Isabel Dos Santos v Unitel S.A.

This appeal concerns two issues arising out of a Worldwide Freezing Order (“WFO”) made by Bright J against the appellant, Ms dos Santos: (1) the meaning of the “good arguable case” element of the test required for the grant of a freezing injunction and whether (whatever the test) the judge was right to find that the respondent Unitel had a good arguable case; and (2) whether there is a general rule that the costs of a WFO application should be reserved.

Ms dos Santos is an engineer and entrepreneur, and daughter of the former President of Angola, José Eduardo dos Santos. She founded Unitel, Angola’s largest mobile telecoms company, in 1998. She was a director of Unitel from its inception until August 2020; and was also beneficial owner of 25% of the company via Vidatel Limited (a BVI company), until late 2020. Her stake (and various other stakes) have since been appropriated by the Angolan state. The First Defendant, Unitel International Holdings B.V. (“UIH”) was incorporated in 2012 in the Netherlands, and is owned and controlled by Ms dos Santos. Between May 2012 and August 2013, Unitel made a series of loans to UIH amounting to €322,979,711 and US$43 million (the “Facilities”) secured by promissory security. From late 2019/early 2020, UIH stopped paying interest on the Facilities. On the basis of these and other alleged defaults, Unitel gave notice of acceleration on 1 September 2020 and demanded repayment.

On 26 October 2020, Unitel issued proceedings against UIH in this jurisdiction (“the UIH Claim”). Sometime in March/April 2022 (the precise date is disputed), Unitel’s shareholders passed a resolution to make a claim against Ms dos Santos personally (the “Approval Resolution” – dated 4 March 2022, but said by Unitel to have been passed on 6 April 2022). On 3 October 2022, Unitel applied to join Ms dos Santos to the UIH Claim and amend its statements of case and also applied on notice for a WFO against her. On 16 January 2023, it applied for permission to serve the proceedings out of the jurisdiction. The joinder, amendment and service applications were granted (in part) by HHJ Pelling KC on 25 May 2023. The claim against Ms dos Santos was then served on 1 June 2023. Permission to appeal against the order of HHJ Pelling KC was refused by Males LJ on 4 October 2023.

It was agreed between the solicitors that the WFO application would be listed at least three months after the hearing of the joinder application. The WFO application was heard by Bright J on 29 and 30 November 2023. On 20 December 2023, the judge handed down judgment granting the WFO application and ordered that the costs of the application be paid by Ms dos Santos with a payment on account. He refused permission to appeal, but permission was granted by Arnold LJ on 12 March 2024 on the basis that both grounds of appeal raise important issues of law on which there has been divergence among High Court judges, and that the appellant has a real prospect of success.

Having set out details of the parties, the claim and the procedural background, the judge explained at [21] to [24] that Ms dos Santos’ assets are already affected by several other freezing orders obtained in other jurisdictions by the Angolan state, by PT Ventures SGPS S.A., a Portuguese company owned by the Angolan state oil company Sonangol, and by Unitel.

The judge then set out the legal principles. At [25] he said that it was common ground that an applicant for a freezing order had to show: (1) a good arguable case on the merits; (2) a real risk, judged objectively, that a future judgment would not be met because of an unjustified dissipation of assets; and (3) that it would be just and convenient in all the circumstances to grant the freezing order.

Much of the dispute on the application related to the first limb. The judge set out the rival arguments. At [27] he summarised Unitel’s case, presented before him by Mr Knowles, that the meaning of “good arguable case” is as explained by Mustill J in Ninemia Maritime Corp v Trave Schiffahrtsgesellschaft GmbH (“The Niedersachsen”) [1983] 2 Lloyd’s Rep 600 at 605, as: “… one which is more than barely capable of serious argument, but not necessarily one which the judge considers would have a better than 50 per cent chance of success.” Mr Knowles acknowledged that some recent cases suggest that the effect of the decision of the Court of Appeal in Lakatamia Shipping Co. Ltd. v Morimoto [2019] EWCA Civ 2203; [2020] 2 All ER (Comm) 359 (at [38] per Haddon-Cave LJ) was to equiparate the “good arguable case” test as applied in the freezing order context to that applied in the context of jurisdiction, most recently in Kaefer Aislamientos SA de CV v AMS Drilling Mexico SA de CV [2019] EWCA Civ 10. However, he submitted that this was not the Court of Appeal’s intention in Morimoto and that Mustill J’s well-known formulation in The Niedersachsen remains good law. Unitel relied on the decision of the Court of Appeal in Kazakhstan Kagazy plc v Arip [2014] EWCA Civ 381; [2014] 1 CLC 451, where Longmore LJ at [25] said that the test in the freezing order context was different from that in the context of jurisdiction; see also at [67] per Elias LJ.

At [28] the judge noted that Ms dos Santos’ case was that Morimoto has indeed changed the law, and that the Niedersachsen approach is no longer appropriate. It was submitted that Haddon-Cave LJ’s judgment in Morimoto at [38] is based entirely on Kaefer, such that the three-limb approach first propounded by Lord Sumption JSC in Brownlie v Four Seasons Holdings Inc [2017] UKSC 80 (and repeated in Goldman Sachs International v Novo Banco SA [2018] UKSC 34) should be adopted. This is the approach which was considered by the Court of Appeal in Kaefer where Green LJ gave guidance about how to apply the test at [72] to [80] and made it clear that it is, essentially, a relative test. The court must try to form a view as to which party has the better of the argument. Only if the court “finds itself simply unable to form a decided conclusion on the evidence before it” does limb (iii) arise, in which event the court can fall back on considering whether there is “a plausible (albeit contested) evidential basis” for the applicant’s case.

The judge noted at [29] the decisions of Edwin Johnson J in Harrington & Charles Trading Co. Ltd. v Mehta [2022] EWHC 2960 (Ch) and of Dias J in Chowgule & Co Pte. Ltd. v Shire [2023] EWHC 2815 (Comm) that the “good arguable case” test in the context of freezing orders is now the same as in the context of jurisdiction. At [30] he noted on the other side the decision of Butcher J (handed down after the hearing before Bright J) in Magomedov v TGP Group Holdings (SBS) LP [2023] EWHC 3134 (Comm), where Butcher J came to the opposite conclusion to that of Edwin Johnson J and decided the test in freezing injunction cases remained that in The Niedersachsen.

He also noted that most of the cases where the Court had had to apply the test in the freezing injunction context since Morimoto had not considered the effect of Haddon-Cave LJ’s judgment and had continued to follow the Mustill J approach. It was only in those three cases that the point had been considered head on.

His conclusion on this part of the case was at [34] to [37]:

“34.

I find it striking that Haddon-Cave LJ’s judgment in… Morimoto makes no reference to Kazakhstan Kagazy plc v Arip (making it unclear whether the views of Longmore and Elias LJJ were cited); and that Haddon-Cave LJ said at [35] that the “good arguable case” test was “not a particularly onerous one” and referred to Gee on Commercial Injunctions (6^th ed., 2016) at §12-026, which endorsed the Mustill J approach.

35.

It therefore seems to me not at all clear that Haddon-Cave LJ intended to have the transformative effect for which Ms dos Santos contends. Notwithstanding Harrington & Charles Trading, I am not aware that many observers have come to this conclusion. Subsequent editions of the White Book and of Gee on Commercial Injunctions (7^th ed., 2020, inc. 1^st supp. 2022) both still endorse the formulation in The ‘Niedersachsen’. Indeed, in its current edition, not only does Gee on Commercial Injunctions §12-033 state at some length that the jurisdictional approach of Brownlie, Goldman Sachs and Kaefer is not applicable in the freezing order context, it positively cites Lakatamia Shipping Co. Ltd. v Morimoto as supporting this conclusion.

36.

My own view is aligned with that of Butcher J; essentially for the reasons that he gives, but especially because of Haddon-Cave LJ’s reference to Gee on Commercial Injunctions in his judgment at [35] – which cannot easily be reconciled with an intention to approve a test different from the one espoused in that text.

37.

However, I cannot help but note that the law is in a confused state, which cries out for a definitive answer from the Court of Appeal. The reality is that Haddon-Cave LJ’s judgment deals with this point briefly, elliptically and ambiguously. This is not at all surprising: the Court of Appeal in that case was not being asked to decide a live point on “good arguable case” on the merits (this was not one of the grounds of appeal – see at [39]). The real focus of the case was all on risk of dissipation. When first instance judges are asked to work out the significance of Haddon-Cave LJ’s judgment in relation to “good arguable case”, we are inevitably drawn into opining as to what Haddon-Cave LJ ‘really meant’ at [35] and [38]. This leads to much effort being spent on whatever minute linguistic indicia can be found within those two paragraphs. In my view, it is not productive to keep trying to squeeze more meaning from the few words uttered by Haddon-Cave LJ on the subject. It would be much better to consider the point afresh and from first principles. But only the Court of Appeal can have the luxury of doing this.”

As a result, the judge said at [38] that he should consider whether Unitel satisfies the “good arguable case” test on both the possible approaches to that test. It would then be apparent whether the difference between Unitel’s case and Ms dos Santos’ case on this point of law is significant to the outcome, or not. If the issue was determinative of the outcome, he would feel bound to grant permission to appeal.

The judge first considered one aspect of the three-limb test in Brownlie which he said was highly pertinent to this case and other similar freezing order applications. He set out the three-limb test and noted at [41] to [42] the explanation of Green LJ in Kaefer of when limb (iii) may come into play. He concluded at [43]:

“…under the three-limb Brownlie test, the court must not merely try to decide who has the better of the argument. If it can decide who has the better of the argument, it must also try to gauge the reliability of its conclusion on that point. This is a feature of the three-limb test that (in my view) makes it difficult to apply satisfactorily to any question going to the merits of the claim, as opposed to a question that will not arise at trial. This has been my experience in this case, hence the observations at the end of this judgment.”

The judge went on to consider Ms dos Santos’ arguments on the merits of the claim, noting that she took issue with nearly every aspect of Unitel’s case against her, but he noted at [46] that her primary point was that the claim against her was brought too late under the law of Angola, as provided for in the Angolan Commercial Companies Law (“ACCL”). Ms dos Santos relied, in particular, on Article 80(1) of the ACCL. The judge introduced this as follows:

“47.

The ACCL is the general source of law for Angolan commercial companies. Accordingly, among other things, it provides the framework for the duties owed by directors to such companies, as well as for the claims that companies can make against their directors for breach of those duties and the circumstances in which such claims can be brought.”

He considered first Article 175 of the ACCL concerned with limitation, noting that Ms dos Santos did not contend that Unitel did not have a good arguable case that the claim was not time barred. He noted that she relied in particular on Article 80(1) which has a very different focus from that of Article 175. It is in a part of the ACCL concerned with the obligations of directors and their liability for breach of those obligations. He set out the text of Article 80 in translation at [52]:

“ARTICLE 80

(Indemnity action)

1.

The company can only file an indemnity action after a shareholder resolution is passed on it and it must be filed within a period of six months counting from the date of the approval of the said resolution, with the shareholders being permitted to nominate special representatives for that purpose.

2.

During the meeting at which the yearend accounts are appraised, resolutions can be approved on an indemnity action and the dismissal of managers or directors whom the meeting considers responsible, even though these matters do not appear on the convening notice, with the managers or directors whom the meeting considers responsible being prevented from voting on those resolutions.

3.

The approval of the resolution referred to in the previous number prevents those managers or directors from being elected again while the indemnity action is pending.”

The judge considered the effect of Article 80(1) at [53] to [55]:

“53.

Article 80 as a whole is concerned with the circumstances in which the company can file an action for an indemnity claim against a manager or directors, and the effect of this on the manager or directors affected and their capacity to act for the company.

54.

The company can only file such a claim after a shareholder resolution is passed, and must do so within six months of the shareholder resolution. The company would normally act by its directors (or persons to whom they have delegated), but in this situation must be enabled to act through persons other than the defendants to the indemnity claim. The last two lines of Article 80(1) therefore empower the shareholders to nominate special representatives for the purpose of filing the claim.

55.

Article 80(1) does not look like a limitation provision. It does not seem likely that it is intended to cut across or otherwise affect the limitation period provided in Article 175. It is concerned, rather, with the company's capacity to file a claim against the manager and/or directors. The company only has such capacity if it is created by a shareholder resolution; and it then has a shelf-life of six months from the date of the shareholder resolution.”

The judge said at [56] that for the purposes of the application he proceeded on the basis that Unitel issued and filed its application to join Ms dos Santos to the UIH Claim on 3 October 2022, within 6 months of the shareholder resolution, assuming that was passed on 6 April 2022. However, the joinder application was not decided until 25 May 2023 and she was only formally joined as a defendant when service was effected on 1 June 2023, more than six months after the shareholder resolution.

At [57] to [61] the judge summarised the parties’ rival cases on this issue and the expert evidence of Angolan law (which was all in writing, there being no oral evidence or cross-examination on the application before the judge):

“57.

