Alford v AMP Superannuation Limited

Solicitor for the Respondents: King & Wood Mallesons

BETWEEN: DALE ROBERT ALFORD First Applicant

ANNE COOPER (and another named in the Schedule) Third Applicant

AND: AMP SUPERANNUATION LIMITED (ACN 008 414 104) First Respondent

N.M. SUPERANNUATION PTY LTD (ACN 008 428 322) Second Respondent

AMP LIFE LIMITED (ACN 079 300 379) (and others named in the Schedule) Third Respondent

order made by: MURPHY J DATE OF ORDER: 5 April 2024

1. The interlocutory application by the First and Second Respondents is dismissed.
2. The First and Second Respondents pay the Applicants' costs of and incidental to the application. Note: Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

INTRODUCTION 1 This proceeding is a superannuation class action brought by the applicants, Dale Alford, Sebastian Smith, Anne Cooper and Jodie Mitchell against the respondents, AMP Superannuation Limited (ASL) and N.M. Superannuation Pty Ltd (together, the Trustee Respondents) and four other companies in the AMP group of companies. The applicants bring the class action on their own behalf and on behalf of a broad class of persons who were members of one or more of 21 superannuation products (Impugned Products) across a time period of up to 12 years (from 1 July 2008 to 15 May 2020 for certain Impugned Products and 30 March 2011 to 15 May 2020 for others) (Relevant Periods). The proceeding alleges that the Trustee Respondents failed to comply with their obligations under the Superannuation Industry (Supervision) Act 1993 (Cth) and under the general law, which caused group members' fees to be higher and investment returns to be lower than they otherwise would have been had the Trustee Respondents complied with their obligations. 2 Before the Court is an application by the Trustee Respondents seeking orders in relation to the form, timing, and distribution of an opt out and registration notice, a registration process prior to a pending mediation, and what has become known as a 'soft' class closure order, pursuant to ss 33J, 33X, 33Y and 33ZF of the Federal Court of Australia Act 1976 (Cth) (FCA Act). For convenience I will describe the Trustee Respondents as "the respondents". 3 The main dispute between the parties concerned the proposed orders to require group members to register and to be subject to a soft class closure order, which were strenuously opposed by the applicants. 4 I heard the application on 5 April 2024 and made orders that day to dismiss the application. I now provide reasons for doing so. In my view the respondents' application for orders to require group members to register and for soft class closure had little or no merit. In broad summary, I considered it appropriate to dismiss the application because: (a) the proposed registration and class closure orders are strenuously opposed by the applicants, and on cogent grounds. The applicants have obligations to represent group members' interests and the respondents' interests are inimical to those interests. Real caution is appropriate when an application for class closure orders is opposed by the applicant; (b) the evidence does not establish that the respondents' records are insufficient to allow the respondents to compile a representative sample of the particular products and investment options typically taken up by group members, and to thereby obtain a sufficient understanding of group members' claims to participate efficiently and effectively in the pending mediation. At least in part, the application for a registration regime is a cloak to disguise the fact the respondents wish to confine the class to just those group members who register; (c) because many group members are disengaged from their superannuation the proposed registration regime is likely to result in a very low level of registration, and is unlikely to generate a representative sample of group members' claims (as distinct from a sample of the claims of those group members who register); (d) the proposed registration regime is likely to cost the applicants in the order of $2 million, and to divert the applicants' legal team from preparing for the mediation and trial; (e) the relationship of the respondents to the group members is that of trustee and beneficiary. The fiduciary nature of that relationship points away from making the proposed class closure order; and (f) this is not a case where the great majority of the class will need to register at some point in order to be identified and paid their share of any settlement or judgment that may be achieved. 5 The application for orders in relation to opt out has more merit, but I considered it appropriate to dismiss that aspect of the application too. It is likely to be cheaper and more efficient if opt out is ordered to occur after the pending mediation.

THE FACTS 6 The respondents rely upon the affidavit of Natalie Tatasciore, a partner of King & Wood Mallesons (KWM), one of the partners with the conduct of the proceeding for the respondents, sworn 2 April 2024. Ms Tatasciore is an experienced lawyer, having been admitted in 2003 and a partner of KWM since July 2018. She is experienced in complex litigation, including in the conduct of class actions and of mediations in such proceedings. 7 The applicants rely upon the following evidence: (a) an affidavit of Rebecca Gilsenan, a partner of Maurice Blackburn which (together with Slater & Gordon) represents the applicants in the proceeding, affirmed 28 March 2024. Ms Gilsenan is an experienced lawyer, with lengthy experience in class action litigation, having practised principally in class action litigation since 1999, and having been a partner in the firm's Class Actions division since 2005 and the National Head of the Class Actions division of the firm since November 2023; and (b) an affidavit of Emma Pelka-Caven, a solicitor in the employ of Slater & Gordon which (together with Maurice Blackburn) represents the applicants in the proceeding, affirmed 28 March 2024. She is an experienced lawyer, with experience in class action litigation, having been admitted in 2005, a Practice Group leader in the Class Actions department of Slater & Gordon since November 2018 and the Head of that department since July 2022. 8 None of the deponents were cross-examined.

The number of group members 9 Based upon an expert's report, Ms Pelka-Caven estimated that there are approximately 2 million to 2.2 million persons who were members of the Impugned Products over the relevant periods. However, she said that limitations in that estimate mean that there might be substantially more group members than that. The respondents referred to the applicants' estimate but did not confirm its accuracy or provide an alternative estimate. For the purposes of the application I accept that there are approximately two million group members.

Presently registered group members 10 Ms Tatasciore stated, based upon an earlier affidavit of Ms Gilsenan, that there are at least 13,691 potential group members who have provided their details to the solicitors for the applicants, comprising: (a) 3,382 persons who have retained Maurice Blackburn. According to the applicants, all of those persons have provided their contact details and all but approximately 500 of them have provided information as to the superannuation product of which they were a member during the Relevant Periods; and (b) 10,309 persons who have registered with Slater & Gordon. According to the applicants, those persons have only provided their contact details and no other information. I accept that evidence.

