By the time of Deutsche Bank v Sebastian Holdings and Alexander Vik  EWCA Civ 23 in 2016 a certain world weariness to the question of the search for principle can be detected by the Court of Appeal. There is a lot in this case which repays careful reading, including the extent to which a non-party can be bound by findings made during a trial, to which he is not a party.
The background to the appeal lies in proceedings between the Bank and Sebastian relating to the operation of accounts maintained by Sebastian with the Bank for trading in foreign currencies, shares and financial products. Except to the limited extent identified later in this judgment, it is unnecessary for present purposes to describe the course of dealing on the accounts or the relationship between the parties, other than to say that in January 2009 the Bank began proceedings against Sebastian in this country to recover the sum of approximately US$250 million due principally in respect of amounts owed on the closing out of various trading positions. Sebastian brought a counterclaim against the Bank for approximately US$8 billion in respect of the losses it alleged it had suffered as a result of the Bank’s breaches of contract in forcing it to close out certain open positions contrary to its wishes. In November 2013, after a trial lasting 44 days, Cooke J. gave judgment for the Bank on its claim in the sum of US$243,023,089 and dismissed Sebastian’s counterclaim. The judge also ordered Sebastian to pay 85% of the Bank’s costs, which are said to amount to about £60 million, on the indemnity basis. We shall refer to the proceedings between the Bank and Sebastian as “the main action”.
The appellant, Mr. Alexander Vik, was at all material times the sole shareholder and sole director of Sebastian, a company incorporated in the Turks and Caicos Islands which he used as a personal investment vehicle. Sebastian, which is now said to have no assets, failed to make any payment in respect of the judgment or the Bank’s costs and the Bank therefore applied to join Mr. Vik as a defendant with a view to obtaining an order that he pay the costs of the proceedings himself. The application was made on the basis that Mr. Vik owned and controlled Sebastian, that he had directed the litigation on its behalf, that he had funded the litigation, or had made funds available to enable Sebastian to pursue it, that it had been conducted for his personal benefit and that therefore he was the “real party” to the litigation. After a two-day hearing, at which Mr. Vik was represented by Leading and Junior Counsel, Cooke J. made the order to which we have referred. The order did not expressly state that the sum of £36,204,891 was to be paid on account of the Bank’s costs, but it is clear from the opening paragraph of his judgment that that was what the judge intended.
In the event, the Court chose to deal with the issues of principle fairly shortly:
It will be apparent from what we have said that Mr. Cogley sought to place great emphasis on the Symphony guidelines to the point of treating them, in particular the third guideline, as laying down requirements that must be satisfied unless the applicant can demonstrate a good reason for failing to do so. In our view that is not the correct approach. When considering those guidelines it is important to bear in mind that they were formulated not very long after the decision in Aiden Shipping v Interbulk, at a time when applications for costs against third parties were relatively uncommon, and that they were intended merely to provide guidance, not to lay down rules. Since then there have been many more applications for orders for costs against third parties under a wide variety of circumstances, as a result of which it has come to be recognised more clearly than perhaps it was at that time that each case turns on its own facts.
As all three members of the court observed in Petromec v Petrobras, the exercise of the discretion is in danger of becoming over-complicated by authority. The decision of the Privy Council in Dymocks, which contains an authoritative statement of the modern law, explains and interprets the Symphony guidelines in a way which reflects the variety of circumstances in which the court is likely to be called upon to exercise the discretion. Thus, the Privy Council has explained that an order of this kind is “exceptional” only in the sense that it is outside the ordinary run of cases where parties pursue or defend claims for their own benefit and at their own expense. Similarly, it has made it clear that the absence of a warning is simply one factor which the court will take into account in an appropriate case when deciding whether, viewed overall, it would be unjust to exercise the discretion in favour of making an order for costs against the third party. We think it important to emphasise that the only immutable principle is that the discretion must be exercised justly. It should also be recognised that, since the decision involves an exercise of discretion, limited assistance is likely to be gained from the citation of other decisions at first instance in which judges have or have not granted an order of this kind.