Ms dos Santos's case is that the requirement in Article 80(1) that the indemnity action be "filed" within six months of the shareholder resolution was not satisfied by Unitel merely issuing and filing its application to join Ms dos Santos; it required her actually to be joined to the proceedings, as Second Defendant, which did not happen until 1 June 2023. Unitel's case is the opposite. This is the principal point on which I have to decide whether Unitel has a "good arguable case".

58.

Unitel's Angolan law expert, Prof. Vicente, stated that proceedings are "filed" under Angolan law/procedure, including for the purposes of Article 80(1), when the claimant's initial application is received at the court registry. He said that the purpose of Article 80(1) is to ensure that it is certain, within six months, whether the company wishes to assert its rights per the shareholder's resolution; such certainty being necessary both for the company and for the manager/directors affected. He said that Unitel's act of issuing and filing the joinder application of 3 October 2022 achieved this certainty and thus fulfilled the purposive requirement of Article 80(1).

59.

He said that the fact that a decision of the Court was necessary, before the joinder could be made effective, does not derogate from this; and that, in some circumstances, even in Angola, the filing of legal proceedings does not mean that the defendant is party to them or that they will proceed against the defendant; they may be rejected by the court.

60.

Ms dos Santos's Angolan law expert, Prof. Dr. Ribeiro, stated Article 80(1) could only be satisfied by issuing a joinder application in England if this were procedurally equivalent to filing an action in Angola. She addressed this by considering whether a judge would have to authorise it, and whether the company would then have to take any further steps for the action to be considered as filed. She said that, because the joinder of Ms dos Santos required both the decision of the Court (i.e., the Order of HHJ Pelling KC of 25 May 2023), and then the service of the Amended Claim Form and Amended Particulars of Claim on 1 June 2023, it was not procedurally equivalent to the filing of an action in Angola.

61.

Both the experts then served further reports. Prof. Vicente accepted that it was legitimate to consider whether the joinder application was equivalent to filing an action in Angola, and opined that it was. Prof. Ribeiro repeated her view that it was not.”

At [62] the judge concluded that if the test to be applied was that in The Niedersachsen he had no doubt that Unitel has satisfied the test. At [63] he said that if the test was the three-limb test from Brownlie, in his view, Unitel has the better of the argument. He concluded:

“…I agree with Prof. Vicente's view that the joinder application of 3 October 2022 was equivalent to filing an action in Angola, for the purposes of Article 80(1). In reaching that view, I note Prof. Vicente's evidence that when an action is filed in Angola, the Court does not have to accept it, and further acts may be required of the company.

64.

It also strikes me that, in considering the question of procedural equivalence, it is necessary to have in mind the purpose of Article 80(1), because this must shed light on how such equivalence is to be measured. On the basis that Article 80(1) is concerned with the capacity of the company to act on the shareholder’s resolution, issuing and filing a joinder application in England seems to me precisely the decisive exercise of this capacity that Article 80(1) requires. Furthermore, because foreign companies effectively have to litigate in England via solicitors, who have to be instructed to go onto the record and act for the company, issuing and filing the joinder application meant that Unitel’s solicitors, Addleshaw Goddard LLP, then had actual and/or ostensible authority to conduct the litigation thereafter – even after the expiry of the six-month period under Article 80(1), their appointment having been made before this. Looking at the question in this way supports the view of Prof. Vicente.

65.

I therefore have a clear view as to which party has the better of the argument: Unitel does.”

The judge went on to say at [66] to [70]:

“66.

What I find much more difficult is gauging the reliability of this decision. It is essentially a question of foreign law, on which I am largely but not wholly dependent on information from the two experts. I need the experts to say what the relevant foreign law provisions are, and to explain their meaning and effect, but am able to use my own critical faculties as well. However, I have not had the benefit of seeing and hearing the experts give oral evidence. At trial, they will have to deal with the cut-and-thrust of cross-examination; and they will develop their respective positions, in the way that invariably happens during the trial process.

67.

I know that I have been deprived of the benefit of seeing this dynamic process unwind, but I cannot say what if any difference it would have made. I am aware of some important questions that Counsel and I would have wished them to address, if the experts had been giving evidence before me, and have in mind that I do not know what their answers might have been. I am also conscious that the trial process invariably throws up further evidence and fresh points, which cannot be predicted in advance but which may turn out to be significant.

68.

I therefore feel uncomfortable saying whether this is a limb (ii) case or a limb (iii) case, because there is no metric by which to measure the reliability of my conclusion on Article 80(1). I know that it is not as reliable as the conclusion that the trial judge will in due course reach on the same point, but I cannot say by how much. I would feel less inhibited in assessing the reliability of my conclusion if I knew that it concerned a point that will not fall to be determined finally at trial, when there is bound to be more information.

69.

What I can say is that, if this is a limb (iii) case, Unitel has a plausible (albeit contested) evidential basis for its case.

70.

It follows that, on either approach to the meaning of “good arguable case” in this context, Unitel’s case on Article 80(1) of the ACCL is a “good arguable case”.”

The judge then considered Ms dos Santos’ other arguments on the merits which he concluded were makeweights and which have not featured in this appeal. He turned to (i) the risk of dissipation (at [76] to [101]) and (ii) whether it would be just and convenient to grant the order (at [102] to [108]). The judge’s decision on these issues is also not the subject of this appeal. The judge held in favour of Unitel that there was a real risk of dissipation of Ms dos Santos’ assets, and that the order sought was both just and convenient notwithstanding the existence of other freezing orders over her assets.

Accordingly, the judge granted the freezing order in Unitel’s favour. At [111] to [117], he made some “final observations” on Article 80(1) and the three-limb test:

“111.

I noted above that I would feel less inhibited in assessing the reliability of my conclusion on Article 80(1) if I knew that it concerned a point that will not fall to be determined finally at trial, when there is bound to be more information. The fact that it concerns a merits point, which will undoubtedly need to be determined at trial, has troubled me. This prompts the following observations.

112.

First, if I were to express a settled view as to the reliability of the evidence I have received, and particularly if I were to say that it is so reliable that the strength of Unitel’s case on the point is above 50%, I would be trampling over turf that should be left pristine for the trial judge.

113.

Second, if applicants for freezing orders are told that they must provide evidence that reliably demonstrates that their prospects are above 50%, they will feel obliged to give the court as much evidence as they can muster. In a case like the present, which turns on a point of foreign law, they will say (with some justification) that the court should hear oral evidence from the rival experts, and decide the point – in effect, as a preliminary issue. I suspect that Mr Sinclair KC might well have asked me to allow oral evidence from the Angolan law experts, if he had known in advance what Mr Hill KC would say about the meaning of “good arguable case”. On the basis that the three-limb test requires Unitel to provide me with the most reliable evidence available to it, I might have found this difficult to refuse. Yet, applications like this are not supposed to become mini-trials.

114.

Third, the overall effect will be to lengthen hearings of this kind. As it was, this hearing took two days. If there had been more evidence of Angolan law, especially oral evidence, it would have taken at least three days and possibly four. If a pattern were to develop of hearings like this encompassing more evidence, and so taking longer, that would have a real effect on listing. This would be detrimental to other court users.

115.

Fourth, one answer would be to say that the court does not have to assess reliability in all cases, it can simply fall back on limb (iii). But this risks making limb (iii) the court’s route home in every case, rather than the exceptional safety-net which, it seems to me, Lord Sumption JSC had in mind. I regret having had to fall back on limb (iii) in this case, rather than dealing properly with limb (ii). That is why, if the three-limb test represents the law, I would probably have permitted oral evidence, if I had been asked to do so. If the law raises a question for the parties to address, they should be permitted to answer it with the best evidence then available to them.

116.

Fifth, another way of addressing this might be to say that the three-limb test applies only where the relevant point will not arise at trial. However, this is not straightforward. At first sight, it might seem tempting to distinguish between the context of jurisdiction and the context of freezing orders. But it is not uncommon for disputes on jurisdiction to depend on points that will arise at trial; and some points that arise in relation to freezing orders will not arise at trial.

(1)

Brownlie itself illustrates this. In the next phase of the litigation it became apparent that jurisdiction depended on a question of Egyptian law going to the ultimate liability of the Defendant, which ordinarily would have been pleaded out, explored in expert evidence and tested at trial: FS Cairo (Nile Plaza) LLC v Lady Brownlie [2021] UKSC 45.

(2)

The converse is illustrated in the context of freezing orders by Lakatamia Shipping Co. Ltd. v Morimoto. Haddon-Cave LJ considered the meaning of “good arguable case” not in so far as it applied to the merits, but in so far as it applied to risk of dissipation: see at [33], referring to the judgment of Peer Gibson LJ in Thane Investments Ltd v Tomlinson [2003] EWCA Civ 1272 at [21]. Risk of dissipation will not arise again at trial, which perhaps makes it appropriate to treat it more stringently than points on the merits.

117.

Finally, fragmenting the phrase “good arguable case” so as to give it two different meanings, depending on the context, seems tortuous. If that is going to be the ultimate outcome, it would seem preferable for different tests to be expressed by different words, not the same words.”

The judge issued a separate short judgment on the costs of the application. He considered the principles at [1] to [7]:

“1.

In relation to interim injunctions generally, costs are reserved when an injunction is granted in favour of the applicant. That is summarising the law somewhat unfairly broadly for two reasons. One is that in that context it may be difficult to say who the winner and the loser is because the merits are not considered in any detail but on the basis of the balance of convenience. And the second is because winning is in any event provisional because what happens at the interlocutory stage may be undone at trial.

2.

In the context of freezing injunctions, the position is not quite the same. The merits are looked at in a different way, not on the basis of a serious case to be tried but on the basis of good arguable case, whatever that means, and I will come back to that point. But also because the provisional nature of the relief that is granted is also different. The freezing injunction, even if the claimant wins, will often not be turned into a final injunction; conversely, if the claimant loses, the freezing injunction will be discharged, because it no longer has any purpose. By contrast, the basis on which ordinary interim injunctions are either made into permanent injunctions or not is more directly related to the merits when decided at trial. So, it is both easier to say, even at the interim stage, who the winner is and who the loser is in the context of a freezing injunction. And also not quite right to say that the success is only provisional.

3.

All that having been said, it also seems to me not right for anyone to suggest that in the context of a freezing injunction there is any strict or general rule that the winner should get their costs because the circumstances in which the matter comes before the court may well vary.

4.

As I briefly said in the course of submissions, it may depend on whether the application is made without notice, how much time the respondent has to consider it at the return date if that is the point at which issue is joined. Or whether as here, the application comes about on notice or indeed separately, whether as in some circumstances an injunction is made and is held at the return date and then there is a subsequent and discreet application to discharge it.

5.

All of those situations might be different and the responsibility that the unsuccessful respondent has for costs being incurred will vary according to those different circumstances. In this case, because (unusually) this was an application for a freezing order on notice, this respondent has had a very long time to evaluate the case that was being made against it on the application and then to prepare its own case in response, and then a further period within which to consider its position before the hearing occurred.

6.

For that reason, this respondent has a greater responsibility for costs being incurred in an application where it has lost than in many other freezing order cases. But it is also significant that, as part of the argument that I have had to deal with, the case of the respondent has been that the meaning of the good arguable case test should be approached in a way that means that the court has to conduct a relative assessment of the two parties’ positions as on the merits. For the reasons that I explained in my judgment, I have approached it on that basis as well as also considering what the outcome would be on the other basis.

7.

I have therefore conducted a more searching enquiry into the merits of the case, as best I could, than would happen in an ordinary application for an interim injunction. No matter whether (1) the law is as Ms dos Santos has argued in relation to the meaning of good arguable case or (2) it is not but because that was the argument that she made, there has been both a clearer and a more demanding examination of the merits, which justifies doing something different from what would be the norm in the standard interim injunction situation.”

He then reached the conclusion at [8]:

“So, I will make an order of costs in favour of the claimant but I emphasise that I do so because of the circumstances in which this application has come about rather than because of general views about what is appropriate in the general category of freezing orders.”

Ms dos Santos pursues two grounds of appeal:

Unitel seeks to uphold the judge’s judgment on two additional grounds:

On behalf of Ms dos Santos Mr Margolin KC took the Court through the authorities on “good arguable case” submitting that there was only ever one test, derived from the decision of the House of Lords in Vitkovice Horni a Hutni Tezirstvo v Korner [1951] AC 869 (“Korner”), where the phrase was first used by Lord Simonds. That case concerned whether the plaintiff could bring himself within one of the gateways for leave to serve out of the jurisdiction under what was then RSC Order 11 rule 1. Under rule 4 leave would not be granted unless “it shall be made sufficiently to appear to the court…that the case is a proper one for service out of the jurisdiction…”

Mr Margolin KC referred to several of the speeches. Lord Simonds at pages 879 to 880 said:

“It is, no doubt, difficult to say precisely what test must be passed for an applicant to make it sufficiently appear that the case is a proper one. I do not wholly like the expression "a prima facie case", for, where leave to serve has been given ex parte under Ord. 11 and application is then made under Ord. 12 to set the proceedings aside, a conflict may arise in which the question is not so much whether a prima facie case has been made out as whether upon all the materials then before him, the judge is of opinion that the case – I can find no better word – is a proper one to be heard in our courts. The description "a good arguable case" has been suggested and I do not quarrel with it.” 