The adequacy of the respondents' records 11 The respondents' central argument is that orders for registration and class closure are appropriate because they do not have sufficient information in relation to group members' claims to ensure that the pending mediation is efficiently and effectively conducted, and capable of leading to a settlement. They contended that the proposed registration process is likely to provide them with sufficient detail of the particular products and investment options taken up by group members to enable the respondents to compile a representative sample of group members' claims (or at least of those group members who wish to participate in any settlement) and enable the respondents to prepare a methodology and undertake some analysis of the aggregate quantum of group members' claims in respect of particular products and investment options. 12 Ms Tatasciore stated, and I accept, that the Further Amended Consolidated Statement of Claim (FACSOC) advances various different claims in respect of the 21 Impugned Products, which are made across a time period of up to 12 years, and that there are hundreds of permutations of Impugned Products and different investment options available to group members within each of the Impugned Products in respect of which the claims pleaded in the FACSOC are advanced. 13 Ms Tatasciore said that it will be necessary for the respondents to identify the specific combination of Impugned Product and investment option applicable to each particular group member, which may have changed from time to time during the Relevant Periods, to understand the particular claims advanced in the proceeding on behalf of that group member. I am not persuaded as to that. Ms Gilsenan stated that the applicants' expert loss evidence, which is to be filed well prior to the mediation, will assess the loss incurred by group members due to the respondents' conduct at the fund level, rather than by reference to each individual group member, and will show the way in which the lead applicants and sample group members incurred a proportion of the loss suffered at the fund level. That is, the applicants and group members' loss claims will be advanced using a top-down methodology, rather than pursuant to a bottom-up methodology which involves the aggregation of individual group members' asserted losses. 14 In support of the respondents' central contention Ms Tatasciore said the following in relation to the information that can be obtained from the respondents' records: [17] The group member definition in paragraph 3 of the FACSOC captures persons who held an "ASL Impugned Product" as far back as 1 July 2008 and persons who held an "NMS Impugned Product" as far back as 30 March 2011 (FACSOC paragraphs 3.1 and 3.2); and an indeterminate cohort of other persons who will not necessarily be members of the superannuation funds the subject of the Proceeding (FACSOC paragraphs 3.3 and 3.6). [18] Insofar as the group members in the Proceeding are persons who held one of the Impugned Products during the Relevant Periods, the Trustee Respondents have proposed in the Proposed Orders that they will use their best endeavours to compile an Excel spreadsheet from their business records which contains (to the extent such information is available) the names and best available email addresses of all persons who held an Impugned Product as at a particular date in each year during the Relevant Periods. I am informed by Ms Rodgie [Senior Legal Counsel, Commercial Litigation, AMP) and believe that the Trustee Respondents anticipate they will face significant data issues seeking to identify the names and best available email addresses for all of these persons across the Relevant Periods, in particular because: (a) the data is held on four product administration systems, two of which have now been archived; (b) many of the Impugned Products are products that were historically offered to members via employer superannuation plans. In these cases, the employer was principally responsible for collecting personal information for their employees who participated in the plan and providing it to the Trustee Respondents. The Trustee Respondents did not have direct contact with those employee members on commencement and had no vetting capability to verify the data at the time and, therefore, historic deficiencies exist in that member data; (c) tax file numbers may not have been provided in all cases as there was no obligation for a prospective member to provide one; (d) there is an inherent reliance on members to proactively update their contact details; and (e) all group members who held accounts in the ERF Product and some other group members (particularly those that became members via an employer superannuation plan), are no longer members of the superannuation funds the subject of the Proceeding, and have not been for some time. Therefore, the best available email address that the Trustee Respondents have for those group members may not be current. [19] Insofar as the group members in the Proceeding are persons who did not hold any of the Impugned Products during the Relevant Periods (i.e., persons who fall within paragraphs 3.3 - 3.6 of the FACSOC group member definition) (and who are not already a Registered Group Member), the parties are unlikely to be able to identify those persons absent a registration process. 15 For the reasons I later explain, I do not accept the thrust of that evidence.

The asserted requirements for a successful mediation 16 Ms Tatasciore then deposed: [22] I am informed by Ms Rodgie and believe that any settlement offer to be made or accepted in this Proceeding will require approval from the Boards of the Trustee Respondents as well as the Board of AMP Limited. I am further instructed by Ms Rodgie and believe that, in order approve any such settlement in this Proceeding, each of these Boards will require a Board paper which will be expected to set out, among other things: (a) the recommended range for negotiating settlement; (b) which members are likely to be paid out of any settlement; and (c) the methodology utilised to calculate the recommended range for settlement, which must include the potential liability if the Trustee Respondents were to lose on liability on the claims advanced in the FACSOC in respect of particular Impugned Products, and how much each group member is likely to receive from a distribution within the recommended settlement range. In order to provide this information to the Boards with a level of certainty, I require some knowledge of (at least) the number of group members in the Proceeding who are likely to participate in the settlement, the particular Impugned Products of which they were members and the period for which they held those products, so that modelling of the quantum of these potential claims can be undertaken ahead of the mediation. [23] As a result of the matters set out in paragraphs 10 to 14 above, and based on my experience as set out in paragraphs 6 and 7 above, my view is that: (a) the Proceeding is not one which can be mediated on the basis that the claims of all group members are the same (given the individualised nature of the pleaded claims in respect of different products and investment options), nor could a mediation be conducted without separating out the sub-classes of group members in each product and in each category of claim in respect of choice investment options, MySuper investment options and cash options; and (b) for those reasons, it would not be appropriate to structure a settlement based upon an estimate of the total number of group members and the Applicants' expert evidence on aggregate loss across all of the Impugned Products once this evidence is served by 2 August 2024 (unless these were each broken down across each particular product). In particular, such a methodology would not have regard to the proportion of group members who hold a product in respect of which there is no allegation that the fees were high and no pleaded counterfactual or causation case (and the approach would be to treat their claims in the same way as those for whom there is such an allegation). [24] Further, based on my experience as set out in paragraphs 6 and 7 above, my view is that, without the information that would be provided through the proposed registration process: (a) it will not be possible for the parties to know the number of group members who might later make a claim against any settlement fund and the estimated quantum of their claims. This will necessarily require assumptions to be made about the likely participation rate in the context of any settlement discussions, in order for the parties to assess the value of any settlement offer; (b) the parties will be constrained in assessing terms for resolution, and are likely to adopt materially different views about the number of potential group members, the value of the claims of potential group members, and the likely participation rate of group members. In my experience, this disparity will make it significantly less likely that the Proceeding will be capable of resolution at mediation; and (c) the Boards of the Trustee Respondents and AMP Limited will not be able to properly assess the quantum of the commercial resolution required to provide those group members who wish to participate in the settlement process a sensible compromise of their claims, and to otherwise settle the Proceedings. [25] In those circumstances, my view is that, in order to properly advise the Trustee Respondents in relation to the appropriateness of any settlement offer or settlement range, the solicitors and counsel advising the Trustee Respondents will require information as to: (a) the number of group members who intend to participate in any settlement as at the date of the mediation (noting that further group members would also be able to register to partake in a settlement after it is reached with leave of the Court); (b) which of the Impugned Products were held by those group members; and (c) a methodology that could be used to estimate the claims of those group members. 17 For the reasons I later explain, I am not persuaded that declining to make the proposed registration and class closure orders will result in lower prospects of the parties reaching a reasonable settlement at the mediation. I consider the respondents' own records are likely to provide them with sufficient information to efficiently and effectively engage in the mediation, and that given the likely low level of registration the proposed registration regime is unlikely to generate a representative sample of group members' claims. 18 Ms Tatasciore stated that the respondents are not opposed to structuring the registration process so as to allow group members who have not registered by the proposed 6 September 2024 deadline a further opportunity to do so after any settlement is reached, although she said that would be "subject to the terms of any settlement agreement" which carried with it some uncertainty. I accept that evidence.

