After-the-event insurance and security for costs: inconsistency in the Australian approach
The below is the text of my article which was published in the Journal of Civil Litigation and Practice as (2021) 9 JCivLP 122
Security for costs is required in situations where a plaintiff is likely to be unable to pay a costs order in favour of a defendant. The most common forms of security are presently the payment of cash into court or the provision of an unconditional bank guarantee. In recent years, there have been increasing attempts in Australia to provide, as security for costs, a policy of insurance obtained by the plaintiff which insures against an adverse costs order (“after the event” or “ATE” insurance).
This article analyses the origins and history of ATE insurance and security for costs, as well as the British and Australian decisions concerning the use of ATE insurance as security. The analysis reveals inconsistent approaches to this question, both between Australia and the UK and within Australia. However, there are a number of principles that can be discerned in relation to when an ATE policy will suffice as a form of security.
1 In the English-derived common law and equity systems, the courts have an ancient jurisdiction to order a plaintiff to provide security for the defendant’s costs of the proceeding, failing which the plaintiff’s claim would be stayed. The power to award security for costs (“SFC”) arises from the courts’ inherent jurisdiction to control their own procedure. The purpose of SFC is “to ensure that a successful respondent to a claim will have a fund available within the jurisdiction of the Court against which the respondent, if successful in defence, can enforce a judgment for costs in the respondent’s favour.”
2 As explained below, in modern times SFC is ordinarily provided by way of payment into court or unconditional bank guarantee. However, in recent years there have been increasing attempts to introduce a new form of security, being a policy of insurance obtained by the plaintiff against an adverse costs order (commonly referred to as “after the event” or “ATE” insurance).
3 This article reviews the case law regarding the attempted provision of SFC by way of ATE policy, in the context of the origin and purpose of the SFC doctrine. The article is structured as follows:
(a) first, there is a brief overview of the history and nature of the jurisdiction to order SFC;
(b) second, there is a discussion of the principles relating to the form in which SFC may be ordered;
(c) third, there is a discussion of the early forms of adverse costs insurance in Australia;
(d) fourth, there is a discussion of the rise of the ATE insurance market in the UK and the decisions of the courts there;
(e) finally, there is a review of the recent Australian decisions on the subject.
4 The essay concludes that the state of the law in Australia is presently unsatisfactory, in that there does not appear to have been a consistent approach to the circumstances in which an ATE policy may be acceptable as SFC and, further, there is a tension between the position in the UK compared with Australia. That said, there are some principles which may be distilled from the decisions handed down to date, which may be of utility for any litigants attempting to provide security by way of ATE policy.
The origins of the power to award security for costs
5 In order to understand SFC, it is necessary to begin with the courts’ jurisdiction to order that an unsuccessful party to litigation pay the successful party’s costs of the proceeding.
6 At common law, there was no inherent power for the court to award costs. Such power as did exist was limited and was conferred by statute. Conversely, in the courts of equity a doctrine developed whereby the court was able to order costs in a manner that was “intirely discretionary”. The discretion was exercised based on conscience, and was derived not from statute but from the Lord Chancellor’s general and inherent authority.
7 In equity, the development of the power to order SFC followed closely after the power to award costs in the late 14th Century. While the discretion to order SFC was initially at large, over time categories developed in which security would regularly be ordered. Two of these categories bear noting in particular.
8 The most common category appears initially to have been where the plaintiff resided permanently abroad. SFC was justified in such cases because a costs order could not be enforced against the plaintiff should the defendant be successful.
9 For present purposes it is relevant that, with the advent in the 20th Century of reciprocal enforcement of judgment treaties, the traditional obstacles to enforcing judgments against some foreign residents no longer applied, meaning that the rationale for ordering security against them had diminished. Thus in Connop v Varena, Rath J held that a plaintiff who was domiciled in New Zealand was required only to give security for the costs of registering and enforcing a judgment for costs of the Supreme Court of NSW, having regard to the reciprocal enforcement of judgment arrangements between Australia and New Zealand. This was followed by Morling J in Barton v Minister for Foreign Affairs, with respect to a plaintiff domiciled in London who was an undischarged bankrupt in Australia.
10 Another important category of case in which SFC could be ordered was where an impecunious natural person was suing for the benefit of some third party. In Perkins v Adcock, Pollock CB explained that “the principle is that, where the nominal plaintiff is bankrupt or insolvent, or has assigned the debt, and is suing for the benefit of the assignee, he ought to give security for the costs.”
11 Compared with equity, the circumstances in which SFC would be ordered at common law were more limited. Thus in Selby v Alston, Buller J observed that “There are only three instances in which the Court will … oblige the plaintiff to give SFC”, namely, where an infant sues by next friend, where the plaintiff resides abroad, or where proceedings were brought in ejectment where there had been a previous unsuccessful ejectment proceeding the costs of which had not been paid.
12 An important development occurred with the passage of the Joint Stock Companies Act 1856 (UK), s 69 of which enabled SFC to be ordered against impecunious companies. This provision and the ones that followed it reflect a policy decision by the legislature that persons who find themselves involuntarily litigating against limited liability companies should have some protection from the consequences of the limited liability. It is in that sense analogous to the “nominal plaintiff” rule, in that an impecunious company will always be suing for the benefit of those standing behind the company, and if security is not granted then the situation may be that those persons can take the benefit of the litigation without the risk of paying the other party’s costs if unsuccessful.
13 The jurisdiction under rules of this type to order SFC is enlivened if it is shown that there is reason to believe that a company is impecunious (the “threshold question”). Once that is shown, the court has a discretion as to whether security should be required.
14 Another important statutory development, both in relation to SFC and the jurisdiction to order costs more broadly, was the passage of the Supreme Court of Judicature Act 1873 (UK) and its Australian equivalents, pursuant to which superior courts were given a broad discretion to order costs, and the previously disparate approaches in the courts of common law and equity were unified.
15 Today, while the courts retain a general discretion to order SFC as an incident of their power to control their own processes or to stay proceedings, SFC is predominantly governed by statute and court rules. However, such rules generally codify rather than modify the common law position. Thus where it is shown that a plaintiff who is an overseas resident, or a corporation, or who is suing for the benefit of some other person, may not be able to satisfy a costs order in the defendant’s favour, the court’s power to order SFC will be enlivened. The court will, however, retain a discretion as to whether to order SFC, as well as the form it should take and the amount that should be required.
