Perera v GetSwift Limited

1. The applicant have leave to appeal: (a) the case continuation orders made on 23 May 2018, as identified in the applicant's amended draft notice of appeal; and (b) the solicitor restraint orders, being orders 1 to 7 made on 25 June 2018, as identified in the applicant's amended draft notice of appeal.
2. Otherwise, the applicant's application for leave to appeal concerning the common fund orders made on 20 June 2018 as identified in the applicant's amended draft notice of appeal be dismissed.
3. The appeal concerning the case continuation orders be dismissed.
4. The appeal concerning the solicitor restraint orders be allowed in part.
5. The solicitor restraint orders made by Lee J on 25 June 2018 be set aside.
6. Within 14 days hereof, the parties file and serve short submissions (limited to 3 pages) on the question of substitute solicitor restraint orders and on the question of costs. Note: Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

BETWEEN: SHAUN MCTAGGART First Applicant SAMANTHA MCTAGGART Second Applicant

AND: GETSWIFT LIMITED (ACN 604 611 556) First Respondent

JOEL MACDONALD Second Respondent BANE HUNTER Third Respondent

JUDGES: MIDDLETON, MURPHY AND BEACH JJ DATE OF ORDER: 20 NOVEMBER 2018

1. The applicants have leave to appeal.
2. The appeal be dismissed.
3. Within 14 days hereof, the parties file and serve short submissions (limited to 3 pages) on the question of costs. Note: Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

THE COURT: 1 In this matter three applicants represented by three different law firms and supported by three different litigation funders commenced three competing 'open' shareholder class actions, within a short time of each other, against the respondent, GetSwift Limited (GetSwift) (and in two of the actions also against officers of GetSwift), doing so in respect of substantially the same claims and on behalf of the same group members. The class actions in order of commencement are conveniently referred to as the Perera Proceeding, the McTaggart Proceeding and the Webb Proceeding. The primary judge ordered that the Perera Proceeding and the McTaggart Proceeding be permanently stayed and the Webb Proceeding continue as an open class action, and made other orders intended to facilitate that outcome. The applicants in the Perera Proceeding and the McTaggart Proceeding have applied for leave to appeal from those orders. 2 Before the primary judge GetSwift argued that the Court should stay two of the open class actions as an abuse of process. In broad summary it contended that: (a) it should not be vexed by multiple proceedings in respect of the same issues; (b) a multiplicity of proceedings was not in the interests of group members when one open class action was capable of vindicating all of their rights and interests; and (c) only one of the open class actions should be advanced, with the other two class actions being permanently stayed, although it was agnostic as to which class action should continue as the vehicle by which group members' claims are pursued. 3 The preferred position of each of the applicants was that only one open class action should proceed, but they made alternative proposals as to how best to achieve that objective. In this respect: (a) each applicant contended that their class action should be the only open class action, and their submissions assumed that there should be no multiple open class actions on a long term basis; (b) the applicants made different contentions as to how to deal with the other two class actions. Mr Perera contended that the McTaggart Proceeding should be temporarily stayed or adjourned, and after a prompt opt out process be either declassed or closed, and that the Webb Proceeding should be declassed. The McTaggart applicants contended that the Perera Proceeding should be closed and the Webb Proceeding should be declassed. Mr Webb contended that both the Perera Proceeding and the McTaggart Proceeding should be stayed, declassed or adjourned; and (c) none of the applicants contended that their class action should continue in any lesser fashion other than on an open class basis, for example, on a closed class basis. Accordingly, as the primary judge recognised, the situation could be distinguished from the competing open class actions scenario discussed by Beach J in McKay Super Solutions Pty Ltd (Trustee) v Bellamy's Australia Ltd [2017] FCA 947 (Bellamy's), where each competing applicant was, as a fall-back position, quite content to proceed as a closed class. 4 The applicants also took different positions regarding the Court's power to permanently stay two of the three competing class actions. Mr Perera submitted that it was doubtful that there was power absent specific legislative warrant but conceded that much the same result could be achieved by using other tools. The McTaggart applicants submitted that it was doubtful there was power to do so under s 33ZF of the Federal Court of Australia Act 1976 (Cth) (the Act). Contrastingly, Mr Webb did not contend that the Court lacked power. 5 In summary, the primary judge decided: (a) first, that only one proceeding should continue as an open class action on a long-term basis; (b) second, that the Webb Proceeding should be the open class action in preference to the Perera Proceeding and the McTaggart Proceeding; (c) third, in respect of the Perera Proceeding and the McTaggart Proceeding that there was power in the circumstances to permanently stay those proceedings as an abuse of process; and (d) fourth, that in all of the circumstances it was appropriate to exercise the power of a permanent stay over any of the other alternatives propounded by the parties, whether a temporary stay pending later declassing or class closure, immediate declassing or immediate class closure. 6 The primary judge reached his findings (at [305] to [324]), his assessment of what was to be done (at [325] to [330]), his preference for a permanent stay (at [331] to [349]) and his consideration of alternative relief (at [350] to [355]) by deciding not just that there should not be more than one class action proceeding on a long term basis, but more specifically that only one open class action proceeding for all claims was the preferable case management response. 7 The Perera and McTaggart applicants seek leave to appeal from his Honour's decision to order a permanent stay of their proceedings. Mr Webb and GetSwift submit that we should refuse such leave to appeal. 8 In broad summary Mr Webb and GetSwift submit that many of the proposed grounds of appeal are no more than an attempt to have a re-run of the discretionary exercise conducted below, and do not establish any House v The King (1936) 55 CLR 499 error. They argue that it is well-established that the Full Court must exercise rigour before deciding to grant leave to appeal from procedural or case management decisions which involve the exercise of discretion. They point out that with the advent of open class actions and common fund orders, it will often be the case that there will be competition between solicitors and funders for who gets to run the action which goes forward, and that the first instance judge in such a competitive scenario will not have an easy decision. The judge will have to synthesise and balance many incommensurable and conflicting factors, there may be no one right answer, and he or she will have to craft the appropriate order in circumstances where to allow one action to go forward will disappoint the other solicitors and funders. 9 Mr Webb and GetSwift further submit that whilst the Perera and McTaggart applicants make vague assertions of substantial injustice if we were to refuse leave, they failed to precisely articulate what such injustice would be or who would suffer it. They contend that the individual applicants would not suffer any relevant injustice as they could opt out of the class action going forward, namely the Webb Proceeding, and could commence their own proceeding with their own choice of lawyer. Nor would the group members in the Perera and McTaggart Proceedings suffer any injustice as they could have their rights vindicated in the Webb Proceeding without taking any further step, or if they wished they could opt out of that proceeding and commence their own case with their own choice of lawyer. They note that there was no evidence from either the Perera or McTaggart applicants, or from any group member, substantiating that their position would be compromised by being left to vindicate their claims through the Webb Proceeding. 10 The submissions by Mr Webb and GetSwift have some force in terms of justifying a refusal of leave, but the applications raise important matters of principle in relation to how the Court should deal with competing class actions. A Full Court has not yet ruled upon the matters of principle and the factors to consider in dealing with and resolving problems arising from competing class actions. It is important that some general guidance be given as to the range of factors to consider and the possible options open to a docket judge in dealing with competing open class actions. 11 Moreover, we cannot say that the primary judge's decision is not attended with sufficient doubt either as to the source of the relevant statutory or other power that his Honour used to justify a permanent stay of the two proceedings or as to its manner of exercise. Further, although the decision involved a matter of practice and procedure and an exercise of discretion, the question and choice of which competing open class action was to go forward arguably has some consequences for the applicants and group members in the cases that are stayed. 12 In all the circumstances it is appropriate that we grant leave to appeal from the decision to grant a permanent stay of the Perera Proceeding and the McTaggart Proceeding, whether or not substantial prejudice would have been caused to any applicant or group member if we had chosen to refuse leave. However, for the reasons we set out we dismiss that appeal. 13 There are two further issues that require mention at this point. 14 First, in the Webb Proceeding the primary judge granted an injunction against the solicitors for Mr Perera and 103 group members who had retained the firm, restraining the solicitors from communicating with their clients who were also group members in the ongoing Webb Proceeding. Mr Perera seeks leave to appeal against that injunction. In our respectful view the primary judge ought not to have imposed a restraint to prevent solicitors from communicating with their clients. We grant leave to appeal against that order and allow that appeal so as to remove the restraint. 15 Second, Mr Perera also seeks leave to appeal against the common fund order made in the Webb Proceeding. We refuse such leave, given that we have left in place his Honour's orders permanently staying the Perera Proceeding and the McTaggart Proceeding and for other reasons we later explain.

BACKGROUND AND PROCEDURAL HISTORY 16 We have drawn the following description from the primary judge's reasons. 17 GetSwift is a technology company founded in 2015, listed in 2016, and which relevantly provides software for businesses to manage what are described as 'last-mile' delivery functions. After listing, GetSwift, at various times, made announcements to the ASX about agreements and 'partnerships' signed with clients. On 11 December 2017, the company announced a $75 million capital raising, at $4.00 per share. On 19 January 2018, the Australian Financial Review reported that GetSwift had allegedly failed to inform the market that its agreements with some customers had been terminated, and further alleged that it had announced the revenue forecasts tied to an agreement with the Commonwealth Bank of Australia prematurely. Shortly thereafter, on 22 January 2018, GetSwift shares were placed in a trading halt, and two days later were suspended from official quotation, pending the company's response to questions from the ASX. A number of announcements relevant to the company's suspension from official quotation (including questions raised by the ASX) were then apparently released up to reinstatement to official quotation on 19 February 2018. 18 Before entering the trading halt on 22 January 2018, GetSwift's share price was $2.92, however, following reinstatement, it declined to $0.51 as at the close of trade on 21 February 2018. 19 Ms Banton, a partner of the law firm Squire Patton Boggs (SPB) gave evidence that SPB began investigating possible claims against GetSwift on 19 January 2018 and published a 'Notice of Investigation' into a potential class action on 2 February 2018 on the SPB website. Its purpose was to seek to bring the potential class action to the attention of shareholders who acquired shares in GetSwift in the period 24 February 2017 to 19 January 2018. A dedicated 'GetSwift Class Action' webpage and email address were established so that interested investors could register their interest online and obtain more information. Eventually, 103 shareholders chose to retain SPB (Perera Funded GMs) and also to enter into a litigation funding agreement with a litigation funder, International Litigation Partners No 18 Pte Ltd (ILP18). 20 SPB filed the Perera Proceeding (Dwayne Perera v GetSwift and Anor NSD 226 of 2018) on 20 February 2018 against GetSwift and Mr Joel Macdonald, the CEO of GetSwift. The proceeding was listed before the primary judge for a first case management hearing on 23 March 2018, which was then adjourned to 29 March 2018. 21 Mr Pagent, a partner of the law firm Corrs Chambers Westgarth (Corrs) gave evidence that on 20 February 2018, a litigation funder, Vannin Capital Operations Limited (Vannin) announced it was investigating a potential class action against GetSwift and its executive directors and invited shareholders who purchased shares in GetSwift between 24 February 2017 and 19 January 2018 to register their interest in the proposed class action. 441 investors registered their interest in participating in the class action via the website www.getswiftclassaction.com.au, and 208 investors signed litigation funding agreements with Vannin (McTaggart Funded GMs). 22 Corrs filed the McTaggart Proceeding (Shaun and Samantha McTaggart v GetSwift and Others NSD 440 of 2018) on 26 March 2018 against GetSwift, Mr Macdonald and Mr Bane Hunter, the Executive Chairman of GetSwift. This proceeding was also listed for a first case management hearing on 29 March 2018. 23 On 19 February 2018 two further potential class actions were publicly announced by two other law firms, Gadens and MC Lawyers & Advisers. Nothing further has been heard by any party as to those two proposed class actions. 24 At the case management hearing on 29 March 2018, the primary judge indicated that he was disposed to make orders facilitating each of the applicants putting before the Court their proposals for how the issue of the two competing class actions should be dealt with. There was no opposition to this course by either of the applicants or GetSwift, and orders were made in both proceedings for the parties, by 4 pm on 9 April 2018, to exchange and file affidavits and submissions directed to: (a) the manner and form by which security for costs was to be provided in the respective proceedings; (b) the details, including any proposed percentage subject to further order, of the terms of any common fund order to be sought in the proceedings; (c) whether, subject to the filing of a defence, any issues in the proceedings were suitable for reference to a special referee or whether a court appointed expert or joint expert be appointed or retained in relation to any issue; (d) an estimate by the applicants' solicitors, on affidavit, of the costs that were likely to be incurred on the assumption that issues of contravening conduct and causally related loss were in dispute; (e) the number of funded group members and the aggregate number of shares held by all funded group members; (f) what, if anything, was proposed to deal with potential overlap between the Perera Proceeding and the McTaggart Proceeding, such as a consolidation of the two proceedings, a permanent stay of one of the proceedings, an order declassing one of the proceedings under either s 33N(1) or s 33ZF of the Act, an order closing the class in one (but not the other) proceeding, or orders allowing a joint trial of both proceedings with each left constituted as open class proceedings; and (g) in the event a temporary or permanent stay or declassing of a proceeding was proposed, the matters relied upon in support of that contention. 25 The primary judge then listed the two proceedings for a further case management hearing on 13 April 2018, with a view to hearing and deciding any applications relating to the matters the subject of submissions filed pursuant to the earlier orders. The primary judge said that the purpose of the process was for each of the applicants to put their 'best foot forward' to come up with a detailed proposal to deal with the competing class actions and, if only one proceeding was to go forward, to explain why it was that their proceeding should be preferred. To preserve the integrity of the process as to the competing proposals, the primary judge further noted in the orders that: there should be no discussion between the applicants, the applicants' lawyers or the funder in this proceeding and the applicant, the applicant's lawyers or the funder in the Related Proceeding as to the content of the material to be filed and served pursuant to [the orders] until after the exchange provided for by that order. 26 Shortly after the exchange of affidavit evidence and submissions between Mr Perera and the McTaggart applicants, another law firm Phi Finney McDonald (PFM) filed an interlocutory application on behalf of Mr Raffaele Webb seeking orders that he be granted leave to intervene in both proceedings, a regime for a notice to group members, and for the case management hearing to be adjourned until late May or June 2018. The application was premised on the basis that Mr Webb intended shortly to file a class action by Mr Webb against GetSwift. 27 To explain any suggested delay in making the application, Mr Ben Phi, a partner of PFM, deposed that prior to 20 February 2018 the firm took preliminary steps to assess the merits of a potential claim against GetSwift, and Mr Phi had preliminary discussions with representatives of a litigation funder, Therium Capital Management Limited (Therium). Mr Phi said that he decided against announcing a potential class action until after GetSwift resumed trading as, in his experience, announcing a class action prior to the resumption of trading (or even shortly after an alleged corrective disclosure) may allow a respondent to argue that any share price fall was caused, at least in part, by the threat of a class action. He also so deferred because his view was that it was appropriate to review analyst reports and media commentary to form a view as to the precise cause of a particular share price reaction, and, further, his preference was not to announce a class action until after he was satisfied that appropriate funding could be secured on terms that he considered were favourable to group members. Despite this initial preference, following the filing of the Perera Proceeding on 20 February 2018, he caused an announcement of PFM's investigation to be published on PFM's website and the potential PFM class action was reported by the media the following day. 28 His Honour immediately listed a case management hearing on 11 April 2018. Following exchanges between his Honour and the parties, Mr Webb did not press for adjournment of the case management hearing. His Honour made orders that Mr Webb have leave nunc pro tunc to intervene in the Perera Proceeding and the McTaggart Proceeding for the limited purpose of appearing at the case management hearing on 11 April 2018, and for the purpose of making any application he wished to make on 13 April 2018. 29 His Honour ordered Mr Webb to exchange with the other applicants any affidavit evidence, submissions, and any other material related to the same issues which were the subject of the earlier exchange between Mr Perera and the McTaggart applicants. His Honour did not allow Mr Webb or his legal advisers to have access to any material already filed, and again noted in the orders that there should be no discussions between the applicants, the applicants' lawyers or the funders. His Honour said that, on 13 April 2018, he would grant leave to Mr Webb to file and have returnable instanter, an originating application commencing a class action with Mr Webb as the representative applicant. 30 We note at this point that the Perera and McTaggart applicants submitted to the primary judge that Mr Webb had obtained some advantage by reason of what occurred. Before us they argue that when his Honour came to weigh up the respective advantages and disadvantages of the competing proposals, he should have taken into account "the fact that the Webb applicant enjoyed a process advantage compared with the McTaggart (and Perera) applicants". They say that arises because, after setting up what the parties said was a "competitive tender" process, Mr Webb's late entry put him at a "clear advantage" apparently because at the time they submitted their respective proposals, the McTaggart applicants (and Corrs and Vannin) and Mr Perera (and SPB and ILP18) "thought they were in a two-horse race, not a three-horse race, and they bid accordingly. But Webb knew it was a three-horse race, and he was able to bid accordingly". They also say that some unfairness was occasioned because, whereas the Perera and the McTaggart applicants were required to submit their evidence and proposals by 4.30 pm on 9 April 2018, Mr Webb was given until 4.30 pm on 12 April 2018 which meant that he was able to incorporate into his proposal developments that occurred between 9 and 12 April 2018. 31 Mr Perera submits that allowing persons such as Mr Webb (and PFM and Therium) to come along "at the heel of the hunt" would "encourage a phenomenon of entrepreneurial lawyers and funders parasitically lying in wait to steal the work product of those who have conscientiously been investigating claims on behalf of real people who have retained them". 32 The primary judge, in our view correctly, rejected these submissions for seven reasons: (a) first, Mr Webb did not have access to any of the material filed by the other applicants. He (and PFM and Therium) like Mr Perera (and SPB and ILP18) and the McTaggart applicants (and Corrs and Vannin), all knew that their responses were to be prepared without contacting or having access to the materials of the other applicants; (b) second, the McTaggart applicants' submissions were silent as to the details of the 'developments' which were asserted to have presented Mr Webb with a competitive advantage between the first exchange (between Mr Perera and the McTaggart applicants on 9 April) and the second exchange (between Mr Perera and the McTaggart applicants on the one hand, and Mr Webb on the other, on 12 April). His Honour said that given Mr Webb did not have access to any of the details of the earlier proposals put forward, no relevant distortion occurred. We agree; (c) third, no complaint was made by the legal representatives of Mr Perera and the McTaggart applicants at the case management hearing on 11 April 2018 as to the orders made on that date in regard to Mr Webb; (d) fourth, his Honour said that what was put in place was not some process of bargaining between the Court (in its protective role as to group members) and scheme promoters and that the analogy to an auction was inapposite. We would also interpolate at this point that US-style auctions are not appropriately adapted to the Australian conditions of Part IVA as Beach J explained in Bellamy's at [23]. His Honour said that each applicant was to put its best foot forward and they did so, and the evidence that ILP18 would have approached the process differently in the event a third funder was to be involved (and not just a competition with Vannin) was not to the point. What the orders sought to elicit, inter alia, was a considered proposal which best served the twin ends of furthering the overarching purpose and protecting the interests of group members, not the optimal commercial 'deal' for the funder pitched at the level to beat (but only just) the commercial 'deal' likely proposed by another funder; (e) fifth, his Honour rejected the notion that Mr Webb had acted in a way which could be accurately described as "parasitically lying in wait to steal the work product" of others. Mr Webb had not filed a class action by the time of the first case management hearing but Mr Phi provided an explanation of why his attention was directed to attempting to negotiate funding terms, which only occurred on 4 April 2018, and as to why he deferred the commencement of proceedings. Mr Phi also said that he did not anticipate orders such as those made at the first case management hearing; (f) sixth, his Honour accepted Mr Webb's submission that the proposal he put forward reflected, at least as to funding rates and budgeted legal costs, those funding terms agreed with Therium and the budget proposed by PFM on 4 April 2018, which was before the orders made on 11 April 2018; and (g) seventh, each of the Perera and McTaggart applicants submitted that their respective proposals were superior and they provided no detail as to how the supposedly superior and well thought through proposals and submissions would have changed. 33 Let us return to the chronology. When all three proceedings came before his Honour on 13 April 2018, detailed evidence was adduced and submissions were made as to the approach to be taken to resolve the issue of the competing class actions, the powers of the Court in this regard and the comparative merits of the competing proposals. Further argument spanned three further hearings, with additional submissions and materials being filed. We would not condone such a complex, elaborate and expensive exercise in other cases when the issue of competing class actions needs to be dealt with. But we mean no criticism of his Honour in the present context given the novelty of some of the legal and forensic issues raised. But, in future cases these questions require to be dealt with less elaborately and more efficiently. 34 The parties put forward various proposals. Their respective proceedings and proposals may be summarised as follows. 35 The Perera Proceeding had the following components: (a) Mr Perera was the representative party having acquired seven parcels of shares in November and December 2017 (each of which were sold at a profit) and he bought his first loss-making shares on 5 January 2018; (b) the proceeding was commenced as an 'open' class action and the class included 103 Perera Funded GMs who acquired 2,575,804 shares in the relevant period; (c) the proceeding was funded by ILP18, a special purpose funding entity. Corporations related to ILP18 had moneys on trust with SPB; (d) the Perera Funded GMs had entered into funding agreements with ILP18 and retainer agreements with SPB; (e) the funding agreement provided for a commission ranging between 25% and 40% (discriminating between group members on the basis of the amount of shares held) but Mr Perera proposed a common fund order which involved payment of the lesser of 25% of net proceeds or 22.5% of gross proceeds, which sum would be capped to ensure that it was an amount not greater than 25% of net proceeds; (f) ILP18 was prepared to abide by any order as to security for costs be it by way of cash deposit, bond or insurance cover; and (g) no suggestion was made that any substantive question could be referred to a special referee or be the subject of evidence from a court appointed expert. 36 Mr Perera submitted that the Perera Proceeding should proceed as an open class action, and the Court should temporarily stay, or adjourn, the other proceedings until shortly after the date for opt out had passed in the Perera Proceeding. Subject to consideration of the number of persons who opted out Mr Perera proposed that the other proceedings then be either closed or declassed. At least when the McTaggart Proceeding was the only competitor, Mr Perera contended that the discriminating feature for deciding whether to close or declass that proceeding (under s 33N of the Act) was whether the McTaggart Proceeding had "insufficient registered group members to justify it continuing as a representative proceeding". Mr Perera also submitted that the Webb Proceeding should be declassed. 37 The McTaggart Proceeding had the following components: (a) the McTaggart applicants were the representative parties having bought 151,208 shares starting in September 2017 and, after selling some at a profit, selling all the remaining shares they then held on 19 February 2018, the day trading resumed, which was alleged to have crystallised a loss of $182,223; (b) the proceeding was commenced as an 'open' class action, and the class included 208 McTaggart Funded GMs who acquired 1,545,374 shares in the relevant period; (c) the proceeding was funded by Vannin which was wholly owned by a substantial corporation known as Vannin Capital PCC, which would meet Vannin's request for funds to finance its litigation funding operation; (d) the McTaggart Funded GMs had entered into funding agreements with Vannin but had not entered into retainer agreements with Corrs; (e) the funding agreement provided that the funding commission would be 10% if proceeds were received before the end of 2018, 20% if received before 26 September 2019 and 30% if received thereafter. At the time they filed their initial submissions the McTaggart applicants did not indicate that they were prepared to make an application for a common fund, preferring to abide further developments. But during the course of oral submissions before the primary judge, they said they would be content with a common fund order which provided for essentially the same terms as in the funding agreement, being 10% of gross proceeds before 31 December 2018, 20% of gross proceeds until a date 42 days prior to the initial trial, and 30% of gross proceeds thereafter; (f) a sum of $250,000 was held in Corrs' trust account and the funding agreement required Vannin to obtain "after the event" (ATE) insurance of an amount in excess of that amount that was likely to be awarded for security for costs. Alternatively Vannin was prepared to provide security for costs by way of a cash deposit; and (g) no suggestion was made that any substantive question could be referred to a special referee or be the subject of evidence from a court appointed expert. 38 The McTaggart applicants submitted that their proceeding should remain an open class proceeding and the Perera Proceeding should be closed under s 33ZF(1) of the Act. They submitted that Mr Webb should be notified that the McTaggart Proceeding would be an open class proceeding, and the Perera Proceeding would be a closed class proceeding. Following this notification, if Mr Webb and Therium wished to continue with the Webb Proceeding, that proceeding should proceed as a single claim heard together with the other proceedings. 39 The Webb Proceeding had the following components: (a) Mr Webb was the representative party having purchased shares in GetSwift during the relevant period (there was no further evidence as to his purchases); (b) the proceeding was commenced as an 'open' class action. No group members had signed a funding agreement and accordingly there were no funded group members; (c) Mr Webb produced a letter from Therium dated 12 April 2018, in which the 'Investment Committee' of another Therium entity represented that it had committed a sum for the purposes of Therium funding the costs incurred by PFM and indemnifying Mr Webb in relation to adverse costs; (d) the only retainer agreement entered into was between Mr Webb and PFM. No funding agreement which provided a 'deadlock' provision to resolve issues which might arise in the settlement context between Mr Webb and Therium existed; (e) Mr Webb proposed a common fund order which would entitle Therium to a funding commission that was the lesser of: (i) 2.2 times the expenses that Therium had paid in the proceedings if the parties entered into a settlement agreement on or before 12 April 2019, or 2.8 times if there was a successful resolution after 12 April 2019; or (ii) 20% of the net litigation proceeds (that is, the settlement sum less approved professional fees and disbursements); (f) initially it was proposed that security for costs be provided by way of an ATE policy together with a deed of indemnity, but during the hearing it was submitted that Therium was prepared to put up cash security should the Court so order; and (g) Mr Webb asked the Court to appoint a special referee to conduct periodic reviews of the reasonableness of Mr Webb's legal costs, with only those costs assessed as reasonable being costs included for part-payment of Mr Webb's legal costs. Additionally, Mr Webb accepted that there was scope for the Court to appoint, or for the parties to jointly engage, a forensic economist to assist the Court in respect of matters of loss causation and the quantification of loss and damage. 40 Mr Webb agreed with the other applicants that only one proceeding should proceed as an open class. He contended that orders should be made allowing the Webb Proceeding to proceed on behalf of an open class and variously submitted that directions should be made to either stay, declass or adjourn the other proceedings. 41 GetSwift sought a stay of all proceedings but one, but was agnostic as to which proceeding should go forward. It submitted that the Court had an inherent jurisdiction to stay its proceedings as an abuse of process or if the proceedings were unjustifiably oppressive or otherwise brought the administration of justice into disrepute. It argued that the categories of abuse of process and vexation or oppression were not closed and were insusceptible of precise formulation, and in the context of Part IVA proceedings that meant that the Court had power to stay a proceeding if it was likely that one party would be subject to vexation and oppression. 42 GetSwift also submitted that there was no issue as to power in relation to declassing any of the class actions. As to the issue of discretion, apart from the fact that GetSwift submitted that the appropriate form of security for costs ought to be cash paid into Court, and that GetSwift wished to be heard as to the timing and quantification of security, it recognised that it was a matter for the Court as to which of the three proceedings should go forward.

