Introduction

This judgment addresses a judicial review of an Ombudsman determination arising from a retail client’s losses under a computer‑driven strategy that traded contracts for differences (CFDs). The Financial Ombudsman Service (FOS) had held Linear Investments Ltd (Linear) liable for misclassifying Professor Leslie Willcocks as an elective professional client and for other regulatory failings, and ordered full compensation calculated by reference to a FTSE UK private investors income total return index plus 8% simple interest.

Linear challenged the determination on three grounds: (1) that it was entitled to rely on the client’s signed application (including tick‑box assertions) when classifying him as an elective professional client under COBS 3.5, such that the Ombudsman’s contrary conclusion was wrong in law and irrational; (2) that the FOS’s choice of a relatively low‑risk benchmark was irrational given the client’s pursuit of a higher‑risk CFD‑based strategy; and (3) that the FOS acted wrongly and irrationally in refusing any reduction for contributory negligence despite the client’s misrepresentations about his CFD experience and financial‑sector work history.

The Court of Appeal dismisses the appeal on Grounds 1 and 2, upholding the Ombudsman’s findings on misclassification and benchmark choice; but allows the appeal on Ground 3, holding that the Ombudsman’s refusal to make any reduction for contributory negligence proceeded on a flawed approach to causation. The issue of reduction is remitted to the Ombudsman for reconsideration.

Summary of the Judgment

• Ground 1 (Client classification): Dismissed. COBS 3.5.3R required Linear to undertake an adequate assessment giving reasonable assurance of the client’s ability to make investment decisions and understand risks. Linear could not rely solely on tick‑box self‑certification in the face of red flags (notably the “blue chip stocks” description and absence of supporting evidence) and failed to obtain clarifying information about CFDs (qualitative test) and key details concerning trade size and market (quantitative test).
• Ground 2 (Benchmark for redress): Dismissed. It was rational to use a diversified, medium‑risk FTSE UK private investors income total return index. Given Linear’s own failure to establish that the client understood high‑risk leveraged CFD trading, it would be illogical to anchor redress to a high‑risk benchmark.
• Ground 3 (Contributory negligence): Allowed. The Ombudsman’s refusal to apply any reduction misapplied causation. The client’s misrepresentations were a concurrent operative cause of the loss because they enabled access to the high‑risk service. The matter is remitted to the FOS to decide the just and equitable reduction.
• Respondent’s Notice (s.31(2A) SCA 1981 “highly likely”): Rejected. The decision‑making history (including the first provisional decision proposing a 25% reduction) undermined confidence that the outcome would have been the same absent the causation error.