It can be observed that although on one reading the Court of Appeal were content to blur the question of a principled exercise of the discretion, the Privy Council before them and the Supreme Court after them, in another case I shall look at in a later post, have not abandoned the search for principle. The danger with any unbridled discretionary power is that the potential range of decisions becomes so wide, and so unpredictable that it imperils a key tenet of the rule of law that laws should be certain and settled in their application.
One issue that had largely disappeared from the jurisprudence made a reappearance in 2018 and, was the question as to whether a failure to give a warning to the proposed non-party of a potential application was fatal to the making of an order. It will be noted that in many contexts a warning will be essentially irrelevant: if someone is contending that they lack control of the litigation, what difference would a warning make? In some contexts, absence of a warning can be very important: particularly in the context of a director and an insolvent company. Thus, in Sony/ATV Music Publishing LLC v WPMC Ltd (In Liquidation)  EWCA Civ 2005:
Just over a year after judgment had been handed down, on 7 July 2016, SATV wrote to Mr David Bailey, the director and majority shareholder of WPMC, intimating for the first time that they intended to seek an order against him under s 51(3) of the Senior Courts Act 1981that he pay most of the costs of SATV’s claim against WPMC and Iambic. An application to that effect was made on 3 August 2016. Mr Bailey resisted the making of the non-party costs order (“NPCO”) on the basis that it would be unjust to make him pay the costs of the action. Amongst the reasons which he gave was that he had not been warned at any stage up to 7 July 2016 that this was a step which SATV intended to take in the event that they succeeded in the action.
The court observed what earlier authority had noted about warnings:
In Dymocks the non-party (“Associated”) relied on the failure to warn of the intention to make the costs application against it. Of this, Lord Brown said at :
“The authorities establish, however, that this [i.e. failure to warn] is no more than a material consideration in the case … and their Lordships are unable to see how an earlier warning could have made any difference to the course of the proceedings here. It is not suggested that Associated would have acted differently in the event of an earlier warning. Nor could they sensibly have been made a party to the litigation at any earlier stage. There is some force, moreover, in Dymocks’ submission that, until the appeal hearings were completed, they were unclear whether or not Associated would stand behind the Todds so as to avoid their bankruptcy.”
In Dymocks itself Lord Brown recorded that it was not suggested that Associated would have acted differently in the event of an earlier warning. In Deutsche Bank the failure to warn was dismissed on the basis that the judge had been “entirely satisfied that a warning at an earlier stage would have made no difference to the conduct of the proceedings”. The judge had held that it did not lie in the non-party’s mouth to say that his evidence would have been different or that he would have conducted the case in a different way. Moore-Bick LJ went on to point out that the judge was “of course, uniquely well placed to understand the way in which the case had been conducted on behalf of Sebastian and to assess [the non-party], both as a litigant and a witness”.
The court therefore concluded:
I consider that where there is credible evidence that a party would have acted differently if he had been warned then that evidence should be given weight in the overall assessment. The weight to be given to it is of course a matter for the judge.
In the instant case, the lack of a warning was decisive:
The absence of warning is a factor whose relevance will vary from case to case. In the present case, however, because the judge has made a finding that Mr Bailey would not have acted any differently if a warning had been given, he has given it no weight at all. Thus, at  he says that none of the factors relied upon by Mr Bailey supported the conclusion that it would be unjust to make the order.
I think the judge did fall into error in concluding that this was a case where no weight could be given to the fact that Mr Bailey was not warned in deciding whether it was just to make a NPCO. In doing so he left out of account a feature which he should have considered, namely the prospect that Mr Bailey would have conducted the defence of the case differently if a warning had been given.