Form of security
16 In early cases SFC generally took the form of a bond supported by solvent sureties (ie a promise by solvent persons to pay a costs order in lieu of payment from the plaintiff), and in at least one case a deposit of money from a foreign plaintiff was found not to be acceptable as a form of security.
17 This position was changed by two important decisions in the mid-1800s. In Cliffe v Wilkinson, Shadwell VC held that it was acceptable for a plaintiff to pay money into court in lieu of a bond, although the amount of security was increased to account for the expense of "bringing the money into Court, and of getting it out again". In Plestow v Johnson, Stuart VC held that a bond from the British Guarantee Association was an acceptable form of security, despite an argument that the Association, as a corporation, was not “some sufficient person” to provide a bond. Thus began the practice of accepting security in the form of a bond from a financial institution. Over the ensuing century and a half the position shifted so markedly that courts began referring to the payment of cash or the provision of a bank guarantee as the “usual form” of security, and, in at least one case, viewing the traditional form of security (an undertaking from a solvent surety) with suspicion.
18 The provision of a bank guarantee as a form of security was considered in Rosengrens v Safe Deposit Centres, although the case concerned security for a judgment debt rather than for costs. The Master at first instance had rejected a proffered bank guarantee from the defendants and held that the security should be provided by payment into court. The England and Wales Court of Appeal overturned that decision and held that a bank guarantee was acceptable. Parker LJ said at 200-201:
“So long as the opposite party can be adequately protected, it is right and proper that the security should be given in a way which is the least disadvantageous to the party giving that security.”
19 Rosengrens was applied to SFC in Blue Oil Energy v Tan. The applicant in that case had offered as SFC a lien over certain shares and an offer of subordination by a company which held security over the shares. It was submitted, in reliance on Rosengrens, that the applicant should be permitted to provide security in the form least disadvantageous or disruptive to it. However, as Beazley P and Tobias JA explained at , Rosengrens did not lay down “some general proposition to the effect that security can only be ordered in favour of a defendant in a manner which is the least disadvantageous to the plaintiff.” Rather, the true issue was “whether the form of security ordered was adequate to protect the party seeking it.”
Early forms of adverse costs insurance in Australia and the growth of litigation funding
20 The first thing resembling ATE insurance in Australia arose in a number of judgments commencing in the mid-1990s in relation to applications by liquidators under what was then s 479(3) of the Corporations Law, seeking directions as to whether they could enter into agreements to fund litigation. These were the decisions that paved the way for the modern-day litigation funding industry. The introduction of adverse costs insurance in Australia was thus inextricably tied to litigation funding, and the approach of the courts was informed by the considerations attending such arrangements.
21 The first relevant decision was Re Movitor. The applicant liquidator sought approval to enter into a “contract of insurance” with an insurer (“Lumley”) in order to prosecute an action against former directors of the company in liquidation for having engaged in alleged voidable transactions. The agreement provided that Lumley would pay 50% of the costs and disbursements associated with the proposed litigation, and would also indemnify the liquidator against 50% of any adverse costs orders. In return, Lumley would receive a “risk premium” of 12% of any recovery made after reimbursement of all expenses. Thus, despite being described as a “contract of insurance”, the agreement resembles what today would be considered a litigation funding arrangement.
22 The main question before the Court was whether the Lumley agreement would be illegal as contrary to public policy because it involved the torts/offences of maintenance (assistance or encouragement to a party to litigation by a person with no interest in the litigation nor any other motive recognised as justifying the interference) and champerty (committing maintenance in return for a share in the proceeds of the action).
23 Drummond J held that the agreement with Lumley involved both maintenance and champerty. Accordingly, the agreement was void unless it fell under a recognised exception. As his Honour noted, one exception to the rule voiding champertous arrangements is that a trustee in bankruptcy may lawfully assign any of the bankrupt’s bare causes of action that have vested in the trustee in return for a share of the proceeds of litigation. His Honour held that the power given to a liquidator under s 477(c) of the Corporations Law to sell the property of the company brought the proposed cause of action within this exception, meaning that the agreement was not void on grounds of champerty.
24 In the course of his judgment in Movitor, Drummond J remarked that it could “be expected that, if there are no legal obstacles to this sort of arrangement, it will become a widespread practice for insolvency practitioners”. Sure enough, in the ensuing years at least two further applications were made by liquidators for the approval of similar arrangements. In Re Tosich, an arrangement was approved which saw the insurer bearing the entire costs of the litigation and insuring against the entire adverse costs risk; and in Re Addstone, an arrangement was approved whereby a bank granted a loan facility, repayable from the proceeds of the litigation, to cover the costs of the litigation plus a guarantee facility specifically to provide security for costs. An insurer then indemnified the bank, the liquidator, and the companies in liquidation against the risk of the facility defaulting and any against adverse costs orders.
25 However, as the market for litigation funding developed in Australia, insurers seemed to lose their appetite for it and bespoke litigation funders took their place. Such funders continued to offer adverse costs insurance as part of the litigation funding arrangements, and were also routinely required to provide SFC by reason of the “nominal plaintiff” rule.
26 Northern Southern v HIH was a claim against an insurer seeking indemnity for damage to a supermarket premises in a fire. The claim was being funded by a commercial litigation funder (“IMF”). The plaintiff offered security in the form of an undertaking from IMF to pay any adverse costs order, and tendered IMF’s half yearly financial reports in order to establish its ability to honour the undertaking. However, Einstein J rejected the undertaking as a form of security, on the basis that the nature of IMF’s business was such that there was a risk to the defendant of IMF becoming unable to meet an adverse costs order.
27 IMF again attempted to provide SFC by way of undertaking in Global Finance, which was a claim against a firm of accountants alleging misleading conduct, negligence, and breach of contract in respect of various audit reports. The plaintiff attempted to distinguish Northern Southern on the basis that IMF’s financial position had significantly improved in the intervening two years. However, Roberts-Smith J held that there was still significant commercial risk and uncertainty in the “developing and still evolving are of business” of litigation funding, and the defendants were entitled to have a “greater degree of financial comfort as to their costs should the plaintiffs’ action fail, than that of an undertaking or guarantee from IMF”. Notably, as IMF’s business and financial standing developed it eventually began routinely providing security by way of undertaking, which undertakings were no longer challenged by defendants.
