Jump to Content Jump to Main Navigation
Signed in as:

4 Governing Law of Sovereign Bonds

From: Sovereign Defaults Before Domestic Courts

Hayk Kupelyants

From: Oxford Legal Research Library (http://olrl.ouplaw.com). (c) Oxford University Press, 2023. All Rights Reserved. Subscriber: null; date: 06 June 2023

Bonds — Debt — Sovereign debt

(p. 111) Governing Law of Sovereign Bonds

I.  Introduction

4.01  In the UK, the governing law of bonds is to be determined pursuant to the Rome I Regulation. The Regulation provides that the parties are free to agree to the governing law of bonds (Art 3) and in the absence of such choice, the governing law is determined under Art 4 of the Regulation. Although it was historically considered that sovereign debt could not be governed by a municipal law other than that of the sovereign debtor,1 there is now no disagreement that States are at liberty to submit sovereign bonds to a foreign law of their choice.2 External sovereign bonds issued by emerging market economies ubiquitously contain choice of law clauses in favour of the law of an established financial centre, often London or New York. A small percentage of sovereign bonds contain choices of Japanese, German, Italian, or Luxembourg laws. A principle permeating these legal systems is that choice of law clauses are binding and enforceable.

4.02  Choice of law clauses need not be formulated in express terms. Buchheit, Gulati, and Tirado examined a clause commonly found in sovereign bonds, which stipulates that all payments are to be subject to the laws, regulations, and directives in the place of payment. Rather unusually, the clause in Cypriot bonds subjects all payments ‘to any applicable fiscal or other laws, regulations and directives’ tout court, without the qualifying caveat of ‘in the place of payment’. The authors argued that this linguistic formulation entitled Cyprus to subject payments under the bonds to the Cypriot law even if the place of payment is outside Cyprus,3 and, as the case may be, to use the clause as a legal basis for restructuring the bonds through legislative amendments. Similarly worded clauses found in Ukrainian (and other) bonds have invited comparable reasoning.4

(p. 112) 4.03  This interpretation is by no means the only reasonable construction of the clause. Understanding the clause as subjecting the bonds to the law of the sovereign debtor eviscerates the enquiry as to why the law of that debtor should apply in the first place. In other words, why is it the ‘applicable’ fiscal or other law? To apply the law of the sovereign debtor requires it to be applicable under the choice of law rules of the prorogated forum or under the contract in question.

4.04  In most cases, the omission of the qualifying ‘at the place of payment’ would not result in a different result. In any case, English choice of law rules associate the applicability of a fiscal or other rule with the place of payment. First, English courts would only apply the mandatory provisions of the place of performance of the contract (Art 9(3) of the Rome I Regulation), discussed below. (Note that US law may adopt a broader view of the applicability of mandatory provisions.5) Second, formalities of the contractual performance are also for the law of the place of performance (Art 12(2) of the Rome I Regulation).

4.05  Moreover, the bonds in question come with a distinct choice of law provision. One should not presume that the parties to the bonds had two governing laws or perhaps the option of depéçage in mind at the time of drafting the bonds. Properly understood, clauses on ‘applicable fiscal and other laws’ do not necessarily conflict in any meaningful way with the choice of law provision.

4.06  Leaving aside the cases where the bonds contain choice of law clauses, either express or tacit, considerable controversy surrounds the determination of the law applicable to sovereign debt in circumstances where the parties have remained silent as to the governing law of the transaction in question. Empirical research shows that choice of law clauses are in no way ubiquitous in internal (domestic) bonds, which constitute the main focus of this chapter (external bonds, as explained, invariably contain a choice of law provision). The chapter thus addresses the question what law applies to domestic bonds in absence of a choice of law provision.

4.07  First a few words of introduction about domestic debt. Alongside international bonds issued in international financial centres, States commonly float domestic debt in their own territory.6 Domestic debt has seen steady growth since the recent financial crisis,7 although the rise of domestic bond (p. 113) markets in developing countries pre-dates the crisis. The domestic sovereign debt, including its governing law, remains one of the least researched topics in the sovereign debt literature.

4.08  To determine how often, if ever, governing law provisions appear in domestic bonds, questionnaires were sent to debt management agencies of EU Member States and major national economies (Brazil, China, India, the Russian Federation, the US, etc).8 Among other questions, the survey asked whether domestic bonds of the particular State contained a choice of law provision. The responses received show that most domestic debt issuance lacks express choice of law clauses. While the bonds issued by many developed countries, such as Italian, Portuguese, Irish, and pre-restructuring Greek bonds,9 are often expressly subjected to their local law, the number of bonds with no choice of law clauses is large. Out of eighteen responses received, domestic bonds of eight States did not have a choice of law clause, namely Albania, Brazil, Bulgaria, Costa Rica, Cyprus, Czech Republic, Lithuania, and Uruguay. Even when choice of law clauses appear in sovereign bonds, they may be lacking in retail bonds (Belgium). In the UK, a governing law provision expressly specifying English law was first introduced in 2007, and only applies to gilts first issued after 20 August 2007.10 The gilts floated before 20 August 2007 were silent on choice of law matters.11

4.09  When the domestic debt issued by the State is subscribed to by its own residents, the law of the sovereign debtor will govern its own domestic debt under any system of the conflict of laws.12 In such a situation, the debt (p. 114) will in all respects be domestic. Nevertheless, with the advent of financial globalisation and the interconnectedness of capital markets, domestic debt may no longer qualify as domestic at all times. The reality is that the domestic debt might and often does contain foreign elements. One of the followings grounds might be relied upon to distinguish domestic from foreign indebtedness:

  • •  the residence of bondholders;

  • •  currency;13

  • •  the place of marketing of bonds; or

  • •  the law governing the debt (jurisdiction of local courts).14

4.10  So far as the first two elements are concerned, the domestic debt may be, and in practice is often held by foreign residents, and to a lesser degree, denominated in foreign currency.15 Thus, in the restructurings of Russian (1998) and Ukrainian (1998) domestic debt, external creditors held a considerable number of bonds restructured (40% in the case of Russia).16 Involvement of foreign residents introduces a foreign (international) element into the legal characterisation of the domestic debt, which, as will be seen, will be decisive in the conflict of laws analysis. With that in mind, the distinction between domestic and external debt needs to be found somewhere else.

