlawful combination.

The one exceptional point is that Prof Mitts identified, from the order data, activity from 8.46 am to 8.51 am on 7 August 2019, before Tweet 2 at 8.53 am and the Muddy Waters website clue a minute earlier that Burford was indeed the target, that could perhaps suggest advance knowledge of the content (and likely effect) of Tweet 2. This, it was said, suggests that any manipulation was coordinated with Muddy Waters. Understandably, Burford would prefer, if it were to pursue a conspiracy claim, to sue all alleged conspirators together. But that does not require that they all be named at the outset.

If there were proper grounds for a conspiracy claim, but as things stand Muddy Waters would be the only alleged conspirator Burford could name, proceedings against Muddy Waters alone, in the first instance, to which additional parties might be added later, would provide a sufficient, just means for the vindication of Burford’s relevant legal rights. One can envisage the possibility, depending on what information Burford obtained from Muddy Waters, and when, through the bringing of such proceedings, that it might yet be necessary to have assistance from the Stock Exchange to bottom out who any (alleged) co-conspirators were. But that could all be managed within the overarching process of a live claim against Muddy Waters. In short, this Norwich Pharmacal claim, so far as it concerns any possible conspiracy claim by Burford, is either bad or premature. The ends of justice do not insist, on account of that possible claim, that the Stock Exchange lend its aid to Burford as now demanded.

There is in any event an insuperable difficulty for the suggested conspiracy claim. There is no basis for proposing that the target of any combination between manipulators, or between manipulators and Muddy Waters, might have gone beyond the Burford share price. By manipulation of the kind posited here by Burford, manipulators intend to profit, at the expense of other investors, from the share price movement induced or aggravated by the manipulative activity. Even if (which may be doubtful here anyway) they could foresee some knock-on consequence harmful to Burford as issuer, that would not found an allegation that Burford was a victim of the conspiracy with a right to sue upon it: see Emerald Supplies Ltd et al. v British Airways plc [2015] EWCA Civ 1024 at [133]-[152], [167]-[174].

Coordinated manipulators could not defend a conspiracy allegation by saying that because their motive was profit they did not intend to harm their investor victims, but that is because the harm to those victims is the flip-side of the intended profit. No loss on the part of Burford of that kind is suggested. The only suggestion of loss came in Mr East’s reply statement (“East 3”) and was that Christopher Bogart, Burford’s CEO, had told Mr East that the 6-7 August 2019 fall in Burford’s share price caused Burford loss in that “(1) the share price decline of that magnitude has eroded investor confidence, reduced the ability of the company to access the equity markets, and has led to massive amounts of management time being incurred to address the short attack and its aftermath; (2) there has likely been an increase in Burford’s cost of debt given the trading performance of its bonds since the short attack; (3) Burford’s employees’ unhappiness with the share price has required Burford to provide further compensation to those employees beyond their shares; and (4) Burford has incurred legal costs as a result of lawsuits prompted by the fall in Burford’s share price.”

How real any of that ever might have been, as loss allegedly caused by unlawful market manipulation (as opposed to the Muddy Waters short attack generally), seems to me open to question. How real or significant any of it might be now, after the unrelated events of Q1 2020, must be highly speculative (and I note that in the 8 months between the short attack and the trial before me, Burford provided no quantification or other particulars, and no supporting evidence). But come what may, it would not be loss inseparably linked to the manipulators’ intended profit such as to found a basis for alleging the intention to injure Burford that would be required to avoid the striking out of a conspiracy claim.

That also means there could be no tort claim by Burford against manipulators for unlawful interference with its economic interests. A suggestion that there might be such a claim was made in Burford’s skeleton, but not developed at all orally, in recognition (as I saw it) that for present purposes it could not add to the suggestion that there might be a conspiracy claim. Burford’s skeleton also mentioned malicious falsehood, but again that suggestion was not developed at the hearing, and I agree with the Stock Exchange’s response, namely that if Muddy Waters’ negative review of Burford involved malicious falsehood, that might be actionable but would not found this Norwich Pharmacal claim as Burford knows who Muddy Waters is and could just get on and sue, whereas it is impossible to spell any malicious falsehood claim out of spoofing or layering, if it occurred.

For the public interest in allowing Burford as claimant to vindicate its legal rights by taking action against alleged manipulators to be of any significance, therefore, it could only be if the type of market manipulation alleged by Burford would be actionable by it as a ‘euro-tort’, in the manner recognised in the Muñoz case, supra.

