Collective action clauses and beyond: Negotiated public debt restructuring in EU Investment Protection Agreements – Enrico Tinti


Collective action clauses and beyond: Negotiated public debt restructuring in EU Investment Protection Agreements

By Enrico Tinti


During the course of history, states have repeatedly defaulted on their sovereign debt.[1] As no single specialised international forum for the settlement of sovereign debt disputes is yet in place, foreign bondholders have increasingly resorted to investment arbitration seeking compensation amounting to full repayment of the bonds. However, uncertainties as to the jurisdiction of investment tribunals remain. At the same time, collective action clauses (CACs) have progressively been introduced in sovereign bond contracts to help facilitate negotiated debt restructurings. The recent Eurozone sovereign debt crisis has shown the importance of such clauses and the practical consequences stemming from the differences in their design. Finally, the investment agreements recently signed by the EU with third parties limit investment protection when public debt restructuring is approved by a qualified majority of bondholders. This article argues that such treaty provisions are a necessary condition for the successful and orderly restructuring of sovereign bonds, while safeguarding creditors equality.

  1. The rationale behind the introduction of CACs in sovereign bonds

Sovereign debt restructuring (SDR) techniques enable states in financial distress to modify original payment terms in order to secure a more sustainable debt burden.[2] This may occur either after a default or under the threat thereof. The use of CACs can be considered an efficient SDR technique. CACs, included in the contractual terms of a debt instrument, allow the issuer and a qualified majority of bondholders to change the contract’s key financial terms.[3] In this way it is possible to postpone the date of payment and reduce the amount payable – so called ‘haircuts’ – including any overdue amount. Most importantly, the agreement produces a modification to the contractual rights of all bondholders and is thus binding on the disagreeing minority as well.[4]

Over the past 30 years sovereign borrowers have shifted towards greater dependence on capital markets financing, resulting in increased heterogeneity of creditors.[5] Yet, despite this evolution, there remains no single specialised international forum for the settlement of disputes between sovereign borrowers and creditors. In the wake of Argentina’s 2001 default crisis, the International Monetary Fund (IMF) advanced a proposal for a statutory sovereign debt restructuring framework, including a quasi-court for public bankruptcy matter, called Sovereign Debt Restructuring Mechanism (SDRM).[6] However, despite the Fund’s efforts, the project could not come into existence due to the lack of support among IMF Member States.[7] As a result, attention was focused on a contractual approach in order to facilitate orderly sovereign debt restructurings in the future. The international community welcomed the practice of including CACs in sovereign bond contracts and advocated for increasing their adoption.[8]

Before the introduction of CACs in sovereign bonds, debt crises have been addressed through a series of ad hoc proceedings.[9] States have typically operated reductions of their outstanding debt either unilaterally or with the voluntary acceptance of creditors, often by way of ‘take-it-or-leave-it’ bond exchange offers. In turn, this approach has often been perceived as coercive by the creditors.[10] In addition, given the voluntary basis of acceptance, creditors experience a conflict of interests. Their collective interest would be to reach an overall satisfactory agreement, on equal terms and with a shared percentage of loss. However, each of them faces an individual incentive to hold out, freeriding on the debt relief granted by other creditors, and enforce full payment on their individual contractual claim against the state.[11] To this end, as no international specialised dispute settlement forum is in place, holdout creditors have sought judicial remedies before domestic courts, foreign claims commissions, and, increasingly, investment arbitration tribunals under the framework of the Convention on the Settlement of Investment Disputes between States and Nationals of Other States (ICSID Convention).[12]

  1. Investor-State arbitration in sovereign bonds disputes: A viable option?

In recent years Investor-State arbitration has attracted growing attention as a potential remedy available to foreign bondholders against sovereign borrowers. Yet, jurisdictional controversies have arisen. While the ICSID Convention empowers investment tribunals to adjudicate ‘any legal dispute arising directly out of an investment’,[13] it does not define the term further.[14] Therefore, it is still disputed whether debt and financial instruments can be considered an ‘investment’ for the purpose of establishing ICSID tribunals’ jurisdiction ratione materiae.[15]  Recourse is often made to the definition of investment provided in bilateral or multilateral investment treaties forming the substantive legal basis of disputes. However, as acknowledged by commentators,[16] a two-step review of jurisdiction is required. First, whether the subject matter of the dispute falls within the outer limits of the term investment, as derived from the interpretation of article 25 of ICSID Convention. This leads to the exclusion of ordinary commercial transactions, outside the ‘objective core’ of the term.[17] Second, whether the dispute relates to an investment included ratione materiae in the ambit of protection of the relevant treaty.