The likely low level of registration 19 Ms Pelka-Caven stated that a feature of the superannuation industry is that superannuation fund members have relatively low engagement in relation to their superannuation. She relied on several documents in support of that statement: (a) an internal memorandum of the respondents dated October 2009 to the directors of ASL titled "Use of General Purpose Reserves - Employee Brochure", which suggested that the expected response rate to communications to members was 1-2%; (b) an internal memorandum of the respondents dated September 2012 to the directors of ASL titled "Personalised Member Benefit Reporting Initiative - Status Update", which discussed the results of that ASL initiative which included the distribution of 900,000 personalised documents to the members of Flexible Lifetime Super and AMP Flexible Super (both Impugned Products). A listed goal of the initiative was to improve the engagement of members with their superannuation. As at the date of the memorandum 4.51% of the 514,859 members who had been mailed a report were considered to be "engaged" under the KPIs set for the initiative; and (c) the 2018 Productivity Commission report titled "Superannuation: Assessing Efficiency and Competitiveness" (December 2018 p18) which found: Some members are highly engaged with the super system - actively comparing products or opening SMSFs. But most are not. Many members simply default, and rely on their funds to manage their super for them. Levels of engagement are especially low among the young and members with low balances. … Low member engagement is not necessarily a problem. For many members, it is rational. Engaging takes time and effort, and trustees are charged with acting on members' behalf and in their best interests. Indeed, low engagement is to be expected in a compulsory and complex system that covers the bulk of the population. In some cases, disengagement can also be a consequence of cognitive constraints and behavioural biases such as myopia, loss aversion, and a tendency to procrastinate. (Emphasis added.) 20 Nothing in Ms Tatasciore's evidence or the respondents' submissions contradicted Ms Pelka-Caven's evidence about the likely low level of engagement of superannuation fund members in relation to their superannuation. I accept that evidence. 21 Ms Pelka-Caven also gave evidence in relation to the opt out rates in three superannuation class actions brought by Slater & Gordon, being: (a) Tracy Ghee v BT Funds Management Limited & Anor (VID 962 of 2019) (BT Funds Class Action), a class action on behalf of members in two superannuation funds of which the first respondent, BT Funds Management Limited, was the trustee. She stated that in this case the opt out notice was sent to a total of 78,739 group members by a variety of distribution methods including email and post. Only 86 group members opted out, which represented only 0.11% of the class; (b) Keith Kayler-Thomson & Ors v Colonial First State Investments Limited & Ors (VID 1313 of 2018), a class action on behalf of members in the respondents' superannuation funds. In this case the opt out notice was sent to a total of 731,246 group members by email, post, and SMS. Only 657 group members opted out, which represented 0.09% of the class; and (c) Marcel Eugene Krieger & Anor v Colonial First State Investments Limited (VID 1141 of 2019) (Colonial First State Fees Class Action), a class action on behalf of members in the respondent's superannuation fund. In this case the opt out notice was sent to a total of 395,454 group members by email and post. Only 815 group members opted out, which represented 0.21% of the class. 22 She said, based on her experience in conducting or overseeing opt out processes across a range of different types of class actions, that the opt out rates in those class actions were markedly lower than in shareholder class actions in which Slater & Gordon has acted. 23 Ms Gilsenan gave similar evidence about the opt out rates in superannuation class actions brought by Maurice Blackburn. She relied on: (a) Lesley Coatman v Colonial First State Investments Limited and Another (Federal Court Proceeding VID 1139 of 2019). In this case the opt out notice was sent to approximately 14,345 group members in hard copy, approximately 77,000 group members via email and approximately 10,500 group members via SMS. Only 102 group members opted out, which represented 0.1% of the class; and (b) David Shimshon v MLC Nominees Pty Ltd and Another (Supreme Court of Victoria Proceeding S ECI 2020 00306). In this case the opt out notice was sent to approximately 220,000 group members by email and SMS, and was published in 10 national broadsheet newspapers. Only 336 group members opted out, which represented 0.1% of the class. Ms Gilsenan stated, based on her experience and on the similarities in class membership between the proceedings, that there was likely to be a very low level of opt out in the present case. 24 Nothing in Ms Tatasciore's evidence or the respondents' submissions said anything to contradict Ms Pelka-Caven's or Ms Gilsenan's evidence in relation to the likelihood that the opt out rate in the present case would be low. I accept that evidence. 25 Ms Pelka-Caven also gave evidence in relation to post-settlement registration processes facilitated or conducted by Slater & Gordon in the Colonial First State Fees Class Action and the BT Funds Class action. She said that: (a) in the Colonial First State Fees Class Action, pursuant to an order of the Court a cohort of group members were notified that they were required to register so that their eligibility for a settlement distribution payment could be determined. The registration period was between 9 February and 15 March 2024. 178,126 group members were given notice of the requirement to register by email and, if no email address was known, by post and newspaper advertisement. By the end of the registration period, 18,684 group members had registered, being approximately 10.5% of the class; and (b) in the BT Funds Class Action on 15 June 2023, pursuant to an order of the Court all group members (including group members who the respondents had contact details for and group members who had given their details to Slater and Gordon to receive updates on the proceeding) were advised by notice of proposed settlement to register in order to receive payments under the proposed settlement, while also noting that some might receive payments directly into their superannuation accounts without registration. Between 17 and 26 July 2023, 78,884 group members were sent the notice of proposed settlement by email and, if no email address was known, by post. After settlement distribution had commenced, all group members for whom there was a telephone number were also sent an SMS and social media advertisements about the requirement to register. As at 9 November 2023, approximately 15,000 group members had registered, being approximately 19% of the class. 26 Again, nothing in Ms Tatasciore's evidence or the respondents' submissions said anything to contradict Ms Pelka-Caven's evidence in this regard. I accept that evidence. 27 I do not accept the respondents' contention that evidence of opt out and registration rates from other class actions should not be received, citing the remarks of Delaney J in Anderson-Vaughan v AAI Limited (No 2) [2024] VSC 65 at [70]-[72]. 28 First, in my view the evidence of opt out and registration rates in the superannuation class actions referred to has some probative value as a proxy in relation to the likely registration rates in this superannuation class action. 29 Second, it is not clear to me, and the respondents did not explain, how the opt out and registration rates in other proceedings is subject to legal professional privilege, without prejudice privilege or a Harman restriction. Nor is there any evidence from which I would so conclude. Generally speaking, opt out notices are filed in the Court by those group members who do not wish to be part of the case, and they are available for public inspection. I do not understand why or how it is said that they attract legal professional privilege or without prejudice privilege in the proceeding in which they are filed, nor why disclosure of the rate of opt out in a proceeding would involve a breach of the Harman obligation in that proceeding. Depending on the process adopted, registration notices may also be filed in the Court by those group members who wish to register, and they too will be available for public inspection. Some registration procedures may be confidential but there is no evidence as to that. Finally, if the registration rates in other class actions are subject to legal professional privilege or without prejudice privilege, such privilege can be waived by the party or parties holding the privilege. There is no evidence as to that either. 30 Ms Pelka-Caven's evidence was supported by Ms Gilsenan who deposed that, having regard to the fact that: (a) most superannuation members are not engaged with the superannuation system; (b) in her experience registration rates in class actions generally are almost always less than half of the class at most, and are often considerably lower, she expected the registration rate for any soft class closure process at this stage of the case to be very low. Nothing in Ms Tatasciore's evidence or the respondents' submissions contradicted Ms Gilsenan's evidence in that regard. I accept that evidence. 31 In my view superannuation fund members typically have a low level of engagement with their superannuation and the proposed registration orders are likely to result in a very low rate of registration. That conclusion is significant to my decision in the application.