Detailed Analysis

Precedents and Authorities Cited and Their Influence

• FSMA and DISP framework:
  • FSMA s.228(2) and DISP 3.6.4R set the FOS’s “fair and reasonable” jurisdiction, under which relevant law, regulations, and good practice are taken into account without binding the Ombudsman to the common law.
  • FSMA s.229 empowers the FOS to award fair compensation and to direct “just and appropriate” steps.
• Ombudsman’s legal remit and reasoning standards:
  • R (Options UK Personal Pensions LLP) v FOS [2024] EWCA Civ 541: confirms the Ombudsman’s broad, non‑court jurisdiction and need for clear reasoning; if departing from law, the Ombudsman must explain why, but need not apply law formulaically.
  • R (Heather Moor & Edgecomb Ltd) v FOS [2008] EWCA Civ 642: Ombudsman must explain any divergence from the “relevant law.”
  • R (Garrison Investment Analysis) v FOS [2006] EWHC 2466 (Admin): redress must have a logical connection to the wrong.
  • R (Gathercole) v Suffolk CC [2020] EWCA Civ 1179 and R (Bradbury) v Brecon Beacons NPA [2025] EWCA Civ 489: set the high bar and process‑focus for s.31(2A) SCA 1981 (“highly likely” no difference) arguments.
• Client classification and reliance on client statements:
  • Spreadex Ltd v Sekhon [2008] EWHC 1136 (Ch): classification under earlier COB rules turns on compliance with procedural steps rather than purely objective client characteristics; supports focusing on whether COBS 3.5.3R was properly complied with.
  • Wilson v MF Global UK [2011] EWHC 138 (QB): a firm may take client statements at face value “unless and until there is some reason to apply further scrutiny”—a principle the Court applies, finding that Linear was on notice to inquire further.
• Contractual binding and estoppel:
  • L’Estrange v Graucob [1934] 2 KB 394; Peekay Intermark v ANZ [2006] EWCA Civ 386: signature binds as to contract terms, but the Court holds those authorities do not control questions of COBS regulatory compliance. The tick‑box assertions were not contract terms.
  • Steria v Hutchinson [2006] EWCA Civ 1551: estoppel by representation requires a clear representation and reasonable reliance. In a regulated context, with firms under a duty to assess clients, reliance on ambiguous or contradictory self‑certification may be unreasonable.
• Contributory negligence and causation:
  • Law Reform (Contributory Negligence) Act 1945, s.1: damages reduced where loss results partly from claimant’s fault; apportionment is “just and equitable.”
  • Caswell v Powell Duffryn [1940] AC 152; Stapley v Gypsum Mines [1953] AC 663: apply a broad, common‑sense approach to causation for both claimant and defendant faults.
  • The Volute [1922] 1 AC 129: when both parties’ faults are operative causes, damages must be apportioned, regardless of sequencing.
• CFDs and their risk profile:
  • City Index v Leslie [1992] QB 98: describes CFDs as “contracts intended… to end in the payment of differences.”
  • Universal Stock Exchange v Strachan [1896] AC 166: historical view of cash‑settled difference contracts as gambling, highlighting speculative nature.

Legal Reasoning of the Court

1) Client classification under COBS 3.5.3R: “adequate assessment” requires more than untested self‑certification when red flags exist

The central regulatory duty was COBS 3.5.3R: before treating a client as an elective professional, a firm must conduct an adequate assessment that gives reasonable assurance the client can make investment decisions and understands the risks (qualitative test), and, where applicable, assess the quantitative criteria (e.g., significant‑size transactions on the relevant market).

Linear relied on the client’s tick‑box statements (including claims of 40‑80 CFD trades per year and “20+ lots” per trade) and the signed declaration. The Court held that, in context, these were insufficient because:

• The firm’s own form anticipated supporting evidence (“Please ensure to attach appropriate evidence…”). No evidence was provided; Linear did not ask for any.
• The open‑text narrative (“Market Knowledge and Individual Objectives”) contained only “Have invested for +15 years in blue chip stocks,” a glaring omission given the strategy’s core use of leveraged CFDs—a materially different, higher‑risk activity.
• There were inconsistencies, notably the SAXO form assertion of a year’s professional financial‑sector work, contradicted by the Account Opening Form’s statement of 12 years at the LSE; no clarification was sought.
• The tick‑box “lots” metric was incapable of demonstrating “significant size” or the “relevant market” without more (lot sizes vary across markets; without market identifiers or lot‑size details, experience cannot be matched to proposed trading).

Applying Wilson v MF Global: while firms need not interrogate client statements absent cause, here Linear was on notice. The omissions and contradictions triggered a duty to inquire, both for the qualitative (knowledge/experience) and quantitative (significant‑size/relevant market) limbs. The Court rejected Linear’s argument that signed statements or contract law principles (L’Estrange/Peekay) insulated reliance; this was a regulatory compliance question, not a contractual construction exercise. Estoppel principles (Steria) underscored that reliance must be reasonable; it was not, given the red flags and the firm’s own designed process demanding supporting evidence.