28 The growth of litigation funding in Australia was accelerated when the High Court in Fostif held that litigation funding arrangements outside of the insolvency context, despite being champertous, were not necessarily void as against public policy.
29 Another significant case was Nylex. The plaintiff was the subsidiary of a company that had recently appointed voluntary administrators and whose assets and subsidiaries, the plaintiff included, had been placed into receivership by secured creditors. The plaintiff’s claim against the defendant was brought in large part as a subrogated claim on behalf of the plaintiff’s insurers. The plaintiff attempted to argue that despite its obvious financial distress, it should not be required to provide SFC because it was indemnified by the subrogated insurers against adverse costs. Mandie J held that this was not sufficient on two grounds. First, there was no evidence of the relevant insurance policies or the extent to which the action was subrogated. Second, given that the plaintiff may be insolvent, there was a risk that the proceeds of any indemnity would be paid to the pool available to all creditors and would not be directly available to the defendant.
30 The plaintiff then proffered security in the form of an undertaking from “QBE Casualty Syndicate 386 at Lloyds”, pursuant to which various underwriters would be liable for any adverse costs in proportion to their respective liability under the insurance policy. Mandie J did not accept the undertaking as a form of security on two grounds. First, the evidence adduced by the plaintiff as to the solvency of the insurers was hearsay only, and his Honour was reluctant to accept such material “[i]n the current global financial climate” (ie March 2009, the height of the “Global Financial Crisis”). Second, his Honour was concerned that the proposed undertaking relied on two warranties of authority from the person who purported to sign it on behalf of the syndicates of insurers, and that it offered only proportional liability by the various insurers.
ATE insurance in the UK
31 The development of adverse costs insurance in the UK took a somewhat different turn from Australia. At the same time as cases like Movitor were permitting litigation funding arrangements to be entered into in the Australian insolvency industry, statutory differences resulted in similar arrangements being refused in the UK on grounds of champerty. Meanwhile, developments in the ways solicitors could charge for their fees had a substantial influence on the market for insurance with respect to adverse costs.
32 In Australia, in 1960 the High Court held that it was lawful in certain circumstances for a solicitor to agree to charge the client only if the client was successful in recovering damages in the litigation. However, such “conditional fee agreements” (“CFAs”) continued to be deemed against public policy in the UK on grounds of champerty. That changed in 1995, with the introduction of regulations permitting CFAs. At about the same time the Law Society of England and Wales, together with insurance brokers, developed a new form of insurance cover, ATE insurance, to provide cover against the risk of adverse costs orders to clients who entered into a CFA. Thus, unlike in Australia, ATE insurance took on a life of its own and was not tied to litigation funding (which in the UK was still considered champertous and therefore unlawful).
33 In order to encourage a varied market for such insurance products, the UK government introduced in 1999 and 2000 express legislation permitting ATE insurance premiums to be recovered as part of adverse costs orders.
34 Soon after those reforms the potential implications of ATE insurance for SFC were considered in Nasser v United Bank of Kuwait, which concerned an application for security against an impecunious USA resident. At , Mance LJ adverted in obiter to the “interesting possibility” of a claimant relying on an ATE policy itself as security. His Lordship stated:
“I would think that defendants would, at the least, be entitled to some assurance as to the scope of the cover, that it was not liable to be avoided for misrepresentation or non-disclosure (it may be that such policies have anti-avoidance provisions) and that its proceeds could not be diverted elsewhere.“
35 The issue was again raised in Al-Koronky v Time-Life Entertainment Group, which concerned SFC application in a defamation claim made by Sudanese claimants against a British newspaper. The claimants had an ATE policy, however it was voidable if their liability for costs was consequent on them having not told the truth. Sedley LJ held accordingly that the ATE policy was effectively useless, because the outcome of the case turned on whether the claimants were telling the truth and thus it was overwhelmingly likely that if the claimants were unsuccessful in the proceeding their insurance cover would be ineffective.
36 In Belco Trading Co v Kondo, the claimant, a corporation domiciled in Russia, sought to offer an ATE insurance policy as security, although no policy had yet been procured. At first instance, Judge Reddihough gave the claimants an option of giving security by way of ATE policy providing “equal or better security” to payment into court or bank guarantee. The claimants appealed on the grounds that the “equal or better security” condition could never be met. The Court of Appeal refused leave to appeal, but indicated that it was unlikely that any standard form of ATE insurance could provide a suitable alternative to payment into court or bank guarantee. The reasons identified for this were that, first, there was likely to be a possibility of the insurer avoiding the policy; and, second, the ATE policy is in favour not of the defendant (whose costs are required to be secured), but of the claimant (who is required to provide the security).
37 These decisions were drawn together in Michael Phillips Architects Ltd v Riklin. The claimant in that case had obtained an ATE policy and argued that SFC was accordingly not required. Akenhead J considered Nasser, Al-Koronky, and Belco, and relevantly extracted from them four propositions, namely that: first, there is no reason in principle why an ATE policy, subject to its terms, could not provide an element of security for the defendant’s costs; second, an ATE policy will rarely provide as good protection as payment into court or a bank bond or guarantee, because, inter alia, the policy is in favour of the insured, not the defendant, and the insurer may be able to void the policy; third, a claimant seeking to rely on an ATE policy as security needs to establish that the policy actually provides security, in that the insurer should not readily be able to avoid liability for the defendant’s costs; and, fourth, there is no reason in principle why the amount fixed for security could not be reduced to take into account a realistic probability that an ATE policy would cover the defendant’s costs.
38 Considering the terms of the policy itself, his Honour was persuaded that it provided no real security for the defendants. The most significant consideration in that respect was that there were a number of provisions pursuant to which the insurer could potentially cancel the policy, including:
(a) where the insured failed to observe a variety of conditions requiring it to, for example, tell the insurer immediately of anything that may alter its assessment of the claim, and try to prevent anything happening that may cause a claim under the policy;
(b) where the insured made a claim which was fraudulent or false (noting that it was “at least within the realms of possibility” that there could be findings of fact to this effect, based on the nature of the proceedings);
(c) where the insurer formed the view that the claim was not more likely than not to succeed; and
(d) at any time by at least 21 days’ written notice to the insured.