THE RELEVANT OPTIONS 43 We begin by discussing the relevant options and the applicable legal principles before analysing the primary judge's reasoning towards the conclusion that it was appropriate to order a permanent stay of the Perera and McTaggart Proceedings. 44 The following realistic options are available to deal with the potential overlap between competing class proceedings: (a) first, the relevant proceedings could be consolidated; (b) second, an order could be made under s 33N(1) in respect of one or more of the proceedings, colloquially known as a declassing order; (c) third, there could be a joint trial of all proceedings with each left as they are presently constituted as open class proceedings i.e. the 'wait and see' approach; (d) fourth, there could be a permanent stay of one or more of the proceedings, the option adopted by the primary judge in the present case; and (e) fifth, an order could be made closing the classes in one or more of the proceedings but leaving one of the proceedings as open class proceedings, with a joint trial of them all. 45 We commence by discussing the three options the primary judge did not adopt, namely consolidation, declassing under s 33N(1) and the 'wait and see' approach discussed by Foster J in Cantor v Audi Australia Pty Ltd (No 2) [2017] FCA 1042 (Cantor).

(a) Consolidation 46 There is no doubt that the Court has power to order consolidation of the competing proceedings. Section 33ZG(c)(iv) of the Act expressly reserves the operation of any law relating to the consolidation of "proceedings" and accordingly the Court's inherent power or the power under r 30.11 of the Federal Court Rules 2011 (Cth) (Rules) to so consolidate, including to facilitate the objective enshrined in s 22 of the Act. 47 In the context of s 33ZG(c)(iv), "proceedings" must encompass a representative proceeding. If otherwise, it has little place in Part IVA. It must embrace consolidation of a representative proceeding with another proceeding, which could either be an individual proceeding or another representative proceeding; see generally Johnson Tiles Pty Ltd v Esso Australia Ltd (1999) ATPR 41-679; [1999] FCA 56 (Johnson Tiles), Ryan v Great Lakes Council (1997) 78 FCR 309 at 312 per Wilcox J, Bray v F Hoffman-La Roche Ltd (2003) 130 FCR 317 at [200], and Grave D, Adams K and Betts J, Class Actions in Australia (2nd ed, Thomson Reuters, 2012) at [9.240] et seq. 48 The difficulty is not so much the existence of the power but rather its exercise. Where there is no agreement between the parties in competing class actions as to consolidation, and different lawyers for the applicants and different funders in each of the actions, there are a number of difficulties in relation to consolidation. For example, what uniform funding model would operate for the consolidated proceedings given the different sets of signed up group members in each of the proceedings, when considered from the perspective of signed up group members, the unsigned group members or the funders? And, absent leave being given for separate representation for each of the applicants, what uniform legal representation for the applicants would be used? 49 In Johnson Tiles two open class actions were consolidated and the two law firms were permitted to jointly represent the plaintiffs and group members. However those law firms had entered into a joint-venture agreement in which they agreed to work cooperatively together, the agreement contained a procedure to resolve disputes and deadlocks between the law firms, and they agreed to engage the one set of counsel. In those cases, because the actions were conducted on a no win-no fee basis and there was no funder, it was not necessary for the consolidation order to deal with the problem of separate funders and how consolidation would operate in the scenario that a significant number of group members in each case may have signed up to different funding arrangements. The order did not deal with how funded group members would be affected by consolidation. The case does not assist to answer questions such as: What would be the position for those who had not signed up? What funding model would apply? And what type of common fund order would be sought, and to reflect what proposed funding mechanism? 50 For a consolidation order to be made in competing funded class proceedings, a mechanism would need to be determined for resolving such issues, including so as to achieve equity between the group members in each of the proceedings. In circumstances where the litigation funders may be jointly and severally liable (at least indirectly) for adverse costs, it may not be possible for consolidation to occur without assurances that each funder has adequate finances or insurance arrangements to meet any order which might be made in respect of the consolidated proceedings. Amongst other things, each litigation funder is likely to require sufficient information to form a view as to the co-funder's financial position and in particular its capacity to meet any order for adverse costs. 51 The result is that consolidation orders are unlikely to be made in the absence of agreement between the different applicants, funders and solicitors. There was no agreement in the present case and the primary judge did not consider consolidation to be a viable option. In circumstances where no party contended for consolidation, this option may be put to one side.

(b) Section 33N 52 Section 33N(1) of the Act provides: Order that proceeding not continue as representative proceeding where costs excessive etc. The Court may, on application by the respondent or of its own motion, order that a proceeding no longer continue under this Part where it is satisfied that it is in the interests of justice to do so because: (a) the costs that would be incurred if the proceeding were to continue as a representative proceeding are likely to exceed the costs that would be incurred if each group member conducted a separate proceeding; or (b) all the relief sought can be obtained by means of a proceeding other than a representative proceeding under this Part; or (c) the representative proceeding will not provide an efficient and effective means of dealing with the claims of group members; or (d) it is otherwise inappropriate that the claims be pursued by means of a representative proceeding. 53 The primary judge considered that, provided it is in the interests of justice to do so, s 33N(1)(c) could be used to declass a competing class action. His Honour said that the words of the subsection seemed directed to the situation in the present case, namely, whether one or more of the current proceedings will not provide an efficient and effective means of dealing with the claims of group members being claims which are also represented in other class actions. Moreover, his Honour considered that that construction was to be preferred when one had regard to its context generally and to s 15AA of the Acts Interpretation Act 1901 (Cth). His Honour said that such a construction would promote s 33N's evident purpose of being a 'control' mechanism to allow for the efficient management of regularly commenced class actions pursuant to Part IVA of the Act. 54 First, his Honour observed that s 33N(1) is part of an integrated scheme, and its purpose includes acting as a safeguard against inapt or maladroit use of the representative proceeding regime in circumstances where that scheme allows class actions to be commenced without the constraints of consent or permission of group members or the Court. Second, his Honour said that the discretion to be exercised is a statutory discretion which requires formation of one or other particular opinions or judgments. His Honour said that at first glance, the discretion created in the opening words of the section may be thought to be a discretion limited only by subject matter, because the controlling discretionary opinion incorporates a concept of some breadth, viz "the interests of justice". But his Honour said that this would be an error, as consideration of this broad evaluative concept only arises if it is established that some or all of the circumstances described in subsections 33N(1)(a) to (c) are present or, if not, s 33N(1)(d) is engaged. 55 His Honour pointed out that the Australian Law Reform Commission in its 1988 report entitled 'Grouped Proceedings in the Federal Court' rejected a certification procedure, concluding that there was "no value in imposing an additional costly procedure, with a strong risk of appeals involving further delay and expense, which will not achieve the aims of protecting parties or ensuring efficiency" (at [147]). This was notwithstanding that the 'gateway' requirements to Part IVA were undemanding. The rejection of certification was considered appropriate because embedded in the proposed Bill introducing Part IVA were protections and safeguards that existed both for group members and more generally. His Honour noted that in addition to specific protections for group members to opt out (s 33J), seek substitution (s 33T), and the right to be notified (s 33Y), there exists the general safeguard, being s 33N. 56 His Honour said that the broader context of the s 33N safeguard is that a policy choice was made to adopt a scheme providing ease of commencement, tempered by a discretionary control mechanism if the preconditions for its exercise were present. In his Honour's view, this control mechanism should not be construed restrictively such that s 33N(1)(c) is only engaged in circumstances of comparison between ordinary proceedings and class actions. It did not seem to his Honour that just because a class action might have been an efficient and effective means of dealing with the claims of group members in quite different circumstances, that is, absent another class action, this is the end of the analysis. As a matter of "logic", so his Honour said, it seemed to him that the existence of a more efficient and effective class action could render a competing class action an inefficient and ineffective means of dealing with the same claims. Accordingly, his Honour did not consider that the notion of relative efficiency is somehow foreign to determining whether the state of affairs described by s 33N(1)(c) exists. 57 His Honour considered the option of making a s 33N(1) order or "declassing" order in relation to one or more of the proceedings, but ultimately put this option to one side. It is therefore unnecessary to decide whether s 33N(1) has operation in the context of competing class actions. 58 However, we would respectfully disagree with his Honour's view in relation to the provision. We accept that his Honour's reading is expedient and consistent with a broad purposive approach if one considers s 33N generally in the totality of Part IVA, but in our view it does not give sufficient consideration to the words of the text when read in context with the other parts of s 33N. 59 We broadly agree with the analysis of Beach J in Bellamy's at [18] to [20]. It is uncontroversial that before a s 33N(1) order can be made, it must be in the interests of justice to do so because one or more of the conditions in subsections (a) to (d) is satisfied. The question is whether those subsections allow the declassing of a competing class action. 60 We consider s 33N(1) requires consideration of the comparator of whether it is in the interests of justice that the proceeding be determined in numerous individual non-representative proceedings. One compares how the factors specified in ss 33N(1)(a) to 33N(1)(d) would apply to such hypothetical proceedings. Such a comparison is expressed in ss 33N(1)(a) and 33N(1)(b) and implied by ss 33N(1)(c) and 33N(1)(d). The implicit focus in s 33N(1)(c) is on the commonality of issues and whether the representative proceedings are efficient and effective to resolve the common issues, rather than resolution by way of individual proceedings. Section 33N(1)(d) is concerned with whether representative proceedings are an appropriate vehicle to pursue the claims. Generally, the focus of ss 33N(1)(c) and 33N(1)(d) is on the efficiency or appropriateness of the group members' claims being pursued in representative proceedings. 61 It seems to us that the requirements of s 33N(1) were not satisfied in relation to any of the three competing proceedings before the primary judge. Sections 33N(1)(a) and 33N(1)(b) are not satisfied if each of the class actions is compared to hypothetical individual actions and regard is had to the interests of justice. Similarly, s 33N(1)(c) is not satisfied because, by comparison to hypothetical individual actions, each of the competing class actions is an efficient and effective process to resolve common issues. Section 33N(1)(c) does not involve a comparison between the representative proceedings and another identical or hypothetical set of representative proceedings. The efficiencies referred to in s 33N(1)(c) are focused on whether the representative proceedings are an efficient mechanism to resolve the claims and common issues, which requires considering the representative applicant's case and a comparison between the representative proceedings and the hypothetical non-representative proceedings. The inquiry is not whether the common issues might be more efficiently resolved by way of some other representative proceedings. 62 Having said this, as Beach J accepted in Bellamy's, the matter is not free from doubt. Section 33N(1)(c) uses the definite article "the representative proceeding[s]" rather than the indefinite article (cf ss 33N(1)(a), 33N(1)(b) and 33N(1)(d)), so that it is open to argue that it could be satisfied if another set of representative proceedings were available and could efficiently deal with group members' claims made in the hypothesised to be "declassed" proceedings. In other words, another available representative proceedings may make it less efficient and less effective to pursue the proposed to be "declassed" proceedings. Even so s 33N(1)(c) is not focused on relative efficiency and effectiveness but rather is expressed in more absolute terms. 63 Generally, the better construction of s 33N(1)(c) is as indicated above. Conceptual coherence dictates that one read subsection (1)(c) through the general lens of s 33N(1) by comparing representative proceedings with non-representative proceedings. Moreover, s 33N(1)(d) is also not satisfied. The claims of the group members in each of the proceedings ought to be resolved by representative proceedings; it cannot be said that it is otherwise inappropriate that such claims be pursued by means of a set of representative proceedings. 64 In our view the primary judge was right not to make a declassing order under s 33N(1)(c) and correct in putting that provision to one side. 65 The primary judge went on to say that, on the assumption that he was wrong and s 33N had no work to do in the context of competing class actions, he accepted s 33ZF as a separate basis for a declassing order, but his Honour did not make a declassing order. There are reasons to doubt that s 33ZF provides an independent basis to s 33N for a declassing order but we express no concluded view on that issue and prefer to leave an analysis of the power under that section to an occasion when it is necessary to decide.

(c) The 'wait and see' approach 66 In some contexts a 'wait and see' approach may be appropriate. However, where there are multiple open class proceedings, numerous group members signed up in each of the proceedings to different funding arrangements, and the prospect of a common fund application being made in each of the proceedings, it may be desirable that clarity be injected sooner rather than later as to the proceedings to go forward and their constitution. This is likely to be in the interests of the respondent who should not be vexed or oppressed by duplicate classes prosecuting duplicate claims. 67 Let us elaborate on one aspect which may justify the need to resolve these issues sooner rather than later. If the applicant in each of the proceedings is proposing to seek a common fund order, whether such an order should be granted and its terms will usually need to be dealt with at an early point. In those circumstances it may be necessary to resolve the appropriate constitution of each of the proceedings first before dealing with common fund questions. It is undesirable that a common fund order be made in more than one of the proceedings as that would create considerable uncertainty for group members. Moreover, the economics of determining whether such an order and its terms are appropriate and in which proceedings would be considerably complicated if one was to endeavour to address the question at the same time in the competing proceedings as they are then constituted. 68 The primary judge did not adopt a 'wait and see' approach, and in circumstances where no party before us contended that his Honour ought to have adopted such an approach, we do not need to consider this further.

(d) A permanent or temporary stay? 69 As we explain below, there can be no doubt that the Court has power to order a permanent stay. The permanent stay option was adopted by the primary judge and we will say more about this later when we analyse his Honour's reasons.

(e) Class closure 70 The Court has the power under s 33ZF(1), and also in the exercise of its inherent power, to make a class closure order that eliminates the existence of overlapping group members in two or more competing class actions. Depending on the circumstances it can be "appropriate … to ensure that justice is done in the proceeding" to close the class in one or more of the proceedings to eliminate the duplication in group membership over all proceedings. This would not involve barring orders as such as group members in each of the proceedings who had entered into a funding agreement would have the choice of remaining as such. As for those group members in all proceedings who had not entered into a funding agreement, they would be captured by the remaining open class proceedings and could choose to maintain membership thereof or opt out.

THE PERMANENT STAY 71 We now turn to consider his Honour's choice of the option to permanently stay the Perera and McTaggart Proceedings and the parties' submissions before us. 72 Before the primary judge GetSwift argued that an abuse of process was plain because the use of the Court's procedures had meant that the Court's processes would, if more than one open class proceeding continued, be being used for an improper purpose. It argued that such a duplicative proceeding would be unjustifiably oppressive and vexatious and/or the use of the Court's procedures in the manner contemplated would bring the administration of justice into disrepute.

(a) Primary judge's reasons 73 The primary judge held that it was uncontroversial that the Court has power to stay a proceeding which constitutes an abuse of process, and that the Court has an implied power to control its own processes. His Honour said that the implied power to control the Court's processes included the power to prevent conduct which, although not inconsistent with the literal application of procedural rules of court, would nevertheless be manifestly unfair or would otherwise bring the administration of justice into disrepute among right-thinking people: see Walton v Gardiner (1993) 177 CLR 378 at 393 per Mason CJ, Deane and Dawson JJ; Jeffery & Katauskas Pty Limited v SST Consulting Pty Ltd (2009) 239 CLR 75 (Jeffery) at [28] per French CJ, Gummow, Hayne and Crennan JJ. 74 His Honour recognised that abuse of process can take many forms, many of which were presently irrelevant, including bringing claims that can no longer be determined justly or where there has been destruction of evidence or in circumstances of re-litigation or attempts to litigate issues which could and should have been raised previously. But his Honour said that some established categories have a closer similarity to the issue that presented itself before him than others, such as the bringing of concurrent proceedings in different courts relating to the same subject matter. 75 His Honour drew attention to Melbourne City Investments Pty Ltd v Treasury Wine Estates Ltd (2016) 243 FCR 474 where Foster J at [97] to [146] surveyed the relevant authorities and at [147] distilled the following propositions, which we accept: (a) Notions of justice and injustice must reflect contemporary values if the courts and the administration of justice are to continue to enjoy the confidence of the public. (b) Abuses of procedure usually fall into one of the following three broad categories: (i) the court's procedures are invoked for an illegitimate or improper purpose; (ii) the use of the court's procedures is unjustifiably oppressive to one of the parties or vexatious; or (iii) the use of the court's procedures in the manner contemplated would bring the administration of justice into disrepute. (c) The onus of proving an abuse in any given case rests upon the party alleging abuse, and that onus is a heavy one. 76 The primary judge said that a good indication of the caution that should guide staying proceedings that have been regularly commenced can be seen in Voth v Manildra Flour Mills Pty Ltd (1990) 171 CLR 538 (Voth) at 554, although that was a case involving principles of forum non conveniens. 77 His Honour also explained that the categories where abuse arises are not closed. In Batistatos v Roads and Traffic Authority of New South Wales (2006) 226 CLR 256 (Batistatos), Gleeson CJ, Gummow, Hayne and Crennan JJ observed that "what amounts to abuse of court process is insusceptible of a formulation comprising closed categories. Development continues" (at [9]). Their Honours emphasised that no "hard and fast definitions" of abuse of process could be laid down, and that this was necessarily so as "notions of justice and injustice, as well as other considerations that bear on public confidence in the administration of justice, must reflect contemporary values and, as well, take account of the circumstances of the case" (at [14]). And as was relevantly explained by the majority in Jeffery at [28]: It has been said repeatedly in the judgments of this Court that the categories of abuse of process are not closed…This does not mean that abuse of process is a term at large or without meaning. Nor does it mean that any conduct of a party or non-party in relation to judicial proceedings is an abuse of process if it can be characterised as in some sense unfair to a party. It is clear, however, that abuse of process extends to proceedings that are "seriously and unfairly burdensome, prejudicial or damaging" or "productive of serious and unjustified trouble and harassment". (Citations omitted.) 78 The primary judge made three further broad points. 79 First, his Honour said that there was no doubt that caution needed to be exercised in staying a proceeding permanently, but that there was an important difference between the context of ordinary inter partes litigation and the context of class actions. In ordinary litigation, leaving aside forum non conveniens stays, the consequence of a permanent stay is that the party whose claim is stayed is shut out from being able to vindicate its rights, and preventing a litigant from vindicating their claim is a serious step. However, his Honour said that this consideration does not apply in the same way in the context of class actions. If a permanent stay of one competing class action is granted, this will not prevent the claim of the stayed applicants, which have an existence separate from, and anterior to, the proceedings they commenced, from being able to be advanced. Those claims can be litigated in the class action that remains, and will be resolved either by settlement approved by the Court or by curial determination. Moreover, each applicant as a group member in any unstayed class action has a statutory right under s 33J of the Act to opt out, and to then maintain their own individual claim if they so wish. 80 Second, his Honour made reference to the concept of a stay being ordered, absent an express finding of abuse of process, utilising either the implied powers of the Court or potentially s 33ZF(1). We will discuss this in more detail later. 81 Third, in the context of considering whether he should invoke this Court's equitable jurisdiction to grant an injunction rather than to order a permanent stay, his Honour discussed at length 'bills of peace'. His Honour recounted that prior to the Supreme Court of Judicature Act 1873 (UK) reforms, Courts of Chancery often considered it necessary to protect or enforce an equity by enjoining proceedings at law or the taking of other steps at law that involved interference with equitable rights. His Honour then narrowed his focus to equity's development of the bill of peace, which according to some scholars is a 'distant mirror' or antecedent of the modern class action: see, for example, William Weiner and Delphine Szyndrowski, "The Class Action, from the English Bill of Peace to Federal Rule of Civil Procedure 23: Is There a Common Thread" (1987) 8 Whittier Law Review 935; Thomas Rowe, "A Distant Mirror: The Bill of Peace in Early American Mass Torts and Its Implications for Modern Class Actions" (1997) 39 Arizona Law Review 711 at 712. 82 By the bill of peace, a plaintiff sought to establish a right which was capable of being or had been disputed in several actions, and claimed a perpetual injunction against future litigation. It could lie where one person claimed or defended a right against many, or where many claimed or defended a right against one. The principle that Chancery would interfere to prevent multiplicity of suits was the foundation of the jurisdiction to entertain bills of peace. In this way, the bill was developed as a means for equity to interpose and prevent unnecessary expense and litigation, and to decide once and for all the validity or invalidity of questions of law or fact common to a larger controversy. The effect of the bill of peace was to provide a mechanism by which equity would overcome rigid common law joinder rules to bring together persons interested in a controversy and then direct a hearing of a suit in equity or a trial at law in a manner which would bind all concerned. But the bill fell into disuse following the Supreme Court of Judicature Act 1873 (UK) reforms, which provided that the equitable rules governing parties, which permitted actions of a representative character, would apply to all actions, even those actions that would once have only been at law.

