key: cord-0049281-wgypsi5o authors: Grund, Sebastian title: The Quest for a European Safe Asset—A Comparative Legal Analysis of Sovereign Bond-Backed Securities, E-Bonds, Purple Bonds, and Coronabonds date: 2020-08-13 journal: nan DOI: 10.1093/jfr/fjaa009 sha: ebb32c9574733c6eb58cab573d70cbbfe3f9d261 doc_id: 49281 cord_uid: wgypsi5o The European sovereign debt crisis and, more recently, the COVID-19 pandemic have revealed the European Economic and Monetary Union’s fragility, which essentially emanates from the inherent tension between a single monetary policy and decentralized fiscal policies. To cushion economic and financial shocks and sever the sovereign-bank doom loop, different proposals to create a common public debt security have been put forward, although none of them has so far seen the light of day. Building on pertinent economic and finance scholarship, this article reviews four promising safe asset proposals from a legal perspective: Sovereign bond-backed securities (SBBS), E-bonds, Purple bonds, and Coronabonds. Rather than focusing on their feasibility under EU law or national constitutional law, this article compares the proposals from an investor perspective against the backdrop of the following formal and functional legal characteristics that render assets ‘safe’: governing law, dispute settlement forum, investor protection, and investor representation in sovereign debt restructurings. Against this backdrop, targeted recommendations on critical design elements of safe assets, with the aim of reconciling the economic policy objectives with the pertinent legal constraints, are advanced. The Global Financial Crisis of 2007-2008 (GFC) has laid bare the importance of safe assets for our modern global financial and economic system. 1 When banks or other financial institutions experience stress, they flee to safety. 2 In other words, they replace risky with safe assets. However, the euro area-the world's largest currency area with a higher gross domestic product (GDP) than the United States-lacks its own safe asset. This divergence between the US and Europe has nurtured a vivid intellectual debateone that has so far outstripped tangible policy action. The euro area consists of 19 European states that share a single currency but lack a common fiscal policy. The GFC and the ensuing European sovereign debt crisis of 2011-2012 underscored the fragility of this set-up: the compounding effects of stress in the financial system and a broad-based economic downturn resulted in a stark divergence of sovereign re-financing conditions, 3 pushing some countries to the edge of default and the euro area to the brink of break-up. 4 The absence of a common debt instrument was identified as a key driver for this dangerous fragmentation. As former European Central Bank (ECB) Executive Board Member Benoît Coeuré noted, 'public debt has been seen as both too safe and too risky' . 5 In good times, there was a widespread belief that the debt of European countries was interchangeable, regardless of existing debt levels and fundamentals. When the crisis hit, investors dropped sovereign bonds of certain sovereigns like hot potatoes-often without any reasonable economic justification. Similar dynamics resurfaced in the first days of the COVID-19 pandemic, when sovereign spreads in hard-hit countries rose rapidly. 6 As soon as markets grasped the potential economic devastation that the pandemic and the necessary public health measures could inflict, spreads in the periphery widened to levels not seen since the peak of the last crisis. 7 Again, the ECB pledged to do 'whatever it takes' to close the spreads and save the common currency. 8 Akin to the last crisis, the absence of a safe asset that foreign as well as domestic investors would retain even in the face of a financial or economic shock became all too obvious. Besides the potential misalignment between (central) monetary and (national) fiscal policy, the stability of the currency union may also be undermined by the interdependence between sovereigns and their banks. As a general rule, euro area banks are heavily 3 exposed to sovereigns, which means that shocks in the banking sector propagate to the sovereign and vice versa. 9 This mechanism has been referred to as 'doom loop' , 'feedback loop' , or 'sovereign-bank nexus' . 10 Notwithstanding major efforts to revamp the supervisory and regulatory architecture, the doom loop seems to be alive and kicking. 11 If anything, the large-scale fiscal expansion which has taken place in an attempt to fend off the economic implications of the COVID-19 pandemic seems to have exacerbated the problem. 12 Against this backdrop, several policy proposals to create a common public debt security for the currency union have been put forward since the heyday of the GFC. They have seen a resurgence amid efforts to reform the Economic and Monetary Union (EMU), 13 but gained particular notoriety in the context of tackling the economic effects of the COVID-19 pandemic. 14 These various proposals differ with regard to the degree of debt mutualization, the type of instrument and various other design elements, as Figure 1 shows. 15 In essence, a safe asset for the euro area could address at least four shortcomings in the euro area's architecture: 16 • Second, a common debt instrument would increase fiscal space in some EU Member States, thereby alleviating the pressure on the ECB to stimulate the economy. 18 • Third, a safe asset could contribute to disentangling sovereign risk from bank risk. 19 • Fourth, if a euro area safe asset were to be eligible for monetary policy operations, either as collateral in standard credit operations or as eligible assets in outright purchases, it could assume a critical role in ensuring financial stability in the euro area. 20 To be sure, hardly any asset is entirely risk free, neither mortgage-backed securities nor government bonds, as history has made all too clear. 21 note, 'the safety [of safe assets] ultimately rests on state capacity to regulate, collect taxes, and issue money, and state willingness to deploy these powers in specific ways for the sake of particular constituents and markets' . 22 Similarly, the International Monetary Fund (IMF) noted that the production of safe assets inter alia requires 'a robust legal framework' . 23 The concept of safe assets thus builds on a legal fiction, namely that the state safeguards their priority in repayment over other creditors, either by means of contract or statute. 24 In other words, policymakers are required to either make assets safe, label them as safe, or guarantee that they are safe. 25 This article will take a closer look at the legal fictions that may be necessary to comfort investors in a safe asset for the euro area. Before delving into the legal appraisal of European safe assets, it is important to understand the dynamics and motivations underlying the policy discussion. 26 Indeed, an outsider may wonder why so many different European safe assets proposals exist and why 'synthetic' instruments have received so much attention. Especially the latter appears counterintuitive at first glance, not least since synthetic financial products have been met with a their fair share of suspicion by the official sector since the financial crisis. 27 Indeed, there is little doubt that the technically most straightforward option for a euro are safe asset would be a jointly-guaranteed debt instrument, epitomizing the Hamiltonian 'one for all and all for one' spirit inherent to federal economic and monetary unions, such as the United States of America. 28 However, the creation of a joint debt instrument in the euro area, which is typically referred to as a 'Eurobond' , has so far enjoyed limited political support in some European capitals. Indeed, the need to avoid a so-called 'transfer union' , where Member States could be forced to step in for each other's financial responsibilities, has become a political rallying cry for some political parties and academics. Moreover, the European sovereign debt crisis, which, among other factors, emanated from an unsustainable build-up of private and public debt in several Member States in the south, has further undermined the willingness to expedite cross-border solidarity. This being said, the experience of the COVID-19 pandemic seems to have erased some previous red lines and the political path towards a genuine European safe asset has become clearer than ever before. 29 This article posits that the creation of a new public debt instrument in the euro area requires us to move beyond the debate economists and financial experts have engaged in for several years and inquire what legal features are necessary to render a euro area safe asset 'safe'? So far, legal questions have often been overlooked or simply pushed aside as 'technicalities' by European policymakers. 30 However, the robustness and indeed the safety of a financial asset is a function of its legal characteristics. Indeed, successfully creating and nurturing the distribution and use of a safe asset requires public and private coordination as well as ordering. 31 This will be no different in the euro area. Forging a public safe asset presents, among others, the following additional challenges in Europe: • First, a European safe asset ought to be 'multi-jurisdictional' , which means that the instrument should have the same legal characteristics and validity across the entire currency area. Indeed, if a new safe asset were to replace (or supplement) government bonds as the archetypical type of safe asset, investors need to trust that their contractual rights are equivalent or at least similar to those of government bondholders. • Second, and related, the EMU's legal foundation are characterized by a mix of EU law, intergovernmental agreements, and national (constitutional) law. Especially in the realm of sovereign debt, the level of harmonization and cross-fertilization of contractual standards remains low across the currency union. 32 Sovereign bond market are, by and large, national. • Third, certain euro area safe asset proposals resort to techniques of structured finance, thus mixing questions of public and private ordering. This, by itself, considerably increases the complexity of the underlying legal framework and is further complicated by the quest to guarantee that the instrument has the same legal value across the entire currency union. • Finally, there are neither many precedents nor statutes to delineate some important aspects of public safe assets in Europe. For instance, the enforceability of sovereign debt in Europe is anything but clear. 