39 On his Honour’s reading, it was at least highly arguable that if the policy was cancelled before the obligation to pay had accrued (such as where a court order was made for the payment of costs), the insurer would have no obligation to pay any costs, and the defendant would be left with no protection. Accordingly, the policy did not provide acceptable security.
40 Riklin was considered by Stuart-Smith J in Geophysical Services Centre Co Ltd v Dowell Schlumberger, in which the claimant was a company domiciled in Jordan. His Honour relevantly doubted Akenhead J’s fourth proposition, ie that the amount of security may be reduced by the amount by reason of the ATE policy. However, his Honour otherwise followed Akenhead J’s summary of the relevant principles, and held accordingly that the court was required to form a view on:
(a) the meaning of the policy and on how readily it may be avoided; and
(b) the likelihood of circumstances arising which will enable the policy to be avoided.
41 It should be noted that the second of these was not expressly stated by Akenhead J, who appeared to consider only whether the policy could readily be avoided and not whether such avoidance was likely.
42 Stuart-Smith J concluded his analysis by stating that the ultimate question was “not whether the assurance provided by an ATE policy is better security than cash or its equivalent, but whether there is reason to believe that the claimant will be unable to pay the defendant’s costs despite the existence of the ATE policy.” In other words, his Honour saw the existence of the ATE policy as going to the threshold question of whether security should be granted at all, rather than the form of security.
43 His Honour proceeded to consider the policy that was before the Court. There were some important differences between that policy and the one in Riklin, including in particular that if the policy was cancelled then the insurer would be liable for the defendant’s costs incurred up to the date of cancellation. Further, there was an “anti-avoidance” provision of the type adverted to by Mance LJ in Nasser, in that the insurer had no right to avoid the policy on the grounds of misrepresentation or non-disclosure, except for fraudulent misrepresentation by the insured.
44 Consistently with his Honour’s earlier observations, in determining whether to order SFC Stuart-Smith J considered not only the ability of the insurer to cancel or avoid the contract, but also the probability of this occurring. Thus, while the policy could be avoided for fraudulent misrepresentation, his Honour found that there was nothing more than a “theoretical chance” that the insurers might actually seek to avoid the policy on this basis, because his Honour was prepared to infer that the information provided by the insured to the insurer was consistent with the claim in the proceeding, and it was not the defendant’s case that the claim was fraudulent.
45 Similarly, his Honour found that there was nothing more than a theoretical possibility that the insured would breach the contract and the insurer would accordingly cancel it, because:
(a) the conditions of the policy were not onerous;
(b) the claimant had no commercial interest in breaching the conditions;
(c) the claimant was represented by experienced and competent lawyers who could advise it on its obligations under the policy; and
(d) there was a long-standing relationship between the solicitors and the insurer, which made it commercially less likely that the insurer would seek to avoid the claim on a spurious basis.
46 Accordingly, his Honour held that there was no reason to believe that the claimant would be unable to pay an adverse costs order, and thus the application failed to satisfy the threshold question.
47 Stuart-Smith J’s decision in Geophysical was followed by Anderson QC (Sitting as a Deputy Judge of the England and Wales High Court) in NGM Sustainable Developments v Wallis. At , his Honour observed that there is always a theoretical possibility that a claimant will become unable to pay a defendant’s costs, but such possibilities do not provide “reason to believe” that it will be unable to do so. As in Geophysical, his Honour considered the clauses of the policy that could result in avoidance or cancellation and determined that the chance of such avoidance or cancellation occurring was no more than theoretical.
48 Another issue that arose in NGM, echoing the consideration in Nylex referred to above, was the possibility that if the claimant became insolvent the insurance would be shared between its creditors. In order to address this, the claimant offered to undertake to direct the insurer to pay all proceeds to the defendants, and to enter into an equitable assignment of its contingent right to the proceeds. Anderson QC held that this satisfied the concern. Accordingly, the power to order security did not arise.
49 In Premier Motorauctions v PWC the England and Wales Court of Appeal had occasion to revisit these authorities. The case concerned a claim by certain companies in liquidation and their liquidator. At first instance Snowden J had declined to order security based on an ATE policy procured by the claimants, in reliance on Geophysical. The defendants appealed that ruling.
50 Longmore LJ (Kitchin and Floyd LJJ agreeing) confirmed that the existence of an ATE policy can be relevant at the threshold stage of a SFC determination. His Lordship largely affirmed Stuart-Smith J’s decision in Geophysical, but distinguished the facts of that case because, unlike in Geophysical, there was no relevant “anti-avoidance” provision. Further, unlike in Geophysical (but similar to Al-Koronky and Riklin),the outcome of the proceeding turned on whether the claimant’s key witness would be believed, meaning that there was a real risk of the policy being avoided if the claimants were unsuccessful. Accordingly, his Lordship held that there was reason to believe that the claimant would be unable to pay the defendant’s costs if ordered to do so.
ATE insurance in Australia up to DIF III
51 With the development of the market for ATE insurance in England it was inevitable that such products would sooner or later make their way to Australia. The first mention of them in a published judgment appears to be Kelly v Willmott Forests, which referred to insurance being obtained from AmTrust Europe Ltd (“AmTrust”) by the applicants’ solicitors in 2010. However, the decision focussed more on the inability to obtain further insurance rather than whether the ATE policy was an adequate form of security.
52 The first published judgment concerning an attempt to provide SFC by way of ATE policy in an Australian court was DIF III Global Co-Investment Fund LP v BBLP LLC (“DIF III (First Instance)”), which was closely followed by Australian Property Custodian Holdings Ltd (in liq) v Pitcher Partners (“APCH”). Both were decisions of associate justices of the Supreme Court of Victoria, In both cases the plaintiffs offered to provide SFC by way of deed of indemnity from AmTrust plus the provision of conventional security in the sum of $20,000 to cover the cost to the defendants of registering a judgment against AmTrust in the UK (in accordance with the decision in Connop, as referred to above). DIF III (First Instance) was subsequently appealed to a justice.