The cost associated with the proposed registration process 32 Based on her experience in overseeing several registration processes in class actions involving large numbers of group members in recent years, including the conduct of such processes in-house, through third-party providers, and by a combination of both, Ms Gilsenan deposed that the proposed registration regime will involve substantial costs, time, and resources. I accept that. 33 She stated that because the cost of sending notices to group members by ordinary mail is significantly more expensive than sending notices via email or SMS, she would ordinarily consider that it would be appropriate to distribute any notice only by email and/or SMS where such contact information is available. Based on her experience she said that distribution of registration notices is likely to be more effective when undertaken through multiple channels. I accept that too. 34 Ms Gilsenan also said that, in the context of the proposed class closure order which could result in group members losing important rights, and in circumstances where the group members have a low level of engagement with their superannuation, it was appropriate to allow for additional notification measures and for follow-up communications to group members reminding them of the requirement to register to be used in a case like this. She said that in a recent large class action in which Maurice Blackburn acts the registration rates effectively doubled following distribution of reminder notices via a dual distribution approach. 35 Nothing in Ms Tatasciore's evidence contradicted Ms Gilsenan's evidence in this regard. Indeed, the respondents accepted that it was important for the proposed notices to be distributed widely and through various means. I broadly accept Ms Gilsenan's evidence as to the appropriate extent of notice in this case. 36 The respondents proposed direct distribution of the notices via email to group members, and then only by way of newspaper publications, and the website pages of the Federal Court, Maurice Blackburn, Slater & Gordon and AMP Superannuation. I do not accept that notice to group members by publication in newspapers and on the nominated websites would be adequate for the notice proposed by the respondents. 37 I have no difficulty in accepting that sending the proposed notice to group members by email is adequate notice for those group members for whom the respondents have email addresses. However, the respondents were unable to say how many, or what percentage of, group members they held email addresses for. It is plainly inadequate to send notices to group members only by email when it is unknown what approximate percentage of group members will be reached by that method. The proposed notice should be provided to group members by email where an email address is held, but where their email addresses are unknown or where the email notice "bounces back", it will be necessary to send the notice to group members by ordinary post. 38 Having regard to Ms Gilsenan's evidence, the cost of the proposed registration process will significantly increase depending upon the percentage of group members to whom notice must be provided by post. 39 Ms Gilsenan provided the following cost estimates: [28] In any event, in my recent experience, the cost of distributing notices is approximately: (a) $1.60 per notice sent by ordinary mail, although the cost can be higher through some third party providers; (b) $0.04 per notice sent by email; and (c) $0.24 per notice sent by SMS ($0.06 per 160 characters). … [35] Using the figures at paragraph 28 above, I estimate the cost of distributing notices to 2 million group members in this proceeding through the following channels as follows: (a) Ordinary mail - $3.2 million. (b) Email - $80,000. (c) Combination of ordinary mail (50%) and email (50%) - $2 million. [36] In light of the potential benefits referred to above that a multiple channel approach and/or reminder notices may have for registration rates, in my view it would be appropriate to send multiple notices to group members in a case such as this. Therefore, in the event that one supplementary / reminder email and one supplementary / reminder SMS is sent to each group member, the cost estimates referred to above substantially increase as follows: (a) Ordinary mail plus additional SMS notice and reminder SMS notice - $4.16 million. (b) Email plus additional email notice and reminder SMS notice - $640,000. (c) Combination of ordinary mail (50%) and email (50%) plus additional email notice (50%) and reminder SMS notice (100%) - $2.16 million. [37] I would also expect the professional fees associated with a registration process (other than the work and cost associated with dispatching the notices) to easily exceed $750,000. The work would include setting up a registration platform, responding to group member enquiries by telephone and email, handling registration processes and managing the data of registered group members. [38] Accordingly, I estimate the cost of any registration process in this case to be between $830,000 and $4.91 million. (Emphasis added.) 40 Nothing in Ms Tatasciore's evidence contradicted Ms Gilsenan's evidence in this regard, but the respondents submitted that a more reasonable estimate was in the order of $1 million. I was not, though, persuaded as to the basis for such a conclusion. 41 Ms Gilsenan said that a combination of ordinary mail (50%) and email (50%) resulted in notice costs of $2 million, but by reference to her own figures that appears to be an overestimate (50% of $3.2 million for ordinary mail ($1.6 million) plus 50% of $80,000 for email ($40,000) totals $1.64 million). But little turns on that. On Ms Gilsenan's figures the cost of notice will be approximately $2.16 million if it is provided by ordinary mail (50%) and email (50%) plus a reminder email notice (50%) and reminder SMS notice (100%). And, as explained below, that total does not take into account the internal costs that will be incurred by the applicants through the proposed registration regime. 42 It is unnecessary to reach a concluded view as to the cost of the proposed registration regime. For the purposes of the application I proceed on the basis that it is likely to cost the applicants in the order of $2 million.

The internal costs of registration for the applicants 43 Ms Gilsenan also said that, in addition to the significant cost associated with conducting a registration process for the purposes of soft class closure, the process will also involve very significant expenditure of time and resources by the applicants' legal team in order to manage group member enquiries and deal with registration issues. 44 She based her evidence on a recent registration process conducted by Maurice Blackburn in three related class actions in the Supreme Court of Victoria (Fox and Anor v Westpac Banking Corporation and Anor (S ECI 2020 02946); O'Brien v Australia and New Zealand Banking Group Limited and Anor (S ECI 2020 03365); and Nathan and Anor v Macquarie Leasing Pty Ltd (S ECI 2020 03924), known as the Flex Commissions class actions. In those related proceedings one million registration notices were distributed to group members as part of an initial round of distributions. She stated that during the three-month registration periods, Maurice Blackburn: (a) received over 12,600 emails, approximately 4,000 voicemails and answered approximately 26,000 telephone calls; (b) required 13 client service officers, three paralegals, four lawyers and two associates, with occasional senior lawyer supervision to work on the registration and opt out process. 45 I accept the thrust of that evidence.

The diversion of key resources 46 Ms Gilsenan also stated that, if soft class closure is ordered, it will divert core resources away from progressing the substantive stages of this case. She noted that the timetable for this proceeding provides that: (a) by 31 May 2024, the respondents file and serve any lay evidence on the question of liability; (b) by 28 June 2024, the respondents file and serve any expert evidence on the question of liability; (c) by 2 August 2024, the applicants file and serve any lay and/or expert evidence on loss; (d) by 23 August 2024, the applicants file and serve any reply expert evidence on the question of liability; (e) mediation must take place by no later than 18 October 2024; (f) the trial is listed to commence on 26 May 2025 on an estimate of eight weeks. 47 She said that the interests of the applicant and group members will be best served if their legal team devotes their resources to the critical work required for preparation for mediation and trial, rather than have core resources diverted by conducting a registration process. 48 She also stated that: (a) the class in this case is approximately double the size of the classes in the Flex Commissions class actions; and (b) most group members in this case are unlikely to know the specific details of the superannuation products they hold (or held) and whether they are Impugned Products, which means that group members are more likely to have difficulties understanding how any registration notice applies to them. In her view this means that the applicants' legal team may be required spend even more significant time in dealing with group member's enquiries. 49 Nothing in Ms Tatasciore's evidence contradicted Ms Gilsenan's evidence in this regard. I accept Ms Gilsenan's evidence.