33 Moreover, as this article shows, 29 the involvement of (synthetic) safe asset holders in sovereign debt restructurings can raise intricate coordination problems. This article makes no claim to be exhaustive. The goal is to present a survey of selected legal questions rather than an all-encompassing review of legal issues that may or may not arise. Importantly, the article will not elaborate on the feasibility of the respective proposals under EU and national constitutional law or discuss the type of debt mutualization-others have done so to a more than satisfying degree. 34 Instead, it approaches the subject from the perspective of investors, investigating the question they are likely to be most interested in: is my asset really safe? The article thus resides closer to the financial law view on safe assets advanced by Gelpern and Gerding, staying away from the rich (EU law) literature focusing on the institutional obstacles to the creation of safe assets. This article posits that the question of investor protection is at least as important for the credibility and the robustness of the European safe asset as its compliance with the EU Treaties. While this practical approach could somewhat compromise the doctrinal contribution of the article, it makes it possible to zoom in on some less covered, yet practically highly relevant questions. Among them are the question of the law governing the safe asset contract, the question of which (national) courts decide creditor-debtor disputes, and the level of legal protection investors enjoy should the safe asset issuer be unable or unwilling to repay. 35 This rest of this article is structured as follows. Section II briefly discusses the four types of European safe assets covered in this article. Section III delves into three selected legal issues that are deemed particularly relevant for the creation of a euro area safe asset, focusing on two formal characteristics (governing law and dispute settlement forum) as well as two functional characteristics (investor protection and investor representation in debt restructurings). Section IV concludes. This article analyses four types of European safe assets: (i) Sovereign Bond-Backed Securities (SBBS); (ii) E-bonds; (iii) Purple bonds; and (iv) Coronabonds. As Figure 1 shows, these proposals have different attributes and emphasis and are thus not equally well-suited to address the objectives of a euro area safe asset. The selection of assets is not arbitrary and reflects the likelihood of their successful implementation. Among the numerous proposals, the four instruments discussed here have received most attention ; Waibel (n 28). Therefore, this article will also refrain from discussing the admittedly important question whether the ECB may purchase the respective safe assets as part of its monetary policy operations. 35 Of course, these questions may also be raised by the owner of the archetypical safe asset-a government bond. However, the reason why they are particularly relevant in the given context is that a European safe asset is, by definition, a common, multi-jurisdictional asset. in both scholarship and the recent policy discussion. 36 Moreover, since the four proposals rely on different legal frameworks, comparing them should yield valuable broader insights into the process of creating a European safe asset. The SBBS proposal is the most developed among the four, particularly since the European Commission issued a proposal for an enabling regulation in 2018. 37 The Commission's proposal has moreover been endorsed by the European Parliament, albeit with some important amendments. 38 The discussion on E-bonds has, so far, almost exclusively taken place in academic and policy-making circles. The same holds essentially true for the Purple bonds proposal. The term 'Coronabonds' , as used in this article, refers to securities issued by the European Commission with the specific purpose of financing fiscal expenditures related to the COVID-19 pandemic. 39 SBBS are multi-tranche debt securities, backed by a diversified portfolio of euro area sovereign bonds. 40 In essence, special purpose vehicles (SPVs) would purchase a portfolio of national government bonds, financed through the issuance of SBBS on capital markets. 41 The different SBBS tranches would have different risk profiles: while the senior tranche would be the actual 'safe' asset, the mezzanine and junior tranches would be subordinated, and would hence take losses first. level, while the remaining 30 per cent of subordinated securities could be divided into a 20 per cent mezzanine security and a 10 per cent junior security. 43 The Commission's SBBS proposal would ensure that SBBS products are considered safe assets in the regulatory sense, by applying the same risk weights as are applicable to sovereign bonds (ie zero). 44 However, recognizing the elevated risk for investors in the subordinated tranches of the SBBS, the European Parliament's amendment proposal assigns positive risk weights to them. 45 To this end, the creation of SBBS would be left entirely to the market-the regulation would only 'enable' their creation, not coerce it. In contrast to the Commission and the European Parliament, the ESRB also discussed a 'public' SBBS option, where the issuer would be a single public entity rather than multiple private ones. 46 The ESRB's High Level Task Force on Safe Assets Report ('ESRB HLTF Report') notably mentions the European Stability Mechanism (ESM) as a potential public arranger, given that it already has subscribed capital that would it allow it to fund the placement and warehousing of sovereign bonds. 47 To that end, the public option is edging closer to the E-bonds proposal, which, too, envisages the ESM as the issuer of the safe asset. 48 2. E-bonds E-bonds can be described as securities backed by a bundle of national senior bonds, possibly with the additional protection of joint guarantees. 49 E-bonds can either be backed by a portfolio of sovereign bonds, similar to SBBS, or by a portfolio of loans made to euro area Member States by the E-bonds issuer. 50 This article will only focus on the second option, given that the purchase of a bond portfolio would give rise to intricate problems of creditor representation in debt restructuring, as explained by Leandro and Zettelmeyer. 51 E-bonds would be issued by a public intermediary 52 that enjoys preferred creditor status (PCS)-in the euro area, the ESM seems to be the obvious choice: due to its (statutory) PCS, holders of national government bonds would be subordinated to the ESM. In other words, the ESM would be paid before all private creditors as well as other euro area governments, rendering it a very creditworthy counterparty from the 43 55 From an economic viewpoint, according to former ECB Vice-President Constâncio this structure could result in 'an increase of the costs of issuing the non-preferential part of national debt that is not included in the securitisation' . 56 However, this this only affects the marginal cost, while average funding costs remain essentially unchanged. This effect is not necessarily unwanted, as it could nurture (welcome) market discipline in sovereign debt markets. 57 In contrast to SBBS and E-bonds, Purple bonds would not be safe assets in the narrow sense. 58 Importantly, they are not a euro area-wide safe asset. Rather, Purple bonds are national government bonds that enjoy one great privilege in comparison to all other public debts: they are 'non-restructurable' in the context of an ESM programme. Given that they are often discussed alongside the two genuine safe asset proposals, Purple bonds are included in the present analysis, with the aim of also sketching out the chief legal differences. They can be described as a vehicle to facilitate the transition from the current state of the world to the one foreseen by the Red/Blue bonds proposal. 59 The Red/Blue bond proposal by Delpla and von Weizsäcker of 2010 envisaged the division of national government bonds into Blue bonds (issued up to a maximum debt-to-GDP ratio of 60 per cent) and Red bonds (when overall debt is beyond 60 per cent of GDP). Red bonds would be subordinated to Blue bonds, and therefore be more expensive to issue for the sovereign. 60 Purple bonds would go further by also addressing transition problems associated with moving to a Red/Blue bonds market. On day one, all debt would be labelled as Purple debt and could thus not be subject to restructuring. States would then have to 53 reduce the excess general government debt above 60 per cent of GDP by 1/20 every year-any debt above the limit would have to be financed with Red bonds, which would be penalized by the market. After 20 years, Member States would ideally have as little Red debt and as much Purple debt as possible, with an overall debt limit close to 60 per cent of GDP. 61 Under the Purple bonds proposal, the legal scaffolding of national government debt markets, ie sovereign debt contracts, would not be touched: according to the inventors of Purple bonds, changes to the ESM Treaty should suffice to achieve the desired policy effects. This proposition will be discussed in greater detail below. 62 On 28 May 2020, the European Commission proposed a 'European Union Recovery Instrument' , aimed at supporting recovery in the aftermath of the COVID-19 pandemic. 63 The proposal states that it would 'authorise the Union to borrow temporarily and exceptionally an amount of EUR 750 billion' , but does not clarify what instruments the Union would rely on to this end. 64 The policy discussion, however, strongly suggests that the Commission would draw on the structure of previous and existing debt issuance programmes, 65 such as the European Financial Stability Mechanism (EFSM), the Balance of Payments (BoP) programme, and the Macro-Financial Assistance (MFA). 66 It should be noted that borrowings of the EU are direct and unconditional obligations of the EU, assumed within the limits of its own resources and the EU budget, which is in turn backed by the EU Member States. 