53 In the DIF III cases the plaintiffs were foreign corporations. The proposed deed of indemnity relevantly provided that:
(a) AmTrust unconditionally and irrevocably undertook to pay the defendants’ costs up to a specified limit;
(b) a certified copy of a relevant court order or a signed agreement was conclusive evidence of the liability to pay without AmTrust enquiring further;
(c) AmTrust was deemed the principal debtor and not a surety of the plaintiffs;
(d) AmTrust’s liability under the deed was not subject to avoidance on the grounds of fraud or misrepresentation by the plaintiffs;
(e) the deed was governed by the laws of Victoria and subject to the exclusive jurisdiction of the courts of Victoria; and
(f) (after initial discussion on this point) AmTrust undertook not to seek security for its costs in respect of any application in the UK to register any costs order.
54 At first instance Landsdowne AsJ began her consideration of the relevant principles by noting that the purpose of security was to provide a fund or asset against which the defendant can enforce an order for costs, and that the court has an unfettered discretion as to what form of security may be acceptable. Her Honour noted that, countervailing considerations in relation to the plaintiff’s absence from the jurisdiction include whether the plaintiff has assets in a foreign jurisdiction in which a costs order could be enforced, and the complexity and costs of such enforcement. Her Honour referred in that respect to the cases establishing that where a foreign plaintiff is in a foreign jurisdiction in which an Australian judgment may be enforced, sufficient SFC may be the costs of enforcement in that jurisdiction.
55 Landsdowne AsJ held that the AmTrust deed did not provide sufficient security for the defendants’ costs because it was inferior to the provision of a bank guarantee. Her Honour’s reasons included that:
(a) enforcement of the deed against AmTrust would be less certain and involve more time and complexity than enforcing a bank guarantee against an Australian bank;
(b) the proposed deeds had not been executed by AmTrust;
(c) there was no direct evidence from AmTrust as to its financial standing (referring to Nylex);
(d) there was no evidence of AmTrust’s willingness to provide further tranches of security; and
(e) there was no evidence that the plaintiffs could not obtain a bank guarantee.
56 In APCH, Ierodiaconou AsJ distinguished DIF III (First Instance) on the grounds that, in that case, there was direct evidence from AmTrust, the proposed deed was executed, and there was agreement on its terms. Her Honour considered that the AmTrust deed before her was appropriate as security because:
(a) the deed was directly enforceable in Victoria as it was governed by the laws of Victoria and subject to the exclusive jurisdiction of the Supreme Court of Victoria;
(b) the deed was directly enforceable against AmTrust;
(c) the deed was irrevocable and unconditional;
(d) the plaintiff offered to pay into court the enforcement costs of the deed in the UK;
(e) AmTrust was an insurer with very substantial assets and was unlikely to default on the deed; and
(f) the terms of the deed were acceptable to the defendant.
57 Accordingly, her Honour accepted the AmTrust deed as adequate SFC.
58 In DIF III (Appeal), Hargrave J agreed with Landsdowne AsJ’s summary of the applicable principles at first instance, save that Landsdowne AsJ had approached the task as one of comparing the form of security proffered by the plaintiffs with the “conventional” forms of security to see which was superior, which Hargrave J considered to be an incorrect test. The question rather was whether the form proffered by the plaintiffs was adequate to protect the defendants’ interests. His Honour also held that many of the matters relied on by Landsdowne AsJ were either erroneous or irrelevant:
(b) the basis on which AmTrust entered into the deed, whether AmTrust was willing to provide further security, that the deed was not executed, and that there was no evidence that the plaintiffs could not obtain a bank guarantee, were irrelevant.
59 Accordingly, based on similar matters to those relied on by Ierodiaconou AsJ in APCH, his Honour held that the AmTrust deed provided adequate security.
Cases following DIF IIII
60 Whilst DIF III (Appeal) and APCH established that a deed of indemnity from an overseas insurer can suffice as SFC in certain cases, it did not end the difficulties confronting Australian litigants seeking to rely on ATE policies as a form of security (or to avoid an order to provide security).
61 In Petersen, a class action funded by a litigation funder, the applicant attempted to rely on an ATE policy from AmTrust as security without proffering a deed of indemnity of the type provided in DIF III and APCH. The applicant relied in particular on the respondent’s ability to bring proceedings against AmTrust directly under the recently enacted Civil Liability (Third Party Claims Against Insurers) Act 2017 (NSW) (“CL(TPCAI) Act”). That Act applied where an insured had a liability to another person which liability was covered by the policy of insurance. It allowed the other person, subject to leave of the court, to bring a claim directly against the insurer for the amount of the indemnity.
62 In Peterson, Yates J was referred to the UK authorities. Based on those authorities his Honour held that an appropriately worded ATE policy might be capable in of providing sufficient SFC. However, his Honour determined that the AmTrust policy in that case did not provide sufficient security for the respondents’ costs, including because:
(a) while the applicant’s principals had undertaken to make claims under the policy to meet any adverse costs order, and to progress such claims expeditiously, those undertakings were not directly enforceable by the respondents and it was not clear that the applicant had the means to progress such claims;
(b) there was no undertaking from the insured to sue AmTrust to enforce any legitimate claim that AmTrust was not prepared to meet, and there was reason to doubt that the applicants could do so, in circumstances where AmTrust had made it clear that it would rely on its legitimate rights;
(c) the CL(TPCAI) Act did not assist because:
(i) it created a local statutory right in circumstances where AmTrust was absent from the jurisdiction;
(ii) the respondents would need to seek leave to proceed against AmTrust; and
(iii) if a claim could be made against AmTrust, Amtrust could rely on the defences and exclusions arising under the policy;
(d) the AmTrust policy could be avoided for fraudulent non-disclosure, and there was no information as to what information AmTrust had;
(e) there were a number of exclusions from liability which were “more extensive and onerous” than those in the English cases, such as exclusions in relation to any negligent act or omission by the applicant’s legal representatives or a decision by the applicant to continue the litigation without AmTrust’s approval;
(f) the premium for an unconditional indemnity in favour of the respondents was about $550,000, which indicated AmTrust considered the risks to it of providing such an indemnity to be materially higher than providing the ATE policy;
(g) it was possible that the policy could be cancelled before any obligation to pay arose, meaning the respondents would have incurred costs but would be left without security;
(h) there was a real question as to whether the proceeds of the policy would be available to the respondents if the applicant went into liquidation; and
(i) the clauses of the policy requiring the applicant to conduct the proceedings in certain ways were in apparent conflict with the applicant’s duties to the Court.