An immediate curiosity about the Muñoz argument here is that, as I have already mentioned, the predecessor to MAR, the Market Abuse Directive, had to be implemented by national legislation in Member States, and that was done here within FSMA 2000. In Hall v Cable and Wireless plc, [2009] EWHC 1793 (Comm), [2011] BCC 542, Teare J held that no personal right of action in tort (for breach of statutory

In particular, Teare J considered, and I agree, that under FSMA, enacting the Market Abuse Directive, the existence and detail of the regime of remedies in favour of private parties affected by market abuse contrary to the Directive was inconsistent with the notion that the statutory wrong created an independent, private cause of action in such parties. I find it difficult to envisage that a different result could be arrived at under MAR and Muñoz. The conclusion essential to Muñoz was that a direct civil claim by a fruit trader against a competitor, to enforce its rights under Council Regulation (EC) No.2200/96 and its predecessor, Council Regulation (EEC) No.1035/72, not to have competition in the fruit trade from seedless grapes not conforming to the EC and, previously, EEC standards set by those Regulations, was necessary to give full practical effect to the obligation under EC law, previously EEC law, to comply with those standards. The remedies provisions of MAR, and the exclusive role of national regulators (here, the FCA) in relation to them, seem to me a very significant obstacle for any such conclusion here.

Therefore, I would not have accepted Mr Dhillon QC’s primary submission that there is a good arguable case (in the sense used in the Norwich Pharmacal jurisdiction) that Article 15 of MAR gives rise to a Muñoz euro-tort. His alternative submission was that the court could not say it was acte clair under EU law that it does not, and that so long as there would be an obligation on the UK to refer the question to the CJEU, under EU law rules still applicable during the post-Brexit transition period, because it was not acte clair, then on this Norwich Pharmacal claim the court had to, or ought to, treat Burford’s desire to pursue a Muñoz euro-tort claim against manipulators, if there were any, as sufficient to bring into play the public interest in allowing claimants to vindicate their legal rights that I am examining.

I disagree with that alternative argument. The Muñoz point is one of pure law, concerning the application of the approach taken by the ECJ in that case to the provisions of MAR. In my judgment, the right way to deal with the point, if it would have determined the outcome of this Claim, would have been to treat it as arising for decision now, and to consider questions of referral to the CJEU, including any contention that the point was acte clair so as not to require referral, in the same way as in any other case where a decisive question of EU law arises. In circumstances, however, where on the evidence there is not a good arguable case that any wrongdoing occurred, I prefer to say no more than I already have about the merits of the Muñoz argument.

The consequence of the foregoing is that even if, contrary to my finding, there were a good arguable case of wrongdoing, there would not be a good arguable case that Burford had any cause of action for it.

That brings me to the question of the prosecution of crime or the stimulation of regulatory action. I assume in Burford’s favour, without deciding, that the desire of a

To take a simple example, a party claiming to have been the victim of a telephone scam, having reported the alleged crime to the police, would have a range of remedies other than a Norwich Pharmacal claim, in the criminal and public law fields, to ensure they were properly informed about the police investigation or could pursue any complaint they might have if they felt it was not being pursued vigorously or competently. Those remedies would exist in the context of the full range of investigatory powers available to the police. A Norwich Pharmacal application against the telephone company, asserting that it was in a position to provide information on the identity of the scammers, if not sought for the purpose of bringing a civil fraud claim against them but for the purpose of the criminal investigation or possible future prosecution, would require, it seems to me, exceptional justification founded upon specific features of the particular facts, for which no obvious ‘for instance’ springs to mind.

In my view, it would not affect that conclusion to note that the victim could bring a private prosecution (if they knew the identity of the scammers). It would be unacceptable to use a Norwich Pharmacal application in substance as a collateral means of pursuing a complaint that the alleged crime had been insufficiently or ineptly dealt with; and if there were no such complaint, yet also no (public) prosecution it is difficult to see how it could ever be said that justice demanded the disclosure by the telephone company of otherwise confidential customer information so a prosecution could be considered or pursued. Indeed, ex hypothesi, if a properly considered decision not to prosecute had been made after a competent police investigation, it is difficult to envisage circumstances in which a private prosecution would not simply be taken over by the DPP, acting by the CPS, and then stopped.