67 When compared to the three other safe asset proposals discussed in this article, Coronabonds differ at least with regard to the following aspects: costs of the COVID-19 pandemic across the EU, thus ensuring a more level fiscal playing field in the aftermath of the crisis. 68 • Second, Coronabonds entail a partial mutualization of debts, given they are guaranteed by the EU budget, which is in turn backed by EU Member States. • Third, Coronabonds would be used to finance COVID-19 related expenditures of all 27 EU Member States, while the other safe assets would be limited to the 19 members of the currency union. • Finally, and similar to Purple bonds-while they are limited in both purpose and time, the (legal) framework underpinning Coronabonds could become the basis for a genuine Eurobond. 69 This section comparatively analyses the four proposals for euro area safe assets from a legal viewpoint, focusing on selected aspects that are, arguably, relevant for the realization of these proposals. Before delving into the analysis, it should be recalled that the respective safe asset proposals are at different stages of (legislative) implementation, although none of the assets analysed has seen the light of day. As pointed out earlier, the legal framework underpinning SBBS instruments is more developed in comparison to the other proposals. Indeed, the Commission's publication of a proposal for an SBBS Regulation allows us to analyse and comment on a concrete legislative proposal. By contrast, the legal review of E-bonds and Purple bonds must rely first and foremost on rough policy papers, all of which put the onus on the financial and economic rather than the legal aspects. 70 The most recent, though not necessarily the least developed, is the Coronabonds proposal. EU institutions have signalled that, legally speaking, they would model Coronabonds on previous bond issuance programmes by the European Commission. 71 For the purposes of this article, Gelpern and Gerding provide a valuable analytical matrix to identify the legal mechanisms employed to render assets 'safe' . 72 They distinguish between legal tools by means of which assets are 'made safe' , 'labelled safe' , and 'guaranteed safe' . 73 Of course, the literature generally distinguishes between public 68 Of course, Coronabonds can also fulfil the safe haven function and mitigate doom loop externalities. 69 Hung Tran, 'The EU recovery plan: A "Merkel" but not a "Hamilton" moment' Atlantic Council Blog (28 May 2020) accessed 31 July 2020. For the issues this may raise under EU law, see Amtenbrink and others (n 26) and Waibel (n 28). 70 In this respect, the legal analysis is, inevitably, more speculative, not least since it warrants some assumptions on legally relevant aspects that the authors of the policy proposals have so far left open. 71 See Horn and others (n 66). 72 Gelpern and Gerding (n 22). 73 Ibid 388. and private safe assets. 74 While the former are produced by governments, the latter are issued by private actors and are thus generally deemed less safe. 75 While drawing on Gelpern and Gerding's seminal piece, this article will employ a modified and more targeted analytical framework, distinguishing between the relevant (i) formal and (ii) functional legal features of euro area safe assets. While the formal characteristics essentially concern the safe asset's governing law as well as the forum for the settlement of disputes, the functional features deal with questions of investor protection as well as investor representation in debt restructurings. Some overlaps between the formal and the functional features cannot be excluded. The governing law may, for example, influence the level of investor protection. 76 At the same time, drawing a distinction between the more procedural and the more substantive characteristics of safe assets should provide a better analytical basis for comparing the respective proposals and introduce the necessary level of nuance to this technical subject matter. The decision to focus on the selected formal (governing law, dispute resolution forum) and functional (investor protection and investor representation in debt restructurings) features is driven by the following broader considerations: • First, the multi-jurisdictional nature of a European safe asset and lack of harmonization of national fiscal laws inevitably raises questions about the applicable (national) legal framework. At the same time, policymakers set out to generate a single instrument with uniform legal features across the entire region. 77 Requiring or encouraging safe asset issuers to use different governing laws may undermine this idea. 78 • Second, investor protection and representation issues have so far received little attention in the pertinent (legal) literature on European safe assets. This is somewhat surprising, since, as Engert and Hornuf point out, 'the main substantive differences in national contract laws governing [government bond] indentures concern the possibility and scope of collective action clauses and of bondholder representation in relation to the issuer' . 79 • Third, this article seeks to make a concrete contribution to the policy discussion, thus focusing on the legal issues that the existing policy papers have raised. 80 When the European Commission discussed the feasibility of a common debt instrument in 2011, it noted that '[c]onsideration must also be given to the 81 Similarly, the ESRB HLTF Report underscored the importance of investor protection and representation in debt restructurings for the marketability of safe assets. i. Governing Law The legal system chosen to govern a security is vital, not only in the world of corporate finance but also in the realm of public debt finance. 82 As the Permanent Court of International Justice famously held in 1929, '[a]ny contract which is not a contract between States in their capacity as subjects of international law is based on the municipal law of some country' . 83 Since the vast majority of government debt contracts are not concluded between countries but between a state and its private creditors, they are either governed by (i) the issuer's own domestic or (ii) a foreign domestic law, ie a law other than the issuer's own law. 84 As a general rule, the less creditworthy a sovereign debtor is, the likelier it is to issue bonds under foreign domestic law. 85 Moreover, the finance literature finds a correlation between foreign-law debt and relatively lower sovereign bond yields. 86 The common explanation of this effect is that sovereign debt instruments issued under the issuer's own domestic rules enjoy the 'local law advantage' , which means that the issuer retains the legislative control over its securities. 87 Investors will thus tend to only accept domestic-law bonds at reasonable interest rate if they have some faith that the sovereign will refrain from making use of this advantage. In the euro area, most countries fall into the category of 'elite issuers' , which can finance much of their public debt under their own laws. As Chamon and others document, the majority of euro area economies have issued less than 10 per cent of their overall government debt under foreign laws. 88 There are some notable exceptions, however. Cyprus and Greece have increasingly relied on foreign law since the Greek restructuring shattered confidence in the country's local-law bonds. Estonia's government debt is almost exclusively governed by foreign law, while the same is true for approximately 40 per cent of Slovenia's debt. 89 Moreover, empirical evidence on the governing law of corporate debt suggests that the introduction of the euro resulted in a significant increase in the proportion of foreign-law instruments. 90 The large share of domestic-law debt in Europe means, in essence, that retroactive changes to the government bond contracts are not excluded. 91 This said, creation of legislative fiat is not without limits. 92 Indeed, the European Court of Human Rights (ECtHR) has delineated the boundaries facing sovereigns in amending government debt contracts ex-post. 93 The discussion on the government's ability to retroactively change its local laws in order to bring about a debt restructuring has recently flared up in the context of Italy's lingering fiscal plight. 94 While most authors agree that the Italian government could unilaterally change the repayment conditions of debt instruments governed by Italian law, 95 it lacks sovereignty over the Italian bonds governed by New York law. The perhaps most compelling evidence to underscore the role of the governing law in sovereign debt markets is the fact that Greece has not been able to sell a single Greeklaw bond to investors following the 2012 restructuring. 96 In other words, Greek and foreign investors demanded that Greece contractually constrains its ability to repeat any retroactive amendments of its government debt instruments. 97 Empirical evidence on the role of the governing law in European sovereign bond contracts is still scarce, but the studies that do exist suggest that this is increasingly important to investors. 98 Like the governing law, the forum where disputes related to the debt security are settled matters. 99 If a foreign court were competent to render judgment on the legality of the sovereign issuer's actions, the risk of an unfavourable decision (from the sovereign issuer's point of view) would, in principle, be higher. 100 In the euro area, most public debt instruments do not include choice-of-forum clauses that transfer jurisdiction to courts abroad-lawsuits are settled by home courts. 101 While emerging market economies usually surrender their courts' sovereignty to access international debt markets, most European countries had no pressure from investors to 'surrender' their courts' power over sovereign indebtedness. 102 The choice of the dispute settlement forum arguably has relevance for the acceptance and marketability of a respective safe asset, especially from the creditor's perspective. Notably, as the sovereign debt literature suggests, foreign courts tend to be less debtor-friendly than domestic courts. 103 Indeed, the very legal structure of the international Eurobond market (ie the market for foreign-currency bonds) can be traced back to investors' trust in New York and English courts. 