(b) The Perera and McTaggart applicants' submissions 83 Before us there was a substantial crossover in the submissions of the Perera and McTaggart applicants, and each relied on the submissions of the other. Although they were not precisely the same it is unnecessary to set them out separately. 84 First, they argue, and we accept, that Part IVA expressly contemplates that a representative proceeding need not encompass the claims of all possible claimants. Section 33C(1) provides that a proceeding may be commenced on behalf of "some or all" persons with claims against the same persons. It is these words that authorise the commencement of closed classes. Section 33J grants liberty to group members to opt out of a proceeding commenced on their behalf, necessarily carrying with it the ability to vindicate their rights outside that proceeding. Accordingly, the legislative intention was not to require persons to participate in any particular proceeding in order to vindicate their rights but rather to enable persons to have common questions affecting their individual claims determined in a particular proceeding unless they did not want that to happen. Accordingly, Part IVA contemplates the possibility of multiple proceedings whether in the form of one representative proceeding and a number of individual proceedings, or in the form of multiple representative proceedings, such as one or more closed classes or an open class representing persons who are not in the closed classes. 85 In this regard they contend that when it is understood that Part IVA does not guarantee that there will not be more than one proceeding against a respondent, it is fallacious to argue that Part IVA is intended to ensure that a respondent is not to be vexed by multiple suits or the costs of defending multiple suits. They submit that the proper way of looking at the matter must acknowledge that the existence of even one representative proceeding saves the respondent from being vexed by even more suits, namely, those of the multiplicity of group members who are represented. But the fact that a respondent is being saved from having to individually defend hundreds of proceedings provides no basis for contending that the respondent should not have to defend more than one proceeding. Generally speaking, we also agree with this contention. 86 To the Webb applicant's and GetSwift's submission that where one proceeding is sufficient to vindicate all proper claims in a matter, any duplicative proceeding must be an abuse, they submit that that proposition is not correct in the context of Part IVA of the Act. If correct, it would mean individual claims would become an abuse, notwithstanding the right of group members to opt out. Part IVA contemplates and provides for multiple claims raising the same issues. 87 Further, to the submission that multiple proceedings are prima facie vexatious they submit that that proposition is contrary to the reasoning in Bellamy's (see also Cantor at [60] (Foster J)). Moreover, they submit that none of the cases relied upon by GetSwift were decided in a Part IVA context. They concern claims which, in substance, were between the same parties, or where the moving parties were the same. That is not the case here. Group members are bound by the resolution of the common questions, but they are not responsible for the commencement of the proceedings and indeed may be ignorant of them. They are not parties in the conventional sense. 88 Further, they contend that there is no inherent conceptual difficulty with the continuation of multiple representative proceedings on behalf of the same group members. In a juridical sense, the only significance of being a group member is that such a person is bound by a judgment given in such a proceeding under s 33ZB of the Act. Unless a person becomes bound by inconsistent judgments no difficulty arises. And given the essentially passive role of group members, if multiple proceedings are managed together, the prospect of inconsistent judgments will rarely arise. They say that for practical reasons the Court has often, but not always, preferred to resolve the "overlap" at some point in the proceedings, so that, for example, notices under ss 33X and 33Y can be sent with minimal prospect of confusion, or a common fund order can be considered, but those decisions are essentially ones of case management. We would also agree with that contention in the generality with which it is expressed. 89 Second, the Perera and McTaggart applicants submit that there is no general principle which requires the Court to prevent all but one open class representative proceeding going forward. They argue that the primary judge approached the competing class actions on the basis that: (a) the overlapping class actions created a problem which "affects the processes of the Court" (at [3] to [4], [94]); and (b) the continuation of duplicative open class proceedings would allow the use of the processes of the Court in a way which would be unfair to GetSwift by countenancing a proceeding being conducted contrary to the overarching purpose, and to group members by increasing costs (at [345]). However, his Honour failed to identify what processes of the Court were affected, or why and submitted that it was not so. In circumstances where Part IVA expressly allows for multiple actions against a respondent (and multiple class actions), they contend that the mere existence of two overlapping open class actions does not affect the processes of the Court in any relevant way. They submit this error led his Honour down a path where he felt he had to resolve the overlap by reducing the multiplicity of proceedings into one proceeding. 90 Third, the Perera and McTaggart applicants say that because the primary judge was driven to select one class action to go forward, this error led his Honour to fail to give any or any sufficient weight to the choices that had been made by significant numbers of group members as to their representation. In the Perera Proceeding the evidence was that over 100 persons had entered into litigation funding agreements with ILP18, many of whom had entered into retainers with SPB, and numerous others had registered an interest in participating in the Perera Proceeding. In the McTaggart Proceeding there were over 200 group members who had signed funding agreements with Vannin. 91 Relatedly they argue that deciding the issue as he did the primary judge gave no weight to the fact that no choice at all was given to other group members as to how, and by whom, they wanted their claims to be conducted. Mr Perera submits that he proposed a different course which involved giving a choice to group members, and was premised upon the Court selecting one proceeding as the 'default open class proceeding', then canvassing the attitude of the group by using the tools conferred on the Court by Part IVA to issue opt out notices under ss 33X and 33Y of the Act, before making a final decision on what to do with the other actions. In this way, where group members had already made a choice this would be respected, other group members would be given an opportunity to choose, and those who remained silent would remain in the proceeding chosen by the Court. Mr Perera submits that such an approach ought to have lead the Court to: (a) select the Perera Proceeding as the default open class and close the McTaggart Proceeding; and (b) declass the Webb Proceeding where the evidence was that no person had registered an interest in participating. 92 The Perera and McTaggart applicants emphasise that there was at least one option available to the primary judge which his Honour did not consider, or adequately consider, namely class closure orders. They say that such orders would have addressed any vexation from GetSwift's point of view as no group member would be permitted to remain in more than one class action: Bellamy's at [62]. GetSwift would be faced with more than one class action, but the scheme of Part IVA permits and contemplates that result. 93 They speculate that the primary judge discarded class closure orders as an option in light of his finding that there were not "a substantial number" of group members signed up to funding agreements, unlike the situation in Bellamy's (see primary judgment at [109] and [217]), and submit that basis is wrong. They argue that the number of group members who entered into funding agreements in each of the Perera and McTaggart Proceedings is far from trivial and there was no principled basis for the primary judge to find it to be insubstantial, compared with, for example, the number of group members signed up in the Basil class in Bellamy's (around 1,000). 94 On their submissions the result of the primary judge's decision is that more than 300 group members (over 200 in the McTaggart class and over 100 in the Perera class) have been deprived of their choice of funder, lawyer and funding arrangements. Those group members may be dissatisfied with the Webb Proceeding and may opt out and commence their own class action proceedings. They argue that, as Beach J found in Bellamy's at [39], such a potential consequence tends against a permanent stay. 95 Fourth, the Perera and McTaggart applicants submit that while all parties accept that the Court has implied power to grant a permanent stay for abuse of process, GetSwift's submissions seek to expand the scope of that power beyond an abuse of process where "the broader interests of justice so require". They contend that such a submission is unsupported by authority and is contrary to basic principle as affirmed in Voth at 554: First, a plaintiff who has regularly invoked the jurisdiction of a court has a prima facie right to insist upon its exercise. Secondly, the traditional power to stay proceedings which have been regularly commenced, on inappropriate forum grounds, is to be exercised in accordance with the general principle empowering a court to dismiss or stay proceedings which are oppressive, vexatious or an abuse of process and the rationale for the exercise of the power to stay is the avoidance of injustice between parties in the particular case. Thirdly, the mere fact that the balance of convenience favours another jurisdiction or that some other jurisdiction would provide a more appropriate forum does not justify the dismissal of the action or the grant of a stay. Finally, the jurisdiction to grant a stay or dismiss the action is to be exercised "with great care" or "extreme caution". 96 They contend that each of those propositions is difficult to reconcile with the existence of a power permanently to stay a proceeding, regularly commenced, merely because the court considers that to do so would be in the "broader interests of justice". This is particularly so given that, as the respondents emphasise, the categories of abuse are not closed. 97 Moreover, they say that in circumstances where all parties accepted that only one open class action would proceed, there could be no question of an abuse of process if the duplication of class could be eliminated. They refer to Beach J's conclusion in Bellamy's at [33] that there was power to stay, and that his Honour went on to observe at [37] that: no authority cited to me persuasively supports the proposition that in circumstances where the duplication of class membership is eliminated in respect of two group proceedings involving the same subject matter against the same respondent, that nevertheless there is still vexation, oppression or an abuse of process in both such proceedings as so modified to eliminate the duplication going forward. (Emphasis in original.) 98 The McTaggart applicants gave an illustrative example in which two identical class actions are commenced, on a closed class basis with no overlap, against the one respondent. Each applicant then makes an application to open their class to include all group members other than those represented by the other group. They submit that the Court in that event would consider which action was the preferable vehicle for prosecuting the claims of those not in either group, and that in such a situation there could be no question of the Court permanently staying one action and allowing the other to expand to the whole class. In substance, but not in form, they argue that was the situation facing the primary judge. 99 In their submissions a permanent stay should only be granted if it is the only fair and practical way of ensuring justice between the parties seeking such an order: Rozenblit v Vainer (2018) 356 ALR 26; [2018] HCA 23 at [34] per Kiefel CJ and Bell J, at [44] per Keane J and at [100] per Gordon and Edelman JJ. On their contentions the primary judge did not give any consideration to issuing class closure orders as a less drastic alternative to a permanent stay. His Honour's reasons on a permanent stay are exclusively concerned with whether the power to grant a stay was enlivened (i.e. whether an abuse of process existed) rather than whether the power, assuming it to be enlivened, should be exercised, having regard to the alternatives: [331] to [349]. Class closure orders are also omitted from the options under consideration in Section E of the reasons: [105] to [167]. 100 Fifth, the Perera and McTaggart applicants generally submit that Part IVA respects group member autonomy, as shown by the fundamental right to opt out and to run one's own proceeding. They said that even if it were desirable for there to be only one proceeding against GetSwift, this could not be achieved prior to group members being given the opportunity to opt out. They contend that for the primary judge to presume that there must be only one representative proceeding against a respondent, and putting forward a 'solution' which does not grapple with opt out (as the primary judge did not) is necessarily inadequate. 101 They also contend that as Part IVA respects the autonomy of group members, when administering Part IVA this Court must respect contracts that group members enter into. They argue that the practical effect of what the primary judge did was to subject that subset of group members who had made a choice to enter a funding contract with ILP18 and to retain SPB to run their case (or alternatively to enter a funding contract with Vannin for it to fund a proceeding in which the McTaggart applicants had retained Corrs to run their case) to a regime where their case was: (a) funded by a funder they had not contracted with but to whom they were bound by a common fund order his Honour foreshadowed making and ultimately did make; and (b) run by lawyers they had not chosen, where his Honour foreshadowed that the lawyers they had chosen would not be permitted to communicate with them about opt out, with those lawyers ultimately being restrained by Court order from so doing. 102 Sixth, they submit that the primary judge appears to have reasoned that, notwithstanding that the Perera Proceeding and the McTaggart Proceeding were regularly commenced under Part IVA bona fide and for a proper purpose, they needed to be prevented from continuing by reason of the existence of the Webb Proceeding, a subsequent open class representative proceeding which his Honour considered to be a preferable vehicle to advance group member claims. His Honour found that because of this the Perera and McTaggart Proceedings had "taken on" the character of using the Court's processes "untethered from advancing the proper purposes" of a representative proceeding (at [347]). It is unclear from [348] whether or not his Honour in fact found the continuation of these proceedings would amount to an abuse of process; he says in that paragraph both that he has, and has not, so found. The juridical basis upon which his Honour considered he was acting when staying the Perera and McTaggart Proceedings is entirely unclear. 103 They contend that there was no juridical basis for the primary judge to stay the Perera Proceeding on the basis identified in [345] to [348]. They argue that his Honour appears to have thought he was acting pursuant to the Court's implied power to control its own processes ([109] to [111]) but the principle in Walton v Gardiner still requires focus on the conduct of the party who is said to be abusing the Court's process and what processes are being interfered with. It does not permit a proceeding to be characterised as abusive in the abstract. 104 Generally, they submit that the primary judge either did not give adequate consideration to the possibility of class closure orders, or discounted such orders on an erroneous assessment of the facts. Either way, there was error in the exercise of discretion. It is said that the same error vitiates the primary judge's hypothesised determination that he would have made s 33N orders "declassing" the McTaggart and Perera Proceedings had a stay and/or equitable restraint been unavailable. 105 Moreover, the Perera and McTaggart applicants submit that the reasons given by his Honour in [345] to [348] were conclusory, and not based on evidence, as follows: (a) contrary to [345] and [348], there was no evidence that that the continuation of more than one open class representative proceeding would unnecessarily increase costs. It may be accepted that additional proceedings involve an additional cost, but: (i) there is no valid comparator by which his Honour could have formed the view that those costs would be unnecessary as it cannot be assumed that the expenditure of more costs will result in a worse outcome for group members either on settlement or judgment; and (ii) the fact that there may be additional costs is irrelevant anyway, unless the evidence showed they would reach such a scale (notwithstanding the Court's ability to control them) as to constitute an abuse of process; (b) contrary to [347], there was no evidence that the Perera Proceeding had become a means of attempting to secure profit (to unidentified persons) while impacting adversely on the progression of group member claims. His Honour did not explain what adverse impact he perceived. There was none; (c) contrary to [345], [347] and [348], no unfairness would be caused to GetSwift by permitting multiple proceedings to continue against it, in circumstances where Part IVA expressly contemplates this. And two class actions is better for a respondent than the alternative of multiple individual proceedings; (d) contrary to [348], it was not necessary to prevent all but one open class proceeding to proceed so as to "safeguard the effective operation of Part IVA". Part IVA can operate effectively regardless of how many class actions are being pursued; and (e) his Honour's finding that the Perera Proceeding had taken on the character attributed to it in [347] was both circular and illogical. It was circular because that finding was a consequence of his Honour's finding that only one representative proceeding should be permitted to continue and so could not be a ground for it. It was illogical because it was not explained how the purpose of the Perera Proceeding changed from being tethered to a proper purpose to untethered from one, simply because of something done by somebody else.

(c) The GetSwift and Webb submissions 106 In brief summary, before us GetSwift and Mr Webb submitted as follows. 107 They point out that there was (and is) no dispute that the Court has power to stay a proceeding that is an abuse of process. The only dispute is whether the present circumstances fall within the concept of abuse of process, and if not whether there is power to grant a stay in circumstances beyond an abuse of process where the broader interests of justice so require. 108 GetSwift and Mr Webb contend that where a conclusion is properly reached, on an evaluation of all the circumstances, that one open class action is sufficient to vindicate the claims of all group members, it is open to a court to grant a permanent stay of other extant open class actions covering the same claims on the ground that they represent an abuse of process or otherwise are contrary to the interests of justice in the proceedings. 109 They submit that the Court has a range of discretionary remedies available to it to address the scenario of competing class actions. One of those remedies is the power to grant a permanent stay of two of the three open class proceedings so that only one proceeding goes forward. They argue that the primary judge was correct to conclude that the Court had power to order such a stay, and that was consistent with the conclusion as to power reached by Beach J in Bellamy's at [33]. 110 Further they contend that it is well-established that the Court possesses an inherent or implied power to stay proceedings as an abuse of process if the proceedings are unjustifiably oppressive or vexatious or manifestly unfair or otherwise bring the administration of justice into disrepute among right-thinking people. The power can be sourced, inter alia, to the terms of ss 23 and 5(2) of the Act. They note that the primary judge also considered other sources of power to impose a stay, beyond an abuse of process, where the broader interests of justice so require, and submit that, if necessary, such power is available to justify the orders made. 111 On their submissions the real question is one of scope or of identifying the precise legal principle(s) by which the power to grant a stay is to be exercised. They say that a stay may be granted to "prevent misuse of [courts'] procedures in a way which, although not inconsistent with the literal application of procedural rules of court, would nevertheless be 'manifestly unfair to a party to litigation … or would otherwise bring the administration of justice into disrepute among right-thinking people'": see Jeffery at [28] per French CJ, Gummow, Hayne and Crennan JJ quoting Lord Diplock in Hunter v Chief Constable of the West Midlands Police [1982] AC 529 at 536. 112 Further, they say that the primary judge correctly emphasised that the categories of abuse are not closed and the flexibility of the principles, while also noting the need for a cautious approach. They argue that the relevant authorities emphasise flexibility, both in respect of the doctrine of abuse of process and the concepts of vexation and oppression, as well as the need for such doctrines to adapt to new circumstances. 113 Further, they submit that the circumstances in which abuse can arise are "very varied". In Batistatos, the plurality observed that "what amounts to abuse of court process is insusceptible of a formulation comprising closed categories. Development continues" (at [9]). Their Honours endorsed (at [14]) a passage from Gaudron J's reasons in Ridgeway v The Queen (1995) 184 CLR 19 at 74 to 75 which emphasised that no "hard and fast definitions" of abuse of process could be laid down, and that this was necessarily so as "notions of justice and injustice, as well as other considerations that bear on public confidence in the administration of justice, must reflect contemporary values and, as well, take account of the circumstances of the case" but that abuse of process extends to proceedings that are "seriously and unfairly burdensome, prejudicial or damaging" or "productive of serious and unjustified trouble and harassment". They contend that the power to stay such proceedings and thereby safeguard the administration of justice transcends the interests of any party to the litigation. 114 GetSwift draws attention to CSR Limited v Cigna Insurance Australia Limited (1997) 189 CLR 345 a decision in a different context, as a good example of where a court, once satisfied that one proceeding is necessary and sufficient to vindicate all proper claims in the matter, may properly conclude that the continuation of another, overlapping proceeding is or becomes vexatious or oppressive so as to be permanently stayed. 115 They note that while the species of stay in question is described as "permanent", the Court always retains the power to reconsider the continuation of a permanent stay should there be a material change in circumstance: CSR v Cigna at 402 to 403; Cooper v Williams [1963] 2 QB 567 at 580 and 582; Lambert v Mainland Market Deliveries Ltd [1977] 2 All ER 826 at 832 to 833. For example, so GetSwift submits, where a court permanently stays one or more open class actions in favour of another, it is possible that the favoured action may later collapse or become unviable because the funder becomes insolvent or the solicitors become incapable of acting and no suitable replacement can be found. In such circumstances the Court can revisit the "permanent" stay. They contend that the primary judge recognised this by making an order in each of the Perera and McTaggart Proceedings that the permanent stay did not prevent either applicant from making an application under s 33T of the Act if for some reason Mr Webb was not able to adequately represent the interests of group members.

ANALYSIS - PERMANENT STAY 116 It is convenient to discuss his Honour's reasons and the parties' criticisms by reference to the following sequence of issues: (a) First, the source of the power identified by his Honour or capable of supporting his Honour's decision to permanently stay two of the three proceedings. (b) Second, whether if the power to stay existed, the selection of that option from the available options rather than one of the other options was a choice free from material error. (c) Third, if it was appropriate to order a stay of two of the three proceedings, whether his Honour's selection of the Webb Proceeding as the action to go forward in preference to either the Perera Proceeding or the McTaggart Proceeding was a choice free from material error. 117 In dealing with these questions we address the key themes in the parties' submissions rather than expressly addressing all of their minutiae. In the following sections we will address the first and second issues above and we will deal with the third issue in a later section.

