Claim No: CL-2023-000687 - [2025] EWHC 1157 (Comm)

That is sufficient to dismiss this application. Had it proceeded further, an issue would have arisen as to whether any order for security for costs should include the costs which RSUK might incur and be ordered to pay to Barclays in the Additional Claim. As I heard argument on this point, I will deal with it.

It is clear that in an appropriate case, such an order can be made. In Sarpd Oil International Ltd v Addax Energy SA [2016] EWCA Civ 120, Sales LJ summarised the position as follows at [24]-[25]:

“Mr Lewis submitted that although these costs are currently Glencore's costs and not Addax's costs, once Addax are ordered to pay them they become Addax's costs and are therefore within CPR r25.12 which entitled a defendant to apply ‘for security for his costs' (underlining supplied). An order for security for costs looks inevitably forward to the time when, at the end of the trial, a party is ordered to pay costs. If Sarpd loses, it is unlikely that Glencore would get an order for costs directly against Sarpd who had not sued it; it was much more likely that Addax would be held liable to pay Glencore's costs. Addax's costs would then include not merely its own costs of defending Sarpd's claim but also its own costs in the Part 20 proceedings and the costs for which it was held liable in those proceedings against Glencore.

We consider this approach to be correct inasmuch as by the time the court makes its final order as to costs it will have determined (if Sarpd loses) that Glencore has won as against Addax. It would inevitably award costs to Glencore to be paid by Addax; it is highly likely that Addax will be entitled to recover those costs from Sarpd (as well as its own costs of defending itself against Sarpd and of suing Glencore). Those costs then become Addax's own costs and thus costs for which, pursuant to CPR r.25.12 , it ought to obtain security if it can come within CPR r25.13”.

Sales LJ noted that, in the “back-to-back” claims before him, the failure of Sarpd’s claim against Addax meant that Addax’s claim against Glencore would fail. Mr Turner submitted that the effect of the Court of Appeal’s decision was that only in such circumstances would security for the costs of the additional claim be appropriate. However, Sales LJ’s reasoning was that the failure of Addax’s additional claim would “inevitably” lead to a costs order in Glencore’s favour, and it was “highly likely” that Sarpd would be ordered to indemnify Addax against those costs. That did not involve, in my view, the Court finding “inevitability” of “back-to-back” liability nor a “high likelihood” of the claimant being ordered to pay the additional party’s costs if its own claim failed before an order could be made requiring the claimant to provide security for those costs.

Whenever a court makes an order for security for costs, which is inherently forward-looking, there may be a number of contingencies which will affect the amount of the defendant’s costs recovery (and hence the level of costs exposure to be secured). Most obviously, there is the amount of costs which will be incurred and awarded on assessment. In Bluewaters Communications Holdings LLC v Bayerische Landesbank Anstalt Des Offentlichen Rechts [2018] EWHC 78 (Comm), [30], Popplewell J held that the test to be applied in this context was as follows:

“The test which I apply is essentially that I should award the sum which the court considers the applicant would be likely to recover in a detailed assessment if awarded costs on a standard basis following trial having regard to the factors set out in the relevant CPR rule 44.5(3). That was the test adumbrated in Vald Nielson Holding AS v Baldorino and is commonly applied in this court.”

That was a reference to a characteristically cogent Commercial Court judgment by the late Robin Dicker QC reported at [2017] EWHC 1033 (Comm), [13], where he stated:

“Under CPR 25.13(1)(a) the court has a discretion to award security in an amount which it considers just having regard to all the circumstances of the case. The appropriate amount will generally be the sum which the court considers that the applicant would be likely to recover in a detailed assessment if awarded costs on a standard basis following the trial, having regard to the factors set out in

CPR 44.5(3).”

That test has been applied in a number of authorities over the years (e.g. Procon (Great Britain) Ltd v Provincial Building Co Ltd [1984] 1 WLR 557, 571B).

As a distinct enquiry within that overall question, there is the issue of whether an order for costs against the claimant will be made on a standard or indemnity basis. It had at one point been suggested that a “reasonable possibility” rather than “speculative possibility” that costs would be awarded on an indemnity basis was sufficient (Danilina v Chenukhin [2018] EWHC 2503 (Comm), [14]), or a “real possibility” (BSG Resources Ltd v Vale SA [2019] EWHC 2456 (Comm), [36]-[37]). In Phones 4U Ltd v EE Ltd [2020] EWHC 1943 (Ch), Roth J addressed the test in greater detail. He held that it was not sufficient that it could not be said at the time of the application for security that an order for indemnity costs was not realistically arguable, saying this “puts the threshold much too low” ([24]). At [31], the test he applied was whether an order for costs on an indemnity basis was “such a significant possibility or, to put it another way, that there is a real prospect of such an order so as to justify a requirement to provide security for costs on that basis.”