63 Another attempt to rely on an ATE policy as security was made in Roo Roofing, a decision of Landsdown AsJ.The plaintiff in that matter was supported by a litigation funder which was domiciled in the Cayman Islands but which maintained liquid funds in the UK. The funder had procured an ATE policy that the defendant was able to call upon directly. The security proffered was an undertaking from the funder to be responsible for any adverse costs, supported by the ATE policy, and the deposit of funds to register any judgment in the UK or the Cayman Islands.
64 Landsdown AsJ held that the applicable principles were those established in DIF III (Appeal), subject to what her Honour referred to as a “possible qualification” arising from the dicta of Tate and Kyrou JJA in Trailer Trash. That case concerned an appeal regarding the quantum of security for costs ordered by the County Court. The primary judge had accepted as security an undertaking from the father of the plaintiff’s sole director and shareholder in lieu of payment of funds into court.
65 While the form of security was not in issue on the appeal, Tate and Kyrou JJA were critical of the primary judge’s decision in that respect. In particular, by reference to the overarching purpose of civil case management “to facilitate the just, efficient, timely, and cost-effective resolution of the real issues in dispute”, as specified in ss 7-9 of the Civil Procedure Act 2010 (Vic), their Honours stated at  that where a court is presented with a choice between a “liquid form” of security and an undertaking by a third party, the court should ordinarily prefer the liquid form. Their Honours went on to say:
“a form of security for costs which does not provide a fund which can be accessed without the cooperation of the opposing party or a person who is connected to that party — and may require the commencement of proceedings to enforce it — has the potential to undermine the overarching purpose. This is because that form of security can give rise to satellite proceedings and additional delay and costs. Such satellite proceedings are contrary to the principle of finality in litigation.”
66 On the basis of those remarks, Landsdown AsJ held that the policy in Roo Roofing was inadequate to provide security because it was “a fund that can only be accessed with the co-operation of the plaintiffs, and may require a proceeding to enforce it”.
67 In Tiaro Coal, Gleeson JA approved a proposed AmTrust deed as a form of security, subject to the resolution of an infelicity of drafting in the use of the words “final costs certificate” and, more significantly, the deletion of a provision that AmTrust would not be liable to the extent that the provision of an undertaking or the making of a payment under the deed “would expose AmTrust to any sanction, prohibition or restriction under United Nations resolutions, and or the trade and economic sanctions, laws and or regulations of the European Union, United Kingdom, United States of America and or of Australia.” As his Honour noted, it was uncertain in what circumstances that clause may come into effect, and this posed an unacceptable risk to the defendant.
68 Most recently, a series of decisions in the Supreme Court of Queensland in relation to proposed AmTrust deeds have resulted in apparently contradictory outcomes. In the Equititrust cases, Bowskill J and then Bond J rejected proposed AmTrust deeds on the grounds that they risked creating satellite litigation regarding whether all of the pre-conditions to establishing AmTrust’s liability under the deeds had been satisfied. Of concern in both cases were the clauses requiring a valid costs determination certificate before AmTrust’s liability was engaged (although Bond J also expressed some concern regarding enforcing a judgment against AmTrust in the UK in light of the COVID-19 pandemic). At  Bond J observed that:
“A truly unconditional undertaking by a surety would be indifferent to the question whether, as between the principal and the obligee, the principal is yet obliged to pay the obligee in a particular amount (and therefore justified in making a demand on the surety in a particular amount).”
69 In contrast, in other cases in the Queensland Supreme Court decided at approximately the same time, deeds of indemnity from AmTrust have been accepted as security, in reliance on DIF III (Appeal) and the cases that followed.
70 The approach taken by Landsdown AsJ in Roo Roofing and by Bowskill J and Bond J in the Equititrust cases warrants further comment. In this author’s respectful opinion, those decisions are in some respects difficult to reconcile with settled authority and principle in relation to SFC, particularly the line of cases commencing with Connop referred to above.
71 It bears noting that the threshold question in relation to a grant of security is whether the plaintiff has sufficient assets in the jurisdiction to satisfy a likely costs order. SFC will not be required if the plaintiff has such assets. It is of course possible that an unsuccessful plaintiff will pay the defendant’s costs without putting up much resistance, but it is equally possible that this will not occur, leaving the defendant in the position of having to enforce the costs order.
72 To enforce a costs order first requires the costs to be assessed and a costs certificate to be issued. The certificate can then be registered and enforced as a judgment debt. In that respect it is curious that Bond J, in the passage quoted above, appeared to indicate that the proposed AmTrust deed was not sufficient as security because it required such a certificate in order to be enforced. No defendant could ever enforce a costs order without obtaining and registering a costs certificate.
73 After the costs certificate has been registered the defendant has a number of tools with which the costs certificate can be enforced, including court processes such as examination orders, writs for levy of property, and garnishee orders; as well as insolvency processes such as the issue of a bankruptcy notice or a creditor’s statutory demand, followed by an application to bankrupt or wind-up the plaintiff. In other words, a defendant who succeeds against a pecunious plaintiff may have to pursue a variety of proceedings against the plaintiff in order to recover its costs. Each such process can be contested by the plaintiff, and accordingly has the capacity to create satellite litigation.
74 It follows that a defendant whose costs are secured by a sum of money paid into court or by an unconditional bank guarantee is in a significantly better position than a defendant facing a pecunious but uncooperative plaintiff. However, pecunious plaintiffs are not required to provide SFC. It seems unlikely that the form of security provided pursuant to an order for SFC should be required to place the defendant in a better position than it would be in if the plaintiff had sufficient assets to satisfy an adverse costs order.
75 It should also be recalled that an undertaking from a pecunious natural person was originally the only acceptable form of security. This may have fallen out of favour in recent times, but those old authorities have not been overruled (and they do not appear to have been drawn to the Court’s attention in Trailer Trash). The provision of security in this manner would place the defendant in a comparable position to one facing a pecunious plaintiff, including the risk of satellite litigation.
76 In view of those matters, the risk of satellite litigation attending the enforcement of a deed of indemnity given by an unquestionably solvent insurer based in a country like the UK, which has reciprocal enforcement arrangements with Australia, may arguably have been given undue weight in the Roo Roofing and Equititrust decisions.