Investment tribunals have come to opposing conclusions as to whether sovereign debt falls within the scope of investment treaty protection. Arbitration proceedings initiated following Argentina’s sovereign debt crisis led to the first ICSID Tribunals confirming their own jurisdiction over SDR disputes.[18] The first was the Tribunal in Abaclat in August 2011.[19] The subsequent Tribunals in Ambiente Ufficio[20] and Alemanni[21] followed the decision. The Argentina–Italy Bilateral Investment Treaty (BIT), at the basis of the three disputes, expressly included ‘bonds, private and public securities’ in the definition of investment.[22] One year later, in the aftermath of the Greek SDR, the Tribunal in Poštová banka[23] came to the opposite conclusion, declaring that the Greek sovereign bonds bought by the claimant did not constitute an investment under the Greece–Slovakia BIT.[24] Notably, the treaty defines investment as ‘every kind of asset’, explicitly though not exclusively including ‘loans, claims to money, or to any performance under contract having a financial value’.[25] The tribunal took a rather restrictive view, observing that sovereign bonds, being held and traded by a multitude of anonymous creditors, substantially differ from loans. Therefore, it held that the claimant bank, having purchased the government bonds on the secondary market, never established a direct contractual relationship with Greece and therefore had no contractual rights against the respondent.[26] As in Abaclat and Ambiente Ufficio, the main argument to deny jurisdiction was also derived from the interpretation of the term ‘investment’ in the BIT at issue. While the explicit inclusion or exclusion of public debt in treaty definitions of investment are likely to remove ambiguities, arbitration case-law reveals tribunals’ discretion on the matter.

  1. Collective action clauses and the eurozone sovereign debt crisis

At the European Council summit held on 24 and 25 March 2011, Member States agreed that from July 2013 CACs would be included in all new euro area government securities with maturity above one year.[27] Such obligation was made legally binding by article 12(3) of the treaty establishing the European Stability Mechanism (ESM),[28] ratified and implemented by all euro area Member States.[29] While CACs were already common in euro area sovereign bonds governed by foreign law – particularly English and New York law – their introduction represented a significant contractual change in bonds governed by domestic law, which in turn account for the vast majority of issued debt.[30]

Developed by the Economic and Financial Committee (EFC),[31] the ‘euro area model CAC 2012’ (euro CAC) includes an aggregation clause whereby a qualified majority of bondholders across multiple bond issues can agree to a single comprehensive SDR project.[32] To achieve such an aim, a ‘two-limb’ voting procedure is envisaged. Specifically, the offer must obtain the consent of creditors holding 66⅔ per cent of the outstanding principal in each bond series and 75 per cent of the aggregated total outstanding principal across all series subject to restructuring. Due to the dual majority requirement holdout creditors can prevent a specific, usually smaller, bond series from being restructured by acquiring a sufficient blocking minority in that series as evidenced in the Greek SDR issue.

Conversely, most recent standard CACs developed in 2014 by the International Capital Market Association (ICMA) include a ‘single-limb’ voting mechanism.[33] Bondholders cast a single vote and the relevant bonds can be restructured if a 75 per cent majority threshold is met with respect to the aggregated amount of outstanding debt securities. Admittedly, this makes it too expensive if not impossible for creditors to hold out by acquiring a blocking position and stop the restructuring or substantially reduce its value.[34] Against this backdrop, at the Eurogroup meeting held in December 2018, euro area finance ministers decided on the inclusion of single limb CACs in new sovereign bonds as of 2022, also with a view to amending the ESM Treaty accordingly.[35]