Whether a registration regime is likely to be necessary at some point 50 One of the respondents' arguments is that (assuming the applicants' success in the proceeding), it will be necessary for group members to register at some point if they are to be identified and paid their share of any settlement or judgment. On this argument, given that the information obtained through registration will be of some assistance to the parties in the pending mediation, registration should be undertaken now. 51 I am not, however, persuaded that in this case it will be necessary for most group members to register before they can be paid their share of any settlement or judgment. The relationship between the respondents and group members is that of trustee and beneficiary. If the Court finds that the respondents have breached their statutory and general law duties to group members then the respondents are likely to have an obligation to ensure that the trust fund is made good, and group members will thereby recover their losses. For most group members that will not involve registration. I accept there are deficiencies in the respondents' records, but for the reasons I explain I expect that it will be possible to identify the great majority of group members and to pay them their share of any settlement or judgment without a registration regime. 52 And for those group members for whom the respondents' records are out of date or unreliable, the evidence shows that the Australian Taxation Office (ATO) has a process for returning funds to former superannuation account holders. Ms Gilsenan said, and I am aware from my involvement as the docket judge in the Coatman superannuation class action, that of the 95,880 group members who received a settlement from the distribution sum, 41,730 (approximately 43.5% of those members) were paid via the trustee voluntary payment (TVP) mechanism administered by the ATO under the Superannuation (Unclaimed Money and Lost Members) Act 1999 (Cth). A further 2,892 group members (approximately 3% of members who received a settlement) received a distribution by way of payment into a rollover superannuation fund. I accept that the settlement monies were successfully distributed notwithstanding that no registration process was conducted. 53 The settlement distribution in Coatman tends to show that the ATO TVP mechanism is an effective process to distribute settlement monies to former superannuation fund members. That is not to suggest it provides a perfect solution. For example, the ATO TVP process requires provision of a tax file number (TFN) and I accept Ms Tatasciore's evidence that the respondents do not have TFNs for all group members. But we are not here searching for perfection. In a class of two million group members dating back for 14 years, the best that can be hoped for is a relatively efficient distribution process which takes into account the data deficiencies that exist.

Whether class closure is necessary for settlement 54 The thrust of Ms Tatasciore's evidence was that the proposed registration and class closure orders are necessary if the pending mediation is to have reasonable prospects of success. She said that without a registration regime it will be impossible to know how many group members might later make a claim against any settlement fund, and that the parties are likely to adopt materially different views about the number of potential group members who might come forward, such that it will be significantly less likely that the proceeding will be capable of resolution at mediation. 55 Against that, Ms Gilsenan provided the following table of cases conducted by Maurice Blackburn in which settlement had been achieved without a soft class closure order. She said while each proceeding is unique, that information shows that it is possible to settle class action litigation without a class closure order: No. Case Settlement date Type of class action 1 Jarra Creek Central Packing Shed Pty Ltd v Amcor Limited (NSD 702 of 2006) 10 March 2011 Competition 2 Casey v DePuy International Ltd & Anor (ACD 10 of 2010) 19 November 2012 Product liability 3 Inabu Pty Ltd v Leighton Holdings Limited (NSD 2244 of 2013) 16 May 2014 Shareholder 4 Gray v Cash Converters Personal Finance (NSD 2089 of 2013); Gray v Cash Converters International Limited (NSD 2090 of 2013) 18 June 2015 (varied on 21 July 2015) Consumer 5 Stanford v DePuy International Ltd (NSD 213 of 2011) 17 June 2016 Product liability 6 McKenzie v Cash Converters International Ltd (NSD 601 of 2016) 22 October 2018 Consumer 7 Simpson v Thorn Australia Pty Ltd trading as Radio Rentals (NSD 448 of 2017) 2 September 2019 Consumer 8 Lynch v Cash Converters Personal Finance Pty Ltd (NSD 900 of 2015) 18 October 2019 Consumer 9 Dalton & Anor v Volkswagen & Anor; Richardson v Audi & Ors; Roe v Skoda & Ors (NSD 1459 of 2015, NSD 1472 of 2015 and NSD 1473 of 2015) December 2019 Consumer 10 Lesley Coatman v Colonial First State Investments Limited and Another (VID 1139 of 2019) June 2022 Superannuation

56 I accept that the table is accurate, and that class actions can be settled without making a class closure order. But that conclusion is not significant to my decision.