4.11  The conventional theory draws the line between domestic and external bonds on the basis of the place of the marketing of bonds.17 According to Borchard, external debt is the debt marketed abroad, whereas domestic (p. 115) debt is the debt floated within the territory of the sovereign debtor.18 In the Canevaro Brothers case, the sovereign debt was characterised as domestic for the simple reason that it was ‘created and payable’ in the territory of the sovereign, and accordingly, the nationality of the current holders of bonds and the currency did not meaningfully affect the characterisation of the debt.19

4.12  Yet, nowadays this position is problematic for two main reasons: (1) the difficulties associated with defining the market; (2) the blurring of the place of issuance in today’s globalised financial world. First, the concept of a market is cloaked with uncertainty, and to be effective, requires additional characterisation. Several additional features can define a market: the place of advertisement of bonds, the place of physical trading, the place of entering into the contract, the location of contracting parties or the place of payment—either by the bondholders to the debtor or vice versa. The adoption of any of these factors as the defining characteristic of the market ultimately would depend upon the policy that the person qualifying the market follows.20

4.13  Second, the current patterns of domestic debt issuance blur the distinction between domestic and external bonds based on the criterion of the place of issuance or marketing—a phenomenon that the issuing State presumably is aware of. Broadly speaking, capital markets are not segmented as they used to be, and the issuance of bonds usually transcends borders. The delocalisation of financial markets challenges the usefulness of markets as a connecting factor.21

4.14  Tellingly, although domestic bonds are often sold through auctions organised within the territory of the sovereign or listed on domestic stock exchanges, domestic bonds are often marketed abroad and targeted to foreign residents. Developed countries (eg, Belgium) market their domestic debt abroad as a general practice. Sovereign bond auctions are usually advertised on the website of the competent State authority (either the (p. 116) ministry of finance or the debt management agency) for direct purchase, and the information is potentially targeting a wider audience than local residents.22 In practice, domestic retail bonds (to be purchased directly by the investor, without the ‘middle man’) are traded through the online platforms of the sovereign debtor (most prominently, in Italy).23

4.15  It is true that certain countries have a system of primary dealers who have the right to participate in sovereign bond auctions to the exclusion of other (foreign) participants.24 Yet, in other jurisdictions, primary dealers may be foreign banks or financial institutions.25 Bypassing the auction format, and instigated by the worsening of the issuing environment brought on by the financial crisis of 2008, some States also use syndication and private placement as a selling technique.26 Certain States permit the direct purchase of their bonds on such international trading systems as Euroclear.27 All of this opens the holdings of domestic debt to foreigners.

4.16  The Norwegian Loans Case, brought before the International Court of Justice (ICJ), highlights the confusion that may arise in defining the market of sovereign bonds.28 Espousing the claims of French nationals, France argued that the bonds issued by Norway were clearly international and thus evaded the application of the domestic legislation of Norway which purported to nullify the gold clauses contained in the bonds. In its pleadings, France relied upon several characteristics of Norwegian bonds to establish their international character: the bonds (p. 117) were placed with beneficiary bondholders in several jurisdictions and were drafted in different languages, the repayment of the debt was stipulated in various optional currencies and jurisdictions, the bonds were listed on stock exchanges located in several jurisdictions, and the holders of the bonds were almost exclusively foreign.29 In its counter-memorial, Norway pursued the argument that the bonds held by the French residents were in reality traded within the territory of Norway as evidenced by various criteria: the loan contracts with intermediary banks were concluded in Norway, drafted in Norwegian, denominated in Norwegian currency, and Norway always figured as one of the alternative jurisdictions of payment. Additionally, the identity of holders of bearer bonds was irrelevant.30 The ICJ did not resolve this complex issue. Instead, it relied on the self-judging Declaration accepting the compulsory jurisdiction of the ICJ that left the determination of whether the dispute fell within the national jurisdiction to the Norwegian government.31 This excluded the competence of the ICJ to hear the matter since Norway had maintained throughout the proceedings its reliance on the domestic character of the dispute.

4.17  The foregoing discussion shows that the dividing line between domestic and international bonds on the grounds of residence, currency, and place of marketing is extremely porous. In light of this, the fulcrum of the dichotomy remains the law governing the debt and the submission to local courts.32 Or so it is argued. As for the jurisdiction of local courts, it is a misleading criterion in and of itself, given that choice of forum clauses are merely procedural.33 Further, the multiplicity of jurisdictional grounds renders this factor somewhat dubious.

4.18  Even the former factor—the application of the law of the sovereign debtor—is at least equivocal in international finance since sovereign bonds may lack a choice of law clause. Further discussion will demonstrate that (p. 118) the quasi-automatic application of the law of the sovereign debtor to its domestic debt is at best doubtful.

II.  Applicability of Rome I Regulation

4.19  It is a common belief that sovereign issuers proceed on the assumption that local law implicitly governs all domestic debt.34 However, it is suggested here that this general assumption often rests on an insecure foundation and ignores the conflict of laws rules applicable in the English (and other) courts. Assuming that sovereign debt litigation is brought before English courts, the common regime of the commercial conflict of laws, as embedded in the Rome I Regulation, would control the determination of the law governing domestic bonds.

4.20  Two threshold matters first need to be overcome to apply the Rome I Regulation. First, the Regulation applies exclusively to ‘civil and commercial matters’ (Art 1(1)). This has been discussed in Chapter 3 in respect of the Brussels Regulation and the same analysis is likely to apply to the Rome I Regulation (Recital 7).

4.21  Second, Art 1(1)(d) of the Rome I Regulation excludes ‘negotiable instruments’ from the scope of its application ‘to the extent that the obligations under such other negotiable instruments arise out of their negotiable character’. Weber comments that bonds are not covered by the then Rome Convention (currently, the Rome I Regulation) as the issuance of bonds is excluded from the coverage of the Rome I Regulation, and instead regulated by the choice of law rules contained in the Bills of Exchange Act of 1882.35 An entire line of English cases, discussed below, which applied common rules of private international law to the determination of the governing law of bonds defies this view. Arguably, Weber’s view might hold true for bonds which serve as means of payment under a related contract and thus are subject to a special conflict of laws regime,36 but the exclusion would not apply to those bonds which are autonomous agreements for the borrowing of money, such as sovereign bonds. Further, the Giuliano–Lagarde Report clarified that ‘[n]either the contracts pursuant to which such [negotiable] instruments are issued nor contracts for the (p. 119) purchase and sale of such instruments are excluded’ from the coverage of the Rome I Regulation.37 Additionally, Plender and Wilderspin argue that ‘negotiable’ in the sense of the Rome I Regulation implies the ability to transfer a title to property, for instance, bills of lading in respect of goods shipped.38 Be that as it may, this book will not address this controversy since obligations specific to the negotiable character of negotiable instruments, for instance, the bearer’s right to bring an action against the drawer irrespective of contractual obligation,39 or the acquisition of title to a bond free from set-offs and defects in title,40 are outside the ambit of this book.

1.  The Law Applicable under Art 4(1)(b)

4.22  Having established the applicability of the Rome I Regulation, the general rule contained therein is that contracts shall be governed by the law of the country where the party required to effect the characteristic performance of the contract has his habitual residence at the time of the conclusion of the contract (Arts 4(2) and 19(3) of the Rome I Regulation).41 The Court of Justice in Kareda v Benkö established that a credit agreement is a type of service, albeit under the Brussels Regulation. Under Art 4(1)(b), in contracts for the provision of services, the law that applies is the law of the service provider. Similarly, the Court of Justice in Kareda v Benkö established that the credit institution provides the characteristic performance, since the borrower’s obligation to repay that sum is derivative from the performance of the service by the lender.42 Unless and until the lender releases the funds, the borrower would not be under an obligation to reimburse the debt.43

4.23  Even if the bonds are not services within the meaning of Art 4(1)(b), it seems that in any case the lender/bondholder provides the characteristic (p. 120) performance. The editors of Dicey, Morris & Collins, when examining the law applicable to a loan contract between a bank and a customer, suggest that the bank provides the characteristic performance and the law of the lender bank should thus govern the loan contract.44

4.24  In Molton Street Capital LLP v Shooters Hill Capital Partners LLP, a US broker dealer sold bonds issued in the US by a third party to a dealer located in the UK. It was conceded that in a contract for the sale of bonds, the seller of the bonds was the characteristic performer.45 Here, one needs to draw a distinction between two situations. One is where the debtor issues bonds to investors. In this case, the subscribing investors provide a ‘service’ by lending the funds. In the second scenario, which was the case in Molton, a trader sells the bonds to the wider public. The trader is not the person raising funds, but he is merely an intermediary selling a chose in action to the investors. In this case, the seller of the bonds is the person providing the characteristic performance under Art 4(1)(a). In light of this distinction, the concession made in Molton should only be taken to apply to the second scenario. This dichotomy allows us to reconcile Molton with the earlier authorities that the lender is the characteristic performer.