104 The Greek debt restructuring 2012, once again, provides anecdotal evidence. While the government was confident that it could successfully defend the sovereign debt workout against holdout creditors in Greek courts, it feared being dragged in front of English courts and having its foreign assets seized. 105 The government thus paid English-law holdouts in full while bailing-in the holders of bonds under the jurisdiction of local courts. Although we will never know if the government's intuition was correct, evidence from the decade-long battle between Argentina and its foreign bondholders suggests that disruptive litigation can be economically costly to the sovereign debtor. 106 So, from an investor perspective, the Greek government bonds subject to the jurisdiction of English courts were 'safer' . iii. Formal legal features and euro area safe assets 1. SBBS SBBS belong to the asset class of asset-backed securities (ABS), and thus differ significantly in their contractual structure from plain vanilla sovereign bonds. Arguably, the choice of governing law is even more important for an SBBS than for normal sovereign bonds, as ABS products are heavily regulated to protect investors and ensure a sufficient degree of transparency-the crux of securities regulation. 107 Moreover, securities laws across the euro area are not fully harmonized, which means that rights of ABS holders, or SBBS holders for that matter, could differ across jurisdictions. The governing law issue is relevant not only for the instrument itself but also the SBBS issuer. First, the SBBS issuer, as a private legal person, has to be incorporated under the domestic corporate laws of a specific Member State. 108 The SBBS issuer is neither a country not an EU body or an organization governed by international law. While certain areas of corporate law have been harmonized to a considerable degree across the Union, no genuine 'European' corporations that are governed by EU law exist. 109 Rather, SPVs used for asset securitization, which are usually limited liability companies, are incorporated under a specific Member State's law. Legal risks stemming from the governing law issue will, with high likelihood, impact the credit rating of an SBBS-the more a jurisdiction accommodates and safeguards methods of securitization and bankruptcy-remoteness of SPVs issuing SBBS, the easier it will be for SBBS investors to gauge the risk they are exposed to. Indeed, the applicable contract law directly affects agency and other transaction costs. 110 The ESRB HLTF Report confirms the importance of governing law, stating that 'an SBBS investor needing to make a claim against an SBBS issuer with regard to nonpayment would need to do so under the law governing those securitised products' . 111 To increase homogeneity, the ESRB stated it may be preferable that all SBBS are subject to the same governing law. 112 However, the Commission proposal for SBBS 113 as well as the European Parliament's legislative resolution 114 remain silent on the issue of governing law and forum, essentially leaving the decision to the private SBBS issuer. 115 Given the objective of creating one common safe asset for the entire euro area, the creation of SBBS governed by different local laws and subject to the jurisdiction of courts in different countries might be counterintuitive. Notably, it could lead to a situation where the pricing of a German-law SBBS may then differ from the pricing of a Spanish-law SBBS as a result of divergent regulatory and legal frameworks for securities and their enforcement. One could also imagine that investors may prefer to purchase an SBBS governed by the laws of their home state, notably because they are more familiar with the domestic legal frameworks and procedures. The best way forward would thus be to agree that the laws of a third-country be applicable. 116 The European Commission has left the question open whether SBBS should be governed by the laws of a specific Member State (such as Luxembourg) and whether disputes should be settled by the courts of this country. If one were to believe in the importance of the governing law, as the author does, this raises the issue whether benefits of achieving legal homogeneity outweighs the obstacles to agreeing on one jurisdiction or choosing the law of a single country. 117 Indeed, it might actually be easier for governments to follow the ESM's and the EIB's established practice of issuing under Luxembourgish law and giving jurisdiction to courts in Luxembourg than to settle on yet another European jurisdiction. Moreover, as mentioned above, the Greek debt restructuring of 2012 illustrated that investors prefer foreign-law debt, especially once the issuer's credibility has come under strain. 118 Given that the SBBS market will have to be created from scratch, credibility is a precious commodity-possibly vital to ensuring that senior SBBS can obtain the highest possible credit rating and therefore fulfil the functions policymakers envisage. Indeed, agreeing on a single governing law could alleviate fears that sovereigns could default on their debts by retroactively amending local-law bonds. 119 ample evidence in the finance literature on the benefits of contract standardization, suggesting that a common governing law will reduce the overall costs of transacting. 120 One should keep in mind that-to the author's best knowledge-there is no Luxembourgish case law on sovereign debt disputes. By contrast, English courts have a long-standing tradition of adjudicating sovereign debt disputes related to instruments governed by English law. 121 Yet, this absence of relevant judicial precedents alone should probably not influence policymakers' choice of governing law and dispute settlement forum. Most European jurisdictions enjoy a high degree of expertise and the courts in Luxembourg are known for their savvy in financial law matters. Finally, Brexit does reduce the attractiveness of English law, at least for political reasons. 122 It would indeed seem odd if European authorities advocate the use of English law after the recent switches by the EIB and the ESM to Luxembourgish law. The question which law ought to govern E-bonds and the E-bonds issuer, respectively, as well as the question where disputes are to be settled, should cause fewer headaches than in the case of an SBBS. In the conception of E-bonds analysed in this article, the issuing entity would be a public body such as the ESM, instead of an SPV in the case of an SBBS. This alleviates some of the legal complications associated with governing law and fora, most notably since the current issuing practice for ESM bonds could be emulated for E-bonds. 123 Similar to SBBS, relying on Luxembourgish law as the single governing law for Ebonds, and courts in Luxembourg City is likely to be met with goodwill from the investor community. This is confirmed by the recent ESM consultation of market participants in the context of switching from English to Luxembourgish law. 124 Moreover, using a different governing law for E-bonds than for normal ESM bond issuances could moreover be understood as a signal that E-bondholders should, in the issuer's view, be subject to divergent standards of protection than holders of 'normal' ESM bonds. Thus, relying on established practices, ie issuance under Luxembourgish law and assigning jurisdiction to Luxembourgish courts, should be the preferable option with a view to achieving the highest possible rating for E-bonds. With regard to Purple bonds, the governing law and dispute settlement forum are of limited relevance. This is because, under the Purple bonds proposal, sovereigns would 120 continue their current practices, issuing government debt mostly under local law and under the control of local courts. Unless one wishes to herald broader changes to the legal structure of euro area sovereign debt markets 125 -an endeavour with merits but politically hardly palatable-Purple bonds may be considered as a useful vehicle to allow for a gradual transition to a fully-mutualized Eurobond. Indeed, Purple bonds are not by themselves safe assets but national government debt instruments that benefit from certain protections under international law. As noted above, the European Commission appears to consider issuing Coronabonds using the Base Prospectus of the 2019 EU Debt Issuance Programme ('EU Base Prospectus'). 126 Akin to bonds issued ESM bonds, newly issued EU debt securities would be governed by Luxembourgish law. English law was also abandoned as of 2019 by the European Commission. 127 For the seasoned observer of sovereign debt markets, this rising importance of Luxembourg will not come as a surprise. As Fontenay and others show in a recent paper, 'the Luxembourg Stock Exchange has been the overwhelming favourite among the global exchanges for international sovereign-bond offerings' . 128 In that sense, Luxembourg seems at the brink of becoming the safe haven for safe havens. When it comes to the dispute settlement forum, however, the EU Base Prospectus conveys jurisdiction to the Court of Justice of the EU (CJEU) rather than the Luxembourgish courts. 129 As is the case with other suits against the Commission, the General Court, which is the CJEU's court of 'first instance' , has jurisdiction over disputes. 130 This being said, the CJEU would still decide on the basis of Luxembourgish law, as the EU's contractual liability is essentially determined by the law applicable to the relevant contract. And it would indeed be surprising if the CJEU takes a markedly different approach than domestic courts in Luxembourg when it comes to the interpretation of Luxembourgish law. The perhaps most important means to achieve a robust level of investor protection in the realm of sovereign debt financing is to take away the sovereign issuer's authority to unilaterally modify the contract-the local-law advantage mentioned earlier. 132 The next step, however, is to identify the substantive legal aspects used to protect safe asset holders. As Pistor observes, 'law determines the degree of investor protection and thereby establishes the rules of the game for a financial marketplace in which actors respond to the incentives law creates' . 133 Put in the words of Gelpern and Gerding, '[g]overnments and private firms deploy financial engineering and legal tools to produce safe assets and reduce risks to payoff ' . 