77 It emerges from the above analysis that courts have taken significantly different approaches to the provision of security for costs by ATE insurance policies. The courts in the UK seem to have been far more liberal in that respect than those in Australia, in that a claimant holding an appropriate ATE policy has been held to be sufficient security for a defendant, while in Australia at the very least a deed of indemnity from the insurer in favour of the defendant has been required. Further, the Australian courts seem more reluctant to consider the ATE policy at the threshold level, and have rather considered whether the policy is an adequate form of security rather than its impact on the question of whether the plaintiff could meet an adverse costs order. On the other hand, British courts have been willing to consider an ATE policy as a factor indicating that the plaintiff would be able to meet a costs order against it.
78 Relatedly, Australian courts have not adopted the approach taken by Stuart-Smith J in Geophysical of considering not only the ways in which the policy could be cancelled or avoided, but the actual likelihood of this eventuating; and neither have they given weight to the commercial interest the insured or its solicitors have in complying with the conditions of the policy. It seems to have been sufficient in Australia to reject the policy as a form of security that there is a risk of the policy being cancelled or avoided over which the defendant has no control. It should be noted in that respect that this risk is not illusory—in at least one British case an ATE insurer has successfully avoided the policy after the defendant was successful.
79 An analysis of the authorities does, however, provide some guidance as to what is required in order that an ATE policy be accepted as satisfactory security for costs. The following in particular is apparent:
(a) it is preferable for the policy to be directly enforceable by the defendant;
(b) there should be an “anti-avoidance provision”, such that the policy cannot be avoided by the insurer;
(c) if the policy can be avoided for fraudulent misrepresentation by the insured, the court will more readily accept the policy if the issues in the case do not have the potential to implicate the claimant in fraudulent conduct;
(d) the policy should provide cover for the costs incurred by the defendant up to the date it is cancelled or avoided, such that its cancellation results in the insurer only avoiding liability going forward and not altogether;
(e) the policy should be subject to the laws and the courts of an Australian jurisdiction; and
(f) if the insurer is domiciled overseas, it should be in a country where reciprocal enforcement of judgments is available, and it should undertake not to seek security for costs in relation to any enforcement proceedings. In such circumstances the plaintiff will also be required to pay an amount into court sufficient to enforce a judgment against the insurer.
80 The extent to which those principles will be consistently applied in Australia is, however, uncertain. As indicated by the above analysis, Australian courts have not always taken the same approach, particularly as regards the risk of satellite litigation. The present uncertainty is such that the area would benefit from a decision at an intermediate appellate level in the near future.
 LLB (Hons I) (UTS), BSc (Sydney). Barrister, New South Wales Bar (University Chambers).  See, eg, Dudley v Born (1658) Sty 322; 82 ER 745; see generally, Stephen Colbran, “The Origin of Security for Costs” (1993) 14 The Queensland Lawyer 44. J H Billington Ltd v Billington  2 KB 106, 109-110 (Lord Alverstone CJ). Commissioner of Taxation v Vasiliades  FCAFC 170; 344 ALR 558,  (Kenny and Edelman JJ).  Stephen Colbran, “The Origin of Security for Costs” (1993) 14 The Queensland Lawyer 44, 46; Oshlack v Richmond River Council (1998) 193 CLR 72,  (Gaudron and Gummow JJ). Jones v Coxeter (1742) 2 Atk 400; 26 ER 642 (Lord Hardwicke). Burford Corporation v Lenthall (1743) 1 Atk 551; 26 ER 731 (Lord Hardwicke); Andrew v Barnes (1888) 39 Ch D 133, 138 (Fry LJ).  Stephen Colbran, Security for Costs (Longman Professional, 1993), [2.14]. Elan v Rees (1784) 3 Doug 382; 99 ER 708; cf, Kinaston v Miller (1762) Dick 773; 21 ER 471; Meliorucchy v Meliorucchy (1750) Ves Sen Supp 260; 28 ER 519; Tenant v Brown (1826) 5 B & C 208; 108 ER 78; and see, Kohn v Rinson & Stafford (Brod) Ltd (1948) 1 KB 327. Connop v Varena Pty Ltd  1 NSWLR 71. Barton v Minister for Foreign Affairs (1984) 2 FCR 463.  See also, DS Parklane Developments Pty Ltd v Korea First Finance Ltd (Unreported, Supreme Court of NSW, Santow J, 20 August 1997, BC9703785); Maxim’s Caterers Limited v Magnona Pty Ltd (No 1)  FCA 450; Dense Medium Separation Powders Pty Ltd v Gondwana Chemicals Pty Ltd  NSWCA 84, - (Young JA; Whealy and Campbell JJA agreeing).  See, Wale v Salter (1728) Sel Cas T King 47; 25 ER 262; Webb v Ward (1797) 7 TR 296; 101 ER 983  (1845) 14 M & W 808; 153 ER 703.  (1786) 1 TR 491; 99 ER 1214.  See generally, Stephen Colbran, Security for Costs (Longman Professional, 1993), Ch 4.  See, Ibid, [3.39]-[3.41], [4.28].  Presently s 1335 of the Corporations Act 2001 (Cth), r 42.21(1)(d) of the Uniform Civil Procedure Rules 2005 (NSW), and its other state cognates. Buckley v Bennell Design & Construction Pty Ltd (1974) 1 ACLR 301, 303 (Street CJ). Bryan E Fencott and Associates Pty Limited v Eretta Pty Limited (1987) 16 FCR 497, 511 (French J).  