4.25  The resulting conclusion is that the law of the habitual residence of the (initial) bondholder (usually, the financial intermediary who purchases the bonds) governs the contractual aspects as between the sovereign and the bondholder.46 This challenges the conventional assumption that the law of the sovereign debtor applies to domestic bonds in all cases. By that, it might awaken the sleeping giant in the sense that sovereign bonds might be governed by a law other than that of the sovereign debtor notwithstanding the latter’s assumptions.

4.26  Although principle suggests that the law of the creditor governs the rights and obligations under sovereign bonds, in practical terms, there are a number of complications to consider. As mentioned above, it is commonplace for domestic bonds to be subscribed to by lenders resident in various jurisdictions. To apply the law of the habitual residence of the lender would result in the undesirable consequence of submitting the domestic debt to a mosaic of governing laws of each and every lender.47 This would (p. 121) give rise to a situation where, as a function of the habitual residence of the creditor, the sovereign debtor might be subject to different legal systems and consequently, varying levels of obligations.

4.27  This drawback becomes all the more acute in collective actions (class actions, representative actions), where a group of bondholders, conceivably from various jurisdictions, are grouped in one single action. In such a scenario, the court seised of the dispute would face a diversity of laws as a function of the variety of habitual residences of the creditors in a collective action. To apply the law of the defendant sovereign, as the law manifestly more closely connected to the dispute, is unlikely to occur in collective actions.48 Nonetheless, the multiplicity of laws is not necessarily an impediment to a collective action—one case management solution available to the court would be to form subgroups of claimants subject to a single law,49 or to locate common issues that although subject to different law may be treated functionally alike.50

4.28  To deal with this ominous legal conundrum on a more principled basis, commentators put forward alternative connecting factors. It has been contended that the most appropriate connecting factor for private loans is the law of the lead (manager) bank. If loans are advanced by more than one leading manager, then the law of one or more of the lead managers which happens to coincide with the law of the fiscal agent or trustee would apply.51 The inadequacy of this solution for domestic bonds lies in the fact that the agent of the sovereign debtor will often be somewhat linked to or indissociable from the sovereign debtor (such as the central bank of the debtor), rather than to the bondholders who provide the characteristic performance. Of course, the aforesaid suggestions are merely de lege ferenda solutions to a situation that is unsatisfactory from a practical point of view.

4.29  Nevertheless, there is no denying that subjecting the sovereign debt to one single law may be a legitimate pursuit since it increases predictability and precludes the subjection of the State to mutually conflicting obligations. The line of reasoning of some old judgments, although few and far (p. 122) between, favours the submission of domestic bonds to one single law.52 This can be achieved by applying the escape clause in Art 4(3).

2.  Escape Clause (Art 4(3))

4.30  Under Art 4(3), the governing law of the contract may be displaced, provided that some other law ‘is manifestly more closely connected’ to the bonds. The threshold is avowedly set very high. The court in Molton Street Capital LLP v Shooters Hill Capital Partners LLP stressed that the Rome I Regulation has raised the threshold under Art 4(3):

The text and architecture of Article 4 of the Rome I Regulation is very different from that of the Rome Convention. In particular, the test is no longer expressed as one of closest connection; the test is that contained in the rules set out in Articles 4.1 and 4.2, which are no longer expressed as presumptions or as being subject to the closest connection test; and the closest connection test has become an ‘escape clause’ to be applied only where it is clear that the connection is manifestly closer to a country other than that dictated by the tests in Articles 4.1 and 4.2 so that they are to be disregarded. The word ‘clear’ reflects what the ECJ had already said was the effect of Article 4.5 of the Rome Convention in the Interfrigo case, but the word ‘manifestly’ suggests a more stringent standard than before, as does the elevation of the criteria in Articles 4.1 and 4.2 to tests from mere presumptions of closest connection. The new language and structure suggests a higher threshold, which requires that the cumulative weight of the factors connecting the contract to another country must clearly and decisively outweigh the desideratum of certainty in applying the relevant test in Article 4.1 or 4.2.53

4.31  There are potentially three legal systems that may satisfy the requirement of manifestly more close connection: the law of the sovereign debtor, the law of the market of issuance, and the law of the majority of bondholders. These are addressed in turn.

Law of the Sovereign Debtor

4.32  Historically, the law of the sovereign debtor applied as a matter of a presumption. An excursus into the history of this presumption is necessary (p. 123) before deciding whether and when the law of the sovereign debtor may be manifestly more closely connected with the bonds.

Historical Presumption of Application of the Law of the Sovereign Debtor

4.33  It was often contended in the past that private contracts with the sovereign, both with the central government and local authorities,54 are subject to the rebuttable presumption that the law of the State governs the contract between a State and a private person.55 Under this paradigm, financial instruments issued by the State in its territory are commonly believed to be ‘creatures’ of the law of that State.56

4.34  The Permanent Court of International Justice (PCIJ) in Serbian and Brazilian Loans applied the presumption to the bonds issued by the Serbian and Brazilian governments and held that the debt contracted by Serbia and Brazil was subject to Serbian and Brazilian laws, respectively (although, French law applied to the determination of the currency of payment). The PCIJ reasoned that the sovereign debtor could not be presumed to subject its obligations to any law other than its own.57

4.35  The presumption, in principle rebuttable, was not rebutted on the facts of the case. The following array of factors strongly pointed towards the law of sovereign debtor:

  • •  the debt was issued under special laws that laid down conditions relating to the bonds;

  • •  these laws were cited in the bonds;

  • (p. 124) •  the bonds were bearer bonds which presupposed that the identity of their holders was irrelevant and the individuality of the issuer only was fixed and unchangeable;

  • •  all creditors were not located in one single State and the bonds were consequently not issued in one single State; and

  • •  even if the entire sovereign debt were to be entirely issued in a single State, that would not suffice to refute the conclusion that the law of the sovereign governed the debt given that the payments of principal and interest were made in diverse States.

4.36  Judge de Bustamante in his Dissenting Opinion explained the presumption of applying the law of the sovereign debtor on the ground that sovereign loans were in essence contracts of adhesion.58 This rationale is somewhat ill-considered. Arguably, the fact that sovereign bonds were contracts of adhesion should lead to the opposite conclusion.59 At the stage of negotiations and drafting, private subscribers of sovereign bonds may often be bereft of the opportunity to negotiate the terms of the contract. Due to the operation of the principle of contract interpretation contra proferentem, expressly referred to in the decision of the PCIJ in Brazilian Loans,60 the adhesive nature of sovereign bonds should have been construed as a factor militating against the law of the sovereign debtor.