2) Benchmarking redress: choosing a medium‑risk proxy was rational where the firm never established client understanding of high‑risk leveraged CFDs

The FOS awarded loss by comparing actual outcomes with a FTSE UK private investors income total return index, plus 8% simple interest. The Court affirmed. Redress must have a logical connection to the wrong (Garrison). The wrong here included misclassification and failure to ensure the client understood the risks of a high‑risk CFD strategy. It would be illogical to assume, for compensation purposes, that the client would otherwise have pursued a high‑risk strategy for professional investors. Selecting a diversified, medium‑risk index was therefore rational and within the FOS’s s.229 FSMA discretion.

3) Contributory negligence: client misstatements were an operative cause; the FOS misapplied causation by treating Linear’s other failings as the sole cause

The FOS initially proposed a 25% reduction for contributory negligence (first provisional decision), then awarded 0% after focusing on Linear’s misleading risk/cost disclosures (second provisional decision and final decision). The Court held this was a legal error. Under the 1945 Act and the authorities (Caswell, Stapley, The Volute), causation is approached broadly and symmetrically for claimant and defendant; sequencing does not matter. Here:

• But for Linear’s misleading information, the client might not have proceeded. However, equally, but for the client’s misrepresentations, he would not have been accepted as an elective professional, would not have accessed the Pembroke CFD strategy, and would not have suffered the losses.
• Both faults were operative causes of the damage. The FOS’s analysis erroneously elevated Linear’s failings as the sole operative cause, contrary to common‑sense causation principles.
• Although the FOS is not bound to apply the common law strictly, DISP 3.6.4R required it to take relevant law into account. Per Options UK and Heather Moor, if departing from law, the Ombudsman should explain why. Neither occurred correctly here.

On the respondent’s s.31(2A) SCA 1981 “highly likely” point, the Court emphasized process: the FOS’s first provisional view (25% reduction) shifted only after the flawed causation analysis was adopted. That procedural history undermined any confidence that the outcome would highly likely be the same absent the error. Relief was therefore appropriate: the quantum determination was quashed and remitted for a fresh apportionment assessment.

Impact and Practical Significance

For regulated firms: onboarding and classification

• Tick‑boxes are not a safe harbour. Where inconsistencies or gaps appear—or where the product is inherently complex/high‑risk (e.g., leveraged CFDs)—firms must seek evidence and clarifications. “Reasonable assurance” in COBS 3.5.3R(1) is objective and may require corroboration.
• Design of forms matters. If a form asks for supporting evidence, the firm should insist on it. Open‑text prompts (e.g., “explain experience”) must be read critically: “blue chip stocks” is not “CFD trading”. Absence of a CFD narrative is a red flag.
• Quantitative test needs specifics. Soliciting “lots per trade” is insufficient. Firms should capture market identifiers, lot sizes, product types, and frequency over the relevant period to test “significant size” on the “relevant market.”
• Contradictions must be resolved. Claims of financial‑sector experience that sit uneasily with disclosed employment histories require follow‑up. Stale experience may be insufficient for modern, algorithmic CFD strategies.
• Record‑keeping and audit trail. Ensure the file shows the enquiries made, evidence obtained, and rationale for classification decisions. These will be scrutinized if a complaint arises.

For the FOS: methodology and reasoning

• Contributory negligence must be squarely addressed where client conduct enabled the loss. Where misstatements are an operative cause of access to a product, an apportionment analysis is required. If the FOS departs from common‑law causation or apportionment principles, it should explain why that is fair and reasonable in the particular case.
• Redress benchmarks should be logically anchored to the wrongdoing found. Where misclassification into a high‑risk product is the fault, medium‑risk comparators may be rational; “but for” the breach, the client would not have been in a high‑risk environment at all.
• Decision‑making discipline matters. Provisional decisions that change materially should reflect correct legal and causation frameworks. Otherwise, s.31(2A) arguments are less likely to succeed on judicial review.