In my judgment, the present case in fact exemplifies all of the above, and the conclusion is the same. The FCA has the exclusive statutory function under FSMA and MAR of investigating and deciding whether to prosecute market abuse contrary to Article 15 of MAR. No private prosecution may be brought. Using the Fraud Act 2006 for a private prosecution, essentially to get round that restriction, strikes me as problematic, at all events where, as here (so far as material), there is no possible Fraud Act offence unless there is a specific market abuse offence under FSMA via Article 15 of MAR. Indeed, there would be a very real question whether, in a public prosecution, a charge under the generic provisions of the 2006 Act would be permitted to run to trial given the availability of the specialist offence. I do not need to take a final view on points of that kind, however. On this Claim, Burford must persuade the court that, because the Stock Exchange would have been mixed up in so

East 3 said (as to fact) that Burford will indeed consider a private prosecution under the Fraud Act 2006, if it can obtain data sufficient to identify the manipulators of its shares, and argued that to refuse Norwich Pharmacal relief would be to pre-empt any decision by the DPP about intervening to stop such a prosecution. As I have just indicated, that misses the point that Burford needs to show affirmatively that justice cries out for it to be given confidential information in the Stock Exchange’s possession. The suggested desire to prosecute privately under the Fraud Act 2006, where a private prosecution is possible in principle, rather than under FSMA, where it is not, does not alter the substance.

That substance is that any prosecution would have to charge facts constituting unlawful market manipulation in a case where the relevant public investigative and prosecutorial body, the FCA, has reached a considered conclusion that there is no basis for such a charge. Furthermore, in doing so it has taken a clear and fairly expressed view that Burford is wrong in its assertion that Prof Mitts’ data analysis provides evidence that wrongdoing occurred. Burford does not need Norwich Pharmacal relief to challenge that conclusion, if there are grounds for doing so, through judicial review. If there is nothing wrong with that conclusion, then the claim for Norwich Pharmacal relief is hollow – justice would not, in that case, demand that the Stock Exchange assist Burford by disclosing otherwise confidential information. (It goes without saying in all of this, obviously, that there is no call for Norwich Pharmacal relief to enable Burford to pass the names of possible wrongdoers to the FCA. The full data available to the FCA is not anonymised in the first place.)

That discussion leads naturally to the further purpose possibly said by Burford to result in a just demand that the Stock Exchange provide the information sought, namely to enable Burford to seek to persuade the FCA to a change of view. It is not entirely clear that Burford does seek relief on that basis. East 3 said that “Burford’s dialogue with the FCA remains ongoing and is not at issue in these proceedings” (my emphasis), but immediately went on to say that Burford hopes the information sought, if provided by the Stock Exchange, “coupled with additional analysis from Professor Mitts, will lead to an FCA investigation and prosecution for market manipulation”.

As to that possible basis for the Claim, the problem is that Burford needs nothing from the Stock Exchange in order to bring a judicial review claim, if it wishes to pursue the contention that the FCA should have concluded, or should now conclude, that some further investigation, or a prosecution, should be undertaken. Furthermore, the FCA, as putative judicial review defendant, is in possession of all the information (and more) that Burford seeks from the Stock Exchange. If Burford were to pursue judicial review, the court sitting in that jurisdiction would have ample power to do any necessary justice to deal with the informational asymmetry. (There might be a real issue whether judicial review is now out of the question because of delay; but I express no view about that, as it does not affect the point just made about leaving any judicial review process to sort itself out without there being any call for Norwich Pharmacal relief against the Stock Exchange.)

I deal for completeness with a variant on that theme raised in argument. There is no suggestion that the Stock Exchange, as regulated entity, has failed in any respect in its regulatory duties in dealing with or responding to Burford’s concern that spoofing or layering occurred. Burford is aggrieved by what the FCA made of the information available to it, including full data provided to it by the Stock Exchange and all of Prof Mitts’ data analysis and arguments, not by any failure on the part of the Stock Exchange in providing information to or co-operating with the FCA. If the case had been, or had included, the latter, though, that would not affect the analysis of the current Norwich Pharmacal factor. It would or might mean that it could not be said Burford’s sufficient remedy was judicial review of the FCA’s conclusions; rather its complaint would be to the FCA about the failure of the Stock Exchange to assist the FCA that could in principle lead to a judicial review of the FCA’s handling of or conclusions upon that complaint. So the route is longer and a little more complex, but the destination is the same. The means by which the law provides for the protection of Burford’s relevant interests is, ultimately, the amenability in principle of the decisions of the FCA to judicial review. There is no just call for the court under its Norwich Pharmacal jurisdiction to add anything or interfere. Justice does not demand that the Stock Exchange give up the confidential information demanded when, on the facts of this case, judicial review provides fair and sufficient protection of Burford’s interests as alleged victim of market manipulation with no private law cause of action against the manipulators.