4.37  The rationale of Serbian and Brazilian Loans was echoed in a 1950 judgment of the Privy Council in Bonython v Commonwealth Australia.61 First, the Board determined that if the loan agreement concluded with the sovereign debtor was denominated in the currency of the sovereign debtor, the implication would be—other factors being considered—that the proper law of the contract is the law of the sovereign debtor.

4.38  It is suggested here that the conclusion of the Privy Council is counter-intuitive to the economic underpinnings of the issuance of sovereign bonds in local currency. In the late 1990s, the economists Eichengreen, Hausmann, and Panizza advanced the concept of ‘original sin’—the general inability of most States to borrow abroad in their own currency irrespective of their (p. 125) ‘credit history’.62 The implication is that developing States are inclined to develop domestic (internal) markets for bonds denominated in local currency,63 to deal with insufficient demand for bonds denominated in their own currency and issued abroad and also to address currency mismatches arising out of borrowings in foreign currency.64 The issuance of bonds in local currency is thus merely a reflection of economic realities and practicality and it ought not to be counted as a major connecting factor.

4.39  Albeit in a commercial context not involving a sovereign, Lord Radcliffe correctly noted that there is not necessarily a correlation between the currency of the obligation and the proper law of the contract.65 No apparent reason exists to reach a different conclusion for the bonds of a sovereign debtor. The choice of currency of sovereign bonds may simply be tied to considerations of currency mismatches, balance of payments, desire to de-dollarise the economy, and more broadly, with other political and economic factors.66

4.40  The Privy Council in Bonython v Commonwealth of Australia then stated more broadly that if the sovereign debtor was a party to a contract that would be of great, if not decisive, weight in determining the proper law of the contract67—a classical expression of the presumption of the application (p. 126) of the law of the sovereign. Conversely, in the case Rex v International Trustee for the Protection of Bondholders A.G.,68 decided thirteen years earlier, the House of Lords underplayed the status of the sovereign in determining the governing law of sovereign bonds. The dispute arose in relation to the bonds issued by the UK government in New York. Following the issuance of the bonds, the US Congress passed a general piece of legislation repealing gold clauses contained in bonds. The crux of the case was thus the identification of the law governing the contracts: were New York law to be found to govern British bonds, it would substantially alleviate the financial obligations of the UK government through the abrogation of gold clauses. The King’s Bench and the Court of Appeal decided that because the UK government was a party to the contract, the bonds were governed by English law despite all the connecting factors to the US (issuance in the US territory, denomination of the bonds in US currency, the principal location of payments in New York based on a value estimated by reference to US coins, and bonds being secured by a pledge agreement with a US bank).

4.41  The House of Lords reversed the judgment of the Court of Appeal holding that New York law governed the bonds. What was crucial in the reasoning of the Law Lords was that common conflict of laws principles governing the determination of the law applicable to contracts were applicable with equal force to the contracts to which the sovereign was a party. The court was persuaded that the fact the sovereign was a party to the contract carried great weight, but it was not decisive and was only one factor in assessing the proper law of the contract. Upon the assessment of all the factors at hand, the House of Lords came to the conclusion that New York law governed the bonds issued by the UK government.

4.42  A certain logical discrepancy is discernible between the two cases. In Rex, the House of Lords stated that the fact the State is a party to the contract carries great weight, but is not conclusive. Instead, it is merely a factor in determining the proper law of bonds, whereas in Bonython, the Privy Council acknowledged that the very same fact had great weight, ‘if not decisive’.

4.43  It has been suggested that the difference between Rex and Bonython (and Serbian and Brazilian Loans) can best be explained by the fact that in the former case, sovereign bonds were issued in a single country which lent itself more easily to the proper law of the contract. Whereas in the latter cases, the bonds were issued in more than one country, giving credence to (p. 127) the submission of the bonds to the law of the debtor as a common denominator.69 Additionally, most connecting factors in Rex pointed towards the law of the US. By contrast, in the Serbian and Brazilian Loans cases, the presumption of the application of the law of the sovereign debtor was strongly supported by other important indices.70

4.44  Yet, the reasons for distinguishing between the two decisions are more profound than these factors. The House of Lords in its decision in Rex was clearly guided by the progressive argument that sovereign bonds should be governed by the common rules for determining the applicable law (although a cynic would claim that the House of Lords rendered a political decision in favour of the UK government and the Treasury). The ruling of the House of Lords in Rex is commercially sound since submitting sovereign bonds to the law of the sovereign debtor and leaving the bondholders to the mercy of the sovereign debtor might often be disquieting.71 Delaume, criticising the outdated decision of the PCIJ in Serbian and Brazilian Loans, argued that the sovereign character of the borrower does not warrant subjecting sovereign bonds to the law of the sovereign debtor.72

4.45  Subjecting sovereign debt contracts to the law of the sovereign debtor, as a matter of presumption albeit rebuttable, appears anachronistic. Properly understood, sovereign bonds are to be subjected to the same common conflict of laws rules as any other commercial contracts. Both English73 and foreign74 progressive case law follows this direction by applying the common choice of law rules to sovereign borrowing.

4.46  Formerly, the presumption in favour of the application of the law of the sovereign debtor was heavily buttressed by the fact that the debt of the (p. 128) sovereign debtor could only be adjudicated in the courts of the sovereign debtor due to the operation of the rules of State immunity. This point strongly implied the application of the law of the sovereign debtor to its debt.75 However, as the dilution of the absolute sovereign immunity has evidenced, the realities of the globalised world and the current climate of opinion are opposed to such a facile presumption of subjecting domestic debt to the volatility of the whims of the law of the sovereign debtor.

The Law of the Sovereign Debtor under Art 4(3)

4.47  Although the presumption of applying the law of the sovereign debtor is a relic of the past, it remains to be seen whether, on a thorough analysis of all relevant connecting factors, the old presumption may return through the back door of the ‘manifestly more closely’ connection.

4.48  The application of the law of the sovereign debtor has the undeniable advantage of assembling the entire sovereign debt under the umbrella of one single legal system.

4.49  In evaluating the law manifestly more closely connected to the domestic debt, the following factors, in addition to the ones referred to in the Serbian and Brazilian Loans cases above, might give cause to favour the law of the sovereign debtor:

  • •  the need to observe the sovereign’s regulatory requirements;

  • •  the habitual residence of beneficial bondholders;76

  • •  the language of the bonds which will almost invariably be the language of the sovereign debtor, though this latter does not carry much weight,77

  • •  the currency; and

  • •  the place of agents and other indicia depending on the circumstances of the particular case.78

4.50  On the other hand, in cases when a foreign party subscribes to the domestic debt of the sovereign debtor, several elements of the transaction tend to be foreign. If the residence of bondholders is outside the territory of the sovereign debtor, the funds are likely to flow from the local banks of bondholders. Furthermore, the fiscal or paying agent of the sovereign debtor (p. 129) may make the payments of interests over the bonds to the foreign bondholder’s foreign bank account, or else to a clearing system which might be either Clearstream Banking (Luxembourg) or Euroclear Bank (Belgium) even for domestic bonds.

4.51  An argument, which clearly resonated with the PCIJ judges in Serbian and Brazilian Loans, concerned the bearer character of the bonds issued by the sovereign. The PCIJ held that the identity of holders of bearer bonds was volatile in that the bonds were easily tradable in the secondary market of sovereign bonds. The only identity that was established and definite was that of the sovereign debtor. In the modern settings of private international law, the argument of the volatility of the identity of bondholders can be easily dispensed with. Under Art 19 of the Rome I Regulation, the relevant point in time for the determination of the governing law is that of the time of the conclusion of the contract (and, accordingly, the trading on the secondary market carries no import).

4.52  If one views all the connecting factors from a distance, commercial reason or commercial significance is the benchmark against which any connecting factor should be measured for the purposes of applying the escape clause under Art 4(3) of the Rome I Regulation.79 The benchmark, however, cannot easily be applied in the sovereign debt context: how can a State’s acts be measured as against their commercial significance if they are, to varying degrees, dictated by the political–economic considerations of a sovereign? Similarly, it might be too far-fetched to say that some connecting factor should be taken into account as commercially significant—for instance, the issuance of bonds under a statute, when it is mandated by the national rules of the sovereign debtor. Such factors denote considerations alien to the commercial nature of bonds and are largely incomparable to such commercially significant connections as the domicile of foreign creditors and the place of repayment of the debt.

4.53  This shows that objective connecting factors in a sovereign context, in their pure and unbent form, do not carry the same commercial significance. Given the specificity of the sector, fact- and context-specific interpretation of objective factors might be called for. It is true that certain objective connecting factors, which translate into reality the political–economic considerations of the sovereign debtor, may link the contract with the local law of the sovereign debtor.

4.54  For example, the issuance of bonds under a statute, and having express references to the statute in the bonds as was the case in the Serbian and (p. 130) Brazilian Loans cases, may seem a factor pointing in favour of the law of the sovereign debtor.80 Arguably, it could evidence an implied intention to submit the bonds to the law of the sovereign debtor (Art 3(1) of the Rome I Regulation). While not absolute,81 this connecting factor at hand can be subject to other, equally reasonable interpretations. First, the citation of a statute in the text of the bonds may merely act as incorporating terms and conditions into the contractual arrangement entered into between the government and private creditors. Thus, in Perry v United States, the bonds issued by the US referred to US circulars ‘for a statement of the further rights of the holders of bonds of said series’.82 Second, the reference to a statute in the text of bond documents often indicates the legitimacy and legality of the issuance of bonds under the constitutional regime of the sovereign,83 and does not necessarily imply that the parties implicitly have chosen the law of the sovereign debtor. Domestically, many governments have a limitation over the amount of the debt that they may borrow.84 This is all the more evident in the practice of those States that guarantee the debt of foreign States: by the operation of their constitutional provisions, States wishing to guarantee loans issued by a foreign State in the territory of the former are compelled to pass statutes authorising the assumption of liability over foreign States’ bonds.85

4.55  The argument sketched here does not deal with the scenario where the sovereign debtor, in accordance with its law, is compelled to contract under its own law. If a sovereign debtor’s local law precludes it from choosing a foreign law to govern its contracts, this should be a strong, though not decisive, factor for the submission of State contracts to its local law.86 In such circumstances, the presumption of application of the law of the sovereign would be difficult to rebut.87

(p. 131) Law of the Market

4.56  The intuitive inclination, as expressed in the mainstream approach, would be to subject domestic bonds to the place of their marketing. Delaume suggests that the law applicable to bonds ought to be the law of the market where the funds were raised and repaid and where the contract was signed.88 Likewise, Kronke argues for the submission of bonds to the law of the market of issuance as that would ensure uniform calculation and administration procedures, as well as equal treatment of private creditors.89

4.57  After all, it may be reasonable to expect a foreigner purchasing bonds targeted to local residents in the territory of a sovereign to submit to the law of the latter sovereign. Indeed, it might come as a surprise to a sovereign that its bonds issued in its own territory purchased by foreigners would for some reason be subjected to a law other than its own municipal law. But one might wonder whether the place of issuance of domestic bonds is as easily discernible nowadays as it was in the early or mid-twentieth century. As shown above, the place of marketing of bonds is no longer so clear-cut as it used to be.

4.58  On a more principled basis, as argued in Chapter 1, in commercial debt placements, the debtor is tied to bondholders through bilateral relations. With the bilateral nature of the sovereign bond issuance affirmed, the market of the issuance of bonds does not have the same unifying character and it could be discarded as a mere lex loci contractus, which is a discredited connecting factor and unlikely to sway the decision under Art 4(3).

4.59  Against this, Batiffol has suggested that the issuance of bonds is a transaction that has a single purpose of raising economic resources and it should be treated as a unitary complex.90 In his view, piecemeal solutions to the law governing various relations of the sovereign debtor with private creditors are incongruent with the anatomy of sovereign bonds. In spite of this, it remains true that the contract for purchase of bonds remains bilateral and autonomous for the purposes of ascertaining the applicable law. The unitary (or multilateral) character of sovereign bonds is only a shorthand for painting the process of the day-to-day raising of the funds by sovereigns.

(p. 132) 4.60  It is not a coincidence that the only instance where the Rome I Regulation submits financial instruments to the law of the market relates to the trading in multilateral systems (including exchanges). Under Art 4(1)(h) of the Rome I Regulation, contracts traded in multilateral systems which bring together or facilitate the bringing together of multiple third-party buying and selling interests in financial instruments will be subject to the law of that multilateral system. Multilateral systems are highly organised and centralised markets that hinge on the application of a single governing law.91 This explains the submission of such trading to the law of the place of exchange. Tellingly, the Rome I Regulation is silent in relation to other markets of financial instruments, including the markets for the placement of bonds.

Law of the Majority of Bondholders

4.61  Recitals 20 and 21 of the Rome I Regulation indicate that to locate the law manifestly more closely connected to the transaction, account should be taken, among other things, of other contracts that have a very close relationship with the contract at hand. The enquiry under Art 4(3) might thus consist in locating the law of the principal chunk of all bondholders (by the percentage of the bonds held), and then linking all other closely connected contracts to it.

4.62  Yet, the argument is misconceived. Sovereign bond contracts are not genuinely connected contracts since they can operate separately and individual bondholders can often bring claims separately from other bondholders, unless the contract provides otherwise.92 More principally, the general principles of conflict of laws suggest that each particular relation of a sovereign debtor with a creditor should be governed by a separate law. The Privy Council in Bonython v Commonwealth of Australia disfavoured the search for one common law to govern the entirety of contractual relations between the sovereign and various bondholders.93

4.63  Returning to the concept of domestic bonds, one might wonder whether the distinction between domestic and external bonds remains feasible in light of the foregoing analysis in this chapter.94 None of the criteria mentioned above which seemingly demarcate domestic bonds from external (p. 133) bonds appear to be stark enough to support the distinction. From a legal perspective, a more precise distinction lies simply between the bonds that contain a choice of law clause and those bonds that have omitted such a clause.

3.  Mandatory Provisions

4.64  It is common practice for parties to a sovereign bond to select a foreign legal system (mostly English or New York law) to regulate their rights and obligations arising under the bonds. Yet, a variety of laws of the sovereign debtor may purport to apply both to the issuance of sovereign bonds and to the process and the content of sovereign debt restructurings. Examples abound: contract formalities, taxation, exchange regulations, currency controls, laws on moratorium, etc.