134 Indeed, multiple contractual, regulatory and statutory tools and techniques exist to render a financial instrument safe. 135 In this article, these will be subsumed under the umbrella term 'investor protection' but the terms 'investor' and 'creditor' will be used interchangeably. For instance, an asset can be made safe by giving the creditor priorities, providing for bankruptcy safe harbours, or by making it acceptable in monetary policy operations. Many safe assets benefit from a specific public-private ordering, 136 where the state safeguards certain contractual arrangements 137 or creates rules that enable private parties to create safe assets. 138 It should, however, be noted that many of the tools to create private safe assets, such as mortgage-backed securities, might not be of much use to create a public safe asset. Ultimately, when looking through the securitization structure, the counterparty to the safe asset holder is a sovereign nation, and therefore enjoys a plethora of legal privileges under both domestic and international law. 139 Paradoxically, the contractual promises they make may thus have less legal value from an investor's perspective than those entered into by a private, non-sovereign actor. As discussed below, these considerations are also relevant in the context of the ESM and the European Commission as (quasi-)immune issuers. 132 The second important function legal issue concerns the representation of safe asset holders in sovereign debt restructurings. The particular importance of investor representation in the context of the safe asset debate stems from the absence of a bankruptcy regime for sovereigns-both in the euro area and at the global level. 140 Since the safe asset issuer itself would be a sovereign (Purple bonds, Coronabonds) or the safe asset be backed by one or more sovereign debt instruments (SBBS, E-bonds), investors' rights in a debt workout are directly relevant to assess recovery values and thus financial risk. To be sure, the two functional categories of 'investor protection' and 'investor representation in debt restructurings' are overlapping, and the former may substantially affect the latter. However, for the purpose of this article, 'investor protection' broadly circumscribes the legal tools to safeguard safe asset holders at all stages of the asset's (legal) existence. By contrast, when discussing 'investor representation' , the specific issues that arise in a sovereign debt restructuring procedure are being addressed. This is because debt a restructuring is, generally speaking, the equivalent of bankruptcy procedure for a country, albeit an imperfect one, as has been pointed out before. 141 So how are investor representation issues addressed in international debt restructurings? To mitigate collective action problems in debt restructurings, most (modern) international sovereign bonds feature collective action clauses (CACs)-clauses that bestow upon bondholders the right to cast a vote on a restructuring proposal by the sovereign issuer, either accepting or rejecting it. 142 Since 2013, all domestic-law government securities as well as supranational bonds with a maturity longer than one year also includes CACs. 143 As discussed below, this relatively recent contractual evolution is likely to also affect the restructuring of safe assets. 144 When sovereign bonds are pooled or bought by an intermediary, thorny principal-agent problems are likely to complicate the relationship between the safe asset issuer and the safe asset holder, such as in the case of SBBS or E-bonds. The issue of investor representation is less relevant in the context of Purple bonds as well as Coronabonds, where there is no intermediary between the safe asset holder and the issuer. iii. Functional legal features and euro area safe assets 1. SBBS In essence, SBBS intentionally establish different levels of investor protection by creating a senior, ie protected, and a junior, ie unprotected, tranche. 145 Only the senior tranche of the SBBS, referred to as 'ESBies' in the original proposal by Brunnermaier and others, 146 would actually qualify as a 'safe' asset. From a legal point of view, the safety of the senior SBBS tranche is achieved by means of contractual subordination. Akin to other ABS products, the junior tranche is contractually more exposed to default events: the senior tranche will only bear losses after all the subordinated tranches of that SBBS issue have done so 147 -and, according to policymakers, the junior tranche would be 'thick enough' to mitigate the possibility of default, although not to eliminate it completely. 148 The envisaged contractual subordination of junior SBBS tranches to senior SBBS tranches should, in principle, not raise particular legal problems. 149 To be sure, the 'contractually agreed priority of payments waterfall' 150 needs to withstand cogent statutory rules of insolvency as well as pertinent case-law. Given that insolvency frameworks in the euro area are-by and large-not harmonized, the choice of governing law would be important to accurately circumscribe SBBS holders' rights, and hence influence the product's rating and pricing. Thus, in order to ensure full bankruptcy-remoteness and enhance legal certainty, changes to national insolvency law and/or the Bank Recovery and Resolution Directive (BRRD) might be warranted. 151 In this context, inspiration may be drawn from the recent proposal by the European Commission for a Directive on the issue of covered bonds; 152 indeed, the ESRB HLTF's SBBS report also draws some parallels between covered bonds and SBBS. While having 'no intention to harmonise national insolvency laws' , 153 the covered bonds proposal requires Member States to include 'specific rules aimed at protecting the cover pool . . . to ensure the fulfilment of the investor's rights including in case of resolution or insolvency of the issuer' . 154 By contrast, the European Commission's SBBS proposal does not entail similar provisions to ensure the bankruptcy remoteness of the SBBS issuer. Rather, it seeks to protect investors from the risk of insolvency of the SBBS issuer by limiting the activities the latter may engage in to the 'issuance and management of SBBS' and imposing strict asset segregation requirements. 155 Thus, to ensure the highest possible degree of bankruptcy remoteness across all jurisdictions in the euro area, legislative clarifications, similar to the ones made in the context of covered bonds, are likely to support the development of an SBBS market. Another important issue from an investor protection perspective pertains to the relationship between holders of (senior) SBBS tranches and holders of (normal) government bonds. The ESRB HLTF recommended that '[i]n a debt restructuring scenario, sovereign bonds in SBBS cover pools must be treated in the same way as those held by investors directly' . 156 The risk of legal arbitrage, usually posed by holdout creditors and aptly captured in the report by the ESRB HLTF, 157 can be mitigated by virtue of incorporating pari passu provisions in euro area sovereign bonds. 158 While the European Commission did not follow this recommendation in its SBBS proposal, the European Parliament suggested including a provision to require that 'Member States need to ensure that holdings of sovereign bonds by SPEs enjoy the same treatment as any other holdings of the same sovereign bond...' . 159 An important legal question in this context is whether Member States would need to amend their sovereign bond contracts, ie include specific pari passu provisions, or whether the European Parliament's proposal would be sufficient to achieve the desired policy aim. The second option would come with two key advantages. First, there would be no transition phase, as equal treatment obligations would enter into force as soon as the first SBBS was issued. Second, due to the direct applicability of EU Regulations in Member States, a consistent interpretation of pari passu could be ensured, ultimately by virtue of CJEU jurisprudence. 154 See ibid, in particular recital (16) and art 5 (stating that 'Member States shall ensure that the payment obligations attached to the covered bonds are not subject to automatic acceleration upon the insolvency or resolution of the credit institution issuing covered bonds'). 155 See recital (11) of the SBBS proposal (n 37). 156 ESRB HLTF Report (n 24) 18. 157 See ESRB HLTF (n 111) 69-71. 158 To this end, the ESRB HLTF recommended that such clause should identify the creditor pool within which discrimination is prohibited, provide a comprehensive definition of actions with which a government might seek to influence returns or value of bonds, and stipulate that if any such action makes some creditors better off, it will automatically apply to all creditors, including SBBS. To be sure, with a few important exceptions 160 and in contrast to international sovereign bonds, euro-denominated and domestic law 161 euro area government bonds lack pari passu clauses. 162 One reason for the absence of pari passu clauses may be that, due to the local-law advantage, 163 the sovereign issuer retains the power to ex-post modify bondholders' rights emanating from a pari passu clause-or indeed from any contractual provision. 164 Another legal consideration is whether the Union legislator could prescribe a certain drafting of national government bonds, or whether that would transgress into Member States' fiscal powers. 165 Indeed, it would need to be argued that the harmonization of sovereign debt practices is essential for the functioning of the EU's internal market and its four freedoms-prima facie a difficult reading. For another, any retroactive change to a contractual position warrants a careful constitutional balancing act both at the national and the Union level. 166 Turning to the issue of investor representation, it should be noted-or perhaps a warning should be given-that the securitization structure underlying SBBS does raise some tricky technical questions. From a legal viewpoint, the SBBS issuer owns the government bond portfolio and thus also has the formal voting rights-the SBBS investor is only represented indirectly should a vote on a debt restructuring proposal take place. The SBBS investor will have a strong economic interest in her representation in a meeting of bondholders, as the outcome likely influences the respective value of the different SBBS tranches. However, the SBBS issuer is not envisaged as acting as the SBBS holders' agent. Rather, according to the ESRB HLTF, SBBS issuers are supposed to be 'algorithmic entities without any vested interest' . 