On the outbreak of the Greek sovereign debt crisis, a small minority of Greek bonds was governed by foreign law – mainly English law – and included CACs. However, the vast majority thereof had been issued under Greek law and without such clauses. Holdout risk is high where there are series-by-series voting procedures, and even higher if no predetermined binding contractual mechanism allowing for restructuring is in place. Therefore, by way of the Bondholder Act 4050/2012 Greece amended its domestic law with retroactive effect, introducing statutory CACs in its outstanding bond contracts.[36] As CACs in foreign-law bonds only provided for series-by-series voting procedures and did not contain aggregation features, restructuring often failed due to holdout creditors’ opposition. On the contrary, Greek statutory clauses featured an aggregation mechanism involving single-limb voting structure with a lower voting threshold, up to just 50 per cent of bondholders.[37] While the retroactive, ex-post legislative introduction of CACs in bond contracts created some controversy, their debtor-friendly design allowed for the successful outcome of the SDR.

  1. Beyond CACs: Negotiated public debt restructuring in EU international investment agreements

This section analyses the protection afforded to sovereign debt holders in the international investment agreements (IIAs) recently signed by the EU and its Member States with third countries, namely the investment chapter of the Comprehensive Economic and Trade Agreement (CETA) with Canada,[38] the EU-Singapore Investment Protection Agreement (IPA),[39] and the EU-Vietnam IPA.[40] At the time of writing, ratification of the agreements is still pending. However, as these are the first IIAs signed by the EU pursuant to its common commercial policy, they show consolidated evidence of the negotiating pattern pursued by the EU in the field of foreign investments protection and are likely to represent a model for future EU IIAs.

4.1 Public debt and the definition of investment

After framing the general characteristics an asset must satisfy to fall under the definition of investment,[41] the IIAs expressly list the forms an investment may take. Under article 8.1 CETA, bonds issued by an enterprise, tangible or intangible moveable property, claims to money, and claims to performance under a contract are included within the definition of investment.[42] Specific reference to public debt issued by a party is subsequently made at article 8.18, paragraph 4. Both the EU-Singapore and the EU-Vietnam agreements comprise ‘bonds, debentures, and loans and other debt instruments, including rights derived therefrom’ as covered forms of investment.[43]

4.2 Investor protection in case of negotiated public debt restructuring

With regard to the substantive protection afforded, all investor protection provisions contained in the relative agreements would in principle be applicable, including fair and equitable treatment (FET) and protection against expropriation.[44]However, each agreement includes an annex narrowing down the standards of treatment available to public debt holders in case an SDR qualifies as a negotiated restructuring.[45] Therefore, it is of utmost importance to assess when an SDR can be considered to be negotiated. This is the case under two alternative circumstances:

First, if the debt restructuring or rescheduling takes place by way of modification or amendment of the debt instruments, as provided for under their terms, including their governing law.[46] This hypothesis certainly includes modifications adopted by way of CAC procedures, whether such clauses are directly inserted in the terms and conditions of the bond or form part of the jurisdiction’s laws applicable to the bond contract.[47]

Alternatively, a bond exchange or other similar restructuring process can still qualify as a negotiated restructuring if a qualified majority of bondholders has consented to it. Both CETA and the EU-Singapore IPA qualify such a majority as 75 per cent of the aggregate principal amount of the outstanding debt subject to restructuring. In the EU-Vietnam IPA the threshold is set at 66 per cent of the amount calculated excluding debt held by the debtor State itself or by entities owned or controlled by it.[48]

This second provision clearly comes to the aid of States that have issued bonds without CACs. With regard to the EU, the legal obligation to introduce euro CACs in new bonds only applies to eurozone Member States, and only those issued after 2013. Therefore, this could be the case of bonds issued by non-eurozone Member States or issued by eurozone States before 2013 and still outstanding at the date of restructuring. But that is not all. In this second circumstance, the texts of the annexes require a qualified majority of the aggregate principal amount of the outstanding debt subject to restructuring. If this were the text of a CAC, it would be a single-limb CAC with full aggregation features. Its requirements are allegedly looser than those of CACs with series-by-series voting procedures and admittedly even more so than those of current euro CACs.