4.65  This section will consider whether these laws could qualify as mandatory and trump the otherwise applicable governing law. It is clear that if the law, which applies to the contract is the law of the foreign sovereign, all the provisions mentioned will apply as part and parcel of the governing law. However, based on the evidence before it, the court may reach the conclusion that a law other than the law of the sovereign debtor applies to the bonds. In this case, certain provisions of the sovereign’s legal system may still purport to apply.

4.66  The preliminary issue to address is that bonds often provide that a declaration of moratorium constitutes an event of default. If the declaration of moratorium is found to be a mandatory provision, it is beyond doubt that the event of default clause may not in itself preclude the application of the declaration of moratorium, which qualifies as an overriding mandatory rule.95

4.67  There are several ways to give effect to what may be loosely defined as foreign mandatory provisions.

Article 3(3) of the Rome I Regulation

4.68  Article 3(3) of the Rome I Regulation provides for the application of non-derogable provisions:

Where all other elements relevant to the situation at the time of the choice are located in a country other than the country whose law has been chosen, the choice of the parties shall not prejudice the application of provisions of the law of that other country which cannot be derogated from by agreement.

(p. 134) 4.69  Article 3(3) allows the application of any provision, which cannot be derogated from by agreement and thus casts the net very widely. However, the non-derogable provisions apply only if all the elements are located in the jurisdiction whose non-derogable provision purports to apply.

4.70  The Court of Appeal in Santander examined the application of non-derogable provisions to interest rate swap agreements entered into between Portuguese public sector transport companies and a Portuguese bank, under ISDA Master Agreements subject to English law and jurisdiction. Portuguese transport companies sought to rely on various provisions of Portuguese law by claiming that they were non-derogable provisions within the meaning of Art 3(3). The court had to identify first whether all the elements relevant to the situation pointed towards Portugal. As a delineating exercise, Sir Terence Etherton MR drew a distinction between ‘elements relevant to the situation’ in Art 3(3) and ‘close connection’ in Art 4 of the Rome Convention and held that ‘elements relevant to the situation’ was a wider concept. It was not constrained exclusively to factors of a kind which connect the contract to a particular country for the purpose of identifying the proper law in the absence of an express choice.

4.71  The Master of the Rolls then approved the evaluation of the ‘elements relevant to the situation’, as conducted by the judge of first instance, Blair J. The judge referred to the following factors to find that Art 3(3) did not engage: the right to assign to a bank outside Portugal; the use of standard international documentation; the practical necessity for the relationship with a bank outside Portugal; the international nature of the swaps market in which the contracts were concluded; and the fact that back-to-back contracts were concluded with a bank outside Portugal in circumstances in which such hedging arrangements are routine.96

4.72  Dexia involved somewhat comparable swap agreements, albeit with a stronger link to Italy. There, only two of the elements from Santander were present, namely the use of standard international documentation, in the form of the ISDA Master Agreement, and the routine back-to-back contracts concluded with banks outside Italy. The Court of Appeal considered that notwithstanding this difference, each of these two factors was enough on its own to demonstrate an international and relevant element in the situation such that it was impossible to say that ‘all elements (other than the choice of law) relevant to the situation’ were located in Italy.97

(p. 135) 4.73  In the context of sovereign bonds, one could imagine situations where the State, having agreed to the choice of English law, would seek to engage its domestic non-derogable provisions through the backdoor of Art 3(3). If the creditor is local to the sovereign debtor, the latter may argue that all the elements of the transaction are within its jurisdiction and therefore non-derogable provisions of its law apply notwithstanding a choice of English or New York law. However, it is likely that plentiful extraneous elements will exist to evade the application of Art 3(3). First, the claimant bondholder has subscribed to the bonds together with other bondholders, some of whom might be foreign. Second, the element of back-to-back contracts that was cited with approval in Santander and Dexia has a corollary in sovereign debt issuance, where the financial intermediary would often purchase the bonds only to resell them to beneficial investors in other jurisdictions. Third, while the Eurobonds documentation is not as uniform as the Master ISDA Agreement, it is at least arguable that by subscribing to Eurobonds, parties intended to insert an element of international finance into their transaction.

Overriding Mandatory Provisions

4.74  Overriding mandatory provisions under Art 9 of the Rome I Regulation is a narrower concept than non-derogable provisions and only comprises those provisions ‘the respect for which is regarded as crucial by a country for safeguarding its public interests, such as its political, social or economic organisation, to such an extent that they are applicable to any situation falling within their scope, irrespective of the law otherwise applicable to the contract under this Regulation’ (Art 9(1)).

4.75  Art 9(2) provides for the application of the overriding mandatory provisions of the forum. The application of those provisions is unlikely to be a common occurrence in sovereign debt litigation.

4.76  Sovereign debtors could avail themselves of Art 9(3) which allows for the application of the overriding mandatory provisions of a foreign State:

Effect may be given to the overriding mandatory provisions of the law of the country where the obligations arising out of the contract have to be or have been performed, in so far as those overriding mandatory provisions render the performance of the contract unlawful. In considering whether to give effect to those provisions, regard shall be had to their nature and purpose and to the consequences of their application or non-application.

4.77  Assuming that the sovereign debtor’s legislation qualifies as an overriding mandatory provision within the sense of Art 9(1), several conditions need to be satisfied to apply the overriding mandatory provisions of the (p. 136) sovereign debtor. First, a condition sine qua non for the application of overriding mandatory provisions is that the place of performance of obligations arising out of the contract has to be in the country whose overriding provisions purport to apply.

4.78  How is the place of performance determined? Harris reasons that the place of performance should be determined in two steps: (1) the place of performance provided in the contract and failing that; (2) it will be a matter for the autonomous meaning of the Regulation.98 The editors of Dicey, Morris & Collins contend that the place of performance is to be determined by the governing law of the contract.99 Bonomi considers that the place of performance is to be determined by the contractual agreement of parties, failing which the place of performance is to be determined by the governing law of the contract.100 Another alternative is to apply the law of the forum to determine the place of performance of the obligation.

4.79  The agreement of parties to a contractually stipulated place of payment will be decisive. However, the ambiguity is whether it denotes the place of payment of funds to the State or the place of repayment by the State. Chapter 3 established that for the jurisdictional purposes, the place of provision of services is the place of payment only. However, for the purposes of applying overriding mandatory provisions, the court will need to identify the exact obligation affected by the mandatory provision and locate its place of performance. If the overriding mandatory provision seeks to affect the obligation to repay, then the court will need to identify the contractually stipulated place of repayment. That place in the sovereign bond documentation is often in the jurisdiction of creditors and therefore outside the territorial scope of the overriding mandatory provisions of the sovereign debtor.

4.80  If, however, the sovereign bond documentation is silent as to the place of payment/repayment, the place of performance can be determined in two ways: applying the autonomous meaning of the place of performance or leaving the matter to the governing law of the contract. In the first scenario, while there is no autonomous meaning under the Rome I Regulation, under Art 7(1) of the Brussels Regulation the place of provision of a credit agreement is the place where the lending institution has its registered office.101 In light of the synergy between the two Regulations (Recital 7), (p. 137) the position under the Brussels Regulation may be extrapolated to apply under the Rome I Regulation.

4.81  Yet, it is at least arguable that the autonomous definition in Art 7(1) developed for jurisdictional purposes may fail to reflect the nuances and purposes of Art 9(3) of the Rome I Regulation.102 Unlike Art 7(1) of the Brussels Regulation, Art 9(3) does not refer to the place of performance of a single obligation (place of provision of services) but to the place of performance of ‘the obligations arising out of the contract’.