CONSIDERATION REGARDING REGISTRATION AND CLASS CLOSURE 57 The respondents' application for registration and class closure orders has little or no merit. 58 First, the respondents' application faces the difficulty that it is strenuously opposed by the applicants, and on cogent grounds. 59 In Melbourne City Investments Pty Ltd v Treasury Wine Estates Ltd (No 2) [2017] FCAFC 98; 252 FCR 1 at [72]-[75] (Jagot, Yates and Murphy) the Full Court said: [72] The Commonwealth Parliament, in implementing a core recommendation of the Australian Law Reform Commission in its report Grouped Proceedings in the Federal Court, Report No 46 (Canberra, 1988) at [127], expressed a legislative intention to adopt an opt out rather than an opt in procedure: Second Reading Speech, Federal Court of Australia Amendment Bill 1991 (Cth), House of Representatives Parliamentary Debates, Hansard, 14 November 1991 p 3,175. It must be accepted that the requirement for class members to take active steps to "register" in order to share in a settlement of a class action undercuts to some extent the opt out rationale underpinning the Part IVA regime. In Mobil Oil Australia Pty Ltd v Victoria (2002) 211 CLR 1; [2002] HCA 27 at [40] (Gaudron, Gummow and Hayne JJ) their Honours said: Group members, however, need take no positive step in the prosecution of the proceeding to judgment to gain whatever benefit its prosecution may bring. [73] Class proceedings are intended to require little or no active involvement by class members and class members participate principally for the limited purpose of taking the benefit or suffering the burden of the findings made on the common questions: P Dawson Nominees Pty Ltd v Brookfield Multiplex Ltd (No 2) [2010] FCA 176 at [16] (P Dawson No 2) (Finkelstein J). As J Forrest J said in Thomas v Powercor Australia Ltd (Ruling No 1) [2010] VSC 489 (Thomas v Powercor No 1) at [30], "one of the consequences of the opt out model, as was clearly intended by the legislature, is the ability of group members to "sit back" and watch the proceeding unfold". There must be a good reason to exercise the discretion to make a class closure order which may operate to deny the benefits of a settlement to class members who do not opt out and who do not take the active step of registering: P Dawson No 2 at [17]. [74] Having said this, if a class closure order operates to facilitate the desirable end of settlement, it may be reasonably adapted to the purpose of seeking or obtaining justice in the proceeding and therefore appropriate under s 33ZF of the Act. The courts have accepted on numerous occasions that, in order to facilitate settlement, it is appropriate to make orders to require class members to come forward and register in order to indicate a willingness to participate in a future settlement, and to make orders that class members be bound into the settlement but barred from sharing in its proceeds unless they register: see for example, Matthews v SPI Electricity & SPI Electricity Pty Ltd v Utility Services Corporation Ltd (Ruling No 13) (2013) 39 VR 255; [2013] VSC 17 at [22]-[80] (Matthews v SPI No 13) (J Forrest J) and the authorities there referred to; Farey v National Australia Bank [2014] FCA 1242 at [11]-[16] (Jacobson J); Inabu Pty Ltd v Leighton Holdings Pty Ltd [2014] FCA 622 at [17]-[22] (Jacobson J); Newstart 123 Pty Ltd v Billabong International Ltd [2016] FCA 1194 at [67]-[68] (Beach J). An important aspect of the utility of a class proceeding is that they may achieve finality not only for class members but also for the respondent. [75] The rationale behind such class closure orders is that a requirement for class members to register their claims will facilitate settlement, because it allows both sides to have a better understanding of the total quantum of class members' claims, permits the settlement amount to be capped by reference to the number of class members, and assists in achieving finality (to the extent the Part IVA regime permits): see Grave D, Adams K and Betts J, Class Actions in Australia (2nd ed, Lawbook Co, 2012) at [14.410]. A class closure order that precludes class members, who neither opt out nor register, from sharing in a subsequent settlement may facilitate settlement, and therefore be reasonably adapted to the purpose of seeking or obtaining justice in the proceeding. 60 There, the Full Court recognised that in enacting Part IVA the legislature intended that class members be entitled to sit back and have little or no active involvement in the proceeding until after the initial trial on the common questions. Plainly, a requirement for group members to register, and for group members to be bound into a settlement but precluded from sharing the benefits of the settlement if they fail to register, cuts across that important aim. 61 As I said in Kuterba v Sirtex Medical Limited [2018] FCA 1467 at [12], class closure orders result in a reduced level of access to justice from that which Parliament intended by choice of an opt out regime. Empirical studies conducted in the United States demonstrate that implementation of an opt in requirement dramatically reduces the size of the class, as does the available empirical evidence in Australia: see Morabito V and Hatcher N, "Security for Costs in Unfunded Federal Class Actions: Back to the Future" (2017) 92 Australian Law Journal 105 at 114-115 and the references cited. 62 That does not mean that it is a "fundamental precept" of Part IVA that group members are entitled to do nothing prior to a settlement and still reap its benefits, nor that there is any absolute rule that group members can never be ordered to take a positive step (such as registration) in a proceeding: Parkin v Boral (Class Closure) [2022] FCAFC 47; 291 FCR 116 at [115]-[123] (Murphy and Lee JJ, with whom Beach J agreed), c.f. Pallas & Pallas (atf Pallas Family Superannuation Fund) v Lendlease Corp Ltd [2024] NSWCA 83. But it must be kept in mind that the effect of a soft class closure order is that, if group members do not register, they will be bound into any settlement that is able to be reached but precluded from sharing in the benefits of the settlement. The plain legislative intent of an opt out regime indicates that the Court should be cautious in making class closure orders and before doing so it must be satisfied that it is in the interests of group members as a whole. There must be a good reason to make such orders. 63 It is a common if not inevitable feature of class actions that the defendant will be faced with uncertainty regarding the quantum of potential class member claims: P Dawson Nominees Pty Ltd v Brookfield Multiplex Ltd (No 2) [2010] FCA 176 at [31] (Finkelstein J); Thomas v Powercor Australia Ltd (No 1) [2010] VSC 489 at [38] (J Forrest J); Winterford v Pfizer Australia Pty Ltd [2012] FCA 1199 at [7] (Bromberg J). Class closure orders and other orders such as discovery orders that require the applicant to provide the respondent with information about group members' claims are not usually made so that the respondents can achieve certainty in relation to the aggregate value of group members' claims. Rather, they are made to allow the respondent to obtain a better understanding of group members' claims including a general sense of aggregate claim value: e.g, Thomas at [54]-[58]. In this case I consider the respondents can obtain a sufficient understanding of group members' claims and their aggregate value from their own records. Indeed, the respondents' own records will be a better source of information in relation to group members claims than the information likely to be generated through the proposed registration regime. 64 In Treasury Wines at [79] the Full Court went on to say: Whether it is appropriate to order class closure is a question of balance and judicial intuition. The Court must take into account the interests of the class as a whole in requiring class members to take steps to facilitate settlement, and consider the surrounding circumstances including the point the case has reached, the attitude of the parties, and the complexity and likely duration of the case: see Matthews v SPI Electricity Pty Ltd (No 13) [2013] VSC 17; 39 VR 255 at [75]-[79] (J Forrest J). 65 Thus, the Full Court recognised that the attitude of the parties is a material consideration in deciding whether to make a class closure order. In the great majority of cases in which class closure orders have been made, the orders were made by consent of the parties, or unopposed. One can readily understand why, in circumstances where parties with divergent interests request that class closure orders be made, the courts have regularly concluded that such an order is likely to facilitate settlement in the proceeding. 66 But here the position is that the respondents propose registration and class closure orders and the applicants strenuously oppose them. The applicants have fiduciary obligations to act in the interest of group members. The respondents are in a quite different position. They are engaged in high-stakes litigation against the applicant and group members and their interests are inimical to the group members' interests. 67 Ordinarily, the applicant (or perhaps more accurately, the applicant's lawyers) will be in the best position to know what is in group members' interests, and the respondent will have its own interests rather than group members' interests at the forefront of its thinking. Generally speaking it is appropriate to accord the applicant's views as to what is in group members' interests substantially greater weight than the views expressed by the respondent in that regard. 68 I agree with Justice Wigney's remarks in Bradgate (Trustee) v Ashley Services Group Ltd [2017] FCA 1591 at [38] where his Honour said: The Court should generally exercise some caution before making a class closure order over the objection of the applicant… It is difficult to see how class closure orders could be seen to be conducive to settlement if they are vigorously opposed by the applicant. I would put it a little more strongly. It is appropriate to exercise real caution in relation to an application for class closure orders brought by the respondent and opposed by the applicant. 69 Even so, the position taken by the parties or their practitioners in relation to the question cannot be determinative. Whether a class closure order is appropriate or necessary in the circumstances of the case is a matter for the Court to decide: Fox v Westpac Banking Corporation; O'Brien v Australia and New Zealand Banking Group Limited; Nathan v Macquarie Leasing Pty Ltd [2023] VSC 414 at [28] (Nichols J). 70 Second, I have no difficulty accepting Ms Tatasciore's evidence that the group definition is broad; that the relevant data is held on four product administration systems, two of which have been archived; that many of the Impugned Products were offered via employer superannuation plans and the employer provided the data which has not been verified by the respondents, and it has historic deficiencies; that not all group members have provided TFNs; that the older records may not be up-to-date because members have not updated their contact details; and that where people are no longer members of the relevant fund their contact details may not be current. 71 But her evidence does not establish that the respondents' records are insufficient to enable the respondents to compile a representative sample of group members' claims, and to use that to obtain a sufficient understanding of group members' claims to participate efficiently and effectively in the mediation. There are a number of problems with that evidence. 72 The first problem is that it tells me nothing about any steps taken by the respondents: (a) to engage experts to bring the archived systems back into operation and to have the systems "talk" to each other to recover the necessary data for the purposes of the proceeding; (b) to take steps to verify the data provided by employers for the purposes of the proceeding; (c) to ascertain how many group members, or what percentage of them, have provided TFNs; or (d) to ascertain the extent of the deficiencies in the records they hold; ie, what number or what percentage of the records are defective, out of date or unreliable. 73 The proceeding has been on foot since May 2019 and one hopes that the respondents began taking such steps some time ago. And if the respondents have not taken such steps I would not make registration and class closure orders on the basis that without such orders the respondents have insufficient information about group members' claims in order to participate efficiently and effectively in the mediation. 74 The second problem is that it reveals little about the scale of the missing information. Amongst other things, Ms Tatasciore does not state: (a) that the data on the four product administration systems is inaccessible; (b) what approximate percentage of the members' personal information that was provided by employers is deficient; (c) what percentage of group members have provided TFNs; or (d) what percentage of group members who held accounts in the ERF Product (one of the 21 Impugned Products) are no longer members of the relevant superannuation funds and whose contact details may be out of date. Order 12 of the proposed orders provides that the respondents shall use their best endeavours to compile a spreadsheet from their business records "which contains (to the extent such information is available) the names and best available email addresses of all persons who held an ASL Impugned Product as at [date in each year] during the period 1 July 2008 to 15 May 2020 and all persons who held an NMS Impugned Product as at [date in each year] during the period 30 March 2011 to 15 May 2020". 75 Until that work is undertaken, the asserted deficiencies in the group member information, or at least the extent of them, is just speculation. 76 The third problem is that the evidence fell a long way short of showing that it is appropriate or necessary to make the proposed orders to allow the respondents to compile a representative sample of group members' claims and to reach a sufficient understanding of their claims so that the respondents can participate efficiently and effectively in the mediation. There are approximately two million group members, each of whom is or was during the Relevant Periods a member of a superannuation fund operated by the respondents. It beggars belief that the respondents do not have the names, addresses, contact information, and product information for the great majority of the members of those funds, even if that information is out of date or otherwise unreliable to some degree. Having regard to the terms of the respondents' proposed Order 12, I infer that they will be able to extract from their records the names, email addresses and product information in relation to a great many group members. 77 I can accept that, even after making reasonable efforts to bring the archived systems back into operation and have the systems talk to each other, taking steps to verify employer provided data, and ascertaining the percentage of group members whose records are out of date or otherwise unreliable, the data held by the respondents may be quite incomplete. But even if, say, a third or a half of the respondents' records are incomplete or somehow unreliable, the respondents will still have a vast amount of information available to them within their own records regarding the particular products and investment options taken up by account-holders in the Impugned Products during the Relevant Periods. 78 I consider the respondents' records are likely to be a much better source from which to compile a representative sample of the particular products and investment options typically taken up by group members when compared to a registration regime which is likely to have a very low take-up rate. In the course of the hearing I pressed senior counsel for the respondents on that issue. Senior counsel was unable to explain how orders for a registration regime which is likely to result in a very low level of registration will mean that the respondents obtain a better representative sample of group members' claims than, for example, utilising the respondents' records to obtain group member information in relation to say, 20,000 - 50,000 account-holders in the Impugned Products during the Relevant Periods. 79 To my mind the evidence indicates that the respondents' real concern is not that the information in their records is insufficient to provide them with a strong representative sample in relation to group members' claims. Rather, it is that the respondents wish to draw the sample from those group members who show an interest in participating in the proceeding, through registration under the proposed regime. 80 But, because there is likely to be a very low level of registration, any such sample is unlikely to be representative of group members' claims. It is relevant too that the relationship between the respondents and the group members is that of trustee and beneficiary, and that the proceeding seeks orders for the respondents to make good the trust fund. It follows that unless the applicants lawyers are prepared to recommend a settlement on the basis that only those beneficiaries who register should be allowed to recover their losses (which it appears that they are not), and unless the Court is prepared to approve a settlement under which all beneficiaries are bound by the settlement but many beneficiaries are excluded from sharing in it, any such sample seems likely have limited value. 81 Third, I do not accept Ms Tatasciore's evidence that the proposed orders are intended to enable the parties to obtain sufficient details of the claims of particular group members to assist in compiling a representative sample of group members' claims and to further the prospects of settlement, and are not intended to confine the number of group members who are able to participate in any settlement. I say that because: (a) it is not the case that "the parties" seek the orders. They are opposed by the applicants on grounds including that such orders will not facilitate settlement; (b) the proposed orders include a class closure order. It would have been straightforward for the application to seek orders to notify group members they should register their claim without an associated class closure order, and they did not do so. Having said that, I note that during the course of oral submissions the respondents advanced a fallback proposal. They proposed that an opt out notice to be sent to group members, accompanied by a request for registration, without a class closure order. That was not, however, the respondents' primary position; (c) the objective of enabling the respondents to obtain sufficient details of group members' claims does not require a registration regime. The evidence indicates that the respondents have sufficient information in their own records to compile a representative sample of group members claims and their own records are likely to more efficiently and effectively provide information about such claims; and (d) Ms Tatasciore stated that "...without the information... provided through the proposed registration process: (a) it will not be possible for the parties to know the number of group members who might later make a claim against any settlement fund…". That is, the registration regime is intended to confine the class. 82 In my view the evidence shows that one of the purposes of the application is to confine the class so that only those group members who register are permitted to benefit from any settlement. At least in part, the registration process is a cloak to disguise the fact the respondents wish to so confine the class. 83 Fourth, the relationship between the respondents and the group members is that of trustee and beneficiary. The relief sought in the proceeding includes orders for the respondents to make good the trust funds reduced by their alleged breach of statutory and general law duties, so as to compensate group members. The fact that the respondents have fiduciary obligations to the group members points away from making a class closure order which will have the effect that some beneficiaries will not get the benefit of any settlement unless they take the positive step of registration, particularly when I accept that, typically, group members are disengaged from their superannuation, and many are unlikely to register. 84 Fifth, the evidence shows that this is not a case where registration must occur at some stage. In my view the great majority of the class will not need to register in order to be identified and paid their share of any settlement that may be reached or following judgment. The group members are almost all either current members of the superannuation funds administered by AMP, or past members to whom funds can be disbursed using the ATO TVP procedure. Ms Tatasciore is correct in identifying that a small part of the class comprises persons who were not superannuation fund members (those identified in paragraphs [3.3] to [3.6] of the FACSOC), but the number of those persons is likely to be insignificant in the context of the class as a whole. To the extent that those persons may need to register to participate in any settlement, the appropriate time for that to occur is after any settlement is reached. That is not a matter which could justify the making of orders requiring the class generally to register, at huge cost, in order to participate in any settlement at mediation. 85 Sixth, the overarching purpose in s 37M of the FCA Act requires that the orders facilitate the just resolution of disputes as quickly, inexpensively and efficiently as possible. 86 Here, Ms Gilsenan's evidence shows that if notice of the requirement to register is provided to 50% of group members by email and to the remaining 50% by ordinary post, the external costs of distributing notices to group members will be approximately $1.64 million, even without other channels of notification or reminder notices. To that must be added the substantial costs associated with setting up a registration platform, responding to group member enquiries by telephone and email, handling registration processes and managing the data of registered group members which, as Ms Gilsenan said, could easily exceed $750,000. 87 The respondents put on no evidence to contradict those estimates, but in submissions they suggested that the likely costs would more likely be around $1 million. As I have said it is unnecessary to reach a final view, and I proceed on the basis that the registration regime will require the applicants to spend in the order of $2 million. When the significant cost of the proposed registration regime is considered together with the fact that it is likely to result in a low level of registration, that is a strong reason to refuse the application. 88 It is also worth noting that the respondents balked at the idea of themselves paying for the proposed registration regime, notwithstanding that they were the only parties that wanted it. 89 Seventh, the respondents seek to rely upon three recent decisions in which soft class closure orders have been made in circumstances where the orders were opposed by the applicants: see Fox; J Wisbey & Associates Pty Ltd v UBS AG [2024] FCA 147 and Anderson-Vaughan. Each case must, however, turn on its own facts. The facts in those cases are different. 90 In relation to Fox and Anderson-Vaughan it should be kept in mind that not only were those decisions based in a different factual context, they were grounded in the express power in s 33ZG in Part 4A of the Supreme Court Act 1986 (Vic) which is not found in Part IVA. In Matthews v SPI Electricity Pty Ltd (No 13) [2013] VSC 17; 39 VR 255, J Forrest J said (at [15]) that it was of particular note that s 33ZG specifically enables a court to impose a requirement for group members to take a step prior to settlement or judgment on liability. More recently, Nichols J recognised in Fox (at [24]) that the emphasis given by the Full Court in Treasury Wine to Parliament's intention that group members not be required to take a positive step in the proceeding must be read in light of the fact that, unlike the Victorian legislative regime, Part IVA of the FCA Act does not contain the express power found in s 33ZG. 91 In Wisbey at [69]-[70] Beach J explained: The issue for the Court when asked to make registration orders of this kind is thus not whether it has the power to make them, but rather whether it is an appropriate exercise of its power under s 33X. The relevant considerations in this context can include whether it is in the interests of group members as a whole to require registration before any prospective settlement is on the table, the point which the proceeding has reached, the attitude of the parties, the complexity and likely duration of the case, with protracted litigation and greater complexity increasing the interests of group members in avoiding litigation risk through achieving a settlement, whether group members have adequate notice of the change and reasonable time to decide whether to register, and whether an estimate of the size and number of claims can be made. Whilst the significance of these considerations will vary with the circumstances of the case, a paramount factor is the extent to which a registration process is likely to improve the prospects of achieving a reasonable settlement. 92 I accept that the extent to which a registration process is likely to improve the prospects of achieving a reasonable settlement is an important consideration, although I would not describe it as "a paramount factor". But in the circumstances of this case nothing turns on that mild disagreement. 93 Here, I am not satisfied on the evidence that the proposed registration and class closure orders are likely to improve the prospects of achieving a reasonable settlement. The respondents did not establish that their records do not contain sufficient information in relation to group members' claims for them to compile a representative sample of their claims and participate efficiently and effectively in the mediation. In my view it is likely that the respondents can more efficiently and effectively compile a representative sample from their own records rather than through the proposed registration regime. 94 I see it as quite unlikely that the experienced solicitors and counsel acting for the applicants would be so bull-headed as to oppose the proposed registration regime if they thought it would increase the prospects of a reasonable settlement. It seems more likely that they do so because, in their view; (a) the rate of registration is likely to be very low and unlikely to provide a representative sample of group members' claims; (b) notwithstanding its likely ineffectiveness the proposed registration regime will cost the applicants in the order of $2 million and divert the applicant's legal team from preparing for the mediation and trial; (c) the information about group members' claims is already available to the respondents from their own records; and (d) if the proceeding is successful at trial the respondents are likely to be ordered to make good the trust fund, and there will be no requirement for the great majority of group members to register in order to recover their losses. Those features of the case mean I cannot accept either the respondents' primary proposal for registration and class closure orders, nor their fallback position seeking a registration process without class closure.

