Security for appeal costs, judicial notice of enforcement risks, and proportionate legal costs

The Russian Federation v Hulley Enterprises Ltd and others [2026] EWCA Civ 1353

What was the background?

Hulley Enterprises Ltd, Yukos Universal Ltd and Veteran Petroleum Ltd (together, ‘Hulley’) have been seeking since 2015 to recognise and enforce in England and Wales three Netherlands-seated arbitration awards against the Russian Federation (‘Russia’) worth in the region of US$50bn plus interest (now over US$66bn). After Russia’s state immunity challenge failed, further preliminary issues were tried by Bright J, who gave judgment for Hulley on 2 March 2026. He ordered Russia to pay the costs of the trial, with an interim payment of £2.5m (‘the interim costs’) within an unusually generous three months, reflecting Russia’s contention that it needed substantial time to authorise payment. Earlier costs orders (‘the outstanding costs’, agreed at £877,106.34 with interest) also remained unpaid. Bright J granted permission to appeal on one ground only, declined to make it conditional on payment of either sum, and directed that any application for security for costs be made to the Court of Appeal.

On a renewed application, Males LJ granted permission on grounds 7 and 8 and ‘provisionally’ on grounds 2–6, reserving to Hulley the right to apply for conditions relating to the outstanding costs, the interim costs and the costs of the appeal. When the three months expired with nothing paid, Hulley applied on 15 June 2026 under CPR 52.18(1)(c) for conditions on every ground other than ground 1 and, separately, for security for the costs of the appeal under CPR 25.29.

By the hearing on 13 October 2026 the position had moved on. The parties had agreed a payment route involving a US law firm associated with Russia’s English solicitors, the Court Funds Office (CFO) and a solicitors’ undertaking as to the provenance of the funds. The outstanding costs were paid into court on 30 July 2026; the interim costs were paid late, in two instalments in August 2026, with an explanation (the CFO’s due diligence) that the court found ‘rather unconvincing’. Some interest remained unpaid and the outstanding costs had not yet been paid out to Hulley’s solicitors. Russia resisted security on the footing that Hulley had not proved by evidence any real risk of non-payment, and that CPR 52.18(3) deprived the court of power to order security at all. Hulley sought 70% of a costs schedule totalling some £3.64m.

What did the court decide?

Fancourt LJ (with whom Phillips LJ agreed) declined to impose any condition under CPR 52.18(1)(c), for which a ‘compelling reason’ is required (CPR 52.18(2)). The principal sums were in court, the delay in paying them out was an administrative matter for the court, and the unpaid interest was small. Had Russia not paid by the hearing date, however, conditions would ‘very likely’ have been imposed, and Hulley had liberty to apply if the picture materially changed. As Hulley accepted, CPR 52.18(3) precluded any application for conditions on ground 1, because Hulley had been present when Bright J granted permission; whether the bar extended to grounds 7 and 8, permission for which was granted on paper, was left open. Permission on grounds 2–6 was confirmed.

The court rejected as ‘simply wrong’ the argument that CPR 52.18(3) barred a later application for security. An application under CPR 25.29, even one framed (as is usual in the Court of Appeal) as an unless order, is not an application for a condition under CPR 52.18(1)(c): the rules ‘are distinct and are intended to operate distinctly’ (at [70]). Spar Shipping AS v Grand China Logistics Holding (Group) Co Ltd [2016] EWCA Civ 520 and SpiceJet Ltd v De Havilland Aircraft of Canada Ltd [2021] EWCA Civ 1834 had been misread: in neither was security refused because it amounted to a condition. The suggestion that a successful party must be ready at hand-down to make an evidenced application for security for the costs of an appeal on newly disclosed grounds ‘would … come as something of a surprise to most practitioners, as it did to the court’ (at [60]). The court did, however, discourage the ‘twin-track’ approach of seeking security under CPR Part 25 while separately asking for its payment to be made a condition under CPR 52.18 (at [73]).