Lord Radcliffe at page 883 said:

“The phrase is a composite one and it is not elucidated by taking it to pieces; but it seems to me clear that the use of the word "sufficiently" in this context shows that it is not necessary that the judge should be satisfied beyond reasonable doubt as to the existence of the qualifying conditions. Further, a case does not sufficiently appear to be a proper case for the purposes of this Order unless on consideration of all admissible material there remains a strong argument for the opinion that the qualifying conditions are indeed satisfied.”

Lord Tucker at pages 890 to 891 said:

“"Made sufficiently to appear" is an awkward expression which can, I think, be conveniently and accurately expressed as "satisfied" without it being thought that he thereby meant that it had to be "proved" as issues require to be proved at a trial.

My Lords, I agree with the analysis of ord. 11 which has been made by my noble and learned friend Lord Radcliffe and his statement of the principles according to which it should be applied, but I find nothing therein inconsistent with the view I have endeavoured to express that the cogency of the evidence required to justify an order for leave to serve out of the jurisdiction may vary with regard to the different matters which fall to be considered before such an order can be made.”

Mr Margolin KC submitted that this threshold of “good arguable case” was imported into the freezing injunction jurisdiction. He pointed out that, in the early cases, including Mareva itself (Mareva Compania Naviera SA v International Bulkcarriers Ltd [1975] 2 Lloyd’s Rep 509), the Court had only granted an injunction where there was a very strong case on the merits. In Rasu Maritima v Pertamina [1978] 1 QB 644, the judge at first instance, Kerr J, had discharged the injunction on the basis that it would only be granted if the plaintiffs could show they would be successful in obtaining summary judgment, which they could not. The Court of Appeal dismissed the appeal on other grounds, but held that the jurisdiction was not limited to cases where the plaintiffs would obtain summary judgment. At page 661E-H, Lord Denning MR said:

“I would not myself limit the discretion of the court to cases so plain that the plaintiff can get judgment under Order 14. We have all had experience of summonses under Order 14. The defendant may put in an affidavit putting forward a specious defence sufficient to get him leave to defend, conditional or unconditional. But when the case actually comes to the court for trial, he throws his hand in. It is then seen that the affidavit was simply filed in order to gain time. So under this new procedure a defendant may put forward a specious defence, just so as to remove his assets from the jurisdiction. The weakness of the defence may not appear until later. So I would hold that an order restraining removal of assets can be made whenever the plaintiff can show that he has a "good arguable case." That is a test applied for service on a defendant out of the jurisdiction: see Vitkovice Horni a Hutni Tezirstvo v Korner [1951] AC 869: and it is a good test in this procedure which is appropriate when defendants are out of the jurisdiction. It is also in conformity with the test as to the granting of injunctions whenever it is just and convenient as laid down by the House of Lords in American Cyanimid Co v Ethicon Ltd [1975] AC 396.”

Accordingly, Mr Margolin KC submitted that, from the very outset of the freezing injunction jurisdiction, “good arguable case” was intended to be the same test as in jurisdiction cases. He submitted that nothing in The Niedersachsen changed that. Mustill J expressed some doubts about Lord Denning MR’s analogy with applications for permission to serve out of the jurisdiction, but described the Pertamina case as “the foundation authority” and considered that he should apply the “good arguable case” test. Mr Margolin KC submitted that Mustill J was not purporting to set out a different test but to summarise the test as it already applied to applications for permission to serve out of the jurisdiction when he said:

“I consider that the right course is to adopt the test of a good arguable case, in the sense of a case which is more than barely capable of serious argument, but not necessarily one which the judge considers would have a better than 50 per cent chance of success.”

He submitted that the meaning of “good arguable case” was further explained by the Court of Appeal in Canada Trust v Stolzenberg (No. 2) [1998] 1 WLR 547. He referred to a passage in Waller LJ’s judgment at page 554 citing the speeches in Korner as demonstrating that Lord Tucker was not an “outlier” and also referred to the formulation of the test at page 555E-G:

“It is also important to remember that the phrase which reflects the concept ‘good arguable case’ and the other phrases in Korner’s case “a strong argument” and “a case for strong argument” were originally employed in relation to points which related to jurisdiction but which might also be argued about at the trial. The court in such cases must be concerned not even to appear to express some concluded view as to the merits, e.g. as to whether the contract existed or not. It is also right to remember that the ‘good arguable case” test, although obviously applicable to the ex parte stage, becomes of most significance at the inter partes stage where two arguments are being weighed in the interlocutory context which, as I have stressed, must not become a ‘trial’. ‘Good arguable case’ reflects in that context that one side has a much better argument on the material available. It is the concept which the phrase reflects on which it is important to concentrate, i.e. of the court being satisfied or as satisfied as it can be having regard to the limitations which an interlocutory process imposes that factors exist which allow the court to take jurisdiction.”

Mr Margolin KC submitted that, in the years which followed, the test in Canada Trust was adopted in the context of freezing injunctions. He referred to three decisions at first instance: Toulson J in Petroleum Investment v Kantupan Holdings [2002] 1 All ER (Comm) 124 at [36] to [38], Roth J in The Complete Retreats Liquidating Trust v Logue [2010] EWHC 1864 (Ch) at [71] to [72] and Deputy High Court Judge Andrew Sutcliffe QC in OJSC TNK-BP Holding v Beppler & Jacobsen Ltd [2012] EWHC 3286 (Ch) at [82] to [83].

Mr Margolin KC noted that Unitel relied upon the decision of the Court of Appeal in Lakatamia Shipping v Nobu Su [2012] EWCA Civ 1195, which declined to interfere with the view of two Commercial Court judges, Blair J and Beatson J, that the claimant had a “good arguable case” applying the Niedersachsen test and where Longmore LJ said at [27]:

“This is not a matter of deferring to the commercial judge, but it must be remembered that applications for freezing injunctions made on the basis of a good arguable case come before the commercial judges all the time. Derived from their time in practice they have developed what is perhaps best described as an instinct as to what is well arguable and what is not. That instinct should be respected by those in this court without the everyday experience of granting and refusing freezing injunctions unless it is plain that the judge is wrong: see Stuart v Goldberg [2008] 1 WLR 823 paragraph 76 per Sedley LJ and paragraph 81, Sir Anthony Clarke, Master of the Rolls.”

However, Mr Margolin KC submitted, it does not appear that either Canada Trust or The Complete Retreats were cited to the Court of Appeal, although, as Popplewell LJ put to him in argument, any suggestion that Longmore LJ did not know about the Canada Trust case would be unreal.

He also noted that in Kazakhstan Kagazy v Arip the judge at first instance, HHJ Mackie QC, had applied the Canada Trust test of “much the better of the argument”. At [25] of his judgment in the Court of Appeal, Longmore LJ had said that this set the hurdle too high and that it was established by Pertamina that the appropriate test was “good arguable case”. Mr Margolin KC submitted that, in practice, that observation had to be viewed by reference to the subsequent rejection by the Supreme Court of the word “much” in that formulation.

In relation to the second appeal in Lakatamia Shipping v Nobu Su [2014] EWCA Civ 636; [2015] 1 WLR 291, which concerned the Chabra jurisdiction, Tomlinson LJ at [32] adopted the analysis of Popplewell J (as he then was) in PJSC Vseukrainskyi Aktsionernyl Bank v Maksimov & Ors, [2013] EWHC 422 (Comm):

“(1)

The Chabra jurisdiction may be exercised where there is good reason to suppose that assets held in the name of a defendant against whom the claimant asserts no cause of action (the NCAD) would be amenable to some process, ultimately enforceable by the courts, by which the assets would be available to satisfy a judgment against a defendant whom the claimant asserts to be liable upon his substantive claim (the CAD).

(2)

The test of "good reason to suppose" is to be equated with a good arguable case, that is to say one which is more than barely capable of serious argument, but yet not necessarily one which the Judge believes to have a better than 50% chance of success.”

Despite the fact that this clearly reflected the Niedersachsen test, Mr Margolin KC submitted that this was not to be taken as a rejection of Canada Trust, although I note that case was not cited to the Court of Appeal.

He accepted that in Holyoake v Candy [2016] EWHC 970 (Ch); [2018] Ch 297, Nugee J (as he then was) conducted a survey of the authorities on what constituted a “good arguable case” in the context of the freezing injunction jurisdiction at [11] to [15] and applied the Niedersachsen test, saying:

“In the case of purely factual questions, I consider that it is sufficient for the claimant to meet the traditional test laid down by Mustill J in The Niedersachsen that the claimant needs to show a good arguable case in the sense of a case which is more than barely capable of serious argument, and yet not necessarily one which the judge believes to have a better than 50% chance of success. Indeed I would regard it as wholly invidious in a case of this type, which is likely to turn largely on the credibility of the principals on each side and their recollections of oral conversations, for a judge faced with nothing other than the pleaded cases and assertions that each side's pleaded case represents the truth, to have to form a view as to where the better of the argument on such issues might lie, let alone where much the better of the argument might lie. I find myself completely incapable, and indeed I would regard it as wholly inappropriate, to judge such matters on the basis of what are at this stage hotly disputed allegations on each side.”

In jurisdictional gateway cases, the three-stage test was adopted by the Supreme Court in Brownlie and Goldman Sachs and “much” removed from “much the better of the argument”: see per Lord Sumption JSC in Brownlie at [7]. Mr Margolin KC pointed out that there was still uncertainty as to whether it was an absolute or a relative test. That uncertainty was resolved by the judgment of Green LJ in Kaefer. The issue was explained by Green LJ at [61]:

“…it is sensible to provide a word of explanation about the key conceptual dispute between the parties. This concerned the asserted difference between an absolute test and a relative test. It is argued that an absolute test is one where the Claimant, to found jurisdiction, need only surmount a specified evidential threshold which does not involve the Court otherwise assessing the relative merits of the competing arguments. In contrast it is argued that a relative test does involve the court in looking to the merits in a relative sense to see whose arguments are stronger. In this context a test which is set by reference to a fixed standard (e.g. arguability) is an absolute test, because provided the Claimant surmounts this hurdle, it is irrelevant that the Claimant's arguments, even at the interim stage, may be (relatively) weaker than the Defendants arguments: an argument might be arguable but still wrong. It follows that an absolute test is easier to establish and therefore one which claimants will prefer; and a relative test is harder to meet, and one which defendants will prefer.”

Having reviewed the Supreme Court decisions, at [73] Green LJ concluded that they were confirming the relative test and then went to consider at [74] what was the correct name for the test, noting that in Goldman Sachs the Supreme Court did not use the terminology of “good arguable case” save in respect of limb (iii). He concluded:

“In my view, provided it is acknowledged that labels do not matter, and form is not allowed to prevail over substance, it is not significant whether one wraps up the three-limbed test under the heading "good arguable case".” 

Mr Margolin KC also referred to what Davis LJ said at [119]:

“I am in something of a fog as to the difference between an "explication" and a "gloss". But whatever the niceties of language involved, it is sufficiently clear that the ultimate test is one of good arguable case. For that purpose, however, a court may perfectly properly apply the yardstick of "having the better of the argument" (the additional word "much" can now safely be taken as consigned to the outer darkness). That, overall, confers, in my opinion, a desirable degree of flexibility in the evaluation of the court: desirable, just because the standard is, for the purposes of the evidential analysis in each case, between proof on the balance of probabilities (which is not the test) and the mere raising of an issue (which is not the test either).”

Mr Margolin KC placed particular reliance on the judgment of Haddon-Cave LJ in Morimoto. At [33] he set out the legal principles for the grant of a WFO:

“The basic legal principles for the grant of a WFO are well-known and uncontroversial and hardly need re-stating. It nevertheless is useful to remind oneself of the succinct summary of the test by Peter Gibson LJ in Thane Investments Ltd v Tomlinson (No 1) [2003] EWCA Civ 1272 at [21] where he stated that, before making a WFO, the court must be satisfied that:

"… the applicant for the order has a good, arguable case, that there is a real risk that judgment would go unsatisfied by reason of the disposal by the defendant of his assets, unless he is restrained by the court from disposing of them, and that it would be just and convenient in all the circumstances to grant the freezing order."”

At [35] he noted that the test for “good arguable case” in the context of freezing injunctions was not a particularly onerous one, citing [12-026] of the sixth edition of Gee on Commercial Injunctions. He then went on to say at [38] that the test had been comprehensively reviewed by the Court of Appeal recently in Kaefer:

“[Green LJ] observed at [59] that a test intended to be straightforward "had become befuddled by 'glosses', glosses upon gloss, 'explications' and 'reformulations'". The central concept at the heart of the test was "a plausible evidential basis" (see paragraphs [73]-[80]).”

Mr Margolin KC took the Court through the three recent first instance decisions referred to at [9] above where Edwin Johnson J and Dias J applied the same test as in the jurisdictional gateway cases but Butcher J did not. He submitted that the three-limb test was the right test for freezing injunction cases as well. The Court should draw the line at saying if the claimant was more likely to win than not it should get an injunction, if it was not more likely to win than not, an injunction was not justified.