CONSIDERATION REGARDING OPT OUT 95 The dispute in relation to the proposed opt out regime is one of narrow compass. 96 The respondents submitted that it is appropriate for an opt out notice to be distributed to group members after pleadings have closed and before the mediation takes place. They said that it would be particularly efficient from a time and cost perspective to conduct the opt out and registration processes together, if the Court is minded to order a registration process. In their view, in the absence of a registration regime, there is no realistic prospect of a settlement at mediation, and not ordering opt out now would merely defer the distribution of opt out notices until shortly before trial. 97 The applicants submitted that an opt out process in a large class action such as this can be very expensive and labour-intensive, and they said it is in the interests of cost and efficiency that the Court minimises the number of times that notice is given to the class. They accepted that opt out will have to be conducted before trial, but said that while there is a possibility of settlement at mediation the most efficient course is for opt out to occur after that. They said that doing so will accommodate the possibility that opt out and notice of a proposed settlement could be part of a single notice event. Otherwise there is a risk that an opt out notice will be sent to group members prior to mediation, and in the event that settlement is achieved at mediation, a further notice will be required to be sent to group members advising them of the in-principle settlement, which will result in duplication of costs. 98 In my view the opt out process will be less expensive than the proposed registration regime. Amongst other things, it may not be necessary to use multi-channel notification or send reminder notices. Even so, having regard to my reasons above, an opt out process involving notification to a class of two million group members will involve substantial cost and notwithstanding the expense it is likely that there will be a very low rate of opt out. 99 I am not persuaded that it is appropriate to require the applicants to spend substantial money now on providing an opt out notice to group members when there is a chance that the case will resolve at mediation. If the proceeding does settle at mediation there will be a substantial saving by combining the opt out notice with the notice of proposed settlement. And if it does not settle at mediation there will be ample time to send an opt out notice to group members well prior to trial. For example, the opt out notice could be sent out in, say, December 2024, with group members being required to opt out by a deadline of mid-February 2025.

CONCLUSION 100 Accordingly I made orders to dismiss the application and order the first and second respondents to pay the applicants' costs of and incidental to the application. The other respondents did not seek a costs order.

OTHER MATTERS 101 I listed a case management hearing for 9 May 2024 for the purposes of: (a) deciding whether it is appropriate to direct that Maurice Blackburn and Slater & Gordon provide to the respondents, for the purposes of mediation, the information provided to them by the 13,691 group members that have registered with them. My present view, subject to hearing further from the parties, is that it is appropriate to do so on a confidential basis for the purposes of mediation, and subject to an undertaking that the respondents will not contact those group members for the purposes of the proceeding. It may also be appropriate to direct Maurice Blackburn and Slater & Gordon to make attempts to gather further information from those group members for provision to the respondents for the purposes of mediation; and (b) deciding whether it is appropriate to direct the respondents to provide Maurice Blackburn and Slater & Gordon with information relevant to the group member losses for the purposes of the mediation. This issue has been a matter of contention between the parties for a long time, and it may be appropriate to refer the issue to a referee or for the Court to appoint an expert to prepare a report for the purposes of mediation only. I certify that the preceding one hundred and one (101) numbered paragraphs are a true copy of the Reasons for Judgment of the Honourable Justice Murphy.

SCHEDULE OF PARTIES VID 572 of 2019 Applicants Fourth Applicant: JODIE MITCHELL Respondents Fourth Respondent: NMMT LIMITED (ACN 058 835 573) Fifth Respondent: AMP SERVICES LIMITED (ACN 081 143 786) Sixth Respondent: THE NATIONAL MUTUAL LIFE ASSOCIATION OF AUSTRALASIA LIMITED (ACN 004 020 437)