167 The ESRB HLTF therefore concluded that, in order to maximize the chances for a restructuring outcome that is in the interest of SBBS holders collectively, two options ought to be considered: (i) appointing a third-party trustee with a fiduciary duty in the SBBS contract to instruct the SBBS issuer how to vote, or (ii) SBBS investors should give instructions to SBBS issuers. 168 The Commission has left it to the market, and hence the SBBS issuers, to decide which option is preferable-the SBBS proposal merely requires the SBBS issuer to provide the holders with information on 'how voting rights shall be assigned to investors' . 169 Let us first take a closer look at the trust model for bondholder representation. As the ESRB HLTF recommended, 'SBBS product regulation together with its accompanying technical standards and guidelines would need to define the minimal legal and professional qualifications of the trustee' . 170 However, for the exact delineation of the trustee's duties, the governing law of an SBBS is critical. 171 Indeed, the concept of a trust in the narrow sense is unusual in most civil law systems, and thus in the majority of euro area countries. 172 In civil law jurisdictions, the agreements regulating the relationship between issuer and bondholder have more resemblance to fiscal agency agreements. 173 Agency agreements differ fundamentally from trusts, as the agent has, in principle, no fiduciary duties and instead performs a largely administrative role. 174 In the context of sovereign debt restructurings, trust arrangements seem to allow for more orderly procedures, notably because go-alone holdout litigation, which has for instance haunted the Argentine debt restructuring for decades, is not possible. In the context of SBBS, the trust model could serve a similar purpose, and adequate representation for SBBS holders could be ensured by mirroring the voting procedure under CACs at the SBBS holder level. For instance, if a qualified majority of 75 per cent (written resolution) or 66 2/3 per cent (bondholder meeting) approves the restructuring proposal, the trustee would take this position on behalf of the SBBS holders in the CAC vote. If the required majorities were not attained among SBBS holders, the trustee would vote against the proposal-depending on the overall amount of outstanding debt, the SBBS holders' vote could be decisive for the success of the restructuring operations. It should be mentioned in this context that, as Hofmann points out, civil law countries do not necessarily prohibit the appointment of bondholder representatives: it is typically up to the bondholders to define the rights and obligations of the representative. 175 Thus, while the divergence between civil and common law jurisdictions in this regard would support the case for a single governing law for SBBS, it might still be conceivable to ensure appropriate bondholder representation in all euro area jurisdictions. 176 If this is done, representation arrangements should be laid down in a 'term sheet' that governs SBBS products and defines an enforceable common standard that would apply across the entire currency union. 177 Under the second option discussed by the ESRB HLTF, every SBBS holder would give direct instructions to the SBBS issuer. To this end, SBBS contracts could replicate the single-limb CAC model by enabling a (qualified) majority of SBBS holders to instruct the SBBS issuer to accept the restructuring proposal. 178 The respective majority thresholds of CACs in national sovereign bonds could be mirrored at the SBBS level. Of course, diverging and conflicting interests may arise in a debt restructuring process because of different economic incentives among SBBS holders of senior, mezzanine, and junior tranches. Notably, senior SBBS holders-and this is of course the intention of policymakers-would not have any in skin in the game if the restructuring would only affect the junior and the mezzanine tranches. Indeed, in such case, the senior SBBS holders would always vote for a (high) haircut to ensure that losses are borne by the junior and the mezzanine SBBS. 179 The ESRB HLTF Report provides a good sense of the formidable level of technical complexity associated with a model where each SBBS holder directly instructs the SBBS issuer in a debt restructuring. This article will not delve any deeper into the complex incentives holders of different SBBS tranches may face. From a legal viewpoint, it is critical that the bondholder representation structure is as litigation-proof as possible. Protracted lawsuits between holders of different SBBS tranches or between SBBS investors and the SBBS issuer could considerably undermine market confidence in the product, especially as debt restructuring events are often a corollary of already volatile government debt markets. Therefore, the trust model offers clear advantages. To be sure, vesting individual creditor enforcement rights in a trustee entails trade-off, especially if the trustee remains too neutral and fails to deliver on her fiduciary duties. 180 However, past sovereign debt crises have underscored the risk of holdouts interfering with debt restructuring procedures by means of litigation. 181 The creation of a new, complex asset class in euro area sovereign bond markets is not going to alleviate the risk of regulatory and legal arbitrage, which is why policymakers should have an interest in creating robust creditor representation arguments, ideally in the form of trust deeds that clearly allocate responsibilities and define the trustee's fiduciary duties. This section will first discuss issues related to investor protection before turning to questions related to investor representation. In the context of E-bonds, the key legal aim from an investor protection perspective will be to ensure that the loans provided by the E-issuer to Member States rank senior to 'normal' bondholders' claims. In contrast to SBBS, policymakers' objective is not to guarantee the pari passu treatment of safe asset holders and bondholders, respectively, but to make the E-issuer's loan portfolio, which backs the E-bonds, as safe as possible. By subordinating normal bondholders to the ESM, 182 the sovereign issuer would always first repay these loans in full and then service any other financial obligation. 183 Leandro and Zettelmeyer argue that this feature of the E-bonds proposal would necessitate a reinforcement of the ESM's PCS 184 -one that is respected by both the sovereign issuer and bondholders. 185 They advocate (i) contractual and (ii) statutory subordination of normal bondholders to E-bondholders. Both types of subordination are outlined below and the possible legal challenges to achieving them sketched out. Moreover, this section also discusses the legal relationship between senior E-bonds and government loans (government debt structured as a loan rather than a bond), which make up roughly 14 per cent of overall government debt in the euro area. 186 Regarding statutory subordination of bondholders vis-à-vis the ESM, note that the ESM Treaty already confers PCS to the ESM (in recital (13)). The concept of PCS is considered international soft law: while not legally binding, it is recognized that countries borrowing from international organizations enjoying PCS grant them priority over other public and private creditors. 187 ESM's PCS is duly acknowledged, not only in the ESM Treaty's recitals but also in pertinent assessments by credit rating agencies. 188 A relevant practical question is whether the sovereign debtor would actually respect the ESM's PCS. In a deep debt crisis, the government might consider reneging on its obligations vis-à-vis the ESM. The Greek case, and its (temporary) default on the IMF in 2015 illustrated that official sector lenders may also face a risk of non-repayment. 189 So what could be done to reinforce the protection of holders of E-bonds? Akin to pari passu obligations in the SBBS Regulation, the primacy and direct applicability of EU law might be levered to honour and reinforce the ESM's seniority. This would warrant creating a genuine legal basis for E-bonds under EU law-a legislative move that ought to be combined with a broader reform of the ESM, and notably its anchoring in EU law. The new 'ESM Regulation' could thus stipulate that euro area countries must always respect the PCS enjoyed by the ESM-the CJEU could be called upon by disgruntled E-bondholders should the government opt for a confrontational approach vis-à-vis the official sector. At the same time, issuing large amounts of senior E-bonds comes with the risk of crowding out 'normal' bondholders, thereby ultimately increasing the government's refinancing costs. For example, as Steinkamp and Westermann observe, rating agencies downgraded Greece and Portugal during the crisis when commercial creditors were subordinated to the ESM. 190 While the ESM's PCS has since been widely accepted by markets, policymakers would need to carefully communicate with markets and rating agencies to understand the potential (financial) implications of subordinating government bonds to E-bonds. The market reaction is, most likely, a function of the volumes of E-bonds issued in a given period as well as the management of market expectations through intelligent communication. 191 Therefore, a gradual approach should be followed so as to avoid triggering any unwarranted market reactions. The second legal mechanism that could be used to increase the safety of E-bonds is contractual subordination. 192 To this end, bond issuers ought to consider inserting a modified pari passu clauses in all of their bonds 193 with a view to conferring PCS 188 to official sector lenders, such as the ESM. 194 As far as the author is aware, pari passu language that explicitly subordinates private bondholders to official sector entities has not yet been used in international or domestic sovereign bond issuances. Rather, the subordination of bondholders to multilateral official creditors would essentially reflect an international custom 195 and well-established debt restructuring practice. 196 Another related question is whether outstanding government bonds could be subordinated to newly issued E-bonds. As a general rule, such retroactive amendments to existing contractual relationships remain legally delicate and practically arduous. 