There was some debate between Mr Margolin KC and the Court about whether “good arguable case” was in essence the same test as “serious issue to be tried” as it is applied in a number of contexts such as summary judgment and as part of the American Cyanamid test. Mr Margolin KC submitted that “good arguable case” should impose a higher standard because of the invasive nature of freezing injunction relief. However, as Popplewell LJ pointed out, interlocutory injunctions where a serious issue to be tried is the merits test can also be invasive.

Submissions on the second part of the first ground of appeal, that whatever the test to be applied, the judge should have concluded that Unitel did not have a good arguable case, were made by Mr Olliff-Cooper. He submitted that the correct date of the shareholder resolution being assumed for present purposes to be 6 April 2022, Unitel was obliged under Article 80(1) of the ACCL to file its “indemnity action” within 6 months, i.e. by 6 October 2022. Unitel had made its application for permission to bring the claim against Ms dos Santos on 3 October 2022, but did not serve the proceedings until June 2023. He submitted that Article 80(1) was a forfeiture period under Portuguese and Angolan law whose purpose was “to avoid that situations remain undefined for a long period of time.” He submitted that this meant that any step which did not actually result in an indemnity claim being brought against the director or which, once taken, will necessarily result in such a claim being brought cannot be sufficient for the purposes of Article 80(1) as it would not prevent the situation as between the company and the director remaining undefined.

Mr Olliff-Cooper took the Court in some detail through the reports of the two Portuguese law professors who gave expert evidence of Angolan law, Professor Vicente for Unitel and Professor Ribeiro for Ms dos Santos. He submitted that the experts had originally agreed that Article 80(1) required a claim to be filed within 6 months which would be fatal to Unitel’s case, but Professor Vicente had changed his evidence in his second report, saying that not just filing the claim but a procedural act exercising a right to claim, expressing the company’s will to recover the loss from the director, would be sufficient. At [28] to [29] of his second report he expressed the opinion that the application to join Ms dos Santos to the proceedings was sufficient. Mr Olliff-Cooper submitted that this was just wrong and in any event was expressing an opinion about the consequences of an English procedural act.

This opinion was challenged by Professor Ribeiro in her second report. Her view was that the legal security and certainty which were required were only maintained if the act of filing an application for permission to join a party was procedurally equivalent to the filing of an action which it was not. She also made a point about protecting the interests of the passive party (i.e. Ms dos Santos) which would only be achieved if she could be considered as having become a defendant.

Professor Vicente produced a third report which Mr Olliff-Cooper submitted misrepresented what Professor Ribeiro was saying. Professor Vicente said that when faced with an application for a new claim, an Angolan judge is not obliged to accept it but may reject it altogether, for example because the court is not competent, or may invite the claimant to complete it or correct it. In other words, the filing of a new claim in Angola is equivalent to the joinder application here, which is what the judge accepted at [63] of his judgment. Mr Olliff-Cooper submitted that he was wrong to do so.

With regard to the point raised in the Respondent’s Notice, that Article 80(1) could be regarded as a foreign limitation period under the Foreign Limitation Periods Act 1984 (“the 1984 Act”), Mr Olliff-Cooper submitted that the Article is not part of the Angolan law of limitation but a forfeiture period (as was common ground between the experts) so that the 1984 Act does not apply and accordingly there is no question of “relation back” under section 35 of the Limitation Act 1980. In any event, any relation back would be to when the claim against UIH was issued in October 2020 which would be before the commencement of the six-month period referred to in Article 80(1).

He submitted that Unitel was also wrong to suggest that for English conflicts of laws purposes Article 80(1) was a procedural rule of Angolan law. Both experts agree that the expiration of a forfeiture period has the effect of extinguishing the cause of action and Professor Vicente says in terms that forfeiture is part of the substantive law of Angola.

In relation to the second ground of appeal, Mr Margolin KC submitted that the judge had erred in making an order for costs against Ms dos Santos. He submitted that the usual order in these circumstances would be costs reserved unless there were special factors pointing to a different conclusion, of which there were none. Mr Margolin KC very fairly took the Court to all the relevant authorities, both those in his client’s favour and those against her.

He relied in particular on two decisions of the Court of Appeal (albeit of two judge Courts in each case). In Richardson v Desquenne et Giral UK Ltd [2001] FSR 1 the defendant appealed a costs order made on an interlocutory hearing to extend an injunction preventing breach of a confidentiality clause in his employment contract. In a judgment handed down in November 1999, soon after the CPR first came into force, the Court (Morritt LJ and Morison J) allowed the appeal. At [12] to [13] of his judgment, Morritt LJ said:

“12.

I accept of course that the issue was one for the judge's discretion. In my view, this is one of those cases where this Court is entitled and indeed bound to interfere with that exercise. I say so for basically three reasons: the first one is that the decision seems to me to be inherently unjust. It is quite plain from the passage in the judge's judgment from which I quoted that he granted or continued the [in]junction on the basis of the balance of convenience in order to hold the ring until the dispute between the parties could be properly decided at a trial. It is inconsistent with an order such as that, that there should be successful or unsuccessful parties for the purposes of the rules either new or old.

13.

Second, it seems to me that the judge was wrong, therefore, in determining, for the purposes of Rule 44.3.2 [which is now CPR 44.2(2)], that either Mr Richardson was the unsuccessful party, or, alternatively, that the employer was the successful party. He was right to consider within the terms of that rule whether to make an order about costs. That was what he did. But the order that he made was, going back to Rule 44.3.1(a), whether the costs should be made payable by one party to another. That seems to me to have been wrong; there were no successful or unsuccessful parties at that stage and the proper orders to be considered were those under the terms of the practice direction to which I have referred.” [The third reason is not relevant.]

Mr Margolin KC referred to the judgment of Neuberger J in Picnic at Ascot v Derigs [2001] FSR 2 which followed Desquenne in February 2000. The claimant applied for an interim injunction against the defendant, claiming infringement of design right and breach of fiduciary duty. Almost three months after the evidence in support of the application for the interim injunction had been filed and only three clear days before the application was due to be heard, the defendant informed the claimant that it no longer intended to contest the application, and proposed that the costs of the application be reserved to the trial. The claimant sought an order for costs. The judge held that in a case without any other special factors, where a claimant obtains an interlocutory injunction on the basis of the balance of convenience, the court normally reserves the costs. He accepted that Desquenne should not be treated as an authority binding the court’s hands on an issue as to costs but said that an inconsistency of approach to questions of costs between different courts was undesirable. Mr Margolin KC submitted that this approach should hold good for freezing injunctions as well.

He noted that at [12], Neuberger J recognised that there were circumstances where it is right to depart from that general approach, saying:

“Thus there may be cases where the balance of convenience is so clear, and the outcome of the hearing of the application for the interlocutory injunction should be so plain to the parties, that the court should conclude that an order should be made against the defendant for wasting time and money in fighting the issue (whether or not the defendant eventually concedes)”

Mr Margolin KC pointed out that Bright J seems to have thought that this was such a case, but he was wrong to do so.

The second Court of Appeal decision on which he relied was the judgment of Lewison and McCombe LJJ in Melford Capital Partners (Holdings) LLP v Wingfield Digby [2020] EWCA Civ 1647; [2021] 1 WLR 1553. The claimants obtained an interim injunction to restrain use of allegedly confidential information and the defendant’s challenge to its continuation was dismissed. At first instance, Birss J (as he then was) said of Desquenne and Picnic at Ascot, which were cited to him, that he suspected that the reality was that those two cases had been overtaken for some years by the modern approach to costs under the CPR. He considered that the claimants were the successful party on the contested application to continue the injunction and that he should make what he regarded as the usual order, that the costs on the application should be paid by the defendant as the unsuccessful party.

The Court of Appeal varied that order to an order that the costs of the injunction application were reserved to the trial judge. At [36] they said that, in their judgment, the passage in the White Book at [44.6.1] commenting on Desquenne and Picnic at Ascot accurately represents the law:

“Where an interim injunction is granted the court will normally reserve the costs of the application until the determination of the substantive issue ( Desquenne …) However, the court's hands are not tied and if special factors are present an order for costs may be made and those costs summarily assessed (Picnic at Ascot)…”

The Court rejected a submission by the claimants that this no longer represented the law because it had been superseded by the “pay as you go” principle which took precedence. Mr Margolin KC placed particular emphasis on [39] and [41] of the Court’s judgment:

“39.

The quest for the successful and unsuccessful party in such cases is usually fruitless. The respondent to the application, like the appellant in the present case, denies that the claimant is entitled to any relief, because the underlying cases of the parties on disputed facts are diametrically opposed. The applicant for the grant of interim relief, even if the court holds that the claimant has a good arguable case or is more likely to succeed than not, the applicant still has to persuade the court that the balance of convenience makes the grant of an interim injunction or other related relief more appropriate than its refusal.

41.

...Whenever a claimant successfully seeks an interim injunction preventing the defendant from doing something (whether using a right of way, working for a competitor or infringing a patent) the defendant will be stopped from doing whatever it is for the time being. That was precisely the case in both Desquenne and Picnic at Ascot. However, the judge's decision that he was unable to resolve the merits of the disputes means that the basis on which those orders were obtained and continued, without objection from the appellant, may prove in the end to have been unfounded. “Success” of this type is only a provisional one. On the other hand, a “costs reserved” order does not mean that claimants generally, or these respondents in particular, will never recover the proper proportion (if not all) of their claimed costs: the matter is open and the costs have been neither won nor lost by either side at this stage.”

Mr Margolin KC submitted that the same analysis applied here. If, at the conclusion of a trial, the Unitel claim failed then the freezing injunction should not have been granted so that Unitel was not entitled to be regarded as the successful party. He submitted that this was absolutely fundamental.

He accepted that there were a number of freezing injunction cases where a different conclusion had been reached. In Bravo v Amerisur Resources [2020] EWHC 2279 (QB); [2020] Costs LR 1329, Martin Spencer J noted that the question had arisen in relation to the issue of costs whether an application for a freezing injunction was akin to a normal injunction application where the court makes assumptions on the facts and refers to the balance of convenience. At [52]-[53] he said:

“...the decision in Picnic at Ascot is not wholly apposite [in] claims for freezing orders where the balance of convenience is not an issue, and where in relation to the merits of the case the court has regard to the question of whether there is a good arguable case on behalf of the claimants or not. That is sufficient for the court to determine whether a freezing order should be made, and even if at the subsequent trial it turns out that the claims fail on the basis of the evidence due to that trial, it does not at all follow that this means that the court was wrong to find that there was a good arguable case. On the contrary, those two findings are wholly consistent with each other, or maybe wholly consistent with each other. Nor is there any reference to the balance of convenience. The question is whether it is just and convenient to make an order.

53.

Therefore, I agree with Mr Lord that the regime for the making of Freezing orders is different to the general position where interim injunctions are sought based upon balance of convenience and holding the ring pending the trial. There are, obviously, overlapping features, holding the ring being one of them. The purpose of a freezing injunction is to avoid a successful claimant being unable to enjoy the fruits of his success because there are no assets left against which the judgment can be enforced, but that is a different kind of holding of the ring to that which is involved in the usual interim injunction and balance of convenience type case."

Mr Margolin KC submitted that [52] was simply wrong as the same situation may arise in freezing injunction cases as in American Cyanimid cases. For example, it may emerge with the benefit of hindsight at the end of the trial that the Court was wrong to grant the injunction because the trial judge concluded that there was no risk of dissipation. Hence the appropriate order was costs reserved. The distinction made by the judge in [53] was not a relevant one.

Kumar v Sharma [2022] EWHC 1008 (Ch); [2022] Costs LR 1029, a decision of Deputy High Court Judge Jonathan Hilliard QC, was another freezing injunction case. In relation to the question whether costs should be reserved to the trial judge or be in the case he followed [53] of Bravo in concluding that a freezing injunction does not hold the ring in the same sense as other types of interim injunction do. Once again Mr Margolin KC submitted that this was not a relevant ground of distinction. At [12(6) and (7)] the Deputy Judge said:

“(6)

By analogy, where, for example, a defendant brings an application for reverse summary judgment against the claimant and fails, it is no answer to the claimant's claim for costs that the defendant may ultimately be the successful party at trial on the balance of probabilities.

(7)

Indeed, were it otherwise, a defendant would have a free shot at opposing a freezing order continuance on a return date on the good arguable case ground, knowing that it would not have to bear costs if it ultimately succeeded at trial, or unless and until the trial took place and had been decided.”

Mr Margolin KC submitted that the analogy in [12(6)] was not a reliable one and that it was wrong to say at [12(7)] that the defendant got a free shot because the trial judge might decide the defendant should pay the costs of the freezing injunction return date. None of the reasons set out in [12] was a good reason for applying a different test as to costs from that applied in American Cyanimid cases.

In Al Assam v Tsouvelekakis [2022] EWHC 2137 (Ch), HHJ Davis-White QC granted a freezing injunction and ordered that the costs of the two-day hearing should be reserved. Mr Margolin KC placed particular reliance on that case. At [223] the judge set out the starting position derived from Desquenne, Picnic at Ascot and Melford Capital:

“223.

In the case of interim or interlocutory injunctions the starting position is now well established that the costs of the application will usually be reserved, though there may be factors, or as it has been put, special factors (which I do not consider to amount to exceptional factors) justifying some other costs order (see [those three cases]).”

The judge went on at [252]:

“252.

As regards interim injunctions granted under the American Cyanamid principle, it is no answer to an application for the costs of the application to be reserved to say that the respondent failed to establish that there was not a serious issue to be tried and that whatever the position at trial the respondent has failed on the assessment of the merits test as they stand and apply at the interim stage. Indeed, that was the flawed approach adopted in cases such as Melford Capital Partners. The reason is because the claim has not then been established. In my judgment, the same is true in principle as regards a freezing injunction. The court has simply decided that there is an arguable claim, not that the claim succeeds. If the claim fails at trial, then the freezing injunction should (with the benefit of hindsight) not have been made.”

He then concluded at [258] in an analysis which Mr Margolin KC submitted was absolutely right:

“In short, the tests under American Cyanamid and for a freezing order are different but in my view they are sufficiently analogous and are sufficiently dealing with the same position that the starting costs position should be the same notwithstanding (a) the tests are in slightly different terms and (b) the holding of the ring is in a slightly different manner. So far as it is suggested that in a freezing injunction application that succeeds the court can say who has won, the answer is that at that stage the court cannot say who has won. The decision is interim.”

That decision was not followed by Edwin Johnson J in his costs judgment in Harrington [2023] EWHC 609 (Ch). At [11] he said that the starting point was CPR 44.2. Sub-rules (1) and (2), to which he referred, provide as follows:

“(1)

The court has discretion as to—

(a)

whether costs are payable by one party to another;

(b)

the amount of those costs; and

(c)

when they are to be paid.

(2)

If the court decides to make an order about costs—

(a)

the general rule is that the unsuccessful party will be ordered to pay the costs of the successful party; but

(b)

the court may make a different order.”

The judge went on to analyse the judgment of HHJ Davis-White QC explaining at [26] to [36] why he disagreed with that judge and preferred the reasoning of the judges in Bravo and Kumar. In particular, at [27] to [29], he accepted and elaborated on the distinction Martin Spencer J had drawn in terms of holding the ring between American Cyanamid injunctions and freezing injunctions:

“27.

…it seems to me that, while it is correct to say that a freezing order holds the ring, it also seems to me that it is correct to say – and I accept the submission of Mr McQuater in this respect – that a freezing order holds the ring in a different way. In my judgment, in a substantially different way to an interim injunction.

28.

As Mr McQuater pointed out, in the case of an interim injunction what is generally happening is that a court is allowing one party to enforce or rely on a right, or an obligation the existence of which has yet to be established. So, in that sense the court is allowing one party to behave as if the right has been established, in circumstances where the right still has to be established at trial and may not be established at trial.

29.

In the case of a freezing order, things are rather different. The freezing order, as Mr Grant quite correctly pointed out, is an ancillary order in aid of the relief which is sought in the relevant case. There is no such thing as a final freezing order. Once the freezing order has been granted, and subject to any subsequent application to vary or discharge, the freezing order then remains in place until trial. It may well be that the freezing order is obtained on a basis which is found not to be well founded at trial, but that, it seems to me, does not go directly to the question of whether the freezing order was correctly granted; rather it relates to the underlying relief which is sought.”

At [34] to [35] he dealt with the “free shot” point, recording at [34] counsel’s argument, also made here by Mr Margolin KC, that it was not a free shot because the costs of the contested application may be recovered pursuant to a costs order made by the trial judge. At [35] he rejected that argument:

“But that seems to me to miss the essential point, which is that if the general principle is that the costs of an application for a freezing order should be reserved, then the defendant does know that it is going to be able to oppose the freezing order, and possibly cause both parties to run up very considerable costs in relation to the freezing order, without having to face the day of reckoning in relation to those costs, assuming that it is unsuccessful, until a trial, which may come along at a much later stage, or may not come along at all, which may in turn leave the parties to negotiate what is going to happen in relation to the reserved costs. In litigation there is a very substantial difference between a set of costs which must be paid there and then by a party, and a set of costs which are reserved off to an indeterminate date in the future.”

Mr Margolin KC also relied upon the analysis of what a freezing injunction is intended to protect most recently expressed by Lord Leggatt giving the majority judgment of the Privy Council in Broad Idea International v Convoy Collateral [2021] UKPC 24; [2023] AC 389 at [89]:

“The interest protected by a freezing injunction is the (usually prospective) right to enforce through the court’s process a judgment or order for the payment of a sum of money. A freezing injunction protects this right to the extent that it is possible to do so without giving the claimant security for its claim or interfering with the respondent’s right to use its assets for ordinary business purposes. The purpose of the injunction is to prevent the right of enforcement from being rendered ineffective by the dissipation of assets against which the judgment could otherwise be enforced.”

He submitted that there was no principled basis for a different approach to costs in freezing injunction cases as opposed to American Cyanamid cases, since if no judgment is obtained by the claimant at trial, in both instances it can be said that the claimant ought not to have obtained the injunction in the first place.

On behalf of Unitel, Mr Sinclair KC submitted in relation to the first ground of appeal that the judge’s endorsement of The Niedersachsen test was right as a matter of authority. That test had been the orthodox approach to the merits threshold in freezing injunction cases for decades, repeatedly endorsed and never disavowed by the Court of Appeal. The suggestion on behalf of Ms dos Santos that Morimoto quietly changed the law is simply wrong. Her reading of the history of the freezing order jurisdiction is also flawed.

He pointed out that, at the time of Pertamina in 1977, it was thought that a Mareva injunction could only be obtained against a foreign defendant not an English defendant, which may explain why Lord Denning MR drew the analogy with the threshold test for the gateways under what was RSC Order 11. In any event, nothing in his judgment suggests that, whatever the test became for establishing a jurisdictional gateway in service out cases, the freezing order merits test must track it. Mr Sinclair KC also relied upon what Elias LJ said in Kazakhstan Kagazy at [67], which the judge in the present case referred to at [27]:

“It is true that in adopting the good arguable test Mustill J was following the decision of Lord Denning in [Pertamina] [1978] QB 644, and Lord Denning had in turn adopted it in the context of a freezing order because he thought that the jurisdiction test was appropriate, at least where the case involved a foreign defendant (see p.661G). But there have been developments in the law relating to jurisdiction since, and although a claimant in both jurisdiction and freezing order cases must establish a "good arguable case", the policy considerations are different in the two situations and it is far from obvious that this inherently flexible concept must have the same meaning in each context. Indeed, even in jurisdiction cases the good arguable case test only goes to the question whether the claim falls within one of the grounds set out in PD6B para.3.1. We are concerned with the merits of the case, and so far as they are concerned, a claimant in a jurisdiction case has only to show that there is a serious issue to be tried: see Seaconsar Ltd v Bank Markazi [1994] 1 A.C.438, 457 per Lord Goff of Chieveley.”

Mr Sinclair KC pointed out that in The Niedersachsen itself Mustill J had drawn a distinction at page 403e-f between interlocutory and Mareva injunctions and between the jurisdictional gateways and Mareva injunctions:

“Moreover, even if the analogy with ordinary interlocutory injunctions were accepted, it would not lead to ‘a good arguable case’ as the test. The Cyanamid case teaches that the court should not attempt to weigh the merits except to ascertain at the outset that the plaintiff’s case is not derisory, and then to employ it as a factor of last resort if all other considerations are equal. The standard of ‘a good arguable case’ seems to have no place here.

The analogy with proceedings under RSC Ord 11 also seems rather distant. It is true that the service of a writ out of the jurisdiction commands a foreigner to appear, against his will. But he need not comply, if he does not choose. The Mareva injunction, by contrast, bites directly on the defendant’s assets; he must come in and defend, or lose them. Moreover, it is now the law, which was not so when the Pertamina case was decided, that the Mareva injunction applies to persons resident within the United Kingdom, so the relationship with Ord 11 seems even more remote.”

Mr Sinclair KC also referred to the decision of the Court of Appeal in The Niedersachsen (reported after the judgment of Mustill J in the All England Reports and also at [1983] 1 WLR 1412) unanimously endorsing the judge’s judgment. At page 415b-e (page 1417B-F) Kerr LJ said:

“Although other, and perhaps slightly stronger. words have been used in other cases, the defendants did not challenge this formulation [“good arguable case”] of the present appeal. We respectfully agree with it, but would add that this aspect of the evidence before the court should not be looked at in isolation when deciding whether or not to exercise the discretion to grant a Mareva injunction. The ultimate basis for this jurisdiction is now to be found in section 37 of the Supreme Court Act 1981. Subsection (1) provides:

“The High Court may by order (whether interlocutory or final) grant an injunction … in all cases in which it appears to the court to be just and convenient to do so.”

In the context of Mareva injunctions one must now also have regard to subsection (3) :

“The power of High Court under subsection (1) to grant an interlocutory injunction restraining a party to any proceedings from removing from the jurisdiction of the High Court, or otherwise dealing with, assets located within that jurisdiction shall be exercisable in cases where that party is, as well as in cases where he is not, domiciled, resident or present within that jurisdiction.”

It follows that the evidence, including the evidence on the second question posed by the judge (whether there was a real risk of dissipation of assets) to which we turn in a moment, must be looked at as a whole. A “good arguable case” is no doubt the minimum which the plaintiff must show in order to cross what the judge rightly described as the “threshold” for the exercise of the jurisdiction. But at the end of the day the court must consider the evidence as a whole in deciding whether or not to exercise this statutory jurisdiction.”

At the end of his judgment, Kerr LJ said at page 422h (page 1426F): “In the present case the judge correctly applied the two tests which fell to be considered on the facts, and there is no basis for criticising the conclusion which he reached in exercising his discretion.” Accordingly, Mr Sinclair KC submitted that Longmore LJ had been right to say in [25] of the first Lakatamia Shipping v Nobu Su appeal that Kerr LJ had approved the test put forward by Mustill J.

In Kazakhstan Kagazy at [25] Longmore LJ had said in terms that “much the better of the argument” (i.e. the Canada Trust test) set the hurdle too high in freezing injunction cases and the appropriate test was “good arguable case”. Jackson LJ was clearly applying the Niedersachsen test at [53] when he said: “It is only by a narrow margin that KK's case is strong enough to support their entitlement to a freezing injunction.” At [66] Elias LJ referred expressly to what Mustill J had said was the test and then added:

“This "good arguable case" test was accepted by Kerr LJ in the Court of Appeal in that case as the minimum which a claimant must show (see [1983] 1 WLR 1412, 1417) and whilst there was no express approval of Mustill J's explanation of what that test involved, the court expressed no disagreement with it. Moreover, that particular formulation was recently cited with approval by Longmore LJ in the Lakatamia Shipping case (para. 25). I am far from satisfied that it is the same as the "much the better of the argument" test adopted by the judge.”

Likewise, as set out at [37] above, in the second Lakatamia Shipping v Nobu Su appeal, Tomlinson LJ approved the Niedersachsen test. There were thus three Court of Appeal decisions after Canada Trust approving the Niedersachsen test in freezing injunction cases.

Mr Sinclair KC submitted that in Holyoake v Candy Nugee J had drawn a distinction between the “serious issue to be tried” test which was the threshold test for the grant of interlocutory injunctions and the good arguable case test. He said at [9]:

“The next question that was argued was what the Claimants need to show so far as the merits of their substantive claims are concerned. There are actually two parts to this question:

(1)

Is it enough, as Mr Trace submits, for the Claimants to demonstrate that there is a serious issue to be tried (as is the usual threshold test for the grant of any interlocutory injunction: see American Cyanamid Co v Ethicon Ltd [1975] AC 396 at 407G per Lord Diplock); or do the Claimants have to show a good arguable case (as is the case for a freezing injunction: see The Niedersachsen [1983] 1 WLR 1412 at 1417E per Kerr LJ)?

(2)

If the test is a good arguable case, what does this mean?”

In deciding that the test was “good arguable case” rather than “serious issue to be tried” Nugee J said at [10]:

“Although a notification injunction is in principle less invasive than a freezing injunction, it is still an invasive order and I think justifies more than a serious issue to be tried which, as appears from Lord Diplock's judgment in American Cyanamid, only really serves to cut out the frivolous or vexatious case. For the reasons I have already given the principles underlying the grant of a notification injunction are closely tied to the principles underlying the grant of a freezing injunction, and in my judgment what is needed to justify a freezing injunction in terms of the merits of the substantive claim is also needed to justify a notification injunction. I therefore hold that the Claimants need to demonstrate a good arguable case.”

Mr Sinclair KC submitted that the lynchpin of Ms dos Santos’ case was Morimoto and the suggestion that it rewrote the law and reunified the jurisdiction and freezing injunction tests, but if it did so, it was without much attention. He submitted that what Haddon-Cave LJ said at [35] (referred to at [43] above) is inconsistent with the Court of Appeal in that case applying a much more onerous test and what Haddon-Cave LJ said at [38] does not bring in the whole of the Brownlie test, but only limb (iii).

Mr Sinclair KC pointed out that there were three post-Morimoto decisions. First was my decision sitting at first instance in PJSC Bank “Finance and Credit” v Zhevago [2021] EWHC 2522 (Ch). At [170] I said that: “The test of whether a claimant has shown a "good arguable case" for the purposes of obtaining a freezing injunction was recently confirmed by the Court of Appeal in [Morimoto]” and then cited [37] and [38] of the judgment of Haddon-Cave LJ. I went on to say at [171]:

“During the course of argument, I indicated that I was satisfied that, despite the arguments raised by the first to fourth defendants as to the arguability of the claimants' case, the claimants could show a "good arguable case" applying that test. Taking a realistic approach, Mr McGrath QC did not seek to persuade me to the contrary.”

I do not recall anyone in that case arguing that Morimoto had changed the law or that the test was no longer the Niedersachsen test. Had they done so I would almost certainly have recorded that in the judgment since, as a judge at first instance, I had applied that test in freezing injunction cases many times. During the course of argument in the present appeal I referred to my judgment in Madoff Securities Ltd v Raven [2011] EWHC 3102 (Comm) where at [145] I said:

“To justify obtaining a freezing injunction, a claimant has to show a good arguable case on the merits. In Ninemia Maritime Corporation v Trave Schiffahrtsgesellschaft GmbH ("The Niedersachsen") [1983] 2 Lloyd's 600 at 605, Mustill J (as he then was) described a good arguable case for these purposes as "one which is more than barely capable of serious argument, but not necessarily one which the judge considers would have a better than 50 per cent chance of success". It can immediately be seen that this either is the same test as "serious issue to be tried" for the purpose of resisting a strike out application or, if there is any difference between the two tests, it is an imperceptible one.”

As Mr Sinclair KC correctly said of Zhevago it does not move the dial on the merits test. The second case since Morimoto is the decision of Edwin Johnson J in Harrington. He went through the authorities with scrupulous care at [243] to [257] of his judgment, concluding at [255]:

“I accept the submission of Mr Higgo that the law has moved on from a simple 50% test of good arguable case. It seems to me that, in applying the test of good arguable case, I should take account of the analysis of Green LJ in Kaefer, and the three limbed test as reformulated by Lord Sumption in Goldman Sachs.”

The third case is the decision of Dias J in Chowgule. At [43] and [44] she referred to Kaefer, Brownlie and Goldman Sachs. At [48] she referred to the distinction drawn by Gee on Commercial Injunctions 7^th edition [12-033] between an application for freezing relief and a jurisdiction challenge before, nevertheless, concluding at [49]:

“I accept that there is a distinction between jurisdictional challenges, where the question of jurisdiction falls to be decided once and for all, and freezing orders where there is no final determination of the merits and it is always possible to have the order set aside. That said, given the serious consequences of a freezing order – which is after all one of the law's "nuclear weapons" and carries penal sanctions – I have some doubts as to whether it would ever be appropriate to grant a freezing order, particularly in a claim asserting fraud, unless the court was satisfied that the claimant had the better of the argument as compared with the defendant. As it is, I have found myself able to reach a conclusion on the relative merits of the respective cases such that limb (iii) becomes moot.”

Mr Sinclair KC submitted that Harrington and Chowgule were the only freezing injunction cases which had applied the three-limb test as against the pre-Morimoto Court of Appeal decisions and the textbooks. He submitted that Butcher J had been right in Magomedov to conclude at [21] that Harrington and Chowgule were wrong in so far as they applied the three-limb test. Butcher J had then cited in detail the judgments of Longmore LJ and Elias LJ in Kazakhstan Kagazy and the judgment of Haddon-Cave LJ in Morimoto before concluding that the Niedersachsen test was the correct test. Mr Sinclair KC submitted that Butcher J was right.

As to why as a matter of principle the three-limb test should not apply, Mr Sinclair KC endorsed what Nugee J said in Holyoake v Candy at [15] (quoted at [38] above) about the invidiousness of trying to decide which party has the better of the argument when the Court is dealing with factual questions. He also referred to what Gee says at [12-033] of his 7^th edition, citing Derby v Weldon [1990] Ch 48 at 57:

“A requirement that the court must form the provisional view that the claimant will probably succeed at trial would be inconsistent with an approach which enables the court to achieve “its great object viz. abstaining from expressing any opinion upon the merits of the case until the hearing”…”

Mr Sinclair KC also submitted that this Court should not interfere with the decision of a Commercial Court judge as to whether there was a good arguable case. He referred to a number of statements to that effect in the previous decisions of the Court of Appeal, including the statement by Longmore LJ at [27] of his judgment in the first appeal in Lakatamia Shipping v Nobu Su quoted at [35] above. At [16] of Kazakhstan Kagazy Longmore LJ said:

“Normally if a judge of the Commercial Court decides that a claimant has a good arguable case, this court would not interfere with that conclusion unless (which is unlikely) he makes an error of law in coming to that conclusion.”

Longmore LJ came back to this point at [22] saying:

“The judge has here made an evaluation of a mass of material and this court should not interfere with that evaluation unless it is obviously wrong or the judge has misdirected himself in some way. Commercial judges have great experience in assessing what is and is not a good arguable case; indeed this judge has expressly (and justifiably) relied on his own long experience as a solicitor in one of the leading City firms in coming to his conclusions.”

At [62] Elias LJ referred with approval to what Longmore LJ had said at [27] of Lakatamia Shipping v Nobu Su. That paragraph in Longmore LJ’s judgment in Lakatamia Shipping v Nobu Su was also cited with approval by Haddon-Cave LJ in Morimoto (at [78]). This all supported the conclusion that this Court in Morimoto was not intending to alter the merits test of “good arguable case” in freezing injunction cases, which remained the Niedersachsen test.

In relation to the second part of the first ground of appeal, that whatever the test to be applied, the judge should have concluded that Unitel did not have a good arguable case, Mr Sinclair KC submitted that, if this Court concluded that the Niedersachsen test is the correct test of what is required to show a good arguable case, it was impossible to see how it could be said that Unitel did not satisfy that test. The relevant merits threshold was satisfied by the evidence of foreign law from Professor Vicente, whose CV showed that he had experience of both Portuguese and Angolan law, in contrast to Professor Ribeiro who had no experience of Angolan law at all.

Mr Sinclair KC submitted that the serious allegations made by Mr Olliff-Cooper in his submissions that Professor Vicente was not independent and had deliberately changed his evidence so that he could not be relied upon, were allegations that could not be adjudicated on at this interlocutory stage. The Court could not safely reach any conclusion to that effect, certainly without cross-examination and could only proceed on the basis that Professor Vicente was a credible and appropriate expert. Mr Sinclair KC also made the point that Ms dos Santos was not arguing that there was no good arguable case generally, for example that she was not in breach of duty as a director. The only point taken was a narrow technical one of Angolan law as to what steps are required to file an indemnity action.

He took the Court through Professor Vicente’s opinion, in particular in his second report, culminating in his conclusions at [43] which he submitted were a convincing and logical analysis of Angolan law. He contrasted Professor Ribeiro’s report, pointing out that she said at [9] that Article 80(1) was open to interpretation which was not a promising start for an argument that she was clearly right and Professor Vicente was wrong. He noted that Professor Vicente said in his third report that her procedural requirements set out at [25] of her second report were simply wrong. On the basis of their reports, it was not possible to say which expert had the better of the argument. What could be said is that Professor Vicente’s evidence was clear and compelling. Professor Vicente said that the obligation in Article 80 was satisfied by filing an application to join Ms dos Santos in the six month period. It was not necessary to have obtained permission to join her during that period. This was clearly more than barely capable of serious argument and, even if the three stage Brownlie test applied, there was a plausible well-reasoned basis for Unitel’s case.

In relation to the second ground of appeal concerning the judge’s order for costs, Mr Sinclair KC submitted that this Court would rarely interfere with a first instance judge’s decision on costs and would only do so if the judge had committed some error of principle. The judge here had made no error of principle and had exercised his discretion appropriately. It was appropriate that a defendant who had fought the granting of the freezing injunction tooth and nail and lost on every point should have to pay the claimant’s costs. He submitted that, in so far as there was a general rule in interlocutory applications, it was that the unsuccessful party would be ordered to pay the costs of the application, which was borne out by CPR 44.2 (set out at [66] above).

He accepted that in American Cyanamid cases there was some suggestion that the starting point was that costs should be reserved. However, there was an important distinction between interim injunctions which allowed a party to enforce a right which has yet to be established and a freezing injunction where the claimant establishes there and then a right or interest in not having the potential future enforcement of his substantive right claimed in the action undermined. The distinction was recognised by the Privy Council in Mercedes-Benz AG v. Herbert Heinz Horst Leiduck Co (Hong Kong) [1995] UKPC 31; [1996] AC 284, followed and applied by the Privy Council in Convoy Collateral at [83] to [85]:

“…But at this stage of the law’s development it is possible to go further and to recognise that a freezing injunction is not, on a true analysis, ancillary to a cause of action, in the sense of a claim for substantive relief, at all.

84.

It is understandable that the House of Lords should have made that assumption in The Siskina at a time when the Mareva injunction was a novelty and no proper rationale for it had yet been worked out. The assumption was compatible with the rationalisations then advanced which sought to justify the grant of Mareva injunctions on the basis either of an interest in the assets frozen or a right to the sum claimed in the action: see Mercedes Benz [1996] AC 284, 300. In Channel Tunnel the question did not arise, as the interlocutory injunction sought in that case was what might be termed an orthodox interlocutory injunction granting – on a temporary and provisional basis – the substantive relief claimed by the applicant, albeit that the claim for final relief was being pursued before another tribunal. It has been clear, at least since Mercedes Benz, that a freezing injunction is different in character. As Lord Mustill observed in the judgment of the Board in Mercedes Benz, at p 299B, “the Mareva injunction does not enforce anything, but merely prepares the ground for a possible execution by different means in the future”. Furthermore, the applicant “does not claim any interest in the assets and seeks an inhibition of dealings with them simply in order to keep them available for a possible future execution to satisfy an unconnected claim” (p 300F).

85.

Lord Mustill elaborated on these points later in the judgment when he said, at p 302, that, if an application for a Mareva injunction succeeds:

“… the relief granted bears no resemblance to an orthodox interlocutory injunction, which in a provisional and temporary way does seek to enforce rights, or to the kind of interim procedural measure which aims to make more effective the conduct of the action or matter in which the substantive rights of the plaintiff are ascertained. Nor does the Mareva injunction enforce the plaintiff’s rights even when a judgment has ascertained that they exist, for it merely ensures that once the mechanisms of enforcement are set in motion, there is something physically available upon which they can work.” (Emphasis added)

In other words, as Lord Nicholls spelt out more fully in his dissenting judgment, the essential purpose of a freezing injunction is to facilitate the enforcement of a judgment or order for the payment of a sum of money by preventing assets against which such a judgment could potentially be enforced from being dealt with in such a way that insufficient assets are available to meet the judgment.”

Mr Sinclair KC submitted that another distinction between the two types of injunction was that unlike in interim injunction cases, in freezing injunction cases, the balance of convenience does not arise. He submitted that the approach in Picnic at Ascot and Melford Capital was too prescriptive and should not be applied in freezing injunction cases. It was not appropriate that costs should be reserved merely because it was reasonably foreseeable that the exercise of discretion as to costs might be affected by the outcome of proceedings at trial. That argument could be made in relation to many interlocutory applications, for example a disclosure application. Whilst, in a number of interim injunction cases, costs had been reserved, in the freezing injunction cases, the unsuccessful defendant had been ordered to pay costs in all the cases, bar one, the decision of HHJ Davis-White QC in Al Assam. The point made by the judge in that case in the last two sentences of [252] (cited at [64] above) could be made in the case of any interlocutory application, such as a disclosure application. Mr Sinclair KC submitted that HHJ Davis-White QC had been in error and Edwin Johnson J had been right in his costs judgment in Harrington to decline to follow Al Assam. The analysis in Harrington at [26] to [36] was correct.

Mr Sinclair KC submitted that the judge in the present case had exercised his discretion correctly. The application for the freezing injunction had been made on notice and Ms dos Santos had had a long time to decide whether to respond to the application. She had fought every limb of the application and lost on them all. He endorsed what the judge had said at [8] of his costs judgment (quoted at [24] above) about the circumstances of this case justifying an order for costs against the defendant and submitted that this Court should not interfere with the judge’s order.

In my judgment, the correct test as to what constitutes a good arguable case for the purposes of the merits threshold for the grant of a freezing injunction is that formulated by Mustill J in The Niedersachsen, applied by first instance judges many times over the last forty years and endorsed by at least three more recent decisions of this Court. A “good arguable case” in the freezing injunction context is not to be assessed by reference to the three-limb test derived from Brownlie to determine whether a claim falls within one or more of the jurisdictional gateways for the purposes of permission to serve out of the jurisdiction and the recent decisions in Harrington and Chowgule which adopt that test were wrong to do so.

I have reached that clear and firm conclusion for a number of reasons. First, although, in one of the earliest Mareva injunction cases, Pertamina, Lord Denning MR adopted the good arguable case test from jurisdictional gateway cases because he thought it appropriate, at least where the defendant was outside the jurisdiction, which of course was the position at the time, since it was thought such an injunction could not be granted against an English defendant, the law has moved on since then. Section 37(3) of the Senior Courts Act 1981 made it clear that such an injunction could be granted against an English defendant and, as Elias LJ pointed out in Kazakhstan Kagazy at [67] (quoted at [72] above), there have been developments in the law both as to jurisdictional gateways and as to freezing injunctions (including that, as I have just pointed out, they are now available against both foreign and English defendants) and although the same phrase “good arguable case” is used in both situations, there is no reason why what Elias LJ described as an “inherently flexible concept” should have the same meaning in both situations.

The same point was made by Longmore LJ in Kazakhstan Kagazy at [25]. As Butcher J put it in Magomedov at [23]:

“That there have been developments in the law on the test to be applied in relation to jurisdiction gateways does not mean that there has or should have been a change in the law in relation to the test to be applied in relation to freezing injunctions. As Longmore LJ said in Kazakhstan Kagazy v Zhunus at [25]:

'But I see no reason why that test [viz that which was applicable in the jurisdictional gateway context] should apply to freezing injunctions where ex hypothesi (or subject to any jurisdictional challenge) the defendant is properly before the court.'

While it is correct that Longmore LJ was there considering a test in the context of jurisdictional gateways of 'much the better of the argument', which was subsequently refined to 'the better of the argument', that refinement is immaterial to the question here. What is significant is that in Kazakhstan Kagazy v Zhunus Longmore LJ identified that there was no reason why the test for those purposes should be that for freezing orders.”

Second, there are obvious differences between whether the threshold of a jurisdictional gateway has been reached for the purposes of service out and whether the merits threshold for the grant of freezing relief has been reached. Mustill J himself highlighted one aspect of the distinction in The Niedersachsen in the passage quoted at [73] above. In my judgment it is important to keep in mind that, in the case of freezing injunctions, the test is applied to satisfy a merits threshold in circumstances where the Court will (unless the case settles) determine the merits at trial, hence the relatively low threshold set by The Niedersachsen. In contrast, in many cases, although the evidence at trial may include what was before the Court on an application to serve out, the Court will only have to determine, applying the three-limb test, whether the case falls within one or more of the jurisdictional gateways at the stage of the application for permission to serve out. Whether or not the case does fall within a gateway does not involve applying a merits test as such. The merits of the case are addressed by the requirement in CPR 6.37(1) that it is stated by the claimant that it believes that the claim has a reasonable prospect of success. The issue of whether the case falls within a gateway will not be revisited at trial.

The third reason follows on from the second and is that, as has been stated in a number of cases, particularly in cases where there is a dispute on the evidence, it is invidious for the Court, at the early stage at which a freezing injunction is usually sought, to have to determine which party has “the better of the argument”. This was pointed out by Nugee J in Holyoake v Candy at [15] (quoted above at [38]). The point was cogently developed by Butcher J in Magomedov at [27]-[28]:

“27.

There appear to me to be good reasons why the three-fold test applied in those cases should not be applied in the context of freezing orders. That test, at least as to the first two limbs, involves a relative assessment of the parties' positions. The making of such a relative assessment is liable to draw the parties and the court into the conduct of 'mini-trials'. A relative assessment encourages the parties to bring forward at this early stage, every piece of evidence which might suggest that they have the better of the argument. This is likely to lead to more of the court's resources being absorbed in interlocutory hearings brought on, very often, on an urgent basis. This is deprecated in the authorities, and would place an even greater burden on the court, where the number and scale of urgent applications is already causing strains. Moreover, to apply such a test in the context of freezing orders would widen, without apparent reason, the gap between the merits test to be applied in relation to interlocutory applications for proprietary injunctions, which is the American Cyanamid test of a serious issue to be tried (see, for example Haque v Hussain [2020] EWHC 2739 (Ch), and Gee on Commercial Injunctions (7^th ed), 2-022, 12-027), and that applicable to applications for freezing orders.

28.

Further, I apprehend that to adopt a test which involves a relative assessment of the parties' positions, at least at the first two stages is to put the merits bar too high to serve the interests of justice. In the type of cases in which freezing orders are very often sought, including cases of alleged fraud, dishonesty, bad faith and the like, it may be difficult for an applicant to demonstrate, at an early stage and prior to disclosure, that it has the better of the argument on the merits. While I fully recognise that the gravity of a freezing order requires a merits test markedly higher than simple arguability, I consider that there is a danger that the adoption of the Brownlie test in relation of freezing orders may deny to victims of wrongdoing the interim protection which the freezing jurisdiction is designed to provide.”

Furthermore, as I have already indicated, the weight of authority supports the application of The Niedersachsen test in determining the merits threshold for the grant of freezing injunctions. It is the test which has been consistently applied by judges in deciding whether to grant such injunctions in the intervening years, at least until Harrington and Chowgule. It is a test approved by the Court of Appeal in The Niedersachsen itself (see [74] and [75] above) and in several cases since.

In the first appeal in Lakatamia Shipping v Nobu Su Longmore LJ at [25]-[27] concluded that the judge at first instance had been right to apply the Niedersachsen test. As Popplewell LJ rightly said, the suggestion that the case was somehow decided per incuriam because Canada Trust was not cited, is unreal. In Kazakhstan Kagazy, at [25], Longmore LJ said in terms that the judge at first instance had set the hurdle too high by applying the Canada Trust case and that the right test in a freezing injunction case was that set out in The Niedersachsen. Elias LJ was also very clear that the Niedersachsen test was the correct test in the passages at [66] and [67] of his judgment quoted at [76] and [72] above respectively. The Niedersachsen test was also endorsed in the context of the Chabra jurisdiction, by Tomlinson LJ in the second Lakatamia Shipping v Nobu Su appeal in the passage set out at [37] above.

As already noted, Mr Margolin KC placed considerable reliance on the decision of the Court of Appeal in Morimoto. I agree with Mr Sinclair KC that the fact that, at [35], Haddon-Cave LJ cited the passage in Gee on Commercial Injunctions at what is now [12-033] of the 7^th edition, that the test is not a particularly onerous one, does not suggest that the Court was intending to change the law and substitute the three-limb test from Brownlie. That is surely borne out by the fact that in the same paragraph [12-033] Gee says in terms that: “the test on the strength of the merits needed for Mareva relief is not the test used to ascertain whether the claimant has brought itself within a jurisdictional gateway.” Furthermore, in the previous paragraph, [12-032] Gee cites the Niedersachsen test with apparent approval.

I also agree with what Butcher J said about Haddon-Cave LJ’s judgment at [33] to [35] of Magomedov:

“33.

The passage which is principally relevant is that at [33]-[38] in the judgment of Haddon-Cave LJ. That it was not intended there to effect any significant change in the tests applicable to the grant of freezing injunctions is strongly indicated by what Haddon-Cave LJ says at para. [33], namely 'The basic legal principles for the grant of a WFO are well-known and uncontroversial and hardly need restating.'

34.

Furthermore, Haddon-Cave LJ referred, at [35], to Gee on Commercial Injunctions (6^th ed), and to the merits test being 'not a particularly onerous one.' While he referred to the analysis of the concept of a 'good arguable case' 'in the context of jurisdictional gateways' at [38], he did not say that this was applicable to applications for freezing orders; and he did not set out the three-fold test. His concluding sentence, that 'the central concept at the heart of the test was "a plausible evidential basis"', is not, as I understand it, a disapproval of the 'Niedersachsen test'. It is rather, in my view, a recognition that in this context, in relation to disputed questions of fact, a case will not be more than barely arguable if there is no plausible evidential basis for it.

35.

It does not appear that Kazakhstan Kagazy v Zhunus was cited. Furthermore, the issue of whether the merits test for a freezing order was or was not the same as the test of a good arguable case in relation to jurisdictional gateways does not appear to have been the subject of argument, and was certainly not the focus of the appeal. The notice of appeal was concerned with challenging the judge's findings as to risk of dissipation (see [39]). The respondent's notice challenged the judge's findings as to a good arguable case, but Haddon-Cave LJ indicated that those points were probably not open to the respondent, and in any event were disposed of on the basis that appeals on jurisdictional issues are deprecated and the judge had not made a clear error of principle (see [71]-[76])

Accordingly, I do not consider that Haddon-Cave LJ in Morimoto was intending to change the law or to assimilate the merits test for freezing injunctions with the three-limb test for jurisdictional gateways. Indeed, that he was approving the Niedersachsen test is indicated by his citation of that case in [37] of his judgment.

During the hearing of this appeal, there was some discussion between counsel and the Court about whether the test of “good arguable case” in the context of freezing injunctions should be assimilated with “serious issue to be tried” as part of the American Cyanimid test in the context of interlocutory injunctions generally. In Madoff Securities v Raven (quoted at [81] above) I suggested that the difference between the two tests is an “imperceptible one”. Nonetheless, when I read what Nugee J said in Holyoake v Candy at [10] (quoted at [78] above) I was initially inclined to agree with him that “good arguable case” is a slightly more stringent test than “serious issue to be tried” on the basis that this was appropriate given the invasive nature of freezing relief. However, having read the judgment of Popplewell LJ in draft and considered the matter further, I am persuaded that the view I expressed in Madoff Securities v Raven is correct and that, in the context of interlocutory injunctions, the two tests should be equated.

The Court also asked counsel whether any assistance could be derived from the Commonwealth authorities on Mareva injunctions. We were greatly assisted by written submissions from both parties between the two hearings of the appeal. What emerges is that the Niedersachsen test of good arguable case has been adopted in a number of Commonwealth jurisdictions. This is the case in both federal and state courts in Australia (with the possible exception of Victoria where what seems to be a weaker test of real prospect of success has been adopted). The High Court of Australia in Cardile v LED Builders Pty Ltd [1999] HCA 18 at [68] articulated a merits test of: “a reasonably arguable case on legal as well as factual matters.”

Multiple Australian judges have found this to be the same test as the Niedersachsen test or indistinguishable from it: see for example per Edelman J in Guardian Melbourne Pty Ltd v Carlei [2016] FCA 72 at [18], a passage recently approved by the Federal Court in Protelight Pharmaceuticals v Wen [2024] FCA 581:

“The phrase ‘a good arguable case’ has a long history. Its provenance may have been in the judgment of Mustill J, as his Lordship was then, in Ninemia Maritime Corp v Trave Schiffahrtsgesellschaft mbH & Co KG [1984] 1 All ER 398, 404. The phrase has been treated as the equivalent of the general law requirement explained by Gaudron, McHugh, Gummow and Callinan JJ in Cardile v LED Builders Pty Ltd [1999] HCA 18; (1999) 198 CLR 380 at [68] that the applicant must establish that it has a reasonably arguable case on legal and factual matters…”

In 2006, the Council of Chief Justices of Australia and New Zealand produced a harmonised set of procedural rules on freezing orders which have been adopted throughout those jurisdictions and which include an express “good arguable case” merits threshold, which as the jurisprudence both before and after 2006 demonstrates, is based on the same or a similar test as in The Niedersachsen, not based on the jurisdictional gateways test. The same conclusion emerges from the New Zealand jurisprudence, which interprets a good arguable case as one which requires the applicant to establish the allegations are tenable and are supported by evidence, having regard to the stage of the proceedings, and does not require the applicant to establish a prima facie case: see most recently the decision of the New Zealand Court of Appeal in Gao v Deng [2023] NZCA 200 at [26].

The Singapore courts have expressly approved the Niedersachsen test. In Bouvier v Accent Delight International Ltd [2015] 5 SLR 558 at [36], the Singapore Court of Appeal expressly approved and applied the good arguable case merits test in The Niedersachsen. The Hong Kong courts have also approved and applied the Niedersachsen test as well and in doing so in Delco Participation BV v Chiho Tiande Group Ltd [2019] HKCFI 2646, expressly approved the rejection of the jurisdictional gateways test by the Court of Appeal in Kazakhstan Kagazy. In a series of recent decisions the Court of Appeal of the Eastern Caribbean, the appellate court for the British Virgin Islands, has applied the Niedersachsen test: see Multibank FX International Corporation v Von Der Heydt Invest SA (2023) per Webster JA at [44] and Svirsky v Oyekenov (2023) per Ventose JA at [30].

Of Commonwealth jurisdictions where WFOs are regularly granted, Canada appears to be something of an outlier, with courts in most provinces applying a stricter test of “strong prima facie case”, on the basis that a freezing injunction is seen as a harsh and powerful exception to the basic rule against execution before judgment set out in Lister v Stubbs (1890) 45 Ch D 1: see per Estey J giving the judgment of the Supreme Court of Canada in Aetna Financial Services v Feigelman [1985] 1 SCR 2 at [30].

However, as I have said, with the exception of Canada, the other Commonwealth jurisdictions where WFOs have been granted have essentially applied the Niedersachsen test. Given that consistency of approach in this area would seem desirable, this is further support for the clear conclusion I have reached that the correct test of “good arguable case” is that set out in The Niedersachsen.