4.82  If, however, to determine the place of performance, the English court needs to apply the law governing the contract and that law happens to be English law, then the common law position is that the place of the obligation to pay is the place where the money is to be received.103 In practice, the place where the money will be received will be outside the territory of the sovereign debtor and thus outside the reach of its overriding mandatory provisions. If, by contrast, the place of (re)payment is within the territory of the State, its mandatory provisions may apply.104 This is consistent with the practice of several foreign jurisdictions, which in deciding on the enforcement of foreign exchange regulations, refused to apply any foreign exchange regulations when the payments were to be made outside the territory of the sovereign who adopted such rules.105

4.83  The solution, both under the Brussels Regulation and English common law, has the advantage of predictability for credit institutions who may legitimately predict what overriding mandatory rules will apply over the course of performance of the contract. An additional benefit is that the State will not be tempted to frustrate or modify the contract by adopting mandatory rules in the vicinity of default.

4.84  In a case decided under English common law, Ispahani v Bank Melli Iran,106 the funds were withdrawn from Bangladesh in breach of Bangladeshi exchange controls and paid into a London bank account. At the interlocutory stage of proceedings, the Court of Appeal found that it was reasonably arguable that the acts were conducted in Bangladesh whose exchange regulations were breached. The case is unusual in that it may give effect to the mandatory provisions of the country of origin of payments and not (p. 138) the place of final destination of payments. However, the case can be supported since the parties intended to commit or procure the commission of illegal acts in the territory of Bangladesh107 within the meaning of Foster v Driscoll.108

4.85  Second, a foreign mandatory provision will only be given effect if it renders the performance of the contract unlawful. ‘Unlawful’ is wide enough to capture most overriding mandatory provisions in sovereign debt disputes, however the mere unenforceability of the contract may not qualify as ‘unlawful’.109

4.86  Third, the exercise under Art 9(3) is discretionary. The court has to consider the nature and purpose of the foreign mandatory provisions, and the consequences of their application or non-application. The discretion of the court is multi-factored, but one thing that will not pass unnoticed is the attempt to avoid the obligations by relying on (retrospective) mandatory provisions. In Eurobank Ergasias SA v Kalliroi Navigation Company Ltd, the judge was particularly taken by the fact that the party relying on illegality ‘was very keen to have the finance in place’ arguably in breach of mandatory provisions.110

4.87  In parallel to Art 9(3) of the Rome I Regulation, issues of foreign illegality are also governed by English common law. The Court of Justice has confirmed that Art 9(3) does not preclude overriding mandatory provisions ‘from being taken into account as a matter of fact, in so far as this is provided for by a substantive rule of the law that is applicable to the contract pursuant to the regulation’. Article 9(2) and (3) thus does not seek to supersede rules of substantive law which take into account foreign overriding mandatory provisions.111 An example of such a national substantive law in English law is the line of cases following Foster v Driscoll112 and Regazzoni v KC Sethia.113

4.88  Article 9(3) and English common law share certain traits. Principally, to trigger both regimes, the place of performance of obligations has to be in the country whose mandatory provisions purport to apply. Mandatory provisions of any other country cannot be applied or taken into account. Of course, the English courts are not bound to apply the autonomous (p. 139) meaning of the place of performance when applying English substantive law and the place of performance will be determined by English law.

4.89  Unlike the discretionary test in Art 9(3), the common law provides for an almost automatic application of the foreign illegality laws. The English court need not engage in the analysis of the nature, purpose, and consequences of the application of foreign laws. This ostensibly mechanical application of foreign illegality laws explains why defendants would often resort to the common law provisions: the claimant may not defeat the operation of foreign illegality laws by seeking to affect the discretionary decision-making of the judge.

Article VIII(2)(b) of the IMF Agreement

4.90  Article VIII(2)(b) of the Bretton Woods Agreement in effect gives effect to foreign exchange regulations of IMF Member States. It reads:

Exchange contracts which involve the currency of any member [of the International Monetary Fund] and which are contrary to the exchange control regulations of any member maintained or imposed consistently with this agreement shall be unenforceable in the territories of any member.

4.91  The first thing to note is that Art VIII(2)(b) does not supersede the English common law on illegality but merely supplements it.114 However, it is open to debate whether Art VIII(2)(b) is a conflict of laws rule or a procedural mechanism which applies irrespective of the governing law of the contract or alternatively, if it is a rule of the governing law of the contract. It would be hard to imagine that Art VIII(2)(b) would be treated as part of the substantive provisions of a legal system.115 Instead, each court is likely to apply it as a matter of its lex fori, irrespective of the governing law of the contract. In any case, Art 25 of the Rome I Regulation preserves the application of international conventions which lay down conflict of laws rules relating to contractual obligations.

4.92  English courts have given a narrow construction to Art VIII of the IMF: ‘exchange contracts’ are nothing but contracts for exchange of currency.116 US courts have also adopted a narrow interpretation of ‘exchange (p. 140) contracts’.117 Thus, sovereign bonds, loans, deposits, and letters of creditors will not qualify as ‘exchange contracts’ within the meaning of Art VIII(2)(b). Further, Art VIII(2)(b) may not apply when the foreign exchange regulation regulates the currency of repayment in case of a restructuring and the creditor has the discretion to resort to a restructuring.118

4.93  By contrast, European jurisdictions have taken a more expansive approach to ‘exchange contracts’ to include all contracts that ‘in any way affect a country’s exchange resources’, thereby potentially including international loan agreements in the definition of exchange contracts.119


1  Mauro Megliani, Sovereign Debt: Genesis, Restructuring, Litigation (2015) 188.

2  Wolfgang Friedmann, Law in a Changing Society (1972) 472–73.

3  Lee Buchheit, Mitu Gulati, and Ignacio Tirado, ‘The Problem of Holdout Creditors in Eurozone Sovereign Debt Restructurings’ (2013) 4 JIBFL 191, 193.

4  Joseph Cotterill, ‘Ukraine’s Bonds: A Little Local Leverage’ (26 March 2015) FT Alphaville.

5  Restatement (Second) on Conflict of Laws (1969) §§187–88.

6  Federico Sturzenegger and Jeromin Zettelmeyer, Debt Defaults and Lessons from a Decade of Crises (2006) 58.

8  From forty-two questionnaires sent, eighteen responses were received: Albania, Argentina, Australia, Belgium, Brazil, Bulgaria, Costa Rica, Cyprus, Czech Republic, Denmark, Estonia, Finland, Hungary, Latvia, Lithuania, Switzerland, the United Kingdom, and Uruguay.

10  The Information Memorandum relating to the Issue, Stripping and Reconstitution of British Government Stock (published on 20 August 2007 and subsequently updated on 15 August 2011) s 103.

11  See, eg, 2% Index-Linked Treasury Stock 2035, Issue of £675,000,000, United Kingdom, 2 December 2003 (no choice of law clause).

12  See, eg, Kahler v Midland Bank [1950] AC 24, 28; Zivnostenska Banka v Frankman [1950] AC 57, 69–70. See also Re United Railways of Havana and Regla Warehouses Ltd [1961] AC 1007, 1067–68 (per Lord Denning); Clive Schmitthoff, ‘The International Government Loan’ (1937) 19(4) J Comp L & Intl L 179, 193.