On the merits, the threshold condition in CPR 25.27(b)—residence out of the jurisdiction—was ‘self-evidently’ met by a foreign state; the question was whether security was just in all the circumstances (CPR 25.27(a)). No question of stifling arose. Russia’s payment record was equally consistent with it paying ‘if and when it suits its purposes to do so, but not otherwise’ (at [40]); the distinction it drew between the awards and English costs orders was ‘little more than a convenient forensic distinction’ (at [41]); and blaming the delay on money laundering concerns raised by Hulley’s solicitors was ‘wholly unconvincing’ when an SRA-regulated client account had always been available (at [45]). Russia did not suggest it had assets within the jurisdiction, but argued that Hulley bore the burden of proving that it could not enforce in Russia. Ordinarily a party must adduce evidence of a ‘real risk’ of enforcement difficulty rather than a ‘mere possibility’ (Danilina v Chernukhin [2018] EWHC 39 (Comm) at [61]). Here that would be ‘superfluous’: this was ‘a rare case’ in which the court could conclude without evidence that there was self-evidently a real risk that these respondents could not enforce a costs judgment against the Russian state in the Russian courts, given the tribunal’s findings about the Russian judiciary, the 2017 Constitutional Court ruling on the European Court of Human Rights’ Yukos judgment, the 2020 constitutional amendments and the deterioration in the enforcement climate since 2018 (at [48]–[55]). The court was explicit that this ‘does not establish any precedent for what is needed to establish a real risk of non-enforcement in other, different cases’ (at [55]). Two further factors told in favour of security: Russia’s own evidence showed that even a willing payer might take months to pay, entailing further expense (at [56]); and the agreed payment mechanism depended on Russia’s continuing will and on the route remaining available ‘in a rapidly changing world’ (at [75]). Against that, the burden on Russia was ‘very slight’.

On quantum, the sums sought were ‘extraordinarily large for a four day appeal’ (at [82]). Counsel’s fees included £275,000 to adapt the skeleton to grounds 2–8, a £2m brief fee for leading counsel and £375,000 for junior counsel, and £125,000 for a consequentials hearing ‘that will not ordinarily take place’. These were ‘astonishing and cannot be justified’; the excellence and seniority of leading counsel and ‘market forces’ had ‘little relevance’ to the criteria a costs judge applies, and Russia could only be expected to pay what was reasonable and proportionate on the standard basis (at [85]). Declining to apply a percentage to an extravagant schedule, the court allowed £600,000 for counsel. Over 700 hours of solicitors’ time raised ‘obvious questions about unnecessary duplication of work and the number of fee earners involved’ (at [88]); £375,000 was allowed for profit costs and other disbursements.Security was fixed at £975,000—roughly a quarter of the sum scheduled—without limiting the discretion of the costs judge on any detailed assessment (at [83]).

Russia asked for six months to pay without explaining why; Hulley proposed three months, then a stay and strike out four weeks later. Rejecting a two-stage structure (at [79]), the court ordered a single, generous deadline: security to be paid into court by the agreed route, or transferred to Hulley’s solicitors from an SRA-regulated law firm, by 4pm on 1 March 2027, failing which the appeal ‘will stand struck out without further order’ (at [80]–[81]). There is no stay in the meantime. Russia may apply for more time if circumstances change, but no substantial extension will be granted for obstacles that were foreseeable or within its control, nor any that would prejudice orderly preparation. The appeal is not to be listed before 9 June 2027.

What are the practical implications of this case?

This is, first and last, a costs decision, and an important one on three fronts: who can be required to put up security for an appeal, what evidence is needed to obtain it, and how much the Court of Appeal will regard as a reasonable and proportionate sum—not, it turns out, a £2m brief fee. For practitioners enforcing awards and judgments against sovereign and state-linked defendants—the daily fare of the fraud and asset recovery bar—it offers a template; for those acting for such defendants, it is a warning that the price of appealing in London may now be money up front.

On the first front, a foreign state is ‘resident out of the jurisdiction’ for the purposes of CPR 25.27(b) without argument, and the State Immunity Act 1978 is no bar: an order for security is neither an injunction nor process against state property (Ukraine v PAO Tatneft [2019] EWCA Civ 763). The real work is done at the ‘just in all the circumstances’ stage. Since Nasser v United Bank of Kuwait [2001] EWCA Civ 556, residence abroad has justified security only where there is objective justification in the form of obstacles to, or additional burdens of, enforcement; since Bestfort Developments LLP v Ras Al Khaimah Investment Authority [2016] EWCA Civ 1099 the applicant need show only a real risk of such obstacles; and the amount is tailored to the risk shown, so that a real risk of non-enforcement outright, as opposed to extra cost or delay, justifies security for the whole of the proportionate costs (Danilina v Chernukhin [2018] EWHC 39 (Comm)). Nasser itself permits the court to ‘take notice of obvious realities without formal evidence’, and in Tatneft the Court of Appeal did precisely that in ordering security against Ukraine on the strength of its non-payment of a treaty award. Hulley takes the dictum to its limit: not obvious realities about a foreign legal system in general, but judicial notice of a state’s hostility to the particular parties before the court. The disclaimer at [55] should be taken at face value, and evidence of the Danilina type remains the safer course. But where the opposing party is a state in open defiance of international adjudication, a well-assembled bundle of public-domain material—constitutional amendments, prior refusals to enforce, findings in the award itself—may now suffice.