My reasons for coming to that conclusion are as follows. Firstly, as I have explained, this was a case where caution was necessary in making summary findings of fact relevant to the making of the NPCO. Mr Bailey had not been a witness in the action, and the judge had had no opportunity to assess his credibility. Secondly, there was no reason to doubt Mr Bailey’s honesty in giving the evidence as to how he would have behaved in the event of a warning. It was, after all, his behaviour which was in issue, and he could be said to be uniquely placed to assess how he would have behaved. The fact that evidence is given with hindsight, as it often is, does not necessarily mean that it is *691 not reliable. Mr Bailey had given evidence in the clearest possible terms as to how he would have behaved if he had known that he was running the risk of a NPCO. No request had been made to challenge this evidence by applying to cross-examine him. Thirdly, the reasons given by the judge for coming to the conclusion that the warning would not have caused Mr Bailey to act differently were inferential ones, namely that Mr Bailey was motivated by recouping money for his fellow GLE investors, by the advice he had received from his lawyers and by the fact that they were prepared to act on a CFA. Those factors do not mandate a conclusion that Mr Bailey would have acted in the same way if he had known that he would face a costs bill for several hundreds of thousands of pounds. It is one thing to help out fellow investors if there is no appreciable downside, quite another if it is to going to result in a large costs liability if the defence is unsuccessful. No one suggests that the chances of success given to WPMC were such that the downside would not have been considered in the event of a warning. The motivation provided by the fact that Mr Wilson and McFaddens were prepared to act on a CFA might have insulated Mr Bailey from funding WPMC’s costs, but Mr Bailey would have been advised that it could have no impact on his liability for SATV’s costs.
It is true, in a literal sense, that none of the factors which the judge relied on would have changed if a warning had been given. The motive to assist the fellow investors, the advice given by the lawyers and the CFA would have still been in place. However they would have required a radical review if the prospect of a NPCO had been brought to Mr Bailey’s attention. He was an experienced entrepreneur, used to assessing risk against reward. The judge was not entitled, in my judgment, to reject Mr Bailey’s evidence as to how he would have behaved in the event of a warning.
I also consider that it is fair to take into account the fact that Mr Bailey could have protected his position by ATE insurance. The judge rejected this point because there was no evidence that Mr Bailey asked for or was given advice about the remedies which were open to SATV if no ATE cover was obtained and WPMC lost. However, it was obvious that Mr Bailey and SATV were operating on the assumption that any costs order made against WPMC would not be met. It is not clear to me therefore what should have prompted Mr Bailey to ask for such advice. However, if he had been warned that a NPCO was being sought against him, there was every reason for him to ask for and *692 obtain appropriate advice as to how he might protect himself against such an order.
The court went on to allow the appeal:
Thus far I would regard matters as fairly evenly balanced. It might be said that Mr Bailey, standing as he did to benefit from the outcome, was “a real party” if not the only one. However the absence of any form of warning is, in my judgment, fatal to the application for the NPCO. It is plain, as the judge indeed held, that SATV knew or should have appreciated that WPMC would not be able to pay their costs in the event that the claim succeeded, and they knew that WPMC, and Mr Bailey with whom they dealt directly, were operating on the same assumption. In those circumstances the failure to warn until a year after final judgment is given strikes me as manifestly unfair to Mr Bailey. It would be unjust because Mr Bailey was deprived of realistic opportunities to settle the litigation or to protect himself against the adverse effects of a NPCO, or to abandon the defence of the litigation at a much earlier stage.
The issue of a warning, and its significance in any given case, will vary greatly. The presence or absence of a warning will be chiefly significant in a case where a non party acknowledges that they had control of the proceedings: such as a company director for example. Because then a warning may have made a difference. Conversely where a non party does not have control, it might be argued what difference would the lack of a warning have made?
There is an issue curiously similar to the application of Morton’s fork: a non party could logically argue, that they did not have control, but that if they were found to have sufficient control, the supply of a warning would have caused them to do something differently. It must however be the case, that having falsely disclaimed control, a court is unlikely to be unimpressed by the switching of horses by the non-party mid-race. The argument should be left to the realm of hypothesis.