For all those reasons, on this key factor, which is the factor encapsulating the original essence of the Norwich Pharmacal jurisdiction, the strong public interest in the vindication of victims of wrongdoing would not provide any or any substantial weight to the claim in this case that justice required the Stock Exchange to assist Burford by the provision of the Participant Identity details in respect of the Share Order Events, if (contrary to my primary conclusion) there were a good arguable case that spoofing or layering had occurred on 6 or 7 August 2019.

I have not needed to consider, for that conclusion, a particular submission advanced by Mr de la Mare QC upon the public law principle of regularity, namely that “public law acts stand and are to be regarded and relied upon as lawful unless and until quashed as being unlawful by the court”, per Dove J in R (Archway Sheet Metal Works Josif Family Trustees) v Secretary of State for Communities & Local Government et al. [2015] EWHC 794 (Admin), at [44] (and see further the cases quoted and conclusion reached at [45]-[48]). That is because my preferred analysis, set out above, operates at one remove from that principle, namely that, at all events on the facts of this case, the public law arena is sufficiently available for Burford’s vindication, if vindication be called for, that the demands of justice do not extend to the exercise of the exceptional power possessed by this court under the Norwich Pharmacal decision. The principle of regularity invoked by Mr de la Mare is a rule of that arena; but it is the existence of the arena, and its sufficiency for Burford’s relevant purposes, not the existence or consequences of that particular rule within it, that deprive the Norwich Pharmacal claim of merit.

Before moving on, I mention, for completeness only, that Burford made some reference to the possibility of a subsequent application, if this Claim were first to succeed, for permission to pass the information disclosed as a result to any shareholders in Burford (past or present, presumably) who might wish to consider the possibility of bringing a claim alleging manipulation causing them loss. As that summary indicates, this was not advanced, nor could it be, as a purpose sufficient to grant relief now, since it was predicated on the Claim having first succeeded.

East 3 nonetheless exhibited, and I was shown, a joint letter dated 4 February 2020 from the UK Shareholders’ Association and ShareSoc, to Burford, under the names of Peter Parry, UKSA Policy Director, and Paul de Gruchy, a Director of ShareSoc, expressing support for Burford’s application. The letter asserted that there is “a widespread belief among investors that there is little appetite by regulators to investigate allegations of wrongdoing on the AIM market” and referred to a Twitter poll said to have been conducted by a member (whether of UKSA or ShareSoc or both was not said) supposedly suggesting scepticism amongst investors that a belief that there had been suspicious trading activity on the AIM would even be investigated. This was unsatisfactory material:

Deterrence

An order of the court requiring the Stock Exchange to divulge Participant Identity details to an AIM issuer asserting that its shares had been the subject of spoofing or layering, Mr Dhillon QC submitted, would be calculated to deter those who might engage in such unlawful conduct. There is subtlety there, however; and whether it would in practice have any deterrent effect seems to me speculative.

The subtlety is that a conclusion by the court that there was a good arguable case of unlawful market manipulation in a case where the Stock Exchange and the FCA had each, independently, concluded that there was no arguable case at all, could be perceived as an authoritative finding that the systems in place for detecting and/or acting upon possible market abuse are weak. That might be said to be an encouragement, not a deterrent, to those who might be minded to engage in unlawful market manipulation. The court would not be deterred by that possible logic, or influenced by it at all, over whether to find that wrongdoing had occurred (or, in the present context, that there was a good arguable case to that effect) if there was evidence justifying it. However, it would make the question of the possible deterrent effect of granting relief a more complex one than might at first be thought.

In any event, as to the practical impact, it is simply speculative to suppose that intervention by the court under the Norwich Pharmacal jurisdiction here would have any actual effect at all.

Alternative Sources

Other things being equal, this factor would favour the grant of relief, in the sense that the Stock Exchange is amenable to the Norwich Pharmacal remedy and no other party identifiable to Burford is amenable to that remedy or otherwise in a position to provide the Participant Identity details it lacks. But that is all subject to considering again the FCA’s role. The FCA is an alternative source for the information. If (a) there were serious grounds for concluding that it was wrong to reject Prof Mitts’ conclusions as to what his data analysis shows and (b) disclosure of the Participant Identity details were required, in fairness to Burford, the FCA could be ordered to give that disclosure within proceedings for a judicial review that ex hypothesi Burford would be able to pursue.