(a) The source of power to stay 118 We commence by noting that in Money Max Int Pty Ltd v QBE Insurance Group Ltd (2016) 245 FCR 191 (Money Max) at [196] Murphy, Gleeson and Beach JJ observed: Competing class actions in this Court are the subject of active case management but they can still cause increased legal costs for both sides, wastage of court resources, delay, and unfairness to respondents, particularly when they are commenced in different courts (such as in both the Federal Court of Australia and a State Supreme Court). 119 More recently, in Wileypark Pty Ltd v AMP Limited [2018] FCAFC 143 (Wileypark) Allsop CJ (with which Middleton and Beach JJ agreed) said (at [2]) that "the running of multiple actions by different lawyers, with different funders was, in principle, potentially inimical to the administration of justice and, in particular, potentially inimical to the interests of group members, and potentially oppressive to [the respondent]". His Honour's observation was particularly emphasised with respect to multiple open class actions. 120 Similar concerns have been expressed in other class action jurisdictions. In 2002 the US Advisory Committee on Civil Rules said that: …overlapping class actions in federal and state courts threaten the resolution and settlement of such actions on terms that are fair to class members, defeat appropriate judicial supervision, waste judicial resources, lead to forum shopping, burden litigants with expenses and burdens of multiple litigation of the same issues, and place conscientious class counsel at a potential disadvantage. (Committee on Rules of Practice and Procedure, Report of the Civil Rules Advisory Committee 15-16 (2002) as cited in Morabito V, "Clashing Classes Down Under - Evaluating Australia's Competing Class Actions through Empirical and Comparative Perspectives", 27 (2012) Connecticut Journal of International Law 245 at 249).

The Court's power to manage the cases before it 121 Faced as the Court is with such risks to the proper administration of justice we consider it to be plain that the Court has power to order a permanent stay of one or more competing class actions whether in the exercise of its inherent power, or its express and implied powers to manage the cases before it in the interests of justice and the parties and consonant with the requirement under s 37M, or in its equitable jurisdiction. 122 In the present case the primary judge reached the view that allowing the continuance of three competing class actions was likely to be more expensive for the parties and group members and less efficient than staying two of the cases and allowing only one to proceed. We would respectfully agree. In our view three class actions brought primarily against the same respondent, in respect of substantially the same claims, and on behalf of the same group members, are likely to: (a) involve increased use of judicial and Court resources; (b) move more slowly and less efficiently through the interlocutory stages; (c) incur increased legal costs on the applicants' side which (if the cases are successful) will ultimately be paid by group members out of the same pool of available settlement or judgment monies; and (d) incur increased legal costs on the respondent's side through the requirement to defend three proceedings rather than one, including by addressing different case theories, different expert evidence and different tactical approaches. Such increased costs may mean costs become disproportionate to the importance and complexity of the matters in dispute. 123 Once it is accepted that competing class actions may have such effects it is plain that the power to permanently stay a competing class action is readily found in the Court's express and implied power to manage the cases before it. The Court's powers include the power under s 22 of the Act to order remedies to avoid a multiplicity of proceedings and under s 23 to make such orders as the Court thinks appropriate. 124 Importantly, s 37M(1) provides that the overarching purpose of the civil practice and procedure provisions is to "facilitate the just resolution of disputes…according to law and…as quickly, inexpensively and efficiently as possible". Section 37M(3) provides that any power conferred under the civil practice and procedure provisions (which are defined to include any provision of the Act and the Rules) must be exercised or carried out in a way that best promotes the overarching purpose. 125 The Court also has a general power to control its own proceedings, which power extends to ordering either a permanent or temporary stay as the occasion demands. For example, across multiple representative proceedings the interests of justice may support a permanent stay of one of the proceedings. That power can be derived from the Court's broad inherent powers rather than necessarily under s 33ZF. 126 These powers are not constrained by any necessity to fit them into a specific category such as "abuse of process". 127 It may be doubted that s 33ZF(1) provides a source of power to stay a competing representative proceeding because it is difficult to see how a stay of the proceeding could be said to ensure that justice is done "in the proceeding" (our emphasis). It is, however, unnecessary to decide that issue in the present context and we prefer to leave an analysis of the power under s 33ZF to an occasion when it is necessary to decide.

Power in the equitable jurisdiction 128 The primary judge did not invoke the use of 'bills of peace' to restrain multiplicity of actions. It is unnecessary to decide whether a bill of peace provides an alternative source of power, but we consider it doubtful. 129 There has been little reference in Australian case law to bills of peace, although in Re Lethbridge v Mitchell (1887) 8 LR (NSW) 249 at 254, Darley CJ referred to the possibility of a suit in equity in the nature of a bill of peace being instituted and that a court of equity ought to "interfere and grant a perpetual injunction to quiet the possession of the applicant and to suppress future litigation of their right". 130 In our view the equitable bill of peace procedure does not sit well with Part IVA. In Wong v Silkfield Pty Limited (1999) 199 CLR 255 at [15] the High Court discussed representative proceedings in equity and pursuant to Part IVA of the Act and said: In the United States, before the introduction in 1938 of the present Federal Rules of Civil Procedure and in particular r 23, there was much discussion as to whether the bill of peace, as understood in English equity, provided a model for class actions in tort, but without jury trial. In particular, there was a controversy as to what was involved in a requirement of a community of right, title and interest in the subject matter. (Citations omitted.) Their Honours then returned to the bill of peace (at [18]) and after noting its characteristics said: Part IVA establishes a regime which supplants these procedures, but the phrase in par (c) of s 33C(1) "a substantial common issue of law or fact" gives rise to issues of construction which reflect the disparate interests which the old procedures had sought to accommodate. 131 In light of the High Court's view that Part IVA "supplants" bills of peace they may be put to one side. 132 The Court does though have an equitable jurisdiction to restrain multiplicity of actions. In CSR v Cigna the majority recognised two somewhat distinct bases for granting an anti-suit injunction (at 391 to 394). The first basis for granting such relief is the inherent or implied power to protect the integrity of the court's processes once set in motion. So, the court may grant an injunction to restrain a person from commencing or continuing parallel proceedings in another forum if those proceedings interfere with, or have a tendency to interfere with, the proceedings pending in the court. Moreover, such an inherent or implied power to grant anti-suit injunctions is not restricted to defined and closed categories, save for the boundary condition that it may only be exercised when the administration of justice so demands and when necessary for the protection of the court's own proceedings or processes. 133 The second basis invokes equitable jurisdiction to restrain conduct where the bringing of another proceeding involves the unconscientious exercise of legal rights. So, one well-established category of case in which an injunction may be granted in the exercise of that jurisdiction is that involving the commencement of parallel proceedings which are vexatious or oppressive. Relatedly, if a party has made an election as to the forum in which it will proceed, equity may intervene to prevent it pursuing proceedings in relation to the same subject matter in another. Generally, it has been said that the limits of the Court's equitable jurisdiction are determined by the dictates of equity and good conscience. 134 The primary judge pointed out that the categories of conduct which can be characterised as unconscientious are not closed, and in the event that the pursuit of one or other of the class actions was contrary to good conscience, such conduct could be enjoined in equity, leaving aside the separate notion of the implied power to protect the integrity of the Court's processes from an abuse of process.

Power to stay as an abuse of process 135 It is uncontroversial that the Court has power to stay proceedings to prevent an abuse of the Court's process, and the primary judge identified the principal source of the power exercised to stay the Perera and McTaggart Proceedings as that power. 136 Having regard to our view as to the availability of express and implied powers to manage the cases before the Court, including to order a stay, it is strictly unnecessary to address the existence of power to stay through the prism of abuse of process but we doubt that the circumstances in the present case show an abuse. 137 In UBS AG v Tyne [2018] HCA 45 at [1] Kiefel CJ, Bell and Keane JJ said that either of two conditions enliven a court's power to stay proceedings for an abuse of process, namely where the use of the court's procedures occasions unjustifiable oppression to a party or where such use serves to bring the administration of justice into disrepute. 138 In determining whether either condition is satisfied, a consideration of all the relevant circumstances is required, which necessitates "a broad, merits-based judgment which takes account of the public and private interests involved and also takes account of all the facts of the case": Johnson v Gore Wood & Co [2002] 2 AC 1 at 31 per Lord Bingham of Cornhill cited with approval in UBS AG at [7]. 139 In determining whether the bringing or continuance of a proceeding is an abuse of process relevant civil procedure provisions, including s 37M of the Act must also be considered. As Kiefel CJ, Bell and Keane JJ said in UBS AG at [38]: The timely, cost effective and efficient conduct of modern civil litigation takes into account wider public interests than those of the parties to the dispute. These wider interests are reflected in s 37M(2) of the FCA. As the joint reasons in Aon Risk Services Australia Ltd v Australian National University explain, the "just resolution" of a dispute is to be understood in light of the purposes and objectives of provisions such as s 37M of the FCA. Integral to a "just resolution" is the minimisation of delay and expense. (Citations omitted.) 140 The plurality also said that: (a) "the just resolution of a controversy may be the permanent stay of the proceeding notwithstanding that the plaintiff is not at fault and that the merits of his or her claim have not been decided": UBS AG at [40], citing Batistatos at [63]; and (b) it is relevant to consider whether permitting the continuance of proceedings may "give rise to the perception that the administration of justice is inefficient, careless of costs and profligate in its application of public moneys": UBS AG at [59]. 141 In UBS AG Gageler J reinforced the necessity, and perhaps the priority, to consider the public interest when he said the following (at [70] to [72]): Finally, and for present purposes no less importantly, Lord Bingham's reference to the need for the requisite normative judgment to take account of "the public and private interests involved" underscores that the question of whether the claim sought to be brought in the later proceedings "should" have been brought in the earlier proceedings cannot be determined solely by reference to interests of the parties to the action. There is a public interest in the timely and efficient administration of civil justice. The importance of that public interest has only increased in the years since Lord Bingham spoke. Dowsett J sought to capture that public interest when he referred in dissent in the decision under appeal to the abusive character of litigious conduct which, if permitted, would lead "the right-thinking person" to perceive the system for the administration of civil justice to be "inefficient, careless about the incurrence of cost by the parties, and profligate in the application of public moneys". His Honour's anthropomorphic allusion was evidently drawn from the frequently quoted description of the power to prevent an abuse of process in terms of an "inherent power which any court of justice must possess to prevent misuse of its procedure in a way which, although not inconsistent with the literal application of its procedural rules, would nevertheless be manifestly unfair to a party to litigation before it, or would otherwise bring the administration of justice into disrepute among right-thinking people". Although undoubtedly capable of application in circumstances in which use of a court's procedures would be unjustifiably oppressive to a party or would bring the administration of justice into disrepute, the doctrine of abuse of process has repeatedly been recognised to be insusceptible of a formulation which would confine it to closed categories. In the context of the application of the doctrine to the bringing of successive proceedings, consistently with the analysis of Lord Bingham, I think it better in weighing the private and public interests involved to eschew the extremes of private "oppression" and of public "disrepute". The relevant public interest is ordinarily appropriately identified in more general and less emotive terms as the timely and efficient administration of civil justice. (Citations omitted.) 142 Gordon J noted (at [138] and [139]): Where there are attempts to raise substantially similar claims in successive proceedings, the doctrine of abuse of process overlaps with the doctrine of estoppel. The raising of issues in successive proceedings can be simultaneously the subject of an estoppel which has arisen as a consequence of the earlier, final judgment, and conduct which constitutes an abuse of process. However, such issues may also be considered an abuse of process in circumstances where they do not give rise to an estoppel because the doctrine of abuse of process is "inherently broader and more flexible than estoppel". By way of example, estoppel would not preclude a person who was neither a party nor the privy of a party to earlier proceedings from raising similar or related claims in successive proceedings, whereas abuse of process might in appropriate circumstances. Not only is there the overlap with estoppel, but the principles relevant to the exercise of power to stay a proceeding as an abuse of process need to be interpreted and applied in light of the overarching purpose set out in s 37M of the Federal Court of Australia Act 1976 (Cth) ("the Federal Court Act"). Section 37M provides that the overarching purpose of "the civil practice and procedure provisions" is to "facilitate the just resolution of disputes: (a) according to law; and (b) as quickly, inexpensively and efficiently as possible". (Citations omitted, emphasis in original.) 143 Her Honour also said (at [151]): The administration of justice may be brought into disrepute, in such a way as to amount to an abuse of process, if the public perception is that the legal system is unfair, inefficient, ineffective, expensive (both for the parties and in terms of the use of public monies) or contrary to the rule of law. Permitting a proceeding to continue in such circumstances might suggest tolerance of behaviour that is contrary to the just, efficient and timely resolution of disputes, including attempts to re-litigate questions already resolved. (Citations omitted.) 144 The categories of abuse of process are not closed, and in UBS AG (and even more recently in Strickland v Commonwealth Director of Public Prosecutions [2018] HCA 53 albeit in a different context) the High Court elucidated the broad scope of the circumstances that may constitute such abuse. Typically however, an abuse of process involves: (a) an illegitimate purpose; (b) unjustifiable oppression; or (c) the use of court procedure in a way that would bring the administration of justice into disrepute. The primary judge appears to have relied upon the second and third of these justifications. 145 It is not clear to us that the continuation of competing class actions necessarily constitutes an abuse of process (see Bellamy's at [36]; Cantor at [60]) and we find it difficult to see how the present case falls within any of the above categories. We are not inclined to accept that continuance of the Perera and McTaggart Proceedings gives rise to unjustifiable oppression or would bring the administration of justice into disrepute, for a number of reasons. 146 First, Part IVA contemplates that there may be more than one proceeding against the same respondent in respect of the same subject matter and the same cause(s) of action (Bellamy's at [34]). A claimant has a choice whether to bring representative proceedings on behalf of "some or all" persons. If that choice is not made, one or more claimants may bring separate individual proceedings against the one respondent in respect of the same claims or may aggregate their claims into another representative proceeding. Thus the Part contemplates the possibility of a multiplicity of proceedings. 147 Moreover, even if representative proceedings are brought, one or more group members may opt out and bring their own individual proceedings, resulting in multiple proceedings against the same respondent. Indeed, such group members who have opted out may bring their own separate representative proceedings. Nothing in Part IVA precludes such an option. And to do so would neither be invalid nor an abuse of process. In other words, the structure of Part IVA permits of multiple proceedings including multiple representative proceedings. 148 Part IVA does not evince a purpose to restrict or prevent claims involving a common issue being the subject of more than one class action and its object is to permit claims involving a common question to be litigated together rather than separately. Its object is facultative not restrictive. In permitting more efficient dispute resolution through group proceedings Part IVA does not insist on the most efficient means of dispute resolution. Put another way, not using the most efficient means of dispute resolution is not necessarily sufficient to constitute an abuse of process. 149 Second, although it is no doubt vexatious, oppressive and an abuse of process for the same applicant to bring multiple proceedings in respect of the same or similar claims against the same respondent, in form that is not the situation with which the primary judge was faced. It is not oppressive nor would it bring the administration of justice into disrepute to permit multiple class actions which are not all brought on an 'open class' basis to proceed. Part IVA of the Act contemplates that multiple class actions may be brought against the one respondent for the same cause of action. For example, group members may opt out of the Webb Proceeding, aggregate their claims and commence a fresh 'closed' class action(s) and doing so would not be an abuse of process. We do not consider it to be an abuse of process for multiple class actions to continue if all but one are closed and overlap is eliminated. 150 Third, as discussed in Johnson Tiles by Merkel J at [11] to [17], the commencement of a subsequent bona fide set of representative proceedings prior to the Court giving substantive directions in existing but overlapping representative proceedings (i.e. with overlapping group members) does not of itself establish any vexation, oppression or an abuse of process. Such is not established for the representative applicant in each of the proceedings, for they are different. And in respect of the group members in each of the proceedings in relation to the overlap, those overlapping group members are not parties as such. They have not engaged in any conduct with respect to their rights that could sensibly be characterised as amounting to vexation, oppression or an abuse of process. Of course, at a later stage, such overlapping group members can be called upon to either opt out of both proceedings or opt out of one of the proceedings. Overlapping group members ultimately cannot remain in both proceedings. 151 Fourth, no authority expressly supports the proposition that in circumstances where the duplication of group membership is eliminated in respect of multiple group proceedings involving the same subject matter against the same respondent, that nevertheless there is still vexation, oppression or an abuse of process in both such proceedings as so modified to eliminate the duplication going forward. For example, take the situation where one has two closed classes that do not overlap or one open class and a closed class, but where the closed class is excluded from the open class. Alternatively, assume that such results can be achieved by case management directions. It is difficult to see how it could be said that to then allow the two representative proceedings to go forward would constitute an abuse of process by reason of that circumstance alone. Whether allowing the continuation of competing class actions may operate against the overarching principle in s 37M is a different question. 152 Fifth, take the scenario where only one of the group proceedings is allowed to go forward with the other stayed. The unstated premise of this scenario is that the respondent will only face one set of proceedings if there is a stay of one of the proceedings but the group members in the stayed proceedings, who may otherwise have overlapped with the other group proceedings, may then opt out of the proceedings that are allowed to continue and bring their own individual proceedings. We accept that it is likely that few if any group members would choose to do so unless they are supported by a litigation funder, but putting that to one side, in such circumstances the respondent may still be faced with a multiplicity of proceedings. 153 Sixth, his Honour analysed the question of abuse of process principally through the prism of allowing multiple open class actions to continue when the Court had available to it a range of remedies to deal with multiple open class actions, including less drastic remedies than a permanent stay. The less drastic measures including closing the class in one or more of the competing actions so that there was no overlap. We do not suggest that his Honour was restricted to that less drastic choice, but we do not see the continuation of non-overlapping class actions as abusive. It is a different question as to whether allowing the continuation of open class actions may be contrary to the overarching principle. 154 Seventh, we find it hard to see it as unjustifiably oppressive for the respondent to be subjected to the risk of duplicative costs when such duplication can be materially reduced or ameliorated by case management orders. It is also difficult to see how that would be so adverse as to bring the administration of justice into disrepute. One might also ask, duplicated relevant to what scenario? No part of Part IVA explicitly or implicitly enshrines the notion that there should be only one proceeding against a respondent, as we have already discussed. Indeed Part IVA recognises that there may be duplicate proceedings against a respondent; and that would entail duplicate costs. Again, it is a different question as to whether allowing duplicative costs may be against the overarching principle. 155 Eighth, GetSwift submits that a paradigm case of vexation or oppression is the institution of multiple proceedings against a respondent in respect of the same subject matter where one will suffice to vindicate relevant rights. In Walton v Gardiner, Brennan J gave as an instance of abuse of process the institution of proceedings "which unnecessarily duplicate proceedings already pending". We accept that prima facie it is vexatious or oppressive to bring a second proceeding where an existing proceeding is on foot and the subject matter of those proceedings is the same or substantially overlaps. It has also been recognised that the fact that the parties are not identical, or the relief sought is different, does not necessarily disentitle relief under this principle. 156 However, in the present case the applicants are different and it cannot be apposite to describe the Perera or McTaggart Proceedings as an abuse of process at the time they were filed. The Webb Proceeding was not even on foot at that time. The only way in which either proceeding could be reasonably be said to be abusive is through their continuance after the Webb Proceeding was commenced and when (as the primary judge found) it was to be conducted on better costs and funding terms for group members. However we find it difficult to see such continuance as rising to the level of an abuse of process when: (a) the filing of the Webb Proceeding was outside of the control of the Perera and McTaggart applicants; (b) until the 'competitive tender' process the primary judge initiated had taken place there was no basis upon which the Perera and McTaggart applicants could conclude that the Webb Proceeding would be conducted on terms more favourable to group members; (c) the Perera or the McTaggart applicants apparently believe that their lawyers, and funding and legal costs models will be better for group members; and (d) under Part IVA discontinuance of a proceeding is only by leave of the Court rather than under the applicant's control. 157 We accept that in considering whether something is an abuse of process the public interest is to be considered including whether there is any inefficiency in the administration of justice by duplicate proceedings. But we cannot say that any inefficiency in duplicative proceedings arising in the present context would rise to such a level of inefficiency such as to warrant the description of an "abuse of process". In our view the label of "abuse" is not apposite.