197 That is not to say they are impossible if the debtor is a sovereign nation, especially during a crisis when the government may need to amend debt instruments to allow for an orderly debt restructuring. 198 In normal times, however, bondholders may successfully challenge such subordination in local courts or before international arbitral tribunals. 199 Thus, a gradual approach similar to that which policymakers pursued with respect to updating the CAC standard in international sovereign bonds 200 seems warranted with a view at reducing legal risks. Finally, the crowding-out problem mentioned above in the context of statutory subordination would also play a role here. 201 By the end of 2019, euro area governments had loans outstanding with a volume of almost e1.4 trillion (12 per cent of GDP and 14 per cent of overall government debt) outstanding. This raises the (legal) issue whether E-bond holders would also rank senior to creditors of these loans and, if not, whether and how contractual subordination could be engineered? Generally speaking, loan agreements in the euro area are concluded not only by the federal government but also by sub-national entities. Additionally, given that loans are traditionally not traded on stock exchanges, most of the loan documentation is not publicly accessible or at least very difficult to obtain. In light of these constraints, in this discussion it is assumed that a typical government loan document does not include any specific creditor ranking provisions. This section now turns to the issue of E-bondholder representation in debt restructurings. In this context, the insights offered by Leandro and Zettelmeyer are instructive. 202 They note that representation issues are alleviated by the fact that the marginal class almost always consists of bondholders. In other words, the ESM as E-issuer would not participate in the debt restructuring negotiations by virtue of its PCS. It is only if all 'normal' bondholders have been wiped out that the ESM would become a marginal creditor and would have to enter negotiations with the sovereign issuer. In such scenario, the question is whether E-bondholders could instruct the ESM to pursue a certain negotiation strategy, which would include accepting or rejecting a specific restructuring offer. 203 This would be challenging for several reasons. First, the revised ESM Treaty stipulates that that the Managing Director and staff of the ESM are independent, implying that they must not take any directions from third-parties, such as E-bondholders. 204 Second, decisions relating to the restructuring of an ESM loan may even need to also be approved by the ESM Board of Governors, which consists of the 19 finance ministers of euro area countries who are exclusively accountable to their national Parliaments. 205 Finally, and on a more general note, a legal framework in which private individuals may direct an international organization to execute certain actions does not exist and is likely to meet fierce political opposition. This being said, given the ESM's ability to reassure E-bondholders of its own repayment capacity, notably due to its priority over other creditors, E-bondholders are unlikely to shy away from E-bonds. Indeed, at the current juncture, ESM bonds are among the highest-rated securities on the market: the current inability of ESM bondholders to instruct the ESM to adopt a certain (creditor-friendly) position in debt restructuring negotiations seems not to undermine creditors' trust in the ESM. Arguably, E-bondholders could place a similar level of trust in the ESM, and specifically that it would enforce its PCS with respect to Member States. An attempt to convey more certainty and control to E-bondholders is likely to run into serious legal obstacles and does not seem to be justified, especially given the ESM's own inherent interest to apply its priority claim on sovereign debt repayments. Since Purple bonds offer a simpler legal structure, investor protection and representation can be summarized without further distinguishing the two. 206 To recall, Purple bonds essentially achieve investor protection by insulating the Purple/Blue bonds from a debt restructuring, cushioned by Red bonds. In other words, Purple bonds would be 'restructuring-proof' , which renders the issue of investor representation for Purple/Blue bonds redundant. Perhaps the greatest advantage of Purple bonds in comparison to SBBS and Ebonds is that transition problems, which typically arise in the context of contractual change, could be excluded. From day one on, all sovereign debt issued by euro area Member States would be benefit from a 'no-restructuring' guarantee in the context of ESM programmes, thereby reassuring investors that the ESM will not 'pull the plug' by insisting on burden-sharing between the ESM and the private sector. Consequently, at the very minimum, recital (12) of the ESM Treaty would need to be removed. This recital foresees adequate and proportionate private sector involvement (PSI), and hence debt restructuring, when the ESM provides financial stability support. However, this raises questions about the interaction between, on the one hand, Article 125 TFEU, which essentially forbids any cross-border bailouts of ailing EU Member States ('no-bailout' clause), 207 and the 'no-restructuring' guarantee on the other hand. 208 Suppose that the restructuring, ie the reduction of their interest and principal, of Red bonds fails to create the financial space necessary to re-establish debt sustainability. 209 While under normal circumstances, the ESM would now demand a restructuring of the remaining (Purple/Blue) bonds, the 'no-restructuring' guarantee would forbid this course of action. From this, two questions follow: first, who ought to bear the losses if it not the holders of Blue bonds and, second, could such no-restructuring guarantee be compatible with the EU Treaties, and specifically Article 125 TFEU. The objective of the 'no-bailout' clause is that 'Member States remain subject to the logic of the market when they enter into debt, since that ought to prompt them to maintain budgetary discipline' , as held by the European Court of Justice (ECJ) in Pringle. 210 According to the Court, in order to honour this logic, the ESM may never assume the debts of the recipient Member State-neither the existing debts nor the new obligations the country incurs by entering into an ESM programme. 211 Hence, unless the ECJ was to perform a 180-degree turn, the ESM may safely be excluded from the potential resources that may be used in this specific scenario. 212 Given that other private creditors than those owning Blue bonds are not available, the only way of not further stretching the ESM's (overall) financing envelope would be to impose harsher macroeconomic conditionality with a view at creating a greater fiscal surplus and a steeper path back to debt sustainability. This strategy seems ill-fated at best and outright dangerous at worst, especially in light of the experience with Greece's financial assistance programme. 213 Another issue worth discussing is how solid the protection of investors, ie holders of Purple bonds, would be in practice. Importantly, a country may, under international law, restructure its bonds unilaterally, outside an ESM programme. 214 The authors of the Purple bonds proposal assume, and probably rightly so, that a Member State would always restructure its debt under the auspices of an ESM programme. 215 However, at least in theory, the decision to restructure lies not with the ESM but with the issue of the debt instrument-the ESM has no formal control over such decision. 216 It is true that restructuring sovereign debt unilaterally and outside an ESM programme may not merely lead to transactional complications but could even result in the expulsion of a Member State from the currency area. But the redenomination risk that the authors of the Purple bonds proposal seek to eliminate, and which presently contributes to the spread between say German and Italian sovereign debt, might not disappear by eliminating debt restructuring in the context of ESM programmes. Overall, from the perspective of investor protection, the question is whether the statutory immunization of Purple bonds will convince creditors that the respective government bonds are safe. To that end, technical legal aspects are of lesser relevance. Similar to Purple bonds, the protection of investors in Coronabonds as well as their representation in debt workouts is more straightforward than in the case of synthetic safe assets. First, as mentioned above, 217 holders of Coronabonds are likely to enjoy a level of legal protection similar to that of investors in bonds issued by the EU as well as its supranational institutions. For the sake of simplicity, it is assumed in this section that Coronabonds are not backed by several guarantees. 218 Generally speaking, debts issued by the EU are direct and unconditional obligations and the EU has an obligation in primary law to service them under Article 323 TFEU. 219 The current EU Base Prospectus further stipulates that the 'EU's debt service is ensured based on multiple layers of debtservice protection' , 220 noting that lending occurs 'back-to-back' and that 'the EU budget guarantees that the EU timely honours its debt repayment obligations' . 221 In addition, the EU Base Prospectus includes several specific contractual terms that may be relevant from an investor protection perspective. This section will focus on the following two: (i) the pari passu clause and (ii) the (lack of a) negative pledge clause. 222 Pari passu clauses are notorious in sovereign bonds, given the absence of statutory ranking provisions, while sovereign immunity is typically the biggest obstacle to efficient debt enforcement. While it seems unlikely that the EU would ever seriously consider breaching the pari passu clauses or come close to enforcement proceedings, it should be assumed that the inclusion of the respective contractual provisions in the EU Base Prospectus is not without meaning or legal value. The EU's pari passu clause mixes (old) sovereign bond boilerplate with some specific language. 