My conclusion upon this factor, therefore, is a mixed one. The point just made concerning the FCA means that it is wrong to view the Stock Exchange as the only, or primary, source for Burford of the information it seeks. However, to be fair to Burford, that is in substance to repeat, or apply again here, the analysis adopted above in relation to factor (ii), ‘Vindication’. That does not add to its weight when striking the final balance.

Therefore, whilst Burford cannot say it needs the information from the Stock Exchange because there is no other way to get hold of it if it has a proper basis for doing so, equally the Stock Exchange cannot say additionally or separately to the more general argument about the regulatory regime that Burford does not need Norwich Pharmacal relief because it has other sources to tap. As a result, in my judgment, this factor does not add material weight on either side of the scales.

Defendant’s Fault

There is no suggestion at all that, had there been spoofing or layering, the Stock Exchange might have been in any way a joint wrongdoer. By reference to Burford’s criticisms of the Stock Exchange’s systems and evidence, it was submitted, however, that the spoofing or layering Burford alleges may have occurred should have been, but was not, detected, so that there was a case for saying that the Stock Exchange ought to have appreciated that it was facilitating wrongdoing.

In a sense, the Stock Exchange agreed. Its evidence was that, given the nature and scale of layering (in Prof Mitts’ usage) asserted by Burford in the case upon the basis of which it sought Norwich Pharmacal relief, that unlawful conduct would in fact have been detected, in real time, and indeed that the Stock Exchange’s monitoring and abuse detecting systems were designed and intended to pick up inter alia the sorts of abuse said to have occurred here. Having found without reference to that evidence that there was no serious case before the court that layering had occurred, it reinforces that conclusion.

Furthermore, as I said at the outset (paragraph 31 above) I saw no foundation for Burford’s more extreme suggestions concerning the Stock Exchange’s ability or desire to spot and stamp on abuse if and as it occurs. To the extent that Burford sought to pray in aid this current factor of ‘Defendant’s Fault’ upon the basis of those suggestions, the attempt fails.

All that said, in this section of this judgment I am assuming, contrary to my primary conclusion, that there is a good arguable case that layering occurred on a significant scale, negatively affecting Burford’s share price, on 6 and 7 August 2019, and that the Stock Exchange did not spot it or any sign of it in real time. Logically, as it seems to me, had that been my conclusion, it would have been difficult not to find that if wrongdoing were ultimately established, it might well then be seen that the Stock Exchange had been to some degree at fault. There is therefore here some support, but limited and rather qualified support, for the grant of relief.

Collateral Damage

This factor favours the refusal of relief. Even on Burford’s case at its highest, had I accepted it, it was recognised that many, in all likelihood most, of those responsible for or interested in the Share Order Events, all of whom would be identified to Burford indiscriminately as to whether there was any case against them individually, will have been wholly innocent. It is a significant and important feature of the AIM market, as operated by the Stock Exchange and regulated by the FCA, that generally trading activity is anonymous. Providing Participant Identity details in this case to a third party other than as part of the regulatory regime would be a serious invasion into

It is not an answer to that concern, as Mr Dhillon QC submitted, that all such innocent parties were ex hypothesi looking to trade Burford shares on 6 or 7 August 2019, and Burford is entitled to know the identities of its shareholders, so that there is a sense in which genuine buyers and sellers of Burford shares on any given trading day must be content for Burford to be aware they were trading. As Mr Dhillon rightly noted and accepted, in a different part of the argument, a simple list of identities is of no use to Burford. Burford does not ask to know (without more) who was trading or apparently trading Burford on 6 or 7 August 2019. It asks to know who was posting each individual Share Order Event so that their positions and trading patterns and strategies can be pored over, and any algorithms used examined (at least indirectly, by inference from trading patterns). That is a major invasion into the sensitive commercial confidentiality of innocent third parties.

There is obviously overlap, at all events on the facts of this case, between this question of inflicting collateral damage upon innocent third parties’ rights or expectations of confidentiality, and the next factor. I find it convenient to focus here on the appropriateness or otherwise of inflicting that initial injury, and to consider when assessing that next factor the question of any ongoing impact, beyond that initial injury, of the confidential nature of the information relating to such parties.

