CMC class-action period shrank to five years as security process reached A$13.2m

The public description of the Australian class action against CMC Markets has changed in a way that materially alters its time perimeter. CMC’s FY26 results say the claim originally covered trading from 7 November 2011 to 30 April 2021, but an amended pleading filed in March 2026 refined the period to 1 May 2016 through 30 April 2021. The new window is exactly five years. Compared with the original span of roughly nine years and six months, about four years and six months were removed—approximately 47% of the elapsed period. That does not determine the number or value of affected claims, but it is a major contraction in the pleaded historical scope.[1][4]

The procedural record also exposes the cost and data burden before any trial finding. A 2025 Federal Court decision addressed discovery of raw trading data over five years and screening tests that produced about 2,500 failed results requiring attention. The same decision recorded that CMC had consensually received about A$4.6 million in security for costs. A 2026 decision concerned CMC’s request for approximately A$8.6 million of additional security and an unsuccessful attempt by the applicants to obtain 17 legal invoices, which the court held were privileged. Together, the security already provided and the additional amount requested reached A$13.2 million. The figure is not damages, a settlement or a finding of liability.[2][3]

A narrower pleading is not a smaller loss calculation

Time is only one denominator in a class action. Reducing a claim period by 47% does not prove that potential liability fell by 47%. Trading volumes, customer numbers, loss distributions, product mix and the legal causes of action may be concentrated in the years that remain. CMC also says the March 2026 amendment removed certain causes of action and added others. The correct conclusion is therefore procedural: the pleaded window became materially shorter and the legal case was reshaped. A reliable financial estimate would require the amended pleading, group definition, loss methodology and trading records, none of which can be replaced by a calendar ratio.[1]

CMC’s filing says the proceeding was initiated in May 2022 and is listed for trial in October 2027. The company states that it has filed a defence and cannot reliably estimate the financial effect. Those statements are consequential. A provision, contingent-liability estimate or settlement figure is not disclosed in the cited report. The absence of a reliable estimate is not evidence that exposure is zero; equally, the existence of a class action is not evidence that the pleaded misconduct occurred. The claims remain unresolved.[1][5]

The claimant website continued to display the original start date in portions of its public explanation while also publishing updates on the amended proceeding. That creates a practical disclosure problem: a reader can encounter two different time perimeters depending on which section is read. The March 2026 amendment is the later procedural event and CMC’s results describe the five-year window. Prospective group members and journalists should therefore verify the operative pleading rather than rely on an older landing-page description.[1][4]

Discovery shows why the case is expensive before trial

The May 2025 judgment gives an unusually detailed view of the underlying data problem. The applicants sought raw data and trading-loss information for a five-year period. Screening tests designed to identify files or records produced about 2,500 failures. The judgment dealt with how CMC should investigate and explain those failures and with the practical burden of producing usable information. This is not a finding about whether customers were harmed. It is evidence that reconstructing the trading population and transaction history is a large technical exercise in its own right.[2]

The judgment also recorded evidence about CMC’s group execution structure. Customer trades were hedged back-to-back with the UK parent and risk was managed at group level. That fact may be relevant to how the applicants frame causation, pricing or conflicts, but it does not by itself establish unlawful conduct. Back-to-back hedging can be a risk-management arrangement. The legal question depends on the duties, representations, product design, customer circumstances and alleged loss mechanisms pleaded in the case. The documentary chain should not be shortened into the assertion that hedging proves the claim.[2]

For a broker, discovery cost can extend beyond legal review. Historical platform records must be located, extracted, reconciled and explained; data definitions may have changed; accounts can span migrations; and transaction records may need to be tied to client agreements, disclosures and market data. A five-year order can still be operationally demanding even after the pleaded period has been reduced. The 2,500 failed screening tests illustrate a quality-control challenge: a dataset is not complete merely because a query ran. Exceptions have to be investigated.[2]

Security for costs is protection, not a damages proxy

Security for costs is intended to protect a respondent against the risk that an eventual costs order cannot be satisfied. It is not an advance judgment on the merits. The A$4.6 million recorded as consensually provided and the approximately A$8.6 million later requested relate to potential litigation costs. They should not be added to any alleged customer losses or described as money CMC has been ordered to pay claimants. The 2026 decision focused on whether invoices sought to test the security request were protected by legal professional privilege.[2][3]

The combined A$13.2 million is nevertheless informative. It provides a disclosed measure of the scale at which the parties were arguing about prospective costs. It also shows why funding arrangements matter in class litigation. Applicants may depend on litigation funders or adverse-cost insurance; respondents seek assurance that costs can be recovered if the claim fails. The court has to balance that protection against the risk that excessive security prevents a potentially valid group claim from proceeding. The published decisions illuminate the mechanism without resolving the underlying allegations.[2][3]

The privilege ruling adds another boundary. The applicants sought 17 invoices to examine the basis for CMC’s additional-security estimate. The court upheld privilege. That means the public record identifies the requested total but does not expose every underlying task, rate or assumption in those invoices. A journalist cannot responsibly reverse-engineer the A$8.6 million into a forecast trial budget or claim that the court endorsed every component. The amount is a request made in the proceeding, not a verified final cost.[3]

The strongest questions remain empirical

The amended period raises testable questions. How many accounts and trades remain in the group after May 2016? What share of the originally identified losses falls inside the retained window? Which causes of action were removed, which were added, and how do those changes alter the loss model? Do the five years of raw data reconcile to the applicants’ proposed population? Those answers would materially improve any assessment of exposure. CMC’s statement that no reliable estimate can be made is consistent with a case whose group and damages mechanics are still being tested.[1][2]

A response from CMC would also be most useful on matters not settled by the filings: whether the company expects further amendments, whether the discovery exceptions have been resolved, and whether the additional security sought was ultimately ordered, reduced or otherwise satisfied. The published company position—defence filed, trial scheduled and impact not reliably estimable—has been incorporated here. No inference of misconduct is drawn from the procedural orders or from the security request.[1][2][3]

For current and former customers, the public-interest issue is precision. The class action remains listed in the Federal Court, but the operative time window described by CMC is now five years rather than nearly a decade. The case has already required substantial data discovery and a security debate measured in millions of Australian dollars. Those facts establish complexity and changed scope, not liability. The trial and any later orders will determine whether the alleged duties were breached and whether compensable loss can be proved.[1][5]

The defensible finding is therefore two-part. First, the pleaded period contracted by about 47%, a change that should be reflected in every current description of the case. Second, the disclosed security process reached A$13.2 million when the A$4.6 million provided is combined with the A$8.6 million additional request. Neither number can be used as a shortcut to damages. Together they show how much the litigation’s perimeter and cost architecture have evolved before the evidence has been tested at trial.[1][2][3][4]

Report an error or submit an update