223 For one, the clause does not follow the latest International Capital Market Association (ICMA) recommendations, which clarified that the pari passu clause in sovereign bond contracts does not require the issuer to make equal or rateable payments. 224 Given the lack of challenges to sovereign debt workouts in Luxembourgish courts, it would be speculative to conclude that investors could leverage the pari passu clause, as some did in New York courts. 225 The current pari passu clause at least does not completely exclude the possibility. For another, the clause states that the bonds rank pari passu with other present and future unsecured obligations for money borrowed in application of decisions by the Council of the EU. In other words, any new debts created by means of Council decisions, the required legal basis for EU borrowing, would enjoy the same ranking as existing debts. Overall, the EU pari passu clause in the EU Base Prospectus ensures a reasonable level of investor protection from subordination, which is at least equivalent but mostly superior to that provided by domestic-law sovereign bonds. 226 The second relevant contract tool to protect bondholders are negative pledge clauses, which are a promise by the borrower not to encumber its assets for the benefit of other creditors. 227 In contrast to most international sovereign bonds, however, the EU Base Prospectus lacks such a negative pledge clause. To this end, the Commission may want to consider including such provisions to further boost investor protection and align the drafting standard for international government bonds governed by English or New York law. Turning to the issue of investor representation, the EU Base Prospectus entails a CAC to reduce holdout inefficiencies in debt restructurings. As is the case for the pari passu clause, however, the EU relies on an older 'double-limb' CAC version that is no longer recommended by ICMA and the IMF for international sovereign bonds. 228 The old clause is more vulnerable to holdout investors since it does not allow for a single bondholder vote across all outstanding series. 229 For the investor community as a whole and the marketability of Coronabonds, it would be beneficial if the EU Base Prospectus were to follow the international standard and include single-limb CACs with full aggregation features. This article examined four types of safe assets for the euro area, focusing on those legal features that arguably matter most for investors: (i) formal features, ie governing law and dispute settlement forum, and (ii) functional features, ie investor protection and investor representation in debt restructurings. As Table 1 shows, the four types of safe assets differ markedly with regard to these features. Looking at the Commission's SBBS proposal and the European Parliament's report on it, critical legal questions have not yet been answered to a satisfying degree. This article elaborated on some of these questions, without however claiming to be exhaustive. To avoid the emergence of a fragmented SBBS market, the harmonization of governing law and dispute settlement fora was recommended, along with some specific measures to increase investors' confidence in the contractual scaffolding of SBBS. Second, with respect to E-bonds, it was seen that the legal framework underpinning ESM bond issuances alleviates some of the complications that arise in the context of SBBS, such as questions on governing law. In the context of protecting E-bondholders, the onus will lie on the reinforcement of the ESM's PCS, as Leandro have already pointed out. To this end, the article proposes explicitly mentioning the ESM's seniority in its loan documentation, but also discusses other options. With regard to E-bondholders' representation in debt restructurings, it is noted that the ESM would act as an intermediary between E-bondholders and the sovereign and highlight some of the challenges associated with this structure. Third, the analysis of the Purple bonds proposal shows that it would be the least invasive, only requiring minimal adjustments to the current legal framework. However, the Purple bonds concept rests on the premise that the statutory immunization of senior bonds from restructuring (as part of an ESM program) will boost investor confidence. Whether this immunization alone is sufficient to create a 'safe' asset remains an open question. Finally, Coronabonds, a one-off type of partially mutualized EU debt instrument to finance the economic recovery from the COVID-19 pandemic were discussed. It was assumed that the Commission would issue Coronabonds by relying on the updated EU Base Prospectus, and some recommendations to align the bond documentation with international issuance standards were made. Overall, Coronabonds present the simplest structure for a European safe asset, since they do not rely on securitization or an intermediary but constitute direct debt financing by the EU. However, whether Coronabonds will ultimately be issued and whether they can provide the basis for genuine Eurobonds remains difficult to anticipate at the current juncture. Coronabonds were not primarily designed to provide a safe haven asset and delink sovereigns from banks See section III.2.c Proposal for a Council Regulation establishing a European Recovery Fund Instrument to support the recovery in the aftermath of the Could Bail Europe Out, Tie It Together' The Washington Post For an overview of previous bond issuance programmes by the European Commission, see Sebastian Horn Debt Issuance Programme NT7S7k8V3IFuH9GNA/B/I6wGoFjxHXrl55sAfC8EmnzRFo+r//fpT2xLj/KO4g3I=> accessed 31 A mezzanine tranche could be issued as a medium-risk product In this regard, the European Commission's proposal explains that '[t]he various securities issued would bear any losses from the underlying portfolio in a certain sequence (i.e., losses would accrue first to holders of sub-senior, or subordinated, securities and only after such securities have been completely wiped out would they also accrue to the holders of senior claims As further discussed below, with regard to the relationship between SBBS holders and government bond holders, policymakers aim to create equal treatment instead of subordination establishing a framework for the recovery and resolution of credit institutions and investment firms and amending Council Directive 82/891/EEC, and Directives Trust and the Civil Law' (1982) 42(5) Louisiana Law Review 1709 Sovereign-Debt Restructuring in Europe under the New Model Collective Action Clauses ) Capital Markets Law Journal 410, 412. For two earlier papers providing a deeper insight into the trade-offs between trusts and fiscal agency agreements in the realm of sovereign bond financing; see Sönke Häseler, 'Trustees versus Fiscal Agents and Default Risk in International Sovereign Bonds They point out that Italy had prepared to use such modified pari passu clause in its New York law issuances, but ultimately refrained from doing so (finding that multilateral official creditors are the most senior creditors of sovereigns See section III.2.b See Arnold and others (n 91) Strengthening the Contractual Framework to Address Collective Action Problems in Sovereign Debt Restructurings See section III.2.c and II See recital (16) of the revised ESM Treaty. See Draft revised text of the treaty establishing the European Stability Mechanism as agreed by the Eurogroup on 14 At the same time, art 16(3) states that '[t]he financial terms and conditions of each ESM loan shall be specified in a financial assistance facility agreement Their 'safety' feature would be derived from a statutory insurance from debt restructuring by amending the ESM Treaty; see Bini-Smaghi and Marcussen See for the seminal case on the interpretation of the 'no-bailout clause' in the context of financial assistance programmes in Case C-370/12 Pringle v Government of Ireland and Others I owe gratitude to Sam Langfield for pointing me to this issue Note that the sustainability of public debt is a pre-condition for an ESM financial assistance programme 207) para 135 (further noting that '[c]ompliance with such discipline contributes at Union level to the attainment of a higher objective, namely maintaining the financial stability of the monetary union Of course, the very purpose of the ESM is to lend to countries in financial distress. However, we are exploring how a financing gap may be bridged if the restructuring of Red bonds was insufficient The IMF's Independent Evaluation Office openly criticized the IMF Board's decision not to insist on a pre-emptive debt restructuring, noting that '[the decision not to restructure] magnified the required fiscal adjustment, and thereby, at least in part The Right to Restructure Sovereign Debt For an analysis of the sovereign debt restructuring process under ESM programmes both de lege lata and ferenda, see eg Grund and Stenström See section II This was, for instance, proposed by the European Commission (n 81) in 2011 and by Grund and others Ibid (further stating that 'the Commission can make available its cash buffer, re-prioritise budgetary expenditure and if needed draw additional resources from the Member States Similarly, Waibel (n 28) mentions that '[t]he two most important contractual devices are negative pledge and pari passu clauses and the Receipts and Coupons relating to them constitute unsecured, direct, unconditional and general obligations of the Issuer and will at all times rank pari passu without any preference among themselves and with all other present and future unsecured obligations of the Issuer for money borrowed in application of decisions of the Council of the European Union, except for indebtedness (a) incurred for all or part of the purchase price of property purchased by the Issuer and (b) secured by a mortgage, lien Solving the pari passu puzzle: the market still knows best In fact, most euro area government bonds governed by local law do not have any pari passu clause Further noting that by virtue of the negative pledge clause 'borrowers commit not to grant security to other creditors, or in the alternative, to grant equivalent security to the creditors benefitting from the negative pledge For a discussion, see Grosse Steffen and others Journal of Comparative Law 367. Recent efforts to update the Euro CAC by introducing single-limb CACs as part of the ESM Treaty Reform have stalled, however. See 'Italy's econmin says ESM reform likely to be delayed' Reuters