(b) Was it appropriate to permanently stay two of the actions? 158 Having concluded his Honour had power to grant a permanent stay or at least a temporary stay of two of the competing proceedings (cf Sterling Pharmaceuticals Pty Ltd v The Boots Company (Australia) Pty Ltd (1992) 34 FCR 287 at 290 to 291 per Lockhart J) we now turn to address his Honour's exercise of that power in determining to stay the Perera and McTaggart Proceedings. 159 We note that his Honour's decision was made against the backdrop that, since approval of common fund orders in Money Max in October 2016 (which amongst other things removed the requirement for book building), and with the increased availability of litigation funding there appears to have been an increase in the number of competing securities class actions. This case is an example of that phenomenon as are the five competing class actions recently filed against AMP: see Wileypark. 160 Where competing class actions occur the respondent commonly applies for orders to stay all but one of the competing proceedings, and contends, as GetSwift contended before the primary judge, that: (a) the respondent should not vexed with the costs of defending more than one class action with overlapping group members in respect of the same factual matters and causes of action; (b) the respondent should not be subjected to the risk of more than one set of adverse costs if it is unsuccessful in defending the proceeding; (c) it is not in the group members' interests that more than one set of applicants' legal costs are incurred in securing any favourable settlement or in any judgment that may be achieved, as such costs will all come out of the same pool; and (d) it is in the group member's interests that the class action with the lowest legal costs and the most attractive funding commission go forward. 161 On some occasions the competing applicants (or more accurately the competing funders and/or solicitors) are able to reach agreement to consolidate the proceedings. When such agreement occurs the competing cases are subsumed into one and the associated case management problems largely disappear. However, in many cases no agreement is reached. In our view that usually has more to do with the interests of the relevant litigation funders and applicants' solicitors than with the interests of the different applicants or the group members. Notwithstanding the important role that solicitors and funders play in providing access to justice, scarce judicial time and Court resources should not be spent dealing with disputes between service providers, and the Court should be allowed to focus on the interests of the parties and group members consonant with the overarching purpose. 162 Where, as in the present case, there is no agreement between the competing funders and solicitors the Court must decide the best way to deal with the case management difficulties that are thrown up. That decision must be made in the particular circumstances of the case and commonly involves weighing up what may be incommensurable and sometimes conflicting considerations. There can be no one right answer to such questions and different judges may weigh the relevant considerations differently. That this is so can be seen in the variable approaches taken by different judges of the Court, including: (a) in Bellamy's Beach J was faced with two competing open shareholder class actions concerning the same subject matter and causes of action but with different applicants, solicitors and funders. In the circumstance that in each case a large number of group members had signed a funding agreement with a particular funder and a retainer with a particular law firm, his Honour decided not to permanently stay either of the actions and made orders to close the class in one proceeding, to allow the other proceeding to remain as an open class proceeding, and to have a joint trial of both proceedings. His Honour also made directions to minimise any duplication of work and expense by the law firms acting for the applicants in the two cases; (b) in Cantor Foster J was faced with five competing open class actions brought by two different law firms alleging that Volkswagen, Audi and Skoda engaged in misleading and deceptive conduct and made false and misleading representations about the emissions of affected diesel motor vehicles. His Honour did not accept the respondent's submission that allowing the continuation of two groups of class actions would, for the time being, unduly oppress the respondents nor that it was productive of undue costs, confusion or delay, and decided to adopt a 'wait and see' approach; and (c) in the present case the primary judge was faced with three competing open shareholder class actions with different applicants, solicitors and funders. His Honour decided that in the circumstances of the case the best option was to stay two of the proceedings, and after conducting a comparative assessment his Honour made orders for the Perera and McTaggart Proceedings to be stayed and for the Webb Proceeding to go forward. 163 In our view the decisions in Bellamy's and Cantor were clearly within the proper exercise of discretion. We take the same view in relation to the primary judge's decision to permanently stay two proceedings and allow one to go forward. Although there are aspects of his Honour's reasoning that we do not fully endorse, we can discern no material House v The King error that would warrant setting aside his Honour's decision. Moreover, even if contrary to our view the primary judge did fall into error, if we were exercising the discretion for ourselves we would also have concluded that it was appropriate to stay two of the three competing proceedings. 164 In deciding whether to exercise the power to grant a stay the primary judge took into account all the circumstances of the case, including the position with respect to each of the three open class actions, the relevant interests of justice, the interests of the respondents concerning having to deal with multiple class actions over the same matter, the interests of the applicants and group members (including the position of many of them as continuing shareholders in GetSwift), and the broader interests of ensuring that class actions are run expeditiously and in a cost efficient manner. His Honour also considered the available alternative remedies, being declassing and class closure. 165 Reasonable minds may differ as to whether the facts of a given case justify the imposition of a permanent stay, but appellate review of such a decision depends upon demonstrating material House v The King error. It would not be sufficient for us to conclude that we would have reached a different view if we had been exercising the discretion for ourselves. In any event, as we have said, we would not have done so. Moreover, the remedial response in a particular context of dealing with competing class actions depends on its own circumstances, with different cases and facts calling for different responses. It is appropriate to elaborate further by reference to the particular distinguishing features of the present context. 166 First, all applicants proposed to seek a common fund order. In Money Max the Full Court observed that common fund orders "enhance[d] access to justice by encouraging "open class" representative proceedings" (at [14]). Their Honours later observed that the central aims of the Part IVA regime were facilitated "by allowing a common, binding decision to be made in one proceeding instead of in multiple proceedings" (at [177]). With the availability of common fund orders as part of an open class action, it is unnecessary to allow the continuance of duplicative open class actions so as to allow access to justice to group members and to protect group member's interests. 167 Three open class proceedings were commenced in the present case, and the primary judge was required to devise an appropriate remedial response to, amongst other things, GetSwift's submissions that; (a) it was being vexed with three class actions based on the same factual matters, for the same causes of action and covering the same group members when one class action would be sufficient to vindicate the group members rights (if any); (b) it would suffer increased legal costs as a result and face a risk of increased adverse costs; (c) group members would incur more than one set of legal costs when that was avoidable; and (d) the continuation of the competing proceedings created a risk of delay and wastage of the parties' and the Court's resources. 168 Beach J dealt with a similar situation in Bellamy's and his Honour said that there were only two realistic options available, namely, a permanent stay of one of the proceedings or closing the class in one of the proceedings and allowing one open class and one closed class proceeding to go forward side-by-side. His Honour adopted the second option because in one proceeding there were over 1,500 group members who had entered into a funding agreement with the relevant funder and signed retainers with the relevant solicitors, and in the other proceeding over 1,000 group members had entered such agreements. But for that circumstance, his Honour said that he would have ordered a permanent stay of one of the proceedings. 169 In the Perera and McTaggart Proceedings a much lower number of group members had entered into funding agreements. Further, Bellamy's was the first occasion when such an approach was taken and Beach J observed (at [97]) that his reliance on the number of group members who had signed up in each case "should not be seen as an incentive to sign up group members before issue". Yet we have no doubt this is what the funders and/or solicitors did in the Perera and McTaggart Proceedings. They both rapidly sought to sign up as many group members they could despite it being the intention to seek a common fund order and in our view that was unnecessary. 170 Second, the circumstances of the present case were also different from those in Bellamy's in which each of the applicants sought orders to stay the other case, but (because each proceeding had a sufficient number of signed up group members to make such an exercise economic from a standalone closed class perspective) as a fallback position they were content to proceed on a closed class basis. In the present case neither the Perera nor McTaggart applicants unequivocally said that if their proceeding was not chosen as the open class proceeding to go forward they were content for their proceeding to go forward as a closed class rather than be stayed. 171 In submissions both the Perera and McTaggart applicants criticise the primary judge for not sufficiently considering alternative remedies, particularly class closure. Indeed, the McTaggart applicants submit that the primary judge addressed the wrong question, namely, whether "duplicative open class actions should proceed" and failed to sufficiently address class closure as one of the options. We disagree: (a) first, it is unsurprising that the primary judge should have framed the question in terms of whether duplicative open class actions should proceed when that was the situation with which he was confronted; (b) second, the primary judge did consider the matter on the basis that it would not be appropriate to "allow more than one open class proceeding, even if one or more were closed, to continue" (at [307]). Further, the primary judge's reasoning is properly to be read as addressing not just whether there should be only one open class action, but also whether there should be only one continuing action, open in form; and (c) third, the Perera and McTaggart applicants did not squarely put that they wished to continue otherwise than as an open class action. They led no detailed evidence to so assert and they did not unequivocally submit that they wished to continue otherwise than as an open class action. 172 We accept that Mr Patrick Coope, the Regional Managing Director of Vannin which funded the McTaggart Proceeding, gave evidence that if the class in that proceeding was closed, it was his "expectation that Vannin will carefully consider the viability of its ongoing funding of the action". But as GetSwift submits, in the hearing below those representing the Perera and McTaggart applicants knew how many people had signed funding agreements in their respective proceedings, who they were, their share acquisitions in the relevant period and therefore the potential value of their claims. Importantly, they knew that one possible outcome from the hearing was class closure. We accept GetSwift's submission that the Perera and McTaggart applicants therefore each knew all they needed to know to decide whether they wished to proceed on a closed class basis, should they miss out on being the one open class action to be allowed to proceed. Despite that each made a forensic choice and did not tell the primary judge in unequivocal terms that, as an alternative, they were content to proceed as a closed class action. As GetSwift colloquially explained it, the Perera and McTaggart applicants each say: "I want to be the open action; I should be preferred for the first prize over the other applicant and over Mr Webb for so-and-so reasons. One or both of them can have a closed action, if they want; but I am not prepared to undertake to rank second fiddle and be the closed action to someone else's open one". 173 Before us the same problem has emerged. The McTaggart applicants complain that the primary judge gave insufficient consideration to the class closure option, but they again only argue for the closure of the Perera Proceeding and for the declassing of the Webb Proceeding. Mr Perera argues for remedies which would allow the different proceedings to continue in various ways, but ultimately wants the McTaggart Proceeding to be closed and the Webb Proceeding to be declassed. Neither Mr Perera nor the McTaggart applicants unequivocally submit that they are content to proceed with a closed class action or put on evidence to show that such a proceeding would be viable, if they are unsuccessful in being selected as the proceeding to go forward. 174 The submission that the primary judge did not sufficiently consider the alternative remedy of closing the class in either the Perera or McTaggart Proceedings is misconceived. Neither the Perera nor McTaggart applicants put that proposition in a way that was practical or viable, either before the primary judge or before us. 175 Third, we do not accept that a stay of the Perera and McTaggart Proceedings is inconsistent with the interests of the applicants and group members. No evidence was adduced by the Perera and McTaggart applicants suggesting that their interests or the interests of group members more generally would be best served by more than one open class action progressing (or as we have said, by their proceeding going forward as a closed class). No direct evidence was adduced before us by any applicant or group member that they are prejudiced or aggrieved in an identified and explained manner by the orders made by the primary judge. 176 The Perera and McTaggart applicants submit that the primary judge failed to give any or any sufficient weight to the funding agreements and retainers that the applicants and funded group members had entered into. That submission is misconceived. The primary judge did consider this matter but concluded that it did not warrant significant weight. We do not consider that conclusion involves error, and in any event a complaint as to weight does not give rise to a House v The King error. 177 Moreover, both the Perera and McTaggart applicants intended to seek a common fund order which would largely supersede any funding agreements in any event. The parties accepted that before the primary judge. As such, the number of group members who had signed a funding agreement or retainer was not a significant matter to the exercise of his Honour's discretion. We can see no House v The King error in that. 178 Amongst other things, it was appropriate for his Honour to treat the number of group members who signed up to funding agreements (or retainers) as a relevant but not significant consideration when: (a) the level of 'choice' exercised by small claimants that sign up to a funding agreement or retainer was overstated. There was little in the way of evidence to show that any such decision was the product of an informed choice between competing offers. More specifically, the evidence does not show that when a group member signed a funding agreement or retainer they were even aware of the alternative offers of funding or legal representation such that it represented a preference for one solicitor or funder over another. A well-established class action law firm or funder may have a relationship with or be attractive to a repeat claimant in class actions such as an institutional investor, but other than that book building is to a large extent just marketing driven. Further, the first solicitor or funder into the market will sign up a number of group members before there is any competition; (b) as observed in Money Max, there is often a significant information asymmetry between the funder and the group members in relation to the costs and risks associated with the action and for small shareholders the opportunity for negotiation of the funding commission is limited or non-existent. Many retail shareholders will understand the headline funding rate but not the important differences between funding agreements that are offered. For example some funding agreements: (a) transfer costs ordinarily paid by a funder (including the cost of adverse costs cover) onto group members; (b) provide for the applicant's solicitors to conduct part of the case on a conditional fee basis costs and upon success the solicitors therefore become entitled to a 25% uplift fee on the unpaid amount; and (c) remove or reduce important rights such as the right to opt out by putting that decision in the hands of the applicant's solicitors or the funder; (c) in the McTaggart Proceeding, unlike in Bellamy's, the funded group members had not entered a retainer with Corrs and the number of such persons were not large compared to the number in Bellamy's; and (d) as Beach J observed in Bellamy's (at [97]) solicitors and funders should not be incentivised to sign up group members before proceedings are issued. As the primary judge said (at [210] to [213]) in circumstances where a common fund order is being sought, having group members sign funding agreements serves "no useful purpose", involves a "waste of costs" and undermines the policy objectives of Part IVA in having 'open' class actions. 179 Further, the interests of the approximately 300 funded group members cannot weigh as much in the balance as those of participating group members overall (which number is presently unknown but is many times more than 300 persons). If, as his Honour decided, it was less costly, less wasteful for the parties and the Court's resources and more efficient for one class action to go forward to vindicate class member's causes of action rather than three, that is in the interests of all participating group members, and their interests trump the interests of the far fewer Funded Perera and McTaggart GMs. 180 Fourth, the Perera and McTaggart applicants and funded group members are not shutout from the pursuit of their claims by the stay granted. As the primary judge recognised, in an ordinary case where a stay is ordered a person may be shut out from pursuing a claim but the position is different in a class action. The Perera and McTaggart applicants and the funded group members in the Perera and McTaggart Proceedings are also group members in the Webb Proceeding and they can vindicate their claims through that proceeding. If they do not wish to pursue their claims through the Webb Proceeding they can opt out and do so individually or aggregate them into another (closed) class action. If they consider Mr Webb is not adequately representing their interests they can make an application under s 33T of the Act to substitute another group member as the representative party. 181 For completeness, we would note one minor technical matter at this point. We do not treat the permanent stay orders as preventing the Perera or McTaggart applicants from pursuing their personal claims. If it is necessary to do so, we would modify the stay orders accordingly or simply allow them to opt out of the Webb Proceeding and institute new personal claims. If necessary we will hear further submissions from counsel on that point. 182 Fifth, it is in the interests of the applicants and group members in the Perera and McTaggart Proceedings to have their claims advanced in the proceeding that will be the most potentially beneficial vehicle in which to advance such claims. If the most potentially beneficial vehicle is the Webb Proceeding, then their interests are appropriately protected. The primary judge decided that the Webb Proceeding was the most potentially beneficial vehicle and, having made that determination, in making the stay orders that his Honour did, the primary judge acted consistently with the Court's protective role. 183 Sixth, it is not in the interests of the applicants and group members for GetSwift's assets to be unnecessarily depleted by having to defend more than one class action. Indeed, Mr Perera's solicitor, Ms Amanda Banton, gave evidence that "the Applicant is concerned that the continuance of two open class actions will result in delay and additional cost, in circumstances where, for the reasons outlined earlier in this affidavit, there is a need to advance this proceeding efficiently given the asset position of GetSwift". 184 Seventh, it is plainly in GetSwift's interests that it only incur the cost of defending one proceeding rather than three, although we accept that the duplication of expenses can be reduced by appropriate case management directions. It is also in GetSwift's interests that it only face the risk of an adverse costs order in one proceeding. Thus it is in the respondents' interests that two of the three competing proceedings be stayed. For the respondents to be required to defend three open class actions in respect of the same subject matter when one will adequately serve to vindicate the group member's interests gives rise to inefficiency and additional expense. 185 Mr Perera submits that there was no evidence before the primary judge to support any finding as to any increased costs that would be faced by the respondents through more than one class action. We found little force in this contention. First, we were informed that at an early case management hearing it was made clear that such evidence would not be necessary and no party took issue with that. Second, it is so obvious that if three class actions go forward instead of one the respondent will incur additional costs that it did not require evidence. Of course, steps can be taken to lessen or ameliorate this inefficiency and additional expense (Bellamy's at [112] to [118]) but they are unlikely to eliminate the problem. 186 Even so, as we have said Part IVA does not enshrine the notion that a respondent may only face one class action. The provisions of Part IVA expressly recognise that a respondent might face multiple actions and necessarily incur duplicated costs, and even two class actions against a respondent may be constitute a costs saving compared with the costs in multiple individual proceedings in different jurisdictions. 187 Eighth, we accept that the broader interests of justice support a permanent stay of two of the proceedings. The primary judge concluded that, in the circumstances of the present case, it was not in the interests of the administration of justice for three open class actions to continue on foot. Mr Perera criticises the primary judge's observations at [347] but: (a) each applicant having submitted that only one open class action should go forward; and (b) the primary judge having concluded (correctly in our view) that all the applicants' and group members' claims can properly be vindicated within one of the three open class actions, and that the Webb Proceeding was the most potentially beneficial vehicle for those claims to be advanced, it was open to the primary judge to conclude that the continuation of three class actions, whether open or closed, would not be in the interests of justice in the particular circumstances of the case.

A carriage motion? 188 The primary judge also discussed in some detail the way foreign jurisdictions have dealt with competing class actions. For present purposes it is only necessary to discuss the Canadian experience. 189 We have no difficulty in accepting that an application to stay one of several competing class actions can be seen as analogous to the Canadian 'carriage motion', and we have no difficulty in accepting that in some cases of competing class actions it will be appropriate to follow that course. It is though, important to keep in mind that there are some important differences, as Beach J discussed in Bellamy's at [49] et seq. 190 Class actions in Canada are usually filed in the provinces, in their provincial superior courts, rather than in the Federal Court of Canada. In a province that is a common law jurisdiction, one or more of the competing class representatives may bring a carriage motion seeking a stay of all the other present or future class actions relating to the same subject matter. As Professor Morabito explains in "Clashing Classes Down Under - Evaluating Australia's Competing Class Actions Through Empirical and Comparative Perspectives" (2012) 27 Connecticut Journal of International Law 245 at 282, the starting point is that there ought not to be multiple class actions for the same class asserting the same cause of action: see Vitapharm Canada Ltd v F. Hoffmann-La Roche Ltd [2000] OJ No. 4594 (Vitapharm) at [14] per Cumming J. 191 Such a carriage motion can only deal with competing class actions in the one province, and cannot deal competing class actions in different provinces. Overlapping class actions filed in different provinces create issues analogous to overlapping class actions in the Federal Court of Australia and the Supreme Courts of those States which have Part IVA-type regimes. In any event, non-overlapping class actions filed in different provinces are not perceived to give rise to any valid objection per se: Tiboni v Merck Frosst Canada Ltd (2008) 295 DLR (4th) 32 at [40] per Cullity J. 192 It is also relevant that Canada has a certification procedure. A carriage motion (or its analogue, a stay application) may be agitated either before or after certification. There have been instances where before a carriage motion has been heard, the party resisting such a motion has raised a preliminary issue as to whether the provincial court has jurisdiction to grant or deny carriage before certification: see Nelson v Merck Frosst Canada Ltd (2006) 61 BCLR (4th) 157 (Nelson), Joel v Menu Foods Genpar Ltd (2007) 78 BCLR (4th) 112 (Joel), Grasby v Merck Frosst Canada Ltd (2007) 216 ManR (2d) 117 (Grasby). It appears that a carriage motion sourced from provincial class proceedings legislation (for example, in Ontario ss 12 and 13 of the Class Proceedings Act, SO 1992, c 6, but s 13 having no direct Part IVA analogue) can be granted by a court only after certification (Nelson at [29]). 193 Under provincial class proceedings legislation, a carriage motion is granted in respect of a "class proceeding", that is, a proceeding that has been certified: Nelson at [20] to [22] and Grasby at [15] to [16]. As such, the source of the power to so grant derives from the relevant class proceedings legislation, but only after certification. But before certification, the court has the power to stay founded in its inherent jurisdiction (Joel at [38], Nelson at [29], Grasby at [25]) and such inherent jurisdiction does not require certification as a pre-condition to its exercise. In substance, the court has the power to permanently stay competing potential or actual class proceedings before or after certification. 194 If a stay is granted prior to certification, the court's reasons for doing so may be to avoid competing certification proceedings which may be lengthy and costly: see, for example, Giles v Westminster Savings Credit Union [2002] BCJ No. 2567; 2002 BCSC 1583 at [46]. Generally, the factors relevant to pre-certification stay applications and carriage motions overlap. 195 A carriage motion is usually brought by a representative applicant rather than the respondent. On such a motion the predominant theme underpinning the decision as to which proceeding goes forward is what is in the best interests of all group members, although fairness to the respondent is a consideration: Locking v Armtec Infrastructure Inc 2013 ONSC 331 at [7] and [8] per Molloy J. The primary concern is to determine which of the competing actions is more likely to advance the interests of the class: Mancinelli v Barrick Gold Corporation (2015) 126 OR (3d) 296 at [18] per Harvison Young J. The relevant factors to be considered and balanced in a qualitative appraisal include: (a) the nature and scope of the causes of action advanced; (b) the theories advanced by counsel as being supportive of the claims advanced; (c) the state of each class action including preparation; (d) the number, size and extent of involvement of the proposed representative applicant; (e) the relative priority of commencing the class action; (f) the resources, experience and competence of counsel; and (g) the presence of any conflict of interest. (Sharma v Timminco Ltd (2009) 99 OR (3d) 260 at [17] per Perell J referring to Vitapharm at [49]). 196 The Canadian experience and analogue is informative, particularly in relation to the considerations that apply when a choice is to be made between competing proceedings, but the distinctions are important. Under Part IVA there is no certification process; representative proceedings can be issued as of right and continue provided that the conditions in ss 33C and 33H are satisfied and subject to no s 33N order later being made. Each representative proceeding is presumptively entitled to go forward, absent vexation, oppression or an abuse of power, and if one or more of the proceedings is to be stayed some significant reason must be demonstrated. In the Canadian provinces there is no presumptive entitlement to go forward until certification and the Canadian provincial courts accordingly have much greater freedom to choose, in the one province, which class proceeding is to go forward and which others are to be stayed. 197 In circumstances where a substantial number of group members have signed up to funding agreements the carriage motion analogue has less application. For example, in Bellamy's it was plainly open to Beach J to refuse to permanently stay one of the proceedings, as to do so would substantially affect the contractual funding and retainer arrangements of over 1,000 group members in whichever proceedings were stayed. No Canadian carriage motion context has had to deal with such a scenario. Of course in the present case there was not the same concern or at least the same magnitude of the problem with funded group members. His Honour distinguished the position before him with the scenario in Bellamy's, and he was correct to do so.

THE CHOICE OF THE WEBB PROCEEDING 198 We now turn to address to address the primary judge's choice of the Webb Proceeding as the class action to go forward in preference to the Perera or the McTaggart Proceeding.

Reduction of legal costs in the Webb Proceeding 199 The McTaggart applicants argue the primary judge erred in finding that the applicant in the Webb Proceeding had proposed innovative ways of seeking to reduce legal costs (at [328]). His Honour's finding was based upon two matters, namely, the proposal to have costs periodically assessed by an independent referee and the willingness of the Webb applicant to be open to a court appointed or jointly engaged expert. They argue that neither of these was a matter on which the primary judge could properly place any weight. 200 On their submissions the Webb applicant's proposal that costs be periodically assessed was not an innovative way of seeking to reduce legal costs, and instead it was a feature designed to mitigate the moral hazard consequent upon the Webb funding model. They contend that in the ordinary course a litigation funder has an interest in ensuring that legal costs be kept to a reasonable level, whereas under the Webb funding model the opposite was the case. Under the Webb funding model the funding commission is in part based on a multiple of the legal costs (2.2 times or 2.8 times depending on when the case is resolved) and therefore an increase in legal costs may lead to an increase in the funder's commission. The McTaggart applicants argue that to treat the existence of a costs referee as a positive feature, rather than a necessity of the model, was an error. Further, they argue that there was no evidence that six-monthly costs assessments by a referee would reduce legal costs any more than they would be by having a funder incentivised to minimise costs and the applicant being entitled to have costs assessed. 201 As to the primary judge's reliance on the use of a court appointed or jointly engaged expert, the McTaggart applicants argue that was misplaced for two reasons. First, they say that there was no evidence that GetSwift would agree to jointly appoint an expert, and it was too early in the proceedings to draw any conclusion as to whether the Court will order a single expert over the objection of GetSwift. That is, that there was no basis to find that a single expert was likely to be used in the case. Further, they contend that even if a single expert is appointed, it is far from clear that that would lead to an overall reduction in costs. On their submissions it is not unlikely that it would actually increase costs, as each party may feel compelled to retain an expert to assist in challenging any adverse views of the court expert, so that the result would be the employment of three experts not two. Second, they argue that in this respect the proceeding below was affected by procedural unfairness. The applicants in the McTaggart and Perera Proceedings were ordered to put on evidence and submissions in relation to, amongst other things, whether they proposed that a court appointed or joint expert be appointed or retained in relation to any issue. They were ordered to do so in circumstances where neither was to be aware in advance of the position of the other. They say, however, that the Webb applicant was given advance notice by the primary judge of the fact that neither the Perera or McTaggart applicants had put anything forward in that regard.

The partly cost-based funding structure in the Webb Proceeding 202 The McTaggart applicants argue that the primary judge's preference for the partly cost based funding structure in the Webb Proceeding was based upon six matters (set out at [285] to [294]). They contend that none of these matters provides an appropriate basis for concluding that the Webb funding model was superior to the McTaggart funding model. 203 First, the primary judge considered that the Webb funding model "recognises the reality that the risk of a funder increases incrementally as legal costs increase" (at [285]). But as the primary judge later recognised at [294], so too does a funder's commission expressed as a percentage of the resolution sum which increases over time, as legal costs increase. 204 Second, the primary judge considered that basing the funding commission on a multiple of costs increased the Court's ability to be sure that a funder was not receiving above average returns. However the McTaggart applicants contend that there was no evidence on which the Court could conclude that the Webb Proceeding's cost-multiple funding model would prevent above average returns being achieved. They argue that it is difficult to see how any such assessment could be made without knowing the level of risk involved in the proceeding. 205 Third, the primary judge considered that the Webb funding model better reflects that "litigation funders are promoting a particular type of commercial enterprise, which is the provision of legal support and hence the funders have become indirectly engaged in the provision of legal services to a client" (at [288]). The McTaggart applicants contend that there is no reason to think that this conclusion reflects any tangible benefit to group members. 206 Fourth, the primary judge considered that the Webb funding model would "prevent windfalls" (at [289]). The McTaggart applicants however note that, without utilising the Webb funding model, the Court will prevent windfalls because it will only approve reasonable funding charges pursuant to the proposed common fund order. They say that even if the funding rate is set early in the proceedings any such order is interlocutory. 207 Fifth, the primary judge considered that the alternative form of remuneration under the Webb funding model (the lesser of 20% or the costs-multiple) would guard against disproportionate recovery if the proceedings were to settle for a low amount. The McTaggart applicants say that is also true of the funding model in the McTaggart Proceeding, which operates on a percentage of proceeds. In any event, they note that the Court has the power to ensure that there is no disproportionate recovery by only approving reasonable funding charges as part of settlement approval. 208 Sixth, the primary judge considered that a "fixation on 'headline' percentage rates may lead, in the future, to a form of reverse auction where common fund proposals are not struck by reference to mature assessments of risk and proportionate and commensurate return, but by competitive undercutting by press release or twitter" (at [291]). The McTaggart applicants submit that there was no sound basis for the primary judge to draw this conclusion. They say that if early-stage 'auctions' as occurred in the present case become common in competing class actions, then it is obvious that funders will seek to undercut one another. They contend that there is no reason to think that funders will be any less willing to do so on a cost-multiple model than a percentage based model.

Modelling of comparative returns to group members 209 The primary judge concluded that the Webb Proceeding is "very likely, in most scenarios at all stages of the proceedings, to produce a better return for group members" (at [329]). That conclusion was based on modelling done by the parties and in particular the modelling of the Webb applicant which the primary judge set out as Annexure A to the judgment. The McTaggart applicants argue that the modelling process was unsound for various reasons. 210 First, they contend that it became apparent in the course of the hearing below that the quantum of security for costs would be a crucial input into any model. In particular this was because the expense of giving security would form part of the legal costs on which the cost-multiple in the Webb funding model would be calculated. The primary judge directed that the parties prepare models based upon $2.4 million in security being required, in three tranches. However the McTaggart applicants contend that there was no evidence as to the likely quantum of security for costs and that was little more than guesswork. Nor was there any basis to assess whether security would be given by way of cash or ATE insurance, or what the cost of a $2.4 million ATE insurance policy would be. They say that the only evidence as to the cost of ATE insurance concerned a much larger policy. 211 Mr Webb admits that there was no evidence about the likely quantum of security for costs, but submits that it was reasonable for the primary judge to make an estimate based on his own experience in class actions. The McTaggart applicants say that the likely quantum of security for costs was not a matter of which the primary judge could take judicial notice, and that even the broad brush approach adopted in security for costs applications is based upon evidence that can be tested and assessed. 212 Second, the primary judge adopted modelling on the basis of a large number of scenarios including a mediation occurring at various different points in time. His Honour did so because the parties and their funders had entered into arrangements whereby different commissions were payable depending on when the matter concluded. However, the McTaggart applicants submit that this was not known in advance among the parties, and the costs estimates provided by the parties were not sufficiently "granular" to enable reliable estimates of costs to be made at the different points in time. According to the McTaggart applicants, this lead to absurdities in the modelling. 213 Third, the primary judge relied upon modelling on the assumption that the matter would settle "in the very broad range of between $15 million to $300 million" (at [300]). The McTaggart applicants submit there was no basis to make that assumption. They say that Ms Banton gave uncontradicted evidence that the total value of group members' claims was about $74.5 million, and that any scenarios which assumed a recovery greater than the likely settlement range based on that claim value should have been discounted. However, we would note that Ms Banton's estimate was made before the share price had completed its decline and may well have been an underestimate. 214 Fourth, the McTaggart applicants submit that an assumption that the cases to be run by all three sets of applicants were substantially similar and accordingly that the overall costs would be similar, underpinned the modelling. In the case of the Webb Proceeding, which had no pleading, that assumption was based upon evidence from Mr Finney, and the McTaggart applicants submit that evidence was subsequently proved incorrect. They note that on 1 June 2018 a Statement of Claim was filed in the Webb Proceeding which (at 164 pages) is substantially longer than the pleadings in the McTaggart (59 pages) or Perera (50 pages) Proceedings. They contend that the Webb pleading makes numerous additional allegations of contraventions to those in the other proceedings and is likely to significantly expand the ambit of factual enquiry in the proceedings. They argue that this will inevitably have consequences for costs, particularly in relation to discovery and evidence. Accordingly, they contend the comparative exercise undertaken by the primary judge proceeded on a false premise. In addition, they argue that the comparative exercise made no allowance for the improved prospects of recovery through the McTaggart Proceeding because it also made a claim an additional respondent, Mr Hunter. 215 Fifth, the McTaggart applicants submit that the primary judge's assertion that the modelling in Annexure A produced 784 scenarios, the vast majority of which showed the Webb Proceeding producing the highest return the group members, gave equal weight to scenarios in the modelling which are irrelevant as it does to those that are relevant. 216 Annexure A contains 28 tables, each of which contain 28 scenarios at different recovery levels. The McTaggart applicants submit that 18 of the 28 tables (those with the top lines shaded grey) are, on Mr Webb's own case, irrelevant. That leaves 10 tables (Tables A to J) and two of these (Tables E and J) model outcomes at trial that are very unlikely to occur. Further, they say the remaining 8 tables (Tables A to D and F to I) model outcomes on settlement at different stages of the proceedings and it is in respect of those tables that the problem appears. 217 Six of the tables model "early stage" mediation at different points in time, but the costs assumed for each proceeding up to mediation are the same, notwithstanding that the mediations are assumed to occur at different points in time. Those tables also model recoveries up to $300 million, when the only evidence in the proceedings as to the claim size was that it was $74.5 million, and any settlement particularly at an early stage would likely be at a significant discount to that amount. The McTaggart applicants note that Mr Webb criticised this evidence, but submit that it was the only evidence any party produced as to the likely claim value and argue that it was unnecessary to model a $300 million scenario because of a concern to avoid "windfalls". They say that the Court would in any event avoid a windfall to a funder when it came to approving a settlement under s 33V of the Act or making a final common fund order on judgment. 218 Sixth, the McTaggart applicants generally submit that the modelling was unsound. They say it was constrained by the limitations on the parties' capacity to adduce further evidence in the context of the "blind auction" process set up by the primary judge. They contend that in a blind auction process, the first of its kind in Australia, in which the primary judge did not direct the parties in advance to produce evidence that would support comparative modelling on any particular parameters, the evidence produced by all parties was not apt to support reliable modelling. They argue that the proper way to undertake the exercise would have been to: (a) receive evidence from the respondents as to their likely costs; (b) receive evidence about the likely quantum of the claims; and (c) on the basis of common assumptions specified in advance as to matters such as the volume of discovery and numbers of interlocutory applications, ask the parties to model returns to group members based upon the likely range of settlement figures for a mediation at specified stages of the proceedings. They submit that only an orderly process of that kind could produce reliable modelling and such a process did not occur. 219 Generally, they submit that any comparative modelling of expected returns to group members must be done with considerable caution (see Bellamy's at [87] to [88]), and argue that, in any event, the results may offer no more than a guide to the range of possible outcomes for group members (Bellamy's at [88]). 220 In light of the matters outlined above, the McTaggart applicants say that it was not open for the primary judge to conclude that the Webb Proceeding was "very likely" to produce a better return for group members.

Other matters not taken into account 221 The McTaggart applicants also submit that the primary judge failed to take into account a number of other matters which favoured the McTaggart Proceeding over the Webb Proceeding as a preferable vehicle for the resolution of group members' claims. 222 First, there was detailed evidence that the funder of the McTaggart Proceeding was of means, and could provide security for costs and meet adverse costs orders. They say, however, that there was no such evidence in relation to the funder of the Webb Proceeding, Therium Litigation Finance (Australia) Ltd (Therium Australia) which despite its name is incorporated in England and Wales. They note that Therium Australia is a $2 company and the only evidence adduced as to its ability to fund the proceedings were two letters from Therium Australia stating that another entity in the Therium group which was a Jersey-based company would lend it certain funds. There was no evidence about the financial position of that entity. The primary judge said that there was "no evidence to suggest otherwise" than that the Therium group had the necessary funds to conduct the litigation but the McTaggart applicants contend that was no basis for a positive finding that Therium had sufficient means. 223 Second, they submit that there was no funding agreement between Therium and Mr Webb, with the result that there was no mechanism to resolve any conflict between Mr Webb and Therium as to how the litigation should be conducted. The primary judge reasoned that this would be resolved through the making of a common fund order, but the McTaggart applicants contend that reasoning begs the question as to what mechanism would resolve any conflict between Mr Webb and Therium as to the appropriate terms of a common fund order. They say that the primary judge erred in ignoring this as a relevant consideration and a reason not to prefer the Webb Proceeding. 224 Third, they submit that the Webb funding model introduced a conflict of interest between Therium and the group members. On the Webb model, Therium's return would be capped by the cost multiple, giving it no financial incentive to negotiate for a higher settlement, once its maximum return under the costs multiple was reached. The primary judge rejected this submission on the basis that the lawyers for Mr Webb would be bound by their duties to negotiate the best return possible and would "do their job" (see [220] to [223]). However, the McTaggart applicants contend that analysis does not take account of practical reality. They did not doubt that Mr Webb's lawyers would "do their job", but say that there are a range of settlement figures in any case that can be considered reasonable, and reasonable minds may differ as to what is the best settlement that can be achieved. They submit that what matters is not that those with duties "do their job", but that negotiation outcomes under the Webb model may be distorted by the fact that one side does not have an incentive commensurate with the other. 225 Fourth, the McTaggart applicants contend that there were forensic advantages to their proceeding which were wrongly ignored by the primary judge, including an additional respondent, Mr Hunter, a more advanced state of preparation, and a cogent direct reliance case. 226 They contend that the primary judge's error in ignoring the additional respondent was not superficial because Mr Hunter has substantial assets. They argue that the primary judge set to one side the more advanced state of preparation in their proceedings on the basis that his Honour did not want to criticise "the orderly progression of the proceeding and the finalisation of funding terms, prior to pleading work being undertaken [in the Webb Proceeding]" (at [174]). They submit, however, that it was not being critical to observe that two proceedings (the Perera and McTaggart Proceedings) had pleadings filed and one proceeding (the Webb Proceeding) did not. 227 The McTaggart applicants submit that although the primary judge referred to their direct reliance case at [208] to [209] his Honour did not adequately address it. They contend that the primary judge ought to have found that there was tangible utility in having a representative applicant with a direct reliance case. They note that market-based causation is yet to be accepted at an intermediate appellate level and group members with direct reliance claims may have better prospects of success than those without. They also argue that the prospect of establishing claims of group members with direct reliance cases may also be useful in the context of settlement negotiations.

(b) Mr Perera's submissions 228 Mr Perera similarly argues that most of the features which his Honour found favourably distinguished the Webb Proceeding involved errors of approach. Again, there was substantial crossover with the submissions made by the McTaggart applicants and we will only set out some additional submissions, or similar submissions with a different emphasis, to those argued by the McTaggart applicants. 229 First, as the primary judge found (at [186]), Mr Perera was the only applicant who was in a position to put up cash as security for costs, and who did not propose to obtain an ATE policy at cost to group members. Mr Perera submits that his Honour erroneously reasoned, without any evidence, that it was appropriate to proceed on the basis that the other funders could and would put up cash too (at [198]). He says that the primary judge failed to take into account that Mr Perera's proposal was the only one which involved certainty that cash would be put up, and that group members would not be visited with significant additional costs on account of ATE. 230 Second, Mr Perera submits that the primary judge erred in finding that it was irrelevant that no group member in the Webb Proceeding had signed funding agreements with Therium, and that the number of group members who had signed-up to the Perera Proceeding was "small", and in distinguishing Bellamy's on this basis (at [216] to [217]). Mr Perera argues that to say something is small begs the question of what one is comparing it to. What was important in Bellamy's was that the sign-up was relatively equal between the two actions whereas in the present case, the sign-up in the Perera Proceeding dwarfed that of the Webb Proceeding, and by claim value was greater than the McTaggart Proceeding too. 231 Third, Mr Perera says that the primary judge erred in finding that differences in the funding structure in the Webb Proceeding made it more favourable than the Perera Proceeding. Further, he argues that differences in the structure of funding terms are logically irrelevant in circumstances where the Full Court in Money Max held that the Court will approve reasonable funding charges at the stage of settlement approval pursuant to s 33V. At that point, the Court could approve funding charges expressed either as a percentage of recovery or as a multiple of fees. 232 Fourth, Mr Perera contends that the primary judge erred in distinguishing between the Perera Proceeding and Webb Proceeding on the basis that Mr Webb had proposed innovative ways to reduce legal costs by appointing a costs referee, when there was no reason that the Court could not have made such orders in the Perera Proceeding, and no reason to think that this exercise would actually save costs given the cost of undertaking it. 233 Fifth, Mr Perera says that the primary judge wrongly found that the Webb Proceeding was likely to produce better returns for group members. He argues that the primary judge's assessment failed to take into account; (a) that the only probative evidence of the approximate claim size of the group was that put forward by Mr Perera; (b) Exhibit MFI C which compared the modelled outcomes on the actual proposals put by each applicant in their evidence as to commission and legal costs, both of which were integers in the equation; and (c) scenarios in the modelled outcomes his Honour relied on (Annexure A) which did not suggest that the Webb Proceeding was more favourable. 234 To elaborate, Mr Perera submits that the primary judge erroneously proceeded on the basis that it was appropriate to consider only the modelled outcomes contained in Annexure A in circumstances where that involved: (a) permitting Mr Webb and the McTaggart applicants to depart from the projection of legal costs and ATE costs they had put forward as their best estimates of the costs they would each incur in conducting the proceeding; (b) failing to take into account Exhibit MFI C, which contained comparative modelling of the outcomes on the actual proposals as to funding commission and costs put forward by each of the respective applicant in their evidence which reflected how they proposed to run the litigation ([260] to [263]); and (c) erroneously finding that Therium could obtain ATE insurance at the same cost as the funder of the McTaggart Proceeding ([268] to [275]). 235 Mr Perera argues that the primary judge erroneously proceeded on the basis that Mr Webb and the McTaggart applicants could depart from their budget of legal costs and ATE costs, while refusing to permit Mr Perera (or more accurately, the funder of that proceeding) to put on a revised offer of funding terms which matched that put forward by the McTaggart applicants or Mr Webb. 236 Further, he says that the primary judge erred by failing to take into account that the only evidence of the approximate claim size of all group members was the estimate tendered by Mr Perera which had been performed by an expert, namely, $74.5 million, and thereby erroneously failed to find that the settlement of the proceeding would likely be for a fraction of the sum of $74.5 million. He argues that the primary judge should have considered that modelled outcomes that assumed higher settlement sums than a reasonable percentage of $74.5 million were unrealistic and ought not to have been considered as relevant to the likely outcome for group members. 237 Further, Mr Perera contends that having found that it was more likely that any settlement would occur well into 2019 and not earlier, the primary judge erroneously failed to find that the modelled outcomes in Annexure A (if they were appropriate to be considered) which assumed earlier settlements ought not to have been considered as relevant to the likely outcome for group members. 238 Further, he says that the primary judge erred in failing to find that the remaining modelled outcomes in Annexure A did not clearly demonstrate that the Webb Proceeding was very likely, in most scenarios at all stages of the proceeding to provide a better return for group members than the Perera Proceeding, but rather demonstrated that in many of those scenarios the Perera Proceeding was likely to provide better or materially similar returns to group members. 239 Further, he argues that the primary judge erred in failing to find that on the basis of Exhibit MFI C, which contained comparative modelling of the outcomes on the actual proposals as to funding commission and costs put forward by each of the respective applicants that reflected how they proposed to run the litigation, the Perera Proceeding was likely to provide better outcomes to group members in respect of likely settlement scenarios than the Webb Proceeding. 240 In summary Mr Perera submits that had the primary judge not made these errors, his Honour would not have chosen the Webb Proceeding as the proceeding permitted to advance as the only class action against the respondents.

(c) Analysis 241 The primary judge's decision to prefer the Webb Proceeding was an exercise of discretion which involved evaluating competing criteria, and his Honour had put in place a process to choose between the competing class actions on the basis of such criteria. Where such an evaluative exercise is involved, upon which reasonable minds might differ and where there is no one correct conclusion, it is not enough that we may have a preference for a different view to that taken at first instance. Error must be shown. 242 In our view no error has been demonstrated in the primary judge's selection of the Webb Proceeding. 243 The selection of the Webb Proceeding followed the primary judge's consideration of the likely returns to group members under the competing funding proposals. Each of the applicants submitted comparative modelling but the primary judge found Exhibit MFI B, Annexure A to his reasons, prepared by the Webb applicant to be the "most useful comparative document" (at [298]). The primary judge was entitled to do so, and as his Honour said, that modelling produced a "relatively realistic application of the likely results of net returns based upon the security for costs figure that [the judge] proposed and the two most likely modes of providing security in a range of scenarios" (at [299]). We also note that Annexure A was to be read together with the observations of the primary judge at [300] to [304] and the costs assumptions contained in Annexure B to his judgment.