Ibid.  See, Oshlack v Richmond River Council (1998) 193 CLR 72, - (Gaudron and Gummow JJ).  Stephen Colbran, Security for Costs (Longman Professional, 1993), [5.12]-[5.13].  See, Hassoun v Wesfarmers General Insurance Ltd t/as Lumley General  NSWCA 33; 18 ANZ Ins Cas 62-071, -.  See generally John Hamilton QC et al, NSW Civil Procedure Handbook (Lawbook Co, 9th ed, 2019), [42.21.40]-[42.21.100].  Stephen Colbran, Security for Costs (Longman Professional, 1993), [3.42]; see, eg, Goodwin v Archer (1727) 2 P Wms 452; 24 ER 809. Ker v The Dutchess of Munster (1718) Bunb 35; 145 ER 586.  (1830) 4 Sim 122; 58 ER 47.  (1853) 1 Sm & G App 20; 67 ER 1327. Trailer Trash Franchise Systems Pty Ltd v GM Fascia & Gutter Pty Ltd  VSCA 293, - (Tate and Kyrou JJA). Rosengrens Ltd v Safe Deposit Centres Ltd  3 All ER 198 (“Rosengrens”). Blue Oil Energy Pty Limited v Tan  NSWCA 81. Re Movitor Pty Ltd (In Liq) (1996) 64 FCR 380 (“Movitor”). Movitor, 387 (Drummond J); citing Halsbury's Laws of Australia, Vol 6, [110-7135] and [110-7140]. Movitor, 389. Movitor, 386. Re Tosich Construction Pty Ltd; Ex parte Wily (1997) 73 FCR 219. Re Addstone Pty Ltd (In Liq) (1998) 83 FCR 583 ; and see further, Elfic Ltd v Macks  2 Qd R 125; 181 ALR 1; Emanuel Management Pty Ltd & Ors v. Foster’s Brewing Group Ltd & Ors  QSC 205 (2003) 178 FLR 1; Emanuel Management Pty Ltd (in liq) v Foster’s Brewing Group Ltd  QSC 299. Northern Southern Western Supermarkets Pty Limited (subject to DOCA) v HIH Casualty & General Insurance Limited (in Liquidation)  NSWSC 541 (“Northern Southern”). Northern Southern,  (Einstein J). Northern Southern, - (Einstein J). Global Finance Group Pty Ltd (in liq) v Marsden Partners (a firm)  WASC 52. Ibid, - (Roberts-Smith J).  Based on author’s observations and experience. Campbells Cash and Carry Pty Limited v Fostif Pty Ltd  HCA 41; 229 CLR 386. Nylex Corporation Pty Ltd v Basell Australia Pty Ltd  VSC 97 (“Nylex”).  See, Nylex, - (Mandie J). Nylex,  (Mandie J); and cf, Corporations Act 2001 (Cth) s 562 (noting that the plaintiff was not in liquidation). Nylex,  (Mandie J). Nylex, -, and see at  (Mandie J); note that a similar issue arose in CCS Equipment Pty Ltd v Galaxy Resources Limited  WASC 249.  See, eg, Grovewood Holdings Pic v James Capel & Co Ltd  Ch 80, 83-84, 86-88; Re Oasis Merchandising Services Ltd (1995) 2 BCLC 493. Clyne v The New South Wales Bar Association (1960) 104 CLR 186.  See generally, Awwad v Geraghty & Co  3 WLR 1041. Callery v Gray (No 1)  EWCA Civ 1117; 3 All ER 833,  (Lord Woolf CJ). Ibid,  (Lord Woolf CJ). Ibid, -, -,  (Lord Woolf CJ). Nasser v United Bank of Kuwait  EWCA Civ 556;  1 All ER 401 (“Nasser”). Al-Koronky v Time-Life Entertainment Group  EWCA Civ 1123;  1 Costs LR 57 (“Al-Koronky”).  Ibid, . Belco Trading Co v Kondo  EWCA Civ 205 (“Belco”). Belco, . Michael Phillips Architects Ltd v Riklin  EWHC 834 (“Riklin”). Riklin, .  Note that, as explained below, the fourth proposition has not subsequently been followed.  See, Michael Phillips Architects Ltd v Riklin  EWHC 834, -. Riklin, . Geophysical Services Centre Co Ltd v Dowell Schlumberger  EWHC 147 (“Geophysical”). Geophysical,  Geophysical, . Geophysical, .  See, Geophysical, -. Geophysical, . Geophysical, -. Geophysical, -, -. NGM Sustainable Developments Ltd v Wallis  EWHC 461 (“NGM”).  NGM, -.  cf, Nylex,  (Mandie J). NGM. -. Premier Motorauctions Ltd (in liq) v PWC  1 WLR 2955 (“Premier Motorauctions”). Premier Motorauctions,  Premier Motorauctions -. Kelly v Willmott Forests Ltd (in liquidation) (No 3)  FCA 78.  See at , , - DIF III Global Co-Investment Fund LP v BBLP LLC  VSC 484 (“DIF III (First Instance)”). Australian Property Custodian Holdings Ltd (in liq) v Pitcher Partners  VSC 513 (“APCH”). DIF III Global Co-Investment Fund, L.P. & Anor v BBLP LLC & Ors  VSC 401 (“DIF III (Appeal)”). DIF III (Appeal), . DIF III (First Instance), -; citing Yara Australia Pty Ltd v Oswal  VSCA 156; 41 VR 245. DIF III (First Instance), . DIF III (First Instance), -. DIF III (First Instance), -. DIF III (First Instance), -. DIF III (First Instance), . DIF III (First Instance), -. DIF III (First Instance), . APCH, . APCH, . DIF III (Appeal), -. DIF III (Appeal), -.  DIF III (Appeal), -.  DIF III (Appeal), .  DIF III (Appeal), .  Petersen Superannuation Fund Pty Ltd v Bank of Queensland Limited  FCA 699 (“Petersen”).  Petersen, .  See, Petersen, -.  Roo Roofing Pty Ltd v Commonwealth of Australia  VSC 694.  Trailer Trash Franchise Systems Pty Ltd v GM Fascia & Gutter Pty Ltd  VSCA 293 (“Trailer Trash”).  Roo Roofing, .  In the matter of Tiaro Coal Limited (in liq)  NSWSC 746.  Equititrust Limited v Tucker  QSC 51, - (Bowskill J); Equititrust Limited v Tucker  QSC 269, -, Annexure A (Bond J).  See, Equititrust Limited v Tucker  QSC 269, Annexure A, Item 9 (Bond J).  Murphy & Ors v Gladstone Ports Corporation Ltd  QSC 12, - (Crow J); Murphy Operator & Ors v Gladstone Ports Corporation (No 6)  QSC 192 - (Crow J); Mallonland Pty Ltd & Anor v Advanta Seeds Pty Ltd (No 2)  QSC 21 (Jackson J).  See, eg, Legal Profession Uniform Law Application Act 2014 (NSW) s 71.  See, eg, Civil Procedure Act 2005 (NSW) Pt 8.  Bankruptcy Act 1966 (Cth) s 41.  Corporations Act 2001 (Cth) s 459E.  Persimmon Homes Ltd v Great Lakes Reinsurance (UK) plc  Lloyd’s Rep IR 101.