For consumers and advisers

• Misstatements can reduce compensation. Even where a firm has regulatory failings, inaccurate declarations about experience or professional background may lead to a reduction in redress as a matter of fairness and law.
• Understanding derivatives is distinct from equity investing. Experience in “blue chip” equities does not equate to experience in leveraged CFDs; the latter carry materially different risks and require distinct knowledge.

Systemic legal implications

• COBS 3.5.3R clarified. The decision underscores that “adequate assessment” is substantive. A firm that is on notice to inquire further but fails to do so will struggle to justify elective professional classification.
• Contributory negligence within the FOS jurisdiction. While the FOS is not bound by the common law, it must take it into account accurately or explain any fair‑and‑reasonable departure. This judgment elevates the expectation that contributory fault will be evaluated where client conduct is causative.
• Judicial review remains a viable check—albeit limited. Errors of law and irrationality in the FOS’s approach can be corrected. Section 31(2A) is not a cure‑all where process shows the error likely affected the outcome.

Complex Concepts Simplified

• CFDs (Contracts for Differences): Derivatives where parties settle the cash difference between the opening and closing price of an underlying (e.g., a stock or index). No delivery of the underlying occurs. CFDs are typically leveraged, meaning small cash outlays can generate large gains or losses.
• Leverage/Gearing: Borrowed‑like exposure: a 5% move in the underlying can translate to a much larger percentage gain/loss relative to the investor’s margin. This magnifies risk.
• Elective Professional Client (COBS 3.5.3R): A retail client who opts to be treated as professional, after the firm conducts an adequate qualitative assessment (knowledge/experience), considers quantitative criteria (e.g., frequency and size of trades, portfolio size, relevant work experience), and completes procedural warnings and acknowledgments.
• “Significant size” and “relevant market” (Quantitative test): Not merely the number of trades. The firm should be able to evidence the markets traded and the size of positions to show that the client’s experience matches the complexity and risk of the intended service.
• Estoppel by representation: Prevents a person from denying a clear representation upon which another reasonably relied to their detriment. In regulated onboarding, reliance on ambiguous or contradictory self‑certification may be unreasonable, particularly when the firm is under a duty to assess.
• Contributory negligence: If the claimant’s own fault contributes to the loss, compensation is reduced proportionally. Both parties’ faults are assessed using a common‑sense causation approach; the order in which faults occur does not matter.
• FOS “fair and reasonable” vs. strict law: The FOS decides cases informally and is not bound to apply the common law. But it must take relevant law and standards into account and explain any departures.
• Judicial review s.31(2A) SCA (“highly likely”): Even if a decision was flawed, the court will refuse relief if it is highly likely the outcome would have been the same. The test is stringent and focuses on the significance of the error in the decision process.

Conclusion

The Court of Appeal’s decision draws a clear line under two recurring issues in financial services disputes. First, the COBS elective professional client regime demands more than untested tick‑boxes when the file discloses contradictions or gaps—especially in relation to complex, leveraged products like CFDs. Firms must probe and evidence their assessments to achieve the required “reasonable assurance.”

Second, where client misstatements enable access to a high‑risk service, contributory negligence is a live and often inevitable consideration. The Ombudsman’s broad fairness jurisdiction does not permit overlooking concurrent causation; relevant legal principles must be taken into account or any departure must be justified. On redress, the Court affirms that benchmarks must be logically connected to the established wrongdoing: where misclassification is the fault, a medium‑risk proxy may be appropriate.

In practice, compliance teams should revisit client onboarding for complex products, ensure contradiction‑spotting and evidence‑gathering are built into procedures, and record the rationale for client categorisation decisions. For the FOS, the judgment is a reminder to articulate the causation analysis and any apportionment with clarity. For consumers, the ruling underscores that inaccurate declarations can reduce compensation, even when firms err.

By insisting on inquiry‑based classification and balanced causation analysis, the Court has refined both regulatory compliance expectations and the fairness calculus applied in Ombudsman redress—guidance that will resonate across future disputes involving high‑risk trading strategies and client categorisation.