That is not an end to the question of collateral damage, however. There are two further elements to consider.

Firstly, there is the possibility of doing damage to public confidence in the FCA as regulator, an echo of the point I considered in paragraph 186 above. The balance between legitimate concern that it is not in the public interest to be (perceived to be) undermining the FCA and proper concern that the court be fearless to act against serious allegations of wrongdoing, is a delicate one.

That balance is not rendered less delicate because on a Norwich Pharmacal claim the court is not required to reach a final conclusion that there has been wrongdoing. Indeed, that is why the sensitivity arises. Though the court is not finding that wrongdoing occurred, nor even forming a provisional conclusion that it probably occurred (since a good arguable case suffices – see paragraph 33 above), it is being asked to interfere in the normal workings of a regulated market despite the conclusion of the regulator after considering all the material available to the court, and more besides, that those normal workings were not, and do not need to be, disturbed by this case.

The strictly contingent conclusion that if (though the court cannot say this will probably be the position) wrongdoing were ultimately established, it may be that then, looking back, it will be possible to identify that the Stock Exchange, or the FCA, or both, did not do as good a job as they should have, is not much of a counter-weight. As that makes clear, in my judgment this factor, in this case, would only add weight to the demand that the Stock Exchange assist Burford, as alleged victim of wrongdoing, if the claim of wrongdoing passed the good arguable case threshold by a substantial margin. That would not be true of Burford’s claim, even if I were wrong to have found that in fact it fell short of that threshold.

Secondly, the Stock Exchange relied on evidence from Mr Jenkins, and from Miles Celic, Chief Executive of TheCityUK, to the effect that:

So the asserted concern was that for the court to intervene, as demanded by the Claim, would harm, indeed might do serious harm, to ‘UK plc’. Mr Celic, for example, expressed concern for “productivity and wealth creation for the UK, generated through the financial markets”. Burford contended that this is evidence of a Chicken Licken variety, and that the court should be confident the sky would not fall down.

Burford had no evidence to contradict that of Mr Jenkins, where it consisted of fact, e.g. that an order such as Burford seeks is unprecedent across the EU, but strongly disputed his opinions on the likely impact of an order being granted. Mr Nelen’s career has included three years with responsibility for regulatory compliance at two high-frequency algorithmic trading firms, Maven Securities Ltd and Sun Trading International Ltd, and his opinion was, for example, that although of course algorithmic trading firms can be sensitive about rivals being able to ‘reverse engineer’ their strategies, “they are also generally keen to avoid any suggestions that their strategies involve unlawful activity” and that “the ability of an issuer … to obtain relevant extracts of trading data (including the identity of traders) in the context of investigat[ing] potential market manipulation would [not] lead … HFT and other algorithmic trading firms to avoid trading on UK markets”.

However, Burford was able to contradict in part, or at least qualify, Mr Celic’s evidence of fact. Evidence in reply from Corey Worcester, a US partner in Quinn Emanuel, vouched for the availability of discovery under the US Federal Rules of Civil Procedure in support of a private law right of action in respect of market abuse that exists in the US. There was a response to that evidence from Ashley Ebersole, a US partner in Bryan Cave, that joined issue on whether a US Federal court would grant relief in a materially similar case (as Mr Worcester suggested it would). But that was an assessment about the individual case, and Mr Ebersole did not contradict the availability of relief in principle. There were disputes of detail between Messrs Worcester and Abersole, but they were peripheral and I shall not take up time over them in this judgment.

It was submitted by Mr Dhillon QC that, absent evidence (and there is none) that issuers or investors are put off the New York financial markets by the possibility that an order such as Burford seeks here might be made, depending on the facts, in support of a claimant seeking to establish that unlawful market manipulation had occurred, the Stock Exchange’s doom-laden forecasts for the impact if this Claim were upheld could and should be discounted.

I agree, in substance, with Mr Dhillon QC on that point, but to be clear – a theme persists – that is because of the approach I have taken to the regulatory scheme and the position of the FCA, as indicated above. In my judgment it could be a concern if the court were seen to be trigger-happy to intervene, whereas if it were seen to be willing in principle to intervene but only (if at all) in a rare case, if one arose, where it was tolerably clearly that a powerful claim of wrongdoing had slipped through the regulatory cracks, that concern would carry no weight and indeed, if anything, the court’s approach might provide added reassurance about the robustness of UK markets as attractive venues for lawful activity.