Finally, so far as security for the costs of an additional claim is concerned, the relevant test was considered by Cockerill J in Maroil Trading Inc v Cally Shipholdings Inc [2020] EWHC 3041 (Comm). Cockerill J applied a “likelihood” test, while noting at [31]:

“In the end, although this is obviously not an entirely straightforward point, I have come to the view that, as a matter of jurisdiction, the test may well be one of likelihood, as appears to be indicated, albeit in passing, by Sarpd, or it might be said to be something rather lower. Though, given the seriousness of having to provide security, it is unlikely to be very low, as in a mere possibility, and it is likely to be not unadjacent to a likelihood. Alternatively, if the test is somewhat lower, it would seem to me right that the lesser degree of likelihood goes to the exercise of the discretion, given that the authorities seem to be quite clear that there are circumstances in which, in Part 20 claims, costs should be recoverable and there are circumstances in which they should not. So there must be a sliding scale.”

I am satisfied that something distinctly more than a “real possibility” of a costs order against a claimant for the costs of liability of the defendant in an additional claim if the claimant’s claim fails is required before it would be appropriate for the court to order a claimant to provide security for costs of those proceedings. That is consistent with Roth J’s considered assessment of the position so far as ordering security on an indemnity basis is concerned, and with the views expressed in Sarpd and Maroil. For my part, I struggle to see why different tests fall to be applied to different contingencies when working out the amount of security required for the costs order a defendant would obtain if the claim fails, although it is important in all cases to avoid going into the merits of a case unless it can clearly be demonstrated one way or another that there is a high degree of probability of success or failure (Danilina v Chernukhin [2018] EWCA Civ 1802, [69]).

Just as a decision on whether to order security by reference to an indemnity assessment of the recoverable costs is likely to involve looking at the broad nature of the allegations in a case (rather than their individual likelihood of success and failure) and deciding whether, if the claimant fails, there is a sufficient prospect of such an order so as to justify a requirement to provide security for costs on that basis, so too when security is sought for the costs of an additional claim, the court will need to look at the broad nature of the allegations in both the claim and additional claim, and ask whether, if the defence succeeds in the main claim, there is a sufficient prospect of the claimant being ordered to indemnify the defendant against the costs of the additional claim to justify an order for security for costs on that basis.

Mr Turner’s principal reason for suggesting that this requirement is not made out in this case is that there are realistic scenarios in which RSUK could defeat the Claimants’ claim, and yet succeed in the Additional Claim. I accept this – an obvious one being if RSUK establishes that its only obligation under the Escrow Agreement is to pass on an instruction to Barclays, and yet Barclays is in breach of contract for not processing that instruction.

However, the scenario which RSUK needs security for is one in which its claim against Barclays fails. In that scenario, I am satisfied that there is a high likelihood that Barclays would obtain an order for its costs against RSUK, and that RSUK would obtain an order against the Claimants in respect of those costs. Further, it is clear that RSUK’s claim against Barclays has only been brought because of the Claimants’ claim against RSUK, and the Additional Claim does not raise independent issues, only issues raised by way of response to the Claimants’ claims. In these circumstances, had I been persuaded that it was appropriate to exercise the discretion to award security, I would have been persuaded that the security ordered should include security for costs of the Additional Claim.

Finally, I should mention that there was no suggestion that the issues raised by the Balance when considering whether to order security for costs might justify a different approach as between RSUK’s application for security for the costs of defending the Claimants’ claim, and security for the costs of the Additional Claim including those payable to Barclays. I have not therefore considered this issue.

I also heard submissions as to the amount and timing of security which I will also address.

RSUK sought security as follows (with some rounding):

A detailed costs budget was provided for RSUK’s costs. Barclays’ costs are not broken down but provided in headline form in a letter to RSUK.

Mr Turner took two issues with the amount of security sought in respect of RSUK’s costs:

The Grade A rates claimed are £680, £620 and £545 as against the current guideline Grade A rate of £566. The Grade C rate claimed is £460 as against the current guideline rate of £299.

It is right to say that a significant part of the work done is based on a Grade A fee-earner with a rate of £620 and that the involvement of Grade A fee earners is high:

I accept that some adjustment is required to reflect these facts. While this is necessarily a broad brush analysis, I am satisfied that a reduction of 30% will address those two issues. I have also rounded the figures down, reflecting the broad brush nature of the exercise and the scope for challenges on assessment.

So far as Barclays’ costs are concerned, I do not have hourly rates or a breakdown. In my view, it is reasonably likely that its recoverable costs will be lower than RSUK’s costs of defending the Claimants’ claim, given the narrower scope of the Additional Claim, and I have arrived at a conservative estimate on that basis.

On that basis, RSUK’s own costs for the three stages would be as follows:

It was agreed that security should be provided in three tranches. Had I been persuaded to order security for costs, I would have ordered security in three tranches as follows: