Late Quantum-Phase Amendments and Litigation Holds: The Court of Appeal’s Four‑Stage Prejudice Analysis in ENRC v Dechert & SFO

Introduction

This commentary examines the Court of Appeal’s judgment in Eurasian Natural Resources Corporation Ltd v Dechert LLP & Ors [2025] EWCA Civ 1307, handed down on 16 October 2025. The appeal arose from case management decisions made during the quantum phase (“Phase 2”) of consolidated proceedings brought by ENRC against (i) Dechert LLP and its former partner, Neil Gerrard (together, the “Dechert defendants”), and (ii) the Serious Fraud Office (“SFO”). The Commercial Court (Waksman J) had earlier found serious wrongdoing by the Dechert defendants and the SFO in Phase 1 (liability) and Phase 1A (causation and some loss).

The appeal concerned ENRC’s applications (made in November 2024) to amend its loss pleadings at the outset of Phase 2. ENRC had historically advanced a claim for increased borrowing costs said to have been caused by the SFO’s criminal investigation opened in April 2013 (“the CI”). It later realised that most of the increased borrowing costs had been incurred not by ENRC itself, but by subsidiaries across the group. To align its pleadings with that reality, ENRC sought to recast its claim as a diminution in the value of its shareholdings (“dollar-for-dollar” with the subsidiaries’ losses, adjusted for tax), and to regularise other heads of loss incurred at subsidiary level.

The Dechert defendants and SFO did not challenge the legal viability or particularisation of the amended claims at this stage (reserving, for example, any reflective loss points for later), but opposed permission on the basis of delay and resulting prejudice. Because no “litigation hold” had been placed on some subsidiaries’ documents, the defendants argued that relevant materials may have been lost, impairing their ability to test causation and quantum and therefore rendering a fair trial impossible or unfair.

The Commercial Court refused the amendments, but the Court of Appeal allowed ENRC’s appeal. The appellate court’s judgment sets out an important clarification of the principles governing late amendments at the quantum stage of split trials, and, critically, the correct approach to alleged prejudice arising from missing “litigation holds.” It articulates a structured, staged analysis for weighing potential prejudice flowing from the non-retention of documents against the concrete prejudice of shutting out substantial claims, particularly when amendments are sought early in the quantum phase.

Summary of the Judgment

The Court of Appeal (Phillips LJ, with whom Nugee and Jeremy Baker LJJ agreed) allowed ENRC’s appeal and permitted the disputed amendments. The court emphasised:

Lateness is a relative concept. Even though the amendments could have been made earlier and no good reason existed for the delay, they were sought at the very start of Phase 2, before disclosure, factual or expert evidence. In such circumstances, lateness is more readily excused.
The first instance judge erred by treating an uncertain, unquantified risk that relevant subsidiary documents might have been lost due to absent litigation holds as outweighing the certain and substantial injustice of excluding a major portion of ENRC’s loss claim.
The judge should have adopted a staged inquiry to determine whether any real prejudice had been shown by the defendants, asking: (i) whether documents would be disclosable at all; (ii) whether any such documents were unavailable because of delay (as opposed to having been lost earlier); (iii) whether any unavailability causes real prejudice (and to whom); and (iv) whether that prejudice outweighs the injustice of refusing the amendments.
A failure to impose a litigation hold is not, by itself, determinative or automatically prejudicial. The key question is the substantive effect: have disclosable documents in fact been lost in a way that causes real prejudice?
The court can and should manage any disclosure difficulties as the case unfolds, including by reserving the possibility of adverse inferences, case management sanctions, or, in the last resort, striking out elements if a fair trial becomes impossible.

Grounds 1 and 2 of ENRC’s appeal were upheld (misapplication of the test and premature findings of prejudice), ground 5 (that the amendments were not “late”) failed, and the remaining grounds were not addressed as unnecessary to the result. The amendments were allowed, with disclosure and related issues to be addressed in Phase 2, and the case proceeds to a quantum trial listed for April 2026.

Analysis

1. Procedural and Factual Context

Phase 1 (liability): Waksman J found (13 May 2022) that the Dechert defendants, acting via Neil Gerrard, breached duties by disclosing privileged/confidential information to the SFO without authority, and that the SFO committed serious breaches and induced contractual breaches by engaging with Mr Gerrard knowing (or being reckless as to the fact) he acted against ENRC’s interests.
Phase 1A (causation/loss): The judge later found (21 December 2023) that the SFO’s wrongdoing effectively caused certain losses, and that but for the wrongdoing the SFO would not have commenced the 2013 criminal investigation (CI), which ran until August 2023 without charges.
Phase 2 (quantum): ENRC’s consequential loss case included increased borrowing costs allegedly triggered by the CI. Initially pleaded as ENRC’s own increased costs, ENRC later identified (in 2024) that most of those costs were incurred at subsidiary level. It sought to reframe its claim as diminution in share value (using a dollar‑for‑dollar measure) for the 2013–2016 period in the amount of US$91.3m (US$290m with interest), of which US$78.7m was at subsidiary level and US$12.6m at ENRC level, with the balance involving holding companies ENRC Finance Ltd and ENRC NV (US$35.8m).
Preservation: ENRC’s and the two holding companies’ records were preserved from 2013, and ERG SARL’s from its 2014 incorporation. However, some subsidiaries (e.g., operating borrowers like SSGPO JSC and TNC Kazchrome JSC) did not have litigation holds in place.
First instance refusal: The judge refused the amendments on 14 February 2025, reasoning that (i) the amendments were late without good reason; (ii) the absence of litigation holds created real prejudice because subsidiary documents which might be disclosable could now be missing; and (iii) that this prejudiced the defendants’ ability to test both increased borrowing costs and the asserted “dollar‑for‑dollar” diminution methodology.

2. Precedents Cited and How They Shaped the Decision

The Court of Appeal placed the amendment dispute squarely within established case management principles while clarifying their application to split trials and quantum-phase amendments.

Municipio de Mariana v BHP Group (UK) Ltd [2024] EWHC 23 (O’Farrell J), synthesising CIP Properties (AIPT) Ltd v Galliford Try Infrastructure Ltd [2015] EWHC 1345 (TCC) (Coulson J) and Quah Su‑Ling v Goldman Sachs International [2015] EWHC 759 (Comm) (Carr J): The court adopted the familiar balancing exercise informed by the overriding objective. Lateness is relative and calls for explanation, but must be weighed against prejudice and proportionality, and the imperative to decide the real dispute where possible.
ABP Technology Ltd v Voyetra Turtle Beach Inc [2022] EWCA Civ 594: Birss LJ emphasised that lateness requires justification and that prejudice caused by lateness may justify refusal, but also that lateness is relative and context sensitive. The Court of Appeal here applied this approach, adding important colour in the specific setting of quantum-phase amendments.
Azam v University Hospital Birmingham NHS FT [2020] EWHC 3384 (QB): Cited on the appellate standard for interfering with discretionary case management decisions (misdirection, ignoring relevant/irrelevant matters, plain wrong). The Court of Appeal was careful to respect the generous ambit owed to the case management judge, but identified a misdirection in the way potential prejudice from missing litigation holds was assessed.
Republic of Mozambique v Credit Suisse International [2023] EWHC 1650 (Comm): Authority for the proposition that courts can manage disclosure deficits pragmatically in-flight, including by drawing adverse inferences or imposing sanctions, and, in extremis, striking out claims if a fair trial becomes impossible. The Court of Appeal relied on this to reinforce that amendments should not be refused simply because disclosure could become challenging; such issues can be policed later.
Practice Directions PD 31B and PD 57AD: The court referenced PD 31B para 7 and PD 57AD para 3 to underscore that parties must take reasonable steps to preserve documents when litigation is in prospect. Crucially, the Court of Appeal held that breach of this obligation is not an automatic bar to amendments, nor does it automatically translate into injustice to the other side. The substantive effects must be assessed.

3. The Court’s Legal Reasoning

The Court of Appeal’s analysis has three pillars: (i) the relative nature of “lateness” at the quantum-phase case management stage; (ii) a structured prejudice inquiry in cases alleging missing documents due to absent litigation holds; and (iii) the principle that courts should preserve the possibility of determining the real dispute unless concrete prejudice makes a fair trial impossible.

(i) Lateness is relative in split trials, and early quantum-phase amendments are more readily excused

While the court accepted the amendments were “late” because they could have been made earlier and there was no good reason for the delay, it stressed the context: these were pre‑CMC Phase 2 amendments, before quantum disclosure, witness statements, and expert reports. The absence of a fixed trial date at risk, and the ability to program fresh disclosure and expert work at a manageable stage, weighed heavily in favour of allowing the amendments.

The judge below placed significant weight on earlier judicial exhortations (in 2020 and again in March 2024) that amendments should be made promptly. The Court of Appeal accepted those indications but held that they did not debar a timely application at the start of Phase 2. Leaning too heavily on earlier missed opportunities distorted the balancing exercise by diminishing the real prejudice to ENRC of losing a major claim.