Therefore, if CACs are absent from the bond terms or they require different and stricter thresholds, it is conceivably possible to qualify a restructuring as negotiated, for the purpose of the above agreements, even in contrast with or beyond the bond’s contract law framework. Certainly, the effects of such qualification are limited to its legal implications under the relevant investment agreement: in the presence of a negotiated restructuring, investors can only claim a violation of national treatment or most-favoured-nation (MFN) treatment.[49] In this respect, the EU-Singapore IPA limits protection to national treatment, as the agreement does not include MFN. The purpose of these standards is to prevent discrimination between bondholders and other foreign and national holders with regard to the terms offered and payments received.[50]As noted in literature,[51] given the way in which bonds are currently issued and traded in the market, intentional discrimination against some foreign bondholders within a single bond issue, while not impossible, may prove difficult in practice. However, if multiple outstanding bond series are restructured under different terms, it is still possible that differential treatment will fall within the scope of non-discrimination provisions. Nevertheless, the annexes contain an additional significant public policy carve-out: different treatment does not breach non-discrimination standards if in pursuit of legitimate public policy objectives justified by, inter alia, different macroeconomic impact or eligibility for debt restructuring. These may include avoiding systemic risks or spillover effects.[52]


Collective action clauses in EU Member States sovereign bonds, when present, do not show homogeneous features. Their inclusion and design depend on the year of issue, the governing law, and whether the issuing State is part of the eurozone. Such heterogeneity has the potential to prevent coherent and uniform restructuring. In presence of CACs with series-by-series voting procedures, holdout creditors may be able to stop specific bond series from being restructured by acquiring the required blocking minority.

In turn, absent a specialised international court, holdouts are likely to bring a number of claims in different fora, including Investor-State dispute settlement tribunals. Investment tribunals have come to opposite conclusions as to whether sovereign debt falls within the scope of their jurisdiction. This reflects both the differences in treaty formulations and arbitrators’ wide margin of discretion on the matter. In summary, whether investment arbitration represents a viable option largely depends on the substantive provisions of investment agreements.

Under IIAs recently signed by the EU, bondholders’ chances of successfully opposing a negotiated restructuring are effectively restricted as claims are only admissible for breach of national treatment or most-favoured-nation treatment, subject to an extensive public policy carve-out. On the one hand, this may be seen as a deficit in investor protection. On the other, it can constitute an extra push for debtor states to seek a restructuring solution with a large acceptance rate among bondholders – the 75 per cent threshold representing a reasonable safeguard – while being a step towards reducing holdouts’ ability to obstruct SDRs by way of threatening recourse to investment arbitration. As such, by limiting the substantive claims available under an IIA, the examined treaty provisions de facto produce an effect equivalent to that of single-limb CACs. This goes even beyond the weaknesses and asymmetries of CACs included in sovereign bonds.


[1] Udaibir S Das, Michael G Papaioannou, and Christoph Trebesch ‘Sovereign Debt Restructurings 1950–2010: Literature Survey, Data, and Stylized Facts’ (IMF Working Paper 12/203 1 August 2012 <> accessed 5 March 2020.

[2] Risto Nieminen, Mattia Osvaldo Picarelli, ‘Sovereign Debt Restructuring. Main Drivers and Mechanism’ (2017) European Parliamentary Research Service (EPRS) Briefing, <> accessed 9 March 2020.

[3] Patrick S. Kenadjian, Klaus-Albert Bauer and Andreas Cahn (eds), Collective Action Clauses and the Restructuring of Sovereign Debt (De Gruyter 2013) 3-12.

[4] Philip Wood, Law and Practice of International Finance (Sweet & Maxwell 2008) 189-191.

[5] Chanda DeLong and Nikita Aggarwal, ‘Strengthening the contractual framework for sovereign debt restructuring—the IMF’s perspective’ (2016) 11 Capital Markets Law Journal 25, 26.

[6] Anne O Krueger, A New Approach to Sovereign Debt Restructuring (International Monetary Fund 2002) <> accessed 12 March 2020.

See also: IMF, ‘Report of the Managing Director to the International Monetary and Financial Committee on a Statutory Sovereign Debt Restructuring Mechanism’ 8 April 2003, <> accessed 12 March 2020.