Two further points will interest enforcement practitioners. Leggatt LJ’s observation in Tatneft that an investor should not be driven to the respondent state’s own courts to enforce a costs order is a policy consideration that Hulley’s reasoning quietly assumes; and the pro-enforcement policy of the New York Convention cuts the other way where security is sought from the award creditor rather than the award debtor (Gater Assets Ltd v Nak Naftogaz Ukrainiy [2007] EWCA Civ 988). Whether the immunity from execution over non-commercial state property under section 13 of the State Immunity Act 1978 is itself an enforcement obstacle of the Nasser kind has not been decided; it is a point worth taking.

Second, and less obviously, delay in payment, and the expense of extracting it, is a freestanding factor supporting security even where the paying party is willing and able to pay (at [56]). Allied to this is the court’s sceptical reading of payment history: paying only when an application for conditions is looming is treated as strategic compliance rather than good faith (at [40]), and assurances about historic liabilities ‘say nothing about future costs liability’ (at [43]). Respondents to appeals by states, state-owned enterprises or opaque corporate vehicles should build a careful chronology of every late or extracted payment. Appellants resisting security need evidence from an identified source addressing willingness and ability to pay future costs, and ideally a compliant payment route already in operation.

That last point will resonate with fraud and sanctions practitioners. Much of the delay here was attributed to the difficulty of moving money from a Russian state source into an English solicitor’s client account in a form acceptable to the recipient’s compliance function. The court was unimpressed: it found it ‘extraordinary’ that an SRA-regulated firm had not simply been instructed (at [45]), built that route into the order (at [81]), and warned that no substantial extension would be given for foreseeable or self-inflicted obstacles (at [80]). Compliance friction is a problem for the paying party to solve well before the deadline, not an excuse. Equally, the recognition at [75] that a payment mechanism may cease to be available ‘in a rapidly changing world’ is a candid acknowledgment that the sanctions environment is itself a reason to take security now rather than trust to enforcement later.

Third, on procedure, the decision defuses a trap and confirms an escape route. CPR 52.18(3) means that a respondent present when permission to appeal was granted cannot later seek conditions—a rule ‘perhaps not as well-known as it needs to be’ (at [60]). At any consequentials hearing, therefore, decide then and there whether to seek conditions (payment of the judgment debt or of unpaid costs), because the only later route is the ‘material change of circumstances’ discipline of SpiceJet. Hulley confirms—in what appears to be the first appellate decision squarely on the point—that security for the costs of the appeal under CPR 25.29 is not caught by CPR 52.18(3), even with an unless sanction; it also helps to obtain, as Hulley did from Bright J, an express direction that any security application be made to the Court of Appeal (at [71]). Avoid the ‘twin-track’ application the court discouraged (at [73]), but do ask for the order in the form made here—a single deadline, automatic strike-out and no stay—rather than the stay-then-apply model now usual in the Business and Property Courts (at [20]). Appellants will want the opposite.

Fourth, and the headline for most readers, is quantum. A £2m brief fee for a four-day appeal was ‘astonishing’; counsel’s seniority and ‘market forces’ had ‘little relevance’ to what a costs judge will allow; and the court abandoned the usual percentage-of-schedule approach because the schedule was too extravagant to be a safe starting point. Security applications in the Court of Appeal are now a forum in which proportionality is scrutinised line by line, and these dicta will be deployed on detailed assessment whatever the disclaimer at [83]. Solicitors preparing schedules for high-value appeals should expect counsel’s fees to be benchmarked against the length and complexity of the hearing rather than the sum at stake, and close attention to fee-earner duplication (four solicitors for four days, and over 100 hours on a costs schedule, drew pointed comment). Clients must be advised at the outset that the gap between what they spend and what they recover may be very wide—here, on the court’s figures, perhaps three-quarters of the budget—and that security will not close it.

Several questions remain open. The judicial-notice approach is confined to its facts, yet the method is portable: a state that has legislated to disapply adverse international decisions, or has a record of defying awards, now invites the same inference, and it is unclear how far the court will go before insisting on evidence. Whether CPR 52.18(3) bites on grounds for which permission is granted on paper was left undecided (at [26]). Most interesting is the tension with SpiceJet: Phillips LJ there warned against using CPR 52.18(1)(a) to impose by another route a condition barred by CPR 52.18(3), yet agreed here that a CPR 25.29 unless order sits wholly outside that policy. A respondent who missed the moment for conditions can therefore still, in substance, make the appeal conditional on money—provided the money is security for costs rather than the judgment debt. For the enforcement practitioner, Hulley joins a lengthening line of decisions in which the English courts decline to let sovereign defendants litigate on credit.

Case details

• Court: Court of Appeal (Civil Division)
• Judge: Lord Justice Phillips and Lord Justice Fancourt
• Date of judgment: 22 October 2026

Article written by Lauren Godfrey. First published by LexisNexis.