13  Art 6, Law of the Russian Federation on the Particularities of the Issuance and Trading of State and Municipal Securities (1998).

14  IMF, ‘Sovereign Debt Restructuring Mechanism—Further Considerations’ (2002) 12. Similarly, no single criterion is used in defining ‘External Indebtedness’, a definition used by many sovereign bonds. ‘External Indebtedness’ is either defined through the place of issuance (see, eg, Commonwealth of Bahamas, US$100,000,000, 7.125% Notes due 2038, p 32), currency (see, eg, Arab Republic of Egypt, US$1,000,000,000, 5.75% Notes due 2020, p 82), or governing law (see, eg, Republic of Slovenia, €1,500,000,000, 4.700%, Notes due 2016, p 18).

15  Henrik Enderlein, Christoph Trebesch, and Laura von Daniels, ‘Sovereign Debt Disputes: A Database on Government Coerciveness during Debt Crises’ (2011) 30 J Intl Mon & Fin 1, 12 (in six domestic debt restructurings large shares of domestic debt were held by foreign residents); Carmen Reinhart and Kenneth Rogoff, ‘The Forgotten History of Domestic Debt’ (2008) NBER Working Paper 13946, 45–48; Kevin Cowan, Eduardo Levy-Yeyati, Ugo Panizza, and Federico Sturzenegger, ‘Sovereign Debt in the Americas: New Data and Stylized Facts’ (2006) Working Paper of Inter-American Development Bank No 577, 21.

16  Udaibir Das, Michael Papaioannou, and Christoph Trebesch, ‘Sovereign Debt Restructurings 1950–2010: Literature Survey, Data, and Stylized Facts’ (2012) IMF Working Paper WP/12/203, 8, 24–25; Anna Gelpern and Brad Setser, ‘Domestic and External Debt: The Doomed Quest for Equal Treatment’ (2004) 35 Geo J Intl L 795, 800.

17  League of Nations Study Committee Report, II Econ and Finan (1939b) II.A 10, 6–7.

18  Edwin Borchard, State Insolvency and Foreign Bondholders, vol 1 (1951) 75.

19  Canevaro Brothers (Italy v Peru), 3 May 1912 (1912) 6 AJIL 746, 750 (but refraining from inquiring what the decision would have been had the debt been owed to foreign residents at the time of the restructuring of bonds).

20  Herbert Kronke, ‘Capital Markets and Conflict of Laws’ (2000) 286 Recueil des Cours 245, 300–01; Francisco Garcimartín Alférez, ‘Cross-Border Listed Companies’ (2007) 328 Recueil des Cours 9, 78–79.

21  Kronke (n 20) 368; Robert Ahdieh, ‘Making Markets: Network Effects and the Role of Law in the Creation of Strong Securities Markets’ (2003) S Cal L Rev 277, 284, fn 20. See also the finding of the District Court that the admittedly domestic bonds of Argentina, governed by its domestic law, were offered outside Argentina (NML Capital Ltd v The Republic of Argentina, 2015 US Dist LEXIS 30625, 40–41 (SDNY 2015)).

22  Order no 60 of the Ministry of Finance of the Russian Federation on the Emission of Internal Bonds of the Russian Federation on International Capital Markets of Bonds, 25 February 2011.

23  The purchase of retail bonds, although limited in practice, can nevertheless be an attractive option for foreign creditors, for two reasons: (1) retail bonds, generally purchased by domestic retail bondholders, are unlikely to be restructured for the reasons of domestic politics; (2) foreign creditors can easily construct a blocking position in the relatively small-scale issue of retail bonds.

24  Hans Blommestein, ‘Responding to the Crisis: Changes in OECD Primary Market Procedures and Portfolio Risk Management’ (2009) 2 OECD Journal: Financial Market Trends, 12. In the EU, only thirteen Member States use the system of primary dealer (Economic and Financial Committee, ‘Progress Report on Primary Dealership in EU Public Debt Management’ (2000), EFC/ECFIN/665/00-En-fin, 1).

25  Jakob de Haan, Sander Oosterloo, and Dirk Schoenmaker, Financial Markets and Institutions: A European Perspective (2012) 143.

26  Hans Blommestein, ‘Responding to the Crisis: Changes in OECD Primary Market Procedures and Portfolio Risk Management’ (2009) 2 OECD Journal: Financial Market Trends, 3–8; Mark De Broeck and Anastasia Guscina, ‘Government Debt Issuance in the Euro Area: The Impact of the Financial Crisis’ (2010) IMF Working Paper WP/11/21, 9–10.

28  Norwegian Loans Case (France v Norway), 1957 ICJ Rep 9.

29  Mémoire Soumis par le Gouvernement de la République Française, 20 décembre 1955, 29–30; Réplique du Gouvernement Français, 20 février 1957, 386–89 (available at www.icj-cij.org).

30  Contre-Mémoire du Gouvernement du Royaume de Norvège, 20 décembre 1956, 263 (available at www.icj-cij.org).

31  Norwegian Loans Case (n 28) 23.

32  UNCTAD, ‘Domestic and External Public Debt in Developing Countries’ (2008) UNCTAD Discussion Papers, 4–5; Cowan et al (n 15) 8; Carmen Reinhart and Kenneth Rogoff, This Time Is Different: Eight Centuries of Financial Folly (2011) 8 (subjection to national jurisdiction).

33  Abaclat and others v The Argentine Republic, Decision on Jurisdiction and Admissibility of 4 August 2011, ICSID Case No ARB/07/5, para 379 (affirming the procedural character of choice of forum agreements).

34  Georges Delaume, Legal Aspects of International Lending and Economic Development Agreements (1967) 92; ILA, ‘The Hague Conference Report: Sovereign Insolvency Study Group’ (2010b) 9.

35  Stefan Weber, ‘The Law Applicable to Bonds’ in Hans van Houtte (ed), The Law of Cross-Border Securities Transactions (1999) 31.

36  On which see Peter Cresswell, William Blair, and Philip Wood (eds), Butterworths Encyclopaedia of Banking Law (2013) F-228, 251.

37  Giuliano–Lagarde Explanatory Report to the Rome Convention on the Law Applicable on Contractual Obligations [1980] OJ C 282/1, 11.

38  Richard Plender and Michael Wilderspin, The European Private International Law of Obligations (2009) 110. For a contrary view arguing that bonds are indeed negotiable instruments, see Ravi Tennekoon, The Law and Regulation of International Finance (1991) 19, 164–65.

39  Plender and Wilderspin (n 38) 110.

41  Compare the more flexible conflict of laws methodology employed by the US courts in the context of sovereign bonds (Salah Turkmani v The Republic of Bolivia, 193 F Supp 2d 165, 177 (DDC 2002)).

42  Kareda v Benkö, Case C-249/16, ECLI:EU:C:2017:472. See also Kronke (n 20) 349. For a contrary view, see Pietro Franzina, ‘Sovereign Bonds and the Conflict of Laws: A European Perspective’ in Pasquale Nappi et al (eds), Studi in onore di Luigi Costato (2013) 11.

43  For a similar reasoning in the context of supply and promotion of products, see Print Concept v GEW [2002] CLC 352.

44  Lord Collins of Mapesbury et al (eds), Dicey, Morris & Collins on the Conflict of Laws, vol 2