The McTaggart applicants' arguments 244 The primary judge said that the McTaggart applicants' comparative document, Exhibit MFI A, was "a less helpful comparative document than Exhibit MFI B, the latter document having been prepared on the basis of a common ATE policy expense" (at [275]). The McTaggart applicants' position was that both they and Webb should be held to their original estimates for ATE insurance premiums (Webb's estimate being higher), and the McTaggart applicants prepared their comparative document on that basis. However, although the only evidence on the cost of ATE insurance was from the funder of the McTaggart Proceeding, as his Honour said, "there is no reason why [the court] would not draw the same conclusion that the costs would be commensurate for Webb" (at [274]). We respectfully agree. 245 The McTaggart applicants criticise the comparative modelling process as unsound because there was no evidence of the likely quantum for security for costs or whether costs would be given by way of cash or ATE insurance, and describe the primary judge's assumption of $2.4 million in security as "little more than guesswork". But in the absence of evidence about the likely quantum of security for costs, it was in our view reasonable for his Honour to make a comparison based on a hypothetical sum payable in three tranches, given that at an early stage in a proceeding, any estimate of likely security for costs would likely be 'broad brush'. 246 Further, in our view it was open to the primary judge to conclude that the overall effect of the comparison was that "in the vast majority of scenarios, the funding proposal put forward by Webb was more likely to result in a greater return to group members than that put forward by the McTaggart applicants and much more likely to result in a greater return to group members than that put forward by Perera" (at [304]). Annexure A modelled the return to group members from the competing funding proposals in 784 different scenarios, of which: (a) the Webb funding model produced the highest return to group members in 742 cases (94.5%); (b) the McTaggart funding model produced the highest return in less than 5% of the total scenarios modelled; and (c) the Perera proposal was superior in only three scenarios. 247 Further, the Webb funding model generally produced superior returns across the range of possible and reasonably likely litigation outcomes. As the primary judge noted, the comparison was overwhelmingly in favour of Webb in the event, which his Honour considered more probable than not, that an ATE policy was likely to be adequate security and if book build costs were excluded (at [304]). The Webb funding model was also superior in the preponderance of scenarios where cash security was used, at all modelled settlement dates, and on the basis of all permutations of the costs assumptions. 248 Contrastingly, of the few scenarios in Annexure A where the McTaggart funding model produced a higher return to group members than the Webb funding model, almost half were modelled on the basis of different cost assumptions being applied to the applicants. In our view it was open to his Honour to discount these. 249 Further, of the other few scenarios where the McTaggart funding model was superior, all but one assumed a settlement of the proceeding prior to 1 January 2019. In our view the primary judge was correct in finding that "it is far more likely that the matter, if it settles, will settle well into next year [2019] rather than this year" (at [303]). That likelihood has increased even further given the applications for leave to appeal before us and, in our view, even if the McTaggart Proceeding had been selected it is likely that Mr Perera would have appealed that decision. Accordingly, under both the actual and counterfactual scenarios, settlement by 31 December 2018 was unlikely. In our view the primary judge correctly exercised his discretion in attributing limited weight to the McTaggart funding model's prospects of achieving higher returns in scenarios where settlement was assumed before a particular date, and where such a settlement would be an outlier relative to the ordinary timing of resolution of such class actions. 250 In any event, as Mr Webb submits, even if the above unlikely assumptions were applied the funding structure in the Webb Proceeding would still produce a superior return in any settlement scenario where the settlement sum was greater than $50 million, being two-thirds of the amount which Ms Banton said represents a realistic assessment of the value of group members' claims. Consequently, the sole settlement scenario in which the McTaggart funding structure appears attractive is one where group members' claims were settled at an unusually early stage in the life of the proceeding and at a substantial discount. 251 Further, the modelling in Annexure A involved a variety of different assumptions, but it provides the most comprehensive snapshot of the operation of the competing funding models based on the evidence filed by the applicants. This exercise was of course constrained by the limitation on the parties' capacity to adduce further evidence, but in our view the primary judge's consideration of the modelling was an appropriate application of the relevant evaluative exercise. 252 Further, it was appropriate for the primary judge to have regard to a broad settlement range, being modelled at each $5 million increment between $15 million and $140 million, with additional modelling at $37.5 million and $300 million. As his Honour correctly pointed out, an important consideration in deciding between funding models was to minimise 'windfall' possibilities (primary judgment at [289], [293] and [320]). It was therefore relevant to model a very high settlement scenario of $300 million, even if it was unlikely. The $74.5 million 'estimate' by Ms Banton upon which the McTaggart applicants relied was unreliable as a guide to the 'likely settlement range'. It was based on matters that were not disclosed to the Webb or the McTaggart applicants and was not adopted by them. Indeed, the McTaggart applicants informed the primary judge that they had estimated the value of group member claims to be between $120 million and $140 million, the higher figure in this range being the second highest figure modelled in Annexure A. It is also to be recalled that Ms Banton's estimate was made when the share price was $1.31 but still declining. It eventually declined to just over 50 cents around 21 February 2018. 253 In summary, we are not persuaded that the McTaggart applicants demonstrated error in the primary judge's consideration of the modelling. We now turn to address some other issues they raised. 254 First, although Mr Webb had not filed a statement of claim during the selection process, there was evidence that his claims would be substantially the same as those in the two other proceedings. We accept that the statement of claim filed in the Webb Proceeding contains additional allegations beyond those in the other two proceedings, but those allegations arise from the same underlying substratum of facts. The statement of claim includes both a direct reliance case and a market-based causation case in a similar way to those advanced by the McTaggart applicants and the Perera applicant. Moreover, it is arguable that if the additional allegations resulted in increase in costs, they are justified by the potential increase to the damages recoverable in the proceeding. We do not consider that the additional allegations justify any different approach to the comparative modelling from that which his Honour took. 255 Second, the inclusion of a claim against Mr Hunter in the McTaggart Proceeding is not a significant forensic advantage over the other proceedings. That claim is limited to allegations of knowing involvement in the contraventions of GetSwift, which requires proof of actual knowledge of GetSwift's wrongdoing on the part of Mr Hunter, and in any event does not expand the scope of the claimable damages. The McTaggart applicants seek to justify the inclusion of Mr Hunter as a respondent in part on the basis that he has substantial assets, but the only evidence as to his asset position is that he has substantial holdings of GetSwift shares. To the extent that GetSwift becomes insolvent or is unable to meet a judgment, it could not realistically be expected that Mr Hunter's GetSwift shares would offer a significant alternative avenue of recovery for group members. Further, the inclusion of an additional respondent is likely to increase both the applicant's and defence costs, and is more likely to do so than the inclusion of additional allegations against the same respondents. 256 To the extent that the alleged contraventions give rise to joint and several liability for the respondents, the loss could be recovered from GetSwift. Moreover, to the extent that the contraventions are apportionable claims, there is little to suggest that GetSwift could reduce its responsibility by apportioning loss to its directors. The McTaggart applicants do not allege that Mr Hunter acted in any capacity other than as an officer of GetSwift. Accordingly, to the extent that he is liable GetSwift would be vicariously liable. 257 Third, as to the McTaggart applicants' submission that there was no evidence about the funder of the Webb Proceeding, there was evidence from Mr Phi that Therium is an experienced litigation funder based in England and founded in 2009, with US$11 billion of claims funded, in 2015 it raised US$300 million, and in 2018 it completed a first close of a new £300 million fund. We accept that the evidence may have been deficient in relation to the precise Therium entity who was to be funder, but we consider that his Honour was entitled to look at the commercial reality of the entity being part of an associated structure. 258 Fourth, as to the question of any conflict between the Webb applicant and Therium, such a conflict would be resolved (and has been resolved) through the making of a common fund order, alleviating the need for any funding agreement. 259 Fifth, the primary judge was not wrong to regard the periodic assessment of costs by an independent referee as a desirable feature of the Webb Proceeding. It seems to us that it was not designed to mitigate any particular moral hazard consequent upon the Webb funding model. The funding structure for the Webb Proceeding applies the lower of a percentage of the net resolution sum or a costs multiple. The higher the amount of legal costs incurred, the greater the amount of any resolution sum applied to the reimbursement of legal costs, and in turn, the lower the total amount of any net resolution sum to which the 20% commission rate would apply. Consequently, the funder has sufficient incentive to minimise the amount of any settlement expended on legal costs. Further, costs would be paid or part-paid from any settlement under the funding models proposed in each of the proceedings and an independent referee would have been in the interests of group members under each of them. 260 Sixth, and contrary to the McTaggart applicants' submissions, the engagement of a joint or court appointed forensic economic expert may be an appropriate measure to reduce legal costs in an appropriate case, although of course it may not necessarily be so. Reasonable minds may differ on this but we need not elaborate further. Whether the primary judge was correct or not on this aspect does not, in our view, change the result or show any material House v The King error. 261 Seventh, as we said earlier, in our view the primary judge was correct in rejecting the Perera or McTaggart applicants' contention that they suffered procedural unfairness. After receiving Mr Webb's evidence and submissions they were given several opportunities to supplement and clarify their evidence on the comparative merits of the funding proposals and make submissions on Mr Webb's evidence. In our view Mr Webb was not afforded an unfair advantage and he was pressed to make his 'bid' hastily. At a case management hearing on 29 March 2018, the primary judge indicated that he was disposed to make orders for the applicants to put before the Court their proposal to address the issue of multiplicity in the Perera and McTaggart Proceedings (in effect their funding proposal) and listed that issue to be heard on 13 April 2018. There was no public notification of those orders, but Mr Webb having become aware of them and having only very recently obtained funding approval from Therium, promptly applied on 9 April 2018 to adjourn the 13 April 2018 case management hearing. At the hearing of the adjournment application on 11 April 2018, the primary judge refused to adjourn the 13 April 2018 hearing and Mr Webb was given only one day to adduce evidence of his funding proposal if he wished to participate in the multiplicity hearing. Mr Webb met that tight deadline, having his funding confirmed on 12 April 2018, and filing an originating application on 13 April 2018. Mr Webb was unaware of the details of the competing funding proposals prior to exchanging his own evidence. 262 Further, it is noteworthy that the Perera and McTaggart applicants did not complain about Mr Webb's participation until after they had received his evidence. We agree with GetSwift and Mr Webb that it is unpersuasive for the McTaggart applicants to assert that they would have adduced different evidence if they had known about the existence of the third participant. Further, the only material matter of which Mr Webb had advance notice was whether a court appointed or joint expert ought to be appointed and in our view this did not give rise to material procedural unfairness. The McTaggart and Perera applicants were directed by the orders of 4 April 2018 to put on evidence regarding the issue and made a decision not to do so. At the hearing on 13 April 2018, counsel for the Perera applicant said that "we actually have given thought to how to advance this case. It will be a two expert case". Counsel for the McTaggart applicants said it was "too early to commit to that kind of process at this time". 263 Further, as to the McTaggart applicants' submission that they suffered procedural unfairness because they and Mr Perera "can be expected to have formulated their proposals…knowing the other party and their track record and making estimates of their appetite for risk", the true complainant would appear to be not the McTaggart applicants but their funder, Vannin. Its complaint appears to be that had it known about the Webb Proceeding it would have put forward more competitive funding terms. However, as the primary judge explained, "[w]hat the orders sought to elicit, among other things, was a considered proposal which best served the twin ends of furthering the overarching purpose and protecting the interests of group members, not the optimal commercial "deal" for the funder pitched at the level to beat (but only just) the commercial "deal" likely proposed by another funder" (at [59]). Permitting Mr Webb to take part in that process was advantageous to achieving those objectives. 264 Eighth, his Honour was correct in giving lesser weight to the modest number of group members who had been signed up and in distinguishing the case from Bellamy's. The numbers of group members that had signed up to the McTaggart Proceeding (just over 200) and the Perera Proceeding (just over 100) were modest, both when compared with the 2,500 that had signed up in the two Bellamy's class actions (see Bellamy's at [2], [3], [96] and [97]), and when compared with the total number of group members likely to participate in the proceeding. We would be surprised if some thousands of group members did not ultimately participate in the proceeding. The number that might do so is, of course, a function of many factors including the number of persons who acquired shares in GetSwift in the relevant period, but it is perhaps worth noting that in the Newcrest class action 4,442 group members participated in the settlement (Earglow Pty Ltd v Newcrest Mining Ltd [2016] FCA 1433 at [16]) and in the Allco class action 5,130 group members participated in the settlement (Blairgowrie Trading Ltd v Allco Finance Group Ltd (recs and mgrs apptd) (in liq) (No 3) (2017) 343 ALR 476; [2017] FCA 330 at [20] and [70]). Many years ago in the GIO class action (King v AG Australia Holdings Ltd (formerly GIO Australia Holdings Ltd) [2003] FCA 980 at [9]; King v AG Australia Holdings Ltd (formerly GIO Australia Holdings Ltd) [2003] FCA 1420) more than 23,099 group members participated in the settlement. 265 Further, in Bellamy's not only was there a significant number of funded group members but the relevant group included sophisticated institutions, investment managers, global asset managers, sovereign wealth funds and superannuation funds, including investors with large claims by value, investors with repeat experience in participating in funded securities class actions, and investors with sufficient knowledge, sophistication and leverage to properly negotiate the terms of the funding agreements they entered into and to make informed choices between competing funding agreements on offer. Contrastingly, in the present case, there is no evidence that any of the group members that had signed funding agreements with the funders of the Perera or McTaggart Proceedings shared those characteristics. The mere fact that a modest number of investors had entered into funding agreements was rightly given lesser weight by the primary judge.

Mr Perera's arguments 266 We now turn to the other arguments raised by the applicant in the Perera Proceeding. 267 First, the primary judge held that the comparative document prepared on behalf of Mr Perera, Exhibit MFI C, was "not useful in providing a realistic comparative analysis" (at [263]). Mr Perera had objected to the comparative process in its entirety, doing so on the basis that parties should be 'held' to their original differing costs estimates based on different assumptions, rather than by applying a set of common costs assumptions to compare the funding proposals. The primary judge rejected that approach, correctly in our view, (at [260] to [263]). His Honour demonstrated the incoherence of that approach through an illustrative example showing the differences in estimates of the length of mediation (at [281]). The primary judge observed that "these differences do not in any rational way reflect qualitative differences between the two firms. There is no basis for thinking a mediation will take a shorter period of time with [Mr Perera's solicitors] rather than [the McTaggart applicants' solicitors]" (at [281]). Mr Perera's insistence that the parties be held to their original estimates was further undermined by the evidence which he adduced following the exchange of evidence with the McTaggart applicants, and his downward adjustment of his costs assumptions (at [264]). 268 As we have already said, we consider his Honour was correct in concluding that the overall effect of the comparison between the funding models was that in the vast majority of scenarios the funding proposal put forward by Mr Webb was more likely to result in a greater return to group members than the McTaggart applicants' proposal and much more likely to result in a greater return to group members than Mr Perera's proposal. And as we said earlier, of the scenarios modelled in Annexure A, the Webb funding model produced the highest return to group members in 742 instances (94.5%). The Perera proposal was superior to the other proposals in only 3 of the total 784 scenarios modelled, and each of these 3 instances were outliers: they assumed a resolution at trial for the lowest modelled sum of $15 million. By comparison, and as Annexure A makes clear, the Webb funding model was superior to the Perera and McTaggart funding models across the range of possible and likely litigation outcomes. 269 Before us Mr Perera made detailed submissions to the effect that the primary judge erred both on the costs integer in the relevant equation and also on the aggregate claim value integer. It is not necessary for us to engage further with the detail at this point as we are not satisfied that overall they justify in any event a different conclusion to the one reached by his Honour, namely, the modelling generally supported the Webb Proceeding as being more advantageous to group members. 270 Further, Mr Perera criticises the fact that the Webb Proceeding was issued on the day of the argument without any pleading. In our view, to the extent it was relevant at all, the primary judge was correct in giving little weight to the relative priority in the commencement of the competing proceedings. Moreover, the absence of a pleading did not have great significance given the evidence before the primary judge that Mr Webb's claims would be substantially the same as those in the other two proceedings. 271 Further, and as we have already said, throughout the selection process each party was given the same opportunity to adduce evidence and the same opportunity (albeit limited) to respond to other parties' evidence. Contrary to Mr Perera's submissions, neither the McTaggart applicants nor Mr Webb were permitted to alter their positions. Rather, the primary judge accepted that the evidence allowed inferences to be drawn as to the likely cost of the proceedings based on the application of certain reasonable assumptions. In our view those inferences were open to be drawn and we consider them to be correct. For instance, while Mr Webb's cost estimate included a provision for the conduct of a book build of funded group members if required, there was no good reason to conclude that he would in fact incur such costs. It was therefore appropriate for the primary judge to rely on modelling that assumed that these costs would not be incurred. 272 In our view, no material House v The King error has been demonstrated in his Honour's selection of the Webb Proceeding.

Conclusion in regard to the selection process 273 As we have said, we are not persuaded that the primary judge made a material House v The King error in deciding that the Court has power to stay two of the three competing class actions nor in deciding that it was appropriate to stay the Perera and McTaggart Proceedings and allow the Webb Proceeding to go forward. In our view his Honour's decision falls within the proper exercise of discretion. 274 It is though worth reiterating, as the primary judge recognised, that there is no one right answer to the case management questions that arise when dealing with competing class actions. There cannot be a 'one size fits all' and different judges will take a different view of some of the incommensurable and conflicting considerations that may arise. It should also be kept in mind that there is no 'silver bullet' solution to the case management problems of competing class actions and each of the 'solutions' can be said to have some or other problem. For example: (a) even when the parties agree to consolidation of the competing proceedings, depending on the agreement reached, there may be wastage of costs and inefficiency. The most efficient outcome is where the competing funders agree to a percentage division of the funding expenses and any funding commission that becomes payable and there is only one set of applicant's solicitors and counsel running the case. Where the agreement provides for joint solicitors under a joint-venture agreement there can be serious inefficiencies and wastage of costs as solicitors from different firms, often with no love lost between them, incur costs by checking each other's work and arguing about the many forensic decisions that are involved in large and complex litigation; (b) without adopting this criticism, the 'wait and see' approach in Cantor might be said to give too little weight to the fact that respondents are vexed with the injustice and cost of two sets of class actions, and to the fact that (if the case is successful) group members will ultimately bear the costs of having two sets of applicant's solicitors and counsel drawing fees from the same pool of settlement or judgment monies; and (c) without adopting this criticism, notwithstanding the case management orders made to minimise the duplication of costs, the approach taken in Bellamy's has been criticised for giving diminished weight to the increased costs suffered by the respondent and the increased transaction costs ultimately suffered by the group members of having two sets of applicant's solicitors, and also for giving too much weight to the contractual arrangements made with a particular lawyer or funder by group members. 275 Similarly depending upon how it is done, a selection process like that used in the present case may give rise to some or other problem, and the Court should be astute to avoid them where possible. 276 First, where the other relevant considerations are essentially equivalent as in the present case, the decisive consideration of which proceeding should be stayed may be argued to be the lower legal costs and funding charges that apply in one proceeding as compared with the other competing proceedings. As desirable as lower costs and funding charges are, one of the real benefits of litigation funding has been the levelling of the playing field for claimants engaged in expensive class action litigation against large, well-resourced and determined respondents (as the primary judge recognised at [230]). The Court should not give an undue focus to lower costs and funding charges as doing so is likely to promote a "rush to the bottom" by funders and solicitors keen to win the tender. In our view there are already signs of this in the market. 277 While not without deficiencies, a good level of Court supervision of costs and funding charges is already available through the settlement approval process under s 33V, pursuant to which the Court only approves costs and funding charges in a quantum which it considers reasonable. Having regard to that protection the Court should focus less on achieving the lowest possible costs and funding charges in any selection process, and more on selecting the proceeding with a funding and costs models likely to best motivate the applicant's solicitor and funder to work assiduously to achieve the best outcome for the applicant and group members and to take responsible risks in that regard. 278 Second, the single most important determinant of the net recovery achieved by group members is not the quantum of legal costs but the amount of the settlement or judgment achieved, and where the settlement or judgment is large the legal costs are usually not material to net recovery. The Court should be astute to select the proceeding with the legal team that is most likely to achieve the largest settlement or judgment, i.e. the most experienced and capable. We accept that differentiating between legal firms or solicitors will often be difficult but the Court should not dodge that question if there is a basis for differentiation. 279 Third, recently a number of securities class actions have been issued very speedily following a share price collapse. Wileypark is a good example of this. The Court must strongly discourage a rush to the Court in large and complex class proceedings, carrying as it does the consequent risks of insufficient due diligence and the commencement of unmeritorious, or at least weak, cases. Unless the hasty filing of such cases is effectively discouraged even those solicitors or funders who wish to take an appropriately cautious approach are likely to be dragged into the same practice. That is so because the first action filed is likely to obtain a 'first mover' advantage in terms of book building and, once one action is filed, other solicitors or funders are pressed to speedily follow or they may not be included in the mix when the Court considers the competing proceedings. 280 It may be time for the Court to consider a procedure, in relation to securities class actions at least, such that upon the filing of the first proceeding the Court orders a standstill in that proceeding for, say, 90 days to allow a reasonable time for other solicitors or funders to undertake a proper due diligence. In order to reduce the incentive to rush to the Court, and to reduce any incentive to speedily follow another party that does so, any book building that occurs during the standstill period should be given no weight by the Court. We note that a 90 day standstill period is imposed under s 77z-1 of the Private Securities Litigation Reform Act 1995 in the USA. 281 Fourth, a selection process such as that used in the present case is conducted in full view of the respondent and it is likely the respondent will obtain a reasonable understanding of the approximate size of the "war chest" available for the case against it. As the primary judge recognised, experience teaches that respondents sometimes engage in trial by attrition and endeavour to use up an applicant's resources to obtain an advantage. Respondents are also likely to understand that the applicant's solicitors may be less inclined to undertake the necessary work if they are approaching or have exceeded the amount allowed for costs, and/or to understand that the funder may put pressure on the applicant to settle in such circumstances. The Court should be careful to avoid the interests of the applicant and group members being damaged in this regard. 282 Fifth, a selection process such as that in the present case is expensive. It seems likely that the costs for the present three applicants for case investigation and preparation, book building, drafting the pleadings, and opposing the stay application will be in the order of $300,000 to $500,000 each. For the unsuccessful funder (or solicitor if the work was done on a conditional basis) such amounts will be lost upon their case being stayed. Losses of that magnitude are likely to prove unsustainable over the medium to long term and some funders or solicitors may withdraw from the market thereby reducing competition. That is undesirable when competition between funders is perhaps the single most important matter giving rise to the lower funding charges recently seen in the market, and competition between solicitors is giving rise to reduced costs. Alternatively, funders or solicitors are likely to build the losses suffered into their charges in other class actions, which is also undesirable.