The second part of the first ground of appeal concerns whether, whichever test of “good arguable case” is applicable, Unitel satisfies the merits threshold on the material before the Court. This aspect can be dealt with briefly, since I agree with Mr Sinclair KC that the analysis of Untiel’s expert Professor Vicente, that the obligation in Article 80(1) of the ACCL to file an indemnity action was satisfied by the filing of the application to join Ms dos Santos to the UIH Claim in the six month period and that it was not necessary to have obtained permission to join her during that period, is clear and compelling. The suggestion on behalf of Ms dos Santos that Professor Vicente has changed his evidence so that he cannot be said to be independent, is an untested assertion without either expert having given oral evidence or been cross-examined, so that, in my judgment, it would not be fair or appropriate to reach such a conclusion at this interlocutory stage. Although Ms dos Santos’ expert Professor Ribeiro did not accept the analysis of Professor Vicente, she did accept that Article 80(1) was open to interpretation which, as Mr Sinclair KC submitted, was not a promising start for an argument that she was clearly right and Professor Vicente was clearly wrong as to the meaning of the provision. On the material before the Court it is impossible to say that Unitel does not satisfy the Niedersachsen test. Furthermore, even if the correct test were the three-stage Brownlie one, like the judge below, I consider Unitel would satisfy that test, since in my judgment, Unitel has much the better of the argument that, based on Professor Vicente’s analysis, the application to join Ms dos Santos to the UIH Claim is equivalent to the filing of a new claim in Angola and thus satisfies Article 80(1).

Given the conclusion I have reached on the first ground of appeal, it is not necessary to consider the additional grounds upon which Unitel seeks to uphold the judge’s judgment (set out at [26] above).

In relation to the second ground of appeal I am very firmly of the view that this Court should not interfere with the judge’s order that Ms dos Santos should pay the costs of the freezing injunction application.

In so far as there is a general rule as to the costs of contested interlocutory or procedural applications, it is that a party who contests an application and fights it tooth and nail on every point, thereby causing the successful party to incur costs which would not otherwise be incurred, should be ordered to pay the successful party’s costs at the conclusion of the application. This is clear from CPR 44.2(2) and is the general rule applied in the Business and Property Courts in relation to contested interlocutory applications. The Court will not usually reserve costs to the trial judge of, for example, a contested jurisdiction or disclosure application which the defendant has lost, merely because the defendant points out that it might succeed in defeating a claim at trial. Were it otherwise trial judges and, in turn costs judges, would be inundated with having to make rulings on costs of interlocutory applications which had been reserved by the judges who heard the applications.

Of course the Court has a discretion to make a different order on a contested interlocutory application, including reserving the costs to the trial judge, as CPR 44.2(b) provides. One situation in which the Court will usually make an order that the costs be reserved is in the case of an American Cyanamid interim injunction as the authorities from Desquenne onwards establish. However, that is because, on the balance of convenience, the Court is prepared to grant an interim injunction which allows a party to rely upon a right or obligation, the existence of which has yet to be established, effectively holding the ring pending the trial. If at trial the right or obligation is established then the injunction can be made final and permanent or other relief granted. However if the claimant’s case fails at trial, then it can generally be said that the interim injunction should not have been granted, since the right or obligation did not exist or was not established. Hence it is generally more appropriate for the costs of the application for the interim injunction to be reserved to the trial judge.

However, the position is different in the case of a freezing injunction. If the claimant establishes the three criteria referred to in [6] above: (1) a good arguable case on the merits; (2) a real risk that a future judgment would not be met because of an unjustified dissipation of assets; (3) that it would be just and convenient in all the circumstances to grant the freezing injunction, then the Court will grant the injunction. When granted it is not “interim” or dependent on the balance of convenience like an American Cyanamid injunction, nor will the Court make the injunction final at trial, as in the case of an interim American Cyanamid injunction. As Edwin Johnson J pointed out at [29] of his costs judgment in Harrington there is no such thing as a final freezing order. Subject to any subsequent application to vary or discharge it, the freezing injunction remains in place until trial. If the claim succeeds the Court may continue the injunction post judgment but that is not the making of a final injunction. The purpose of the freezing injunction remains as set out at [85] of Convoy Collateral: “to facilitate the enforcement of a judgment or order for the payment of a sum of money by preventing assets against which such a judgment could potentially be enforced from being dealt with in such a way that insufficient assets are available to meet the judgment.”

Another important distinction between a freezing injunction and an American Cyanamid injunction is that whereas, in the case of the latter, if the relevant right or obligation is not established at trial it can generally be said that the interim injunction should not have been granted, in the case of the former even if the claim fails at trial, it does not follow that the freezing order was not correctly granted on the basis that the claimant satisfied the three criteria for the grant of the freezing injunction. This point was made by Martin Spencer J at [52] of Bravo (cited at [60 above) and by Edwin Johnson J at [29] of Harrington (cited at [67] above) and I agree with their analysis, which I consider is to be preferred to that of HHJ Davis-White QC in Al Assam (although I recognise that his decision on costs may have been justified on the facts of that case). I also note that in a recent judgment handed down since the hearing of the appeal, Cancrie Investments Ltd v Haider [2024] EWHC 2302 (Comm), Nigel Cooper KC (sitting as a Deputy High Court Judge) preferred the analysis in Bravo and Harrington to that in Al-Assam: see [20] to [31] of that judgment. In my judgment, Mr Margolin KC’s “fundamental” point set out at [59] above, that if the claim failed at trial, the freezing injunction should not have been granted in the same way as in the case of an interim American Cyanamid injunction, is misconceived.

Accordingly, I consider that Bright J was entitled to reach the conclusion that Ms dos Santos should pay the costs of the freezing injunction application which she had fought and lost and that the judge exercised his discretion appropriately. I would have reached the same conclusion.

In all the circumstances, I consider that the appeal must be dismissed on both grounds.

I agree that the appeal should be dismissed for the reasons given by the Chancellor. I add some words of my own because the time has come, in my view, to recognise that the gateway merits test for a freezing order is and should be the same as that for interim injunctions generally, namely whether there is a serious issue to be tried. That is so both as a matter of principle and because it is no different in substance from the test applicable to freezing orders of ‘good arguable case’, in the sense defined in The Niedersachsen.

The merits test of serious issue to be tried applies in two types of interlocutory application. In jurisdiction applications it is the test applicable to the merits of the substantive claim, rather than whether the claim falls within a jurisdictional gateway (Altimo Holdings v Kyrgyz Mobil Tel Ltd [2011] UKPC 7 [2012] 1 WLR 1804 at [71]); in this context it is the same as the ‘real prospect of success’ test for defeating a summary judgment application (Altimo at [71]; Lungowe v Vedanta Resources plc [2019] UKSC 20 [2020] AC 1045 at [42]).

It is also the merits test in interim injunction applications other than freezing orders (American Cyanamid Co v Ethicon Ltd [1975] AC 396 at p. 407G). In this context too it bears the same meaning as in the context of summary judgment, namely whether the claim has a real prospect of success. At p. 408A-B Lord Diplock said:

“So unless the material available to the court at the hearing of the application for an interlocutory injunction fails to disclose that the plaintiff has any real prospect of succeeding in his claim for a permanent injunction at the trial, the court should go on to consider whether the balance of convenience lies in favour of granting or refusing the interlocutory relief that is sought.”

When addressing the American Cyanamid test of ‘serious issue to be tried’ in Planon Ltd v Gilligan, Nugee LJ said, obiter, at [102]: “This is not a demanding test, and it really only serves to exclude the case where the claim is frivolous or vexatious, or otherwise demonstrably bad.” This reflects the language used by Lord Diplock in the American Cyanamid case at p. 407 but I do not take Nugee LJ there to have meant that it is a lower test than that used for summary judgment applications. A claim which cannot show a real prospect of success, so as to be capable of surviving a reverse summary judgment application, is one which should not be pursued and should not attract any form of interlocutory relief. In American Cyanamid Lord Diplock identified the test at p. 408 as being whether the claimant can establish a real prospect of succeeding in his claim. In Sportsdirect.com Retail Ltd v Newcastle United Football Club Ltd [2024] EWCA Civ 532 Sir Geoffrey Vos MR equated the tests at [29]:

“The question is whether the material available to the court shows that the claimant has a real prospect of succeeding in its claim for a permanent injunction. If so, the court proceeds to consider the balance of convenience.”

What the summary judgment test of ‘real prospect of success’ means has been the subject of considerable jurisprudence which it is not necessary to explore at length. In a summary judgment application by a claimant it applies to the merits of the defence; and in a reverse summary judgment application by a defendant it applies to the claim. It is not enough that the defence or claim respectively is merely arguable; it must carry some degree of conviction: ED & F Man Liquid Products Ltd v Patel [2003] EWCA Civ 472 at [8]; AC Ward & Sons Ltd v Catlin (Five) Ltd [2009] EWCA Civ 1098 approving the principles summarised by Lewison J (as he then was) in Easyair Ltd v Opal Telecom Ltd [2009] EWHC 339 (Ch) at [15]. A claim which is more than merely arguable and carries some degree of conviction is no different in substance from one which is more than barely capable of serious argument, which is the Niedersachsen test. I respectfully agree with the observation made by the Chancellor in Madoff Securities Ltd v Raven, quoted at [81] above, that there is no perceptible difference between the two tests.

In Holyoake v Candy, quoted at [78] above, Nugee J, as he then was, expressed the view, obiter, that it was appropriate for ‘good arguable case’ to be a slightly more stringent test than ‘serious issue to be tried’ given the invasive nature of freezing relief. However for my part, I would not treat the nature of freezing order relief as justifying any such distinction.

I readily accept that freezing order relief is invasive, and no one with experience of how such orders operate on defendants in practice can be unaware that in some cases they can operate harshly, and not only by reason of the disclosure requirements which often accompany them. In the past such orders were famously described by Donaldson LJ in Bank Mellat v Nikpour [1985] F.S.R. 87 at 92 as one of the law’s two ‘nuclear weapons’, although the number which are now granted in the Business and Property Courts suggests that this description is inapt, dating from a time when the jurisdiction was in the infancy of its development. It now has its justification firmly rooted in the protection of prospective rights and interests in what is termed the enforcement principle, as the Privy Council explained in Convoy at [84]-[89], and although there are important differences between freezing orders and other forms of interim injunctions, as the Chancellor has explained when addressing the second ground of appeal in this case, there seems to me no justification for a different merits test in the nature of the interests being protected. For interim injunctions the right being protected is a substantive legal right, assuming it to be a substantive right which will be vindicated at trial by a judgment; for freezing orders it is a legally protected interest in a judgment being enforceable, again on the assumption that there is a substantive right which will be vindicated at trial by a judgment. The same prospect of establishing the substantive right at trial should stand as the merits threshold for both.

Moreover, other forms of interim injunction to which the American Cyanamid threshold applies can operate just as invasively as freezing orders, for example in the context of proprietary injunctions or those restraining employment or use of information.

Nor can I see any logic in seeking to control the grant of freezing orders through a heightened merits test as a gateway. Rather, the invasive nature of the relief should be taken into account in considering the other aspects of the test which are required to be fulfilled; in the safeguards built in to the wording of the orders in the form of exceptions; and in the application of the cross-undertaking in damages. I understand the concern that freezing orders should not be granted too readily, and fully endorse the proposition that care should be taken to ensure that they do not operate unfairly. It is always necessary to give anxious scrutiny not only to the second limb of the test, real risk of dissipation, but also to the third, whether it is just and convenient to make the order. Although this has been expressed as the third limb of the test, it is ultimately the whole test expressed in s. 37 Senior Courts Act 1981, and should be considered in every case, having regard among other things to the effect of granting, or not granting, the order. It may come to the forefront in the context of applications to set aside a freezing order, or to vary it so as to permit particular expenditure or transactional activity, the restraint of which represents the invasive nature of the order. It is by reference to the just and convenient criterion that the apparent strength of the claim may fall again for consideration, as both Mustill J and Kerr LJ recognised in The Niedersachsen at pp. 402j-403a, and 415d-e respectively, just as it does where interim injunctions may be finally determinative: Lansing Linde Ltd v Kerr [1991] 1 WLR 251; National Commercial Bank Jamaica Ltd v Olint Corpn Ltd [2009] UKPC 16; [2009] 1 WLR 1405; Planon v Gilligan at [103], [110].

For these reasons there is not, and should not be, any distinction to be drawn between the ‘good arguable case’ test for freezing orders and the test of ‘serious issue to be tried’ for other forms of interim injunctions. That being so, it would be preferable to use the latter in the context of freezing orders and to restrict the use of the expression ‘good arguable case’ to the context of jurisdictional gateways, where it bears a different meaning in accordance with the principles explained in Brownlie. It is obviously unsatisfactory for an expression used to define a merits test to mean something different in one context (freezing orders) from that in another (jurisdictional gateways); that is likely to give rise to confusion and misunderstanding, as is well illustrated by the cases to which the Chancellor has referred.

I agree with both judgments.