[7] Benu Schneider, ‘Sovereign debt restructuring: further improvements in the market-based approach’ (2018) 13 Capital Markets Law Journal 294. See also Alexis Rieffel, Restructuring Sovereign Debt: the case for ad hoc machinery(Brookings Institution Press 2003) 255-256; Brad Setser, ‘The Political Economy of the SDRM’ in Barry Herman, José Antonio Ocampo, and Shari Spiegel (eds) Overcoming Developing Country Debt Crises (OUP 2010) 317-318, 330.

[8] Kunibert Raffer ‘The IMF’s SDRM – Simple Disastrous Rescheduling Management?’ in Chris Jochnick, Fraser Pretson (eds) Sovereign Debt at the Crossroads: Challenges and Proposals for Resolving the Third World Debt Crisis(UOP 2006) 277; Sean Hagan, ‘Designing a Legal Framework to Restructure Sovereign Debt’ in IMF, Current Developments in Monetary and Financial Law. Volume 4 (International Monetary Fund 2008) 266-267.

[9] Deborah Zandstra, ‘A possible work plan for further reform of sovereign debt restructuring’ (2018) 13 Capital Markets Law Journal 356.

[10] Michael Waibel, ‘Opening Pandora’s Box: Sovereign Bonds in International Arbitration’ (2007) 101 The American Journal of International Law 711.

[11] Christoph Grosse Steffen, Sebastian Grund and Julian Schumacher, ‘Collective action clauses in the euro area: a law and economic analysis of the first five years’ (2019) 14 Capital Markets Law Journal 134.

[12] Sebastian Grund, ‘Enforcing Sovereign Debt in Court – A Comparative Analysis of Litigation and Arbitration Following the Greek Debt Restructuring of 2012’ (2017) 1 University of Vienna Law Review 34.

[13] Convention on the Settlement of Investment Disputes between States and Nationals of Other States (1966), art 25.

[14] Campbell McLachlan, Laurence Shore, Matthew Weiniger, International Investment Arbitration—Substantive Principles (OOP 2017) 218.

[15] Jurisdiction ratione materiae (‘by reason of the matter’) refers to the authority of a judicial body to adjudicate disputes on the basis of their subject matter. Investment tribunals are only competent to decide cases concerning assets falling within the applicable definition of ‘investment’. Zachary Douglas, ‘Jurisdiction Ratione Materiae’, The International Law of Investment Claims (Cambridge University Press 2009).

[16] Michael Waibel, Sovereign Defaults before International Courts and Tribunals (CUP 2011) 215.

[17] ibid.

[18] Belen Olmos Giupponi, ‘ICSID Tribunals and Sovereign Debt Restructuring-Related Litigation: Mapping the Further Implications of the Alemanni Decision’ (2015) 30 ICSID Review 556, 564.

[19] Abaclat and others v Argentine Republic (formerly Giovanna a Beccara and others v Argentine Republic), ICSID Case No ARB/07/5, Decision on Jurisdiction and Admissibility, 4 August 2011.

[20] Ambiente Ufficio SpA and others v Argentine Republic (formerly Giordano Alpi and others v Argentine Republic), ICSID Case No ARB/08/9, Decision on Jurisdiction, 8 February 2013.

[21] Giovanni Alemanni and others v Argentine Republic, ICSID Case No ARB/07/8, Decision on Jurisdiction and Admissibility, 17 November 2014.

[22] Agreement between the Italian Republic and the Republic of Argentina on the Reciprocal Promotion and Protection of Investments (1990), art 1(1)(c), <> accessed 15 March 2020.

[23] Poštová banka, a.s. and Istrokapital SE v Hellenic Republic, ICSID Case No ARB/13/8, Award, 9 April 2015. See Francesco Montanaro, ‘Case comment: Poštová Banka SA and Istrokapital SE v Hellenic Republic. Sovereign Bonds and the Puzzling Definition of “Investment” in International Investment Law’ (2015) 30 ICSID Review 549.

[24] Agreement between the government of the Czech and Slovak Federal Republic and the Government of the Hellenic Republic for the promotion and reciprocal protection of investments (1991), <> accessed 14 March 2020.

[25] Article 1, para 1, let (c) Greece–Slovakia BIT.

[26] Poštová banka (n 23) [337]-[338], [343]-[345].