(ii) A four‑stage prejudice analysis for litigation hold issues

The core jurisprudential contribution of the judgment is a structured set of questions for courts to apply when a late amendment is opposed on the basis that some relevant documents may not have been preserved:

Step 1 – Disclosability: Identify if there are documents or categories which would, in fact, be disclosable in relation to the new issues raised by the proposed amendments.
Step 2 – Causation and timing of loss: If so, assess whether any such disclosable documents have ceased to be available because of the lateness and the absence of litigation holds, distinguishing between losses occurring before and after the time the application should have been made (e.g. since 2020).
Step 3 – Prejudice in substance: If documents were lost during the critical period, evaluate whether their loss causes real forensic prejudice to the defendants (noting that, depending on the issue, the claimant may be the more prejudiced party, as Males LJ observed when granting permission). This is not a box-ticking exercise; it turns on the importance and irreplaceability of the materials in question.
Step 4 – Balancing injustice: Weigh any established prejudice against the certain injustice to the claimant of being denied adjudication of substantial claims that are otherwise arguable and properly particularised.

The first instance judge did not complete this staged analysis. Instead, he reasoned that because he could not exclude the possibility that relevant subsidiary documents might be disclosable and might be missing, the risk of prejudice sufficed to refuse the amendments. The Court of Appeal held this was an error: a mere possibility of prejudice, unquantified and untested, cannot outweigh the certain and significant prejudice of shutting out large parts of an otherwise viable claim.

(iii) Litigation hold failures are not automatically disqualifying

The Court of Appeal explicitly rejected any notion that failing to impose a litigation hold is, by itself, a basis to refuse amendments or a proxy for unfairness. The rules require preservation steps, but the court’s job is to secure a just resolution of the real dispute in a way that is proportionate and efficient. The correct approach is to see what, if anything, has actually been lost that matters to the issues. If problems later emerge, the court can apply tools ranging from targeted disclosure orders, to adjusted evidential burdens, adverse inferences, costs, and, in extreme cases, striking out particular issues if a fair trial proves impossible.

This flexible, consequence‑focused approach sits comfortably within modern disclosure practice under PD 57AD and the court’s robust case management powers.

4. Specific Points From the Judgment

The “relevance” shorthand: The first instance judge used “relevant” as shorthand for “relevant and disclosable by ENRC.” The Court of Appeal proceeded on that basis but indicated that the real issue was not whether relevance could be excluded in the abstract; rather, whether any disclosable materials were actually missing due to delay and whether that caused real prejudice.
Group-level treasury negotiations: ENRC argued that negotiations were conducted at ENRC/ERG treasury level, so meaningful material would be in group files which had been preserved. The Commercial Court was not prepared to resolve that without disclosure lists and a DRD hearing; the Court of Appeal’s point was not to decide that question now, but to insist that such issues should be worked through in the ordinary course following permission to amend.
ENRC Finance Ltd and ENRC NV: Even for the holding companies whose documents were preserved, the Commercial Court considered the same difficulties applied. The Court of Appeal found it unnecessary to decide this point separately because the overarching error infected the decision on all increased borrowing cost amendments; once the proper staged analysis is adopted, those issues should be addressed in disclosure.
“Dollar-for-dollar” diminution claim: The amended pleading anchors ENRC’s loss to the diminution in the value of its shareholdings, matching the subsidiaries’ increased borrowing costs (subject to tax). The court did not endorse the measure as correct; it simply held that whether that measure is right is a matter for expert evidence and trial. Importantly, any reflective loss challenge was not taken at the amendment stage and remains open.

5. Impact and Implications

(a) Clarifying amendments at the quantum phase of split trials

The decision materially clarifies that when amendments are sought at the outset of a quantum phase:

They will be considered “late” if they could have been made earlier, but the label carries less sting. Early quantum-stage amendments, before disclosure and expert evidence, are “far more readily” excused.
The court will be slow to shut out substantial, arguable claims unless concrete prejudice to trial fairness is shown, rather than hypothesised.
The court expects parties to litigate real issues, and will manage disclosure and expert processes accordingly.

(b) Litigation holds: substance over form

The judgment will influence document preservation practice and how parties deploy absent litigation holds in amendment battles. While the duty to preserve remains real and enforceable, tactical reliance on possible gaps will not suffice. Parties resisting amendments should, where possible, particularise:

Which categories of documents would have been disclosable;
When they were likely lost, and why that is attributable to the lateness;
How the absence creates specific, irremediable prejudice to a fair adjudication.

Conversely, amending parties should be ready to demonstrate what materials exist at group level, what was preserved, and how any gaps can be mitigated through alternative sources (e.g. counterparty files, public ratings data, banking covenants, treasury policies) and expert methodology.

(c) Group claims and the path to shareholding diminution

The court’s willingness to permit ENRC to pursue a “dollar-for-dollar” share value diminution theory, without at this stage deciding its correctness, will be noted by parent companies seeking to regularise claims where cash outflows occurred at subsidiary level. This does not relax reflective loss doctrine; rather, it allows arguable damages methodologies to proceed to trial where liability and causation have been found and where the measure of loss requires expert assessment. Reflective loss defences remain available on the merits.