Confidentiality

As I noted when assessing the previous factor, in my judgment the degree of confidentiality of the Participant Identity details – unless all that was provided was an unsorted list of names that would be useless to Burford – is substantial.

Mr Dhillon QC submitted that the confidentiality concerns could be largely overcome because the Norwich Pharmacal claimant here is Burford itself, not a participant or possible participant in the trading market for Burford’s shares. There is thus, he argued, no danger of competitive advantages being lost, or of sensitive information being misused. If necessary, a confidentiality club could be created and the information disclosed relating to Share Order Events that were not ultimately said by Burford to be part of any unlawful manipulative activity could be required not to be disclosed further or used in any proceedings.

No detailed confidentiality club terms were put forward, however, and I doubt that last element would be workable. This factor has to be assessed on the assumption that the Norwich Pharmacal relief sought now would result in proceedings involving allegations of unlawful manipulation against one or more individuals or entities. Whilst it is perhaps possible that each such allegation could be examined satisfactorily without reference to any other party’s market activity, I think it far more likely that either side of the argument, perhaps indeed both sides, would have comparative points that in fairness they ought to be allowed to pursue, i.e. analysis and argument of a compare and contrast nature by reference to the market activity of innocent parties.

That said, however, the sentiment behind this submission for Burford is sound, in my view, and (a) I would expect it to be possible for any comparative points to be made using re-anonymised data, cloaking the identities of parties accepted to be innocent but with the ability, as a result of the Norwich Pharmacal relief, to know for any given Share Order Event which other Share Order Events came from the same source,

In the circumstances, if there had been a powerful case otherwise for the grant of relief, I would have been open to persuasion that ongoing confidentiality concerns (beyond the initial invasion by the court into confidential territory) might be overcome. The burden would have been on Burford, however, to put forward a detailed confidentiality club proposal and to satisfy me that it did enough to tip the final balance in favour of granting Norwich Pharmacal relief on those terms.

Privacy (Article 8)

There was no suggestion that Article 8 ECHR, in this case, materially added to the debate.

Data Protection

I did not understand the Stock Exchange to take any particular point on data protection rights. Nor that I can see do they add anything in this case to the three preceding factors already considered.

Journalistic Sources

The important public interest in maintaining the confidentiality of journalistic sources, as recognised in s.10 of the Contempt of Court Act 1981 and Article 10 ECHR, does not come into play on the facts of this case.

Placing all those factors in the scales – or asking in the round whether, having regard to all of them, justice here requires that the Stock Exchange assist Burford by providing the Participant Identity details demanded by the Claim – on the basis, contrary to my primary conclusion, that Burford had made out a good arguable case that wrongdoing had occurred, my conclusion would still be that relief should not be granted:

In my judgment it would be right to say that the nature and structure of the factors listed by Lord Kerr is such that factor (i), ‘Merits’, and factor (ii), ‘Vindication’, taken together, are by nature a necessary but not sufficient building block. If, other factors aside, the public interest in the vindication of the Norwich Pharmacal claimant, bearing in mind the apparent strength of its claim that it was the victim of wrongdoing, does not seem to require that the Norwich Pharmacal defendant provide the assistance demanded, then really that is that. The other factors, as applicable in and weighed by reference to the particular circumstances of any given case, may mean that it is not right to grant relief, despite it appearing initially from the perspective solely of vindicating the claimant that there might be a just call for assistance from the defendant; or they may confirm or reinforce the justice of saying that the defendant should assist; but I find it hard to envisage that they could make it right to exercise this exceptional jurisdiction if, bearing in mind the apparent strength of the claim of wrongdoing, the public interest in vindicating the claimant would not (other factors aside) justify the imposition upon the defendant of a duty to assist.

That is not a necessary analysis in this case, however, as even if the individual factors are taken to inhabit no particular hierarchy, the scales are in my view heavily tipped against Burford overall.

For the reasons set out above, this Norwich Pharmacal claim fails and is dismissed, because (a) there is no good arguable case that unlawful market manipulation occurred in respect of Burford’s share price on 6 and 7 August 2019, in any event (b) on the facts of this case, justice would not have demanded that the Stock Exchange provide the assistance sought had there been a good arguable case of wrongdoing (or, if this be the preferred analysis still, I would not have considered it just and convenient, in my discretion, to grant relief).