[27] Conclusions of the European Council (24/25 March 2011) EUCO 10/11, 31, <> accessed 2 April 2020.

[28] Treaty establishing the European Stability Mechanism (2012) <> accessed 28 March 2020.

[29] Economic and Financial Committee, Report on the implementation of euro area model Collective Action Clauses (CACs) <> accessed 28 March 2020.

[30] Christoph Grosse Steffen, Sebastian Grund and Julian Schumacher (n 11) 136.

[31] For the activity of the EFC Sub-Committee on EU Sovereign Debt Markets in developing the euro area Model CAC 2012, see: <> accessed 29 March 2020.

[32] The text of the euro area Model CAC 2012 is available at: <> accessed 29 March 2020.

[33]  The text of ICMA standard CACs 2014 is available at <> accessed 29 March 2020.

[34] Chanda DeLong, Nikita Aggarwal (n 5) 30.

[35] Eurogroup report to Leaders on EMU deepening, 4 December 2018, <> accessed 2 April 2020.

[36] Sebastian Grund, ‘Restructuring government debt under local law: The Greek case and implications for investor protection in Europe’ (2017) 12 Capital Markets Law Journal 253.

[37] Yannis Manuelides, ‘Using the local law advantage in today’s eurozone (with some references to the Republic of Arcadia and the Mamatas judgment)’ (2019) 14 Capital Markets Law Journal 469.

[38] Comprehensive Economic and Trade Agreement between Canada, of the one part, and the European Union and its Member States, of the other part (2016), ch 8. <> accessed 15 March 2020.

[39] Investment Protection Agreement between the European Union and its Member States, of the one part, and the Republic of Singapore, of the other part (2018). <> accessed 15 March 2020.

[40] Investment Protection Agreement between the European Union and its Member States, of the one part, and the Socialist Republic of Vietnam, of the other (2019). <> accessed 15 March 2020.

[41] Including ‘the commitment of capital or other resources, the expectation of gain or profit, the assumption of risk and a certain duration’; See art 8.1 CETA, art 1.2 EU-Singapore IPA, art 1.2, let (h) EU-Vietnam IPA.

[42] Article 8.1 let (c), (h) and (i) CETA.

[43] Article 1.2, para 1, let (c) EU-Singapore IPA; art 1.2 let (h), number (iii) EU-Vietnam IPA.

[44] Fair and equitable treatment: art 8.10 CETA; art 2.4 EU-Singapore IPA; art 2.5 EU-Vietnam IPA. Expropriation: art 8.12 CETA; art 2.6 EU-Singapore IPA; art 2.7 EU-Vietnam IPA.

[45] Annex 8-B CETA, <> accessed 15 March 2020; annex 4 EU-Singapore IPA, <> accessed 15 March 2020; annex 5 EU-Vietnam IPA, <> accessed 15 March 2020.

[46] Annex 8-B para 1, let (a) CETA; annex 4, para 3(i) EU-Singapore IPA; annex 5, para 3, let (a)(i) EU-Vietnam IPA.

[47] For the definition of governing law: annex 8-B para 1 CETA; annex 4, para 3 EU-Singapore IPA; annex 5, para 3, let (b) EU-Vietnam IPA.

[48] Annex 8-B para 1, let (b) CETA; annex 4, para 3(ii) EU-Singapore IPA; annex 5, para 3, let (a)(ii) EU-Vietnam IPA.

[49] Annex 8-B, para 2 CETA; annex 4, para 1 EU-Singapore IPA; annex 5, para 1 EU-Vietnam IPA. The EU-Singapore agreement does not include the standard of most-favourite-nation treatment. Therefore, protection is only limited to national treatment.

[50] Waibel, Sovereign Defaults before International Courts and Tribunals (n 16) 274.

[51] ibid.

[52] Annex 8-B, para 3, footnote 7 CETA; annex 4, para 1, footnote 1 EU-Singapore IPA; annex 5, para 2, footnote 1 EU-Vietnam IPA; For the economic rationale behind giving priority to domestic debt during an economic crisis, see: UNCTAD, ‘Sovereign debt restructuring and international investment agreements’ (2011) 2 IIA Issues Note <> accessed 15 March 2020.