(d) Expert evidence and comparators

The Commercial Court was rightly concerned with the necessity for experts to probe subsidiary-specific risks, comparators, mitigation, and pricing. The Court of Appeal accepted this concern, but held it should be addressed through the normal machinery of disclosure and expert instructions, not as a pretext to refuse amendments in limine.

(e) Case management signals

The judgment underscores the court’s preference to keep live, arguable disputes alive and manage uncertainty dynamically, rather than deciding hypothetical disclosure fights at the gate. It also indicates receptiveness to later sanctions if real prejudice emerges, preserving trial timetables and proportionality.

6. Complex Concepts Simplified

Split trials (Phases 1, 1A, 2): The court sometimes separates liability (Phase 1) and quantum (Phase 2) to streamline complex cases. Findings in Phase 1/1A can shape what is relevant in Phase 2.
Lateness: An amendment is “late” if it could have been made earlier. But whether lateness justifies refusal depends on timing, explanation, and real prejudice. Early quantum-stage amendments are less problematic than last-minute trial-phase changes.
Litigation hold: A directive within an organisation to preserve potentially relevant documents once litigation is in reasonable contemplation. Failing to impose one is a breach of practice obligations, but its significance depends on whether important documents were actually lost and whether that causes unfairness.
Relevance vs disclosability: “Relevant” is often used as shorthand, but under disclosure rules, the question is whether documents are within a party’s control and disclosable because they may support or adversely affect a party’s case on the issues.
Diminution in share value (“dollar-for-dollar”): A parent may claim that its shares in a subsidiary are worth less because the subsidiary suffered losses (e.g., increased borrowing costs). Whether a dollar‑for‑dollar translation is valid is a factual and expert question and may be contested by doctrines like reflective loss.
Reflective loss: A rule that generally prevents shareholders from recovering for a diminution in the value of their shares where the company itself has a claim for the loss. The Court of Appeal did not decide reflective loss here because it was not raised at the amendment stage.
Adverse inferences: If a party fails to preserve or disclose documents, the court may infer that missing materials would not have assisted that party’s case.

Key Takeaways

New structured approach: Where late amendments are resisted due to absent litigation holds, courts should apply a four‑stage analysis focused on (1) disclosability, (2) timing/causation of any loss of documents, (3) real prejudice, and (4) balancing injustice.
Lateness is contextual: Early-stage quantum amendments are more readily permitted, even absent a “good reason,” particularly when no trial date is jeopardised.
Substance over form in preservation: Failure to impose litigation holds is not automatically prejudicial; the decisive issue is the substantive impact on a fair trial.
Manage, don’t pre‑empt: Potential disclosure complexities are best handled through case management, not by excluding arguable claims at the gateway. Sanctions and adverse inferences remain available if needed.
Group claims can be re‑cast: Parents may reformulate loss to share value diminution where subsidiaries incurred the cash outflows, though the correctness of “dollar‑for‑dollar” is for trial and subject to reflective loss arguments.

Practical Guidance

For claimants: If you discover at quantum stage that loss sits at subsidiary level, move promptly to amend. Document the preservation status, identify what remains at group/treasury level, and propose practical disclosure routes (counterparty files, public ratings data, treasury approvals, covenants).
For defendants: Particularise the prejudice. Identify specific disclosable categories you say are missing, why their absence is attributable to lateness, and how that impedes expert analysis. Consider whether the gaps can be remedied by counterparties or alternative sources before asking to shut the amendments out.
For both sides: Anticipate that the court may allow the amendment and then police disclosure with tight, issue‑specific orders, staged expert timetables, and, if necessary, adverse inferences and costs consequences.
Preservation policies: In complex groups, institute group‑wide litigation holds where issues likely implicate subsidiaries. While failure is not automatically fatal, it complicates litigation and may shift forensic risks.

Conclusion

ENRC v Dechert & SFO [2025] EWCA Civ 1307 establishes a clear and pragmatic framework for handling late quantum-phase amendments where document preservation at subsidiary level is in issue. The Court of Appeal reaffirmed that lateness is a relative concept and that courts should avoid excluding substantial, arguable claims based on conjectural prejudice. The judgment’s principal contribution is its four‑stage prejudice analysis and its insistence on substance over form regarding litigation holds. It underscores the judiciary’s commitment to determining the real dispute while preserving trial fairness through dynamic case management, including calibrated disclosure orders, adverse inferences, and, where necessary, later sanctions.

In a litigation landscape where complex corporate groups, split trials, and large quantum exercises are common, this decision provides valuable guidance. It instructs first instance courts to distinguish between hypothetical and real prejudice, to recognise the developmental character of quantum proceedings, and to reserve robust sanctions for cases where actual unfairness is demonstrated. In doing so, it materially advances the law on amendments and disclosure management in high‑value commercial litigation.