COMMON FUND ORDER 283 Mr Perera seeks leave to appeal from the common fund order the primary judge made in the Webb Proceeding. We refuse leave to appeal. In our opinion the order is not attended with sufficient doubt and neither Mr Perera nor the group members he represents in the Perera Proceeding will be substantially prejudiced if leave to appeal were to be refused. 284 The proposed ground of appeal is expressed in the following terms: The primary judge erred in making a "common fund order" on 20 June 2018 in circumstances where: a. the Federal Court does not have power to make, or alternatively ought not in its discretion to have made a "common fund order" which binds group members in the Webb Proceeding who have signed funding agreements with ILP18 or retainers with Squire Patton Boggs (the funder and solicitor in the Perera Proceeding) without the consent of those group members, or the consent of ILP18 to the overriding of their contractual rights and obligations; b. the Federal Court ought not in its discretion to have made a "common fund order" which binds group members in the Webb Proceeding who have signed funding agreements with ILP18 or retainers with Squire Patton Boggs (the funder and solicitor in the Perera Proceeding) without giving notice to those group members in advance of the Court's intention to do so. 285 The common fund order is a discretionary interlocutory order that in significant respects concerns practice and procedure. Moreover, the order itself may be varied as the Court sees fit either before or in the course of the approval of any settlement of the Webb Proceeding under s 33V. 286 Before the primary judge neither Mr Perera nor any other party made submissions expressing any doubt as to the power to make a common fund order. Nor did they do so before us. Further, the common fund order made does not seek to restrain ILP18 (the funder of the Perera Proceeding), Mr Perera or the Perera Funded GMs from enforcing their contractual rights vis-à-vis one another, and any such contractual interference could only occur by the making of further orders. 287 Given that we have dismissed the appeal from the primary judge's order permanently staying the Perera Proceeding we do not see how any substantial prejudice could arise to Mr Perera or the Perera Funded GMs from the common fund order made in the Webb Proceeding. 288 First, Mr Perera and the Perera Funded GMs will be free to opt out of the Webb Proceeding if they so choose, and if they do so they will then not be bound by the common fund order. 289 Second, let it be assumed that Mr Perera and the Perera Funded GMs do not opt out of the Webb Proceeding and that they continue to be contractually bound under their funding agreements to pay a separate commission to ILP18. Any such liability would only arise on a successful outcome in the Webb Proceeding. In that event the primary judge could make orders to eliminate any duplicate liability of the Perera Funded GMs to pay commission to both ILP18 (under contract) and Therium (under the common fund order), by modifying the common fund order to exclude the Perera Funded GMs. It may also be possible, but it is unnecessary to express a final view, that the primary judge in exercise of his statutory powers could make an order which in effect overrides ILP18's contractual entitlement. 290 We also note that the opt out notice approved by the primary judge in conjunction with making the common fund order informed Perera Funded GMs that the Court had not yet determined whether the promises to pay ILP18 remain enforceable, it was unclear if they would be enforced, and that they may "work on the assumption" that any total amount paid for funding would not exceed the quantum of what was promised to ILP18, and may be less. 291 Further, pursuant to s 33X of the Act, Mr Perera and the Perera Funded GMs will receive notification of any application for Court approval of any proposed settlement of the Webb Proceeding. They will have the opportunity to object to any settlement, or the application of the common fund order in respect of such a settlement, which would have the effect of interfering with their contractual relationships with ILP18. Indeed, the primary judge's order of 20 June 2018 directed that, inter alia, ILP18 be notified of such an application. In our view, until such time, the common fund order cannot be said to be binding upon the Perera Funded GMs in a manner which substantially interferes with their interests. 292 Third, as to Mr Perera's submissions that the common fund orders could not be made without the consent of ILP18, we note that by orders made on 23 May 2018 the primary judge granted leave to ILP18 to appear and be heard in respect of the contents of the opt out notice. ILP18 could have sought leave to also be heard on the question of the making of the common fund order, but elected neither to do so nor to appear in order to make submissions in respect of the opt out notice. Nor has ILP18 sought leave to appeal the making of the common fund order. It may be inferred that it does not consider that it will suffer any substantial injustice if Mr Perera's application for leave to appeal the common fund order is not allowed. 293 Fourth, Mr Perera also submits that the common fund order should not have made without notice having been given to group members in advance of it being made. We accept that while Mr Perera had notice and was able to object to the common fund order, the group members for whom he acts in a representative capacity in the stayed Perera Proceeding had no such notice. Such notice to group members has usually been given prior to making a common fund order. For example, it was given in Money Max, and also in Bellamy's, and it is usually given when a common fund order is sought as part of a settlement approval application under s 33V. 294 Even so, as the primary judge noted (at [246]), in Pearson v State of Queensland [2017] FCA 1096 Murphy J decided that in the circumstances of that case it was appropriate to make a common fund order early in the proceeding, including that group members would be informed of the requirement to pay the commission before they decided whether to opt out and which would allow group members to avoid the effect of the order if they were unhappy with it. The primary judge also noted that making a common fund order at that stage would avoid the wasted costs associated with book building and the waste of time and effort that would have been needed to explain the funding arrangements against a backdrop of uncertainty as to what the Court would eventually do. 295 In our view it was open to the primary judge to exercise the discretion to make a common fund order in the absence of notice, in circumstances where: (a) group members may opt out if they do not wish to be bound by an order which provides that upon a successful outcome in the proceeding they must pay a reasonable court approved funding commission; (b) doing so avoids wasted costs associated with book building and wasted time and effort in explaining funding arrangements; (c) the common fund order has no final binding effect at this stage; and (d) it would appear that everything that could reasonably be said for and against the common fund order was said. 296 For the foregoing reasons we are not persuaded there is any substantial prejudice to Mr Perera or the Perera Funded GMs by the common fund order in the Webb Proceeding. Moreover, we do not consider there is any substantial prejudice to them even if the decision of the primary judge was wrong. For this reason alone, leave to appeal should be refused. Further, to the extent it is necessary to say so, we do not consider that the primary judge's order was attended with sufficient doubt to warrant the grant of leave.

THE INJUNCTION AGAINST SPB 297 As we have indicated, in Webb v GetSwift Limited (No 2) [2018] FCA 994 (the 'restraint judgment') the primary judge made orders that restrained certain communications between Mr Perera's solicitors SPB and group members in the Webb Proceeding (other than Mr Perera himself), including the Perera Funded GMs who were clients of SPB. 298 The relevant orders made by the primary judge were as follows: 1. Pursuant to s 33ZF of the Federal Court of Australia Act 1976 (Cth) (Act), the respondent to the interlocutory application dated 23 June 2018 (Interlocutory Application), Squire Patton Boggs, and its officers, servants and agents be restrained from communicating with group members in this proceeding, other than Mr Dwayne Cavan Shanahan Perera, solely in relation to the issue of whether group members should exercise their right to opt out of this proceeding pursuant to s 33J of the Act, with such restraint to continue until either: (a) the final determination of the applications for leave to appeal and the appeals in proceedings NSD950/2018, NSD951/2018 and NSD964/2018, which determinations have the effect that one or both of NSD226/2018 and/or NSD440/2018 no longer be subject to a stay of proceedings; or (b) the conclusion of the opt out process following the expiry of the date for opt out in this proceeding. 2. For the avoidance of doubt, Order 1 does not prevent Squire Patton Boggs from responding to any unsolicited enquiry concerning opt out made to them by a group member in this proceeding: (a) prior to the service of the opt out notice - by communicating to that group member that matters relevant to the issue of opt out are currently before the Court and the Court will, in due course, set a date for opt out and will approve an opt out notice which will be sent to the group members in this proceeding; or (b) after the service of the opt out notice - by communicating to that group member that they should refer to the opt out notice and take such advice as they think appropriate relating to their right to opt out in such manner as they think fit by taking independent legal advice or contacting the independent barrister referred to in the opt out notice. 299 The amended application for leave to appeal set out the following relevant ground: The primary judge erred in making orders restraining Squire Patton Boggs (the solicitors for Mr Perera in the Perera Proceeding) from communicating with group members in the Webb Proceeding (including the clients of Squire Patton Boggs) in relation to whether they should exercise their right to opt out of the Webb Proceeding, in circumstances where: (a) the Federal Court does not have power to restrain, or alternatively ought not in its discretion to have restrained, particular solicitors from communicating with group members about a particular subject, and, in particular, where a significant number of group members have expressly chosen to be represented by particular lawyers, and have entered into retainers with those lawyers. (b) The effect of the restraint was to interfere with contracts between Squire Patton Boggs and those group members who have entered into retainers with Squire Patton Boggs in proceedings to which they are not parties.

(a) Legal principles 300 In addition to the specific powers vested in the Court already referred to, the Court has broader powers to "make such other order as the Court thinks just" (under s 33Z(1)(g)) and "make any order the Court thinks appropriate or necessary to ensure that justice is done" in any proceeding conducted under Pt IVA of the Act (under s 33ZF). In addition, the Court has the ability to protect itself against any abuse of its processes, so to ensure its processes are used fairly and not lead to an erosion of public confidence: see e.g. Williams v Spautz (1992) 174 CLR 509 at 520. 301 In Johnstone v HIH Ltd [2004] FCA 190, Tamberlin J contemplated the application of these legislative powers to misleading communications which may induce a misunderstanding on the part of recipients as to the nature and operation of the representative proceeding, and the rights and obligations of the recipients in respect of the proceeding. Specifically at [105], as his Honour stated: The Court has an important and continuing role in managing representative proceedings in the public interest to rectify any potential misleading communications to class members or potential class members, in order to ensure that there is no misunderstanding engendered by such communications, particularly when they emanate from legal advisers, as to rights and obligations and procedures to be followed by recipients of such communications. See also the remarks of Sackville J in Courtney v Medtel Pty Ltd [2002] FCA 957; (2002) 122 FCR 168 and Middleton J in Davaria Pty Limited v 7-Eleven Stores Pty Ltd [2018] FCA 984, in which their Honours in a different context cautioned against an overly expansive interpretation of the Court's powers under ss 33Z(1)(g) and 33ZF.

(b) The primary judge's view 302 The primary judge made the following comments at [49] to [53] of the restraint judgment: 49. … I have already made remarks about the centrality of the opt out procedure to the efficient operation of Part IVA. As I remarked during the course of argument, if anyone is to opt out, they should do so on the basis of information that the Court has approved as to both form and content and that if they are to obtain advice concerning those matters, at least in the first instance, it should be from someone who does not, in a colloquial sense, have a "dog in the fight". 50. This is not a criticism of the funders or solicitors. The Court has a protective role and a duty to protect the integrity of opt out. It does not in any way amount to a conclusion that the solicitors and funders would act inappropriately, but the stark reality is that they may (or there is a reasonable apprehension that they may) have a real commercial interest in maximising the number of persons opting out of the Webb Proceeding to present the possibility of a viable, alternative commercial vehicle. No evidence has been adduced to suggest that another representative proceeding would not be funded or promoted by ILP18 and conducted by SPB, following opt out, if enough group members did opt out. Moreover, no evidence was adduced to suggest that any actual retainer by the Perera Funded GMs of SPB was anything other than as part of becoming a participant in the commercial enterprise to promote and maintain a representative proceeding funded by ILP18 (such as the Perera Proceeding). Indeed, as I indicated in an earlier case management hearing to Mr Edwards who was then appearing on behalf of Mr Perera, I would infer that this was the case absent the tender of the retainers which suggested to the contrary (T-69; T-80, 8 June 2018). This is not a case where there is any evidence to find that SPB have been retained generally to advise the group members independently of the common enterprise. 51. In the principal judgment, I have already explained in detail the nature of the scheme by which the Court's processes are being used to derive financial benefits for participants in the scheme. It is contrary to common sense not to recognise the possibility of conflict arising in a solicitor deriving commercial benefits from the scheme advising a participant in the scheme as to whether they should continue to be involved in another proceeding. The protective and supervisory role of the court to ensure opt out occurs on the basis of non-conflicted and accurate information as to the consequences is not mere rhetoric. Hence my desire to put in place a mechanism by which a group member is entreated to get objective, disinterested advice. 52. Enough has been said to explain that this has no relevant analogy, it seems to me, to a circumstance where I am dealing with the proprietary rights of a litigant and ensuring that all necessary parties are before the court who may have an interest in the relief sought being granted. 53. The relief that I have granted on an interim basis and propose to grant pending the completion of opt out, is under s 33ZF. I have already explained the width of this power in the principal judgment. It very regularly operates to oblige third parties to do (or refrain from doing things). Group members, commonly bound by s 33ZF orders, are, it will be recalled, themselves non-parties. Procedural fairness was amply provided to SPB by granting the right to intervene at all material stages (and indeed encouraging submissions). Also, the relief sought, which is interlocutory in nature in the Webb Proceeding (to protect the opt out process and is appropriate or necessary to ensure that justice is done in the Webb Proceeding), has been made pursuant to a prayer for relief sought in interlocutory process which has been served on the party against whom relief has been sought and that party has appeared and (belatedly) has made submissions as to why that relief ought not be granted. Although, as I have said, no submissions have been made by SPB or ILP18 as to what should be included in any notice, despite invitation. 303 In addition, the primary judge further observed at [63] that: 63. The whole point of the exercise I propose to put in place in relation to opt out is to provide accurate information, including as to this aspect of the decision, to all group members. The fact that this contention could even be made on behalf of SPB in the light of the prior approved notice (which seems to suggest that SPB apparently wants to advise Perera Funded GMs that they could conceivably be exposed to damages if they did not opt out of the Webb Proceeding) graphically illustrates the danger that funding agreements could possibly be used as a spectre to encourage opt out, and reinforces the need that the Court must do all it reasonably can to ensure accurate, complete and independent information be provided to group members. As was clear during argument on approval of the terms of opt out notice and by the notice itself, there is simply no prospect of any funded group member paying two funding fees if an amount was recovered for their claim in the Webb Proceeding (even if this required an order that the common fund order not apply in relation to the Perera Funded GMs and the McTaggart Funded GMs). Insofar as payment for funding is concerned, it will be necessary to leave to another day any issue as to whether the funding agreements in relation to the stayed proceedings can be enforced according to their terms, but one thing is clear: the power of the Court to ensure no funded group member will be worse off by the imposition of any additional funding impost pursuant to Court order.

Mr Perera's position 304 Mr Perera contends that the restraint order in the restraint judgment is an extraordinary one insofar as it purports to prevent a firm of solicitors from communicating with its own clients (namely those group members in the Webb Proceedings who had retained SPB to act for them in the Perera Proceeding). Mr Perera further submits that the restraint order is framed in such a way so as to both prevent SPB from initiating any communication with its existing clients about a subject which one might think was of importance to them (given that, as existing clients, they had already decided they wanted SPB to run their cases through the Perera Proceeding), and to dictate what could be said by SPB to potential prospective clients in response to unsolicited inquiries. 305 Mr Perera argues that, while his own communications with SPB were not affected by the restraint order, the order did affect some of the persons who Mr Perera still represented in the stayed Perera Proceedings (namely the Perera Funded GMs). Mr Perera submits that he feels duty bound to protect those persons' interests in being able to communicate with the solicitor that many of them have retained. 306 Mr Perera contends that the Court's function under the Act does not extend to interfering with retainers between solicitors and their clients. While it can protect and warn group members through notices issued under ss 33X and 33Y to consider obtaining independent legal advice, it cannot and should not prevent them from communicating with a solicitor of their choice. It submits that the restraint order pays no or insufficient heed to the confidential nature of the relationship between a solicitor and a client and the interests of the client in being able to seek legal advice from someone they have chosen, and trust. 307 Mr Perera highlights that the fact that the primary judge had previously also made orders for the appointment of an independent barrister to advise group members about opt out rights (and the liberty to SPB to refer their clients to that person) does not improve the position. Mr Perera says this raises questions such as: According to what instructions is such an independent barrister to act? Who is to pay them? How are they to get the client's files from SPB? Do they have insurance for providing legal advice pursuant to such an arrangement, if they get something wrong?

Mr Webb's position 308 Mr Webb reminds the Court that the restraint order is an interlocutory order concerning practice and procedure, and the primary judge made no relevant House v The King error. Mr Webb also observes that although the restraint was against SPB and not Mr Perera, it was Mr Perera who purports to challenge it. 309 Mr Webb submits that the injunction is of narrow scope, prohibiting SPB from communicating with group members in this proceeding, other than Perera, solely in relation to the issue of whether group members should exercise their right to opt out of this proceeding pursuant to s 33J of the Act. 310 He contends that the restraint does not prevent SPB from responding to any unsolicited enquiry concerning opt out made to them by a group member by communicating, as the case may be, that the matters relating to opt out are currently before the court, or that they should refer to the opt out notice and take advice. 311 Mr Webb asserts that the reason underpinning the restraint is the preservation of the integrity of the process implemented by the Court for the resolution of the issue of multiplicity of class actions. He submits that it remains a reasonable apprehension that, in the absence of such a restraint, SPB could undermine the integrity of the process by recommending to clients to opt out of the Webb Proceeding which the Court had selected to proceed as an open class action. To do so would be to the advantage of SPB and ILP18. 312 Mr Webb further submits that the reasonableness of that apprehension can be established by the position taken by Mr Perera and separately SPB, at hearings where the question of the appropriateness of such a restraint order was raised, and SPB persisted in refusing to give the relevant undertaking. 313 He also submits that Mr Perera has no interest in the restraint order that rises above that of a mere intellectual or emotional concern and no substantial injustice arises from the order. He notes that by letter dated 13 June 2018 sent to Mr Webb's solicitors, SPB confirmed that Perera and "his servants and agents, including Squire Patton Boggs" gave an undertaking in substantially the same terms as the restraint order, and that undertaking has not been withdrawn. He argues that the restraint order was made on the basis that SPB did not accept that the undertaking restrained SPB in capacities other than as servant and agent of Mr Perera, and sought to reserve its rights in this regard. That is to say, he says that the restraint order only impacts on SPB to the extent that it is not acting in the capacity of servant and agent of Mr Perera. 314 Mr Webb then argues that any group member who considered his or her interests to be affected by the restraint order could have applied for leave to appeal the restraint order, or to intervene in the appeal, but none has done so. That is so notwithstanding that the restraint order would not prevent SPB providing to its clients a copy of that order, nor prevent SPB (subject to consideration as to the potential existence of a conflict of interest and as to whether such clients should obtain separate and independent legal advice and representation) advising any such clients that they should consider filing an application for leave to appeal the making of that order. He submits that there was no evidence before the Court to support the conclusion that group members' interests have been or may be materially negatively affected by the restraint order. 315 In the circumstances, Mr Webb submits that the party with sufficient interest in the restraint order potentially to ground an application for leave to appeal the restraint order is not Mr Perera, but rather SPB. SPB could have applied for such leave, but elected not to. Relevantly, if SPB had elected to apply for such leave, then the resulting exposure to an adverse costs order in the event that leave was not granted or the appeal not allowed would have rested with SPB rather than with its client. 316 Finally, Mr Webb contends that leave to appeal ought be refused having regard to the fact that SPB had been named as a concurrent wrongdoer by the respondents in their defence filed in the Webb Proceeding on 16 July 2018. In the circumstances, if leave was granted and the appeal allowed, any communications between SPB and its clients as to whether those clients should opt out of the Webb Proceeding would be infected by an irresolvable conflict of interest, alternatively an appearance of such a conflict, given communications regarding the appropriateness of opt out would invariably touch on the questions of SPB's alleged role in causing the losses suffered by group members which are the subject of the Webb Proceeding. Given in that circumstance it would be inappropriate for SPB to engage in any communications the subject of the restraint order, Mr Webb submits there is no utility in granting leave and allowing the appeal.

Mr Perera's reply 317 In reply Mr Perera submits that Mr Webb is only concerned that the Perera Funded GMs will wish to reaffirm their original choice to litigate their claims through the Perera Proceeding, or with the solicitors they have chosen, and that Mr Webb merely seeks to prevent leakage from the proceeding which the Court has forced group members into. Mr Perera says that Mr Webb wants to prevent those group members who have already indicated their choice from acting in accordance with that choice. 318 Mr Perera also submits that even if there was a reasonable apprehension that SPB might advise clients to opt out of the Webb Proceeding, nothing follows from this, as group members can opt out of a class action for whatever reason they want. 319 Finally, in response to Mr Webb's submission that Mr Perera lacks standing to challenge the order because the restraint does not affect him, Mr Perera argues that as the representative of the class in the stayed Perera Proceeding, which includes Perera Funded GMs who have signed retainers with SPB, it is within the proper scope of Mr Perera's role to protect their interests.

(d) Consideration 320 We need to consider the restraint both from the point of view of the communications with group members in the Webb Proceeding who were not clients of SPB at the time of the restraint order, and with group members who were clients of SPB at that time. 321 The Court clearly has power to make the restraint order in relation to the group members who were not clients of SPB to protect the integrity of the Court processes. In our view, the primary judge properly exercised his discretion to make the restraint order in the unusual circumstances confronting him, so as to ensure the proper communication of information to group members other than the group members who were clients of SPB. We see no basis to interfere with the restraint order in so far as it impacts upon the group members who were not clients of SPB at the time. 322 As to the restraint involving communication with clients of SPB, the situation is different. It is unnecessary to decide whether the primary judge had power to make this restraint order, as we have come to the view that the primary judge erred in making such an order even if within power. 323 It is important to recall that the evidence before the primary judge did not disclose that SPB communicated with or was likely to communicate with their clients in a manner which is misleading or otherwise unfair. There was no evidence that any communications by SPB would be unlawful or involve any ethical constraint. 324 The primary judge was concerned about the lack of independence and the prospect of lobbying on the part of SPB (see [49] of the restraint judgment), but the fact is that SPB had already a solicitor/client relationship with the Perera Funded GMs. Whether the retainer was also part of a "commercial enterprise" to use the primary judge's words (in [50] of the restraint judgment) does not mean that a general retainer to advise and act generally did not arise (with all that entails). The primary judge effectively assumed that SPB had not been retained generally to advise the group members independently of the commercial enterprise. 325 We can accept that the reality in class actions is that the interests of the applicant's solicitors and/or the funder may be in conflict with the interests of the class members. In this case there was a real possibility of conflict because of the competing class actions. A conflict may arise where the interests of the representative applicant and group members diverge and in certain circumstances, the Court may need to intervene to resolve a conflict. 326 In our respectful view, on the evidence that was before the primary judge, the fact that there was a "possibility of conflict" (see [51] of the restraint judgment), was not a sufficient basis for the restraint order interfering with communications between clients of SPB who were group members in the Webb Proceeding. Accepting as we do that the "protective and supervisory role of the court" may be enlivened in circumstances where evidence indicates a need for such protection, this was not one of these circumstances. Assuming the possibility of a conflict, there was no evidence to suggest that SPB would not appropriately manage that conflict, or more importantly, not give professional and non-misleading information to their clients. We accept that the potential conflict of interest exists, but it would not be impossible to remove. It should not be assumed, as the primary judge seems to have done, that SPB would act in its personal interest, or in the interests of ILP18, or that SPB would be unable to act conformably with the traditional duties a law firm owes to its clients. 327 The primary judge failed to differentiate between the group members who were existing clients of SPB and the other group members - and hence failed to exercise his discretion taking into account all the relevant circumstances, including the special position of the clients of SPB who were also group members. 328 Even accepting that the restraint order to the extent it impacts on the clients of SPB is limited, it still interferes with the lawyer/client relationship. This in our view does constitute a substantial injustice which was not justified in the circumstances, albeit the unusual circumstances, confronting the primary judge. We do not see removing the restraint order as being futile because SPB should be given the opportunity to properly manage the conflict, which would not necessarily mean restricting the extent of communication as imposed through the restraint order. We also see no difficulty in Mr Perera having standing to apply for the removal of the restraint as he is obliged to represent the interests of Perera Funded GMs in the stayed Perera Proceeding (who are clients of SPB). 329 Therefore, the order to restrain certain communications between SPB and its clients should be set aside, with the application for leave allowed in part and the appeal allowed in part so as to remove this restraint.

CONCLUSION 330 In summary, we have rejected the challenges to the permanent stay orders in the Perera Proceeding and the McTaggart Proceeding. We grant leave to appeal in each case but dismiss those appeals. As to the solicitor restraint orders in the Webb Proceeding, we grant leave to appeal and allow that appeal. As to the challenge to the common fund order in the Webb Proceeding, we refuse leave to appeal. 331 The parties have leave to make submissions on any necessary consequential orders and on the question of costs. The parties must file any such submissions within 14 days. I certify that the preceding three hundred and thirty-one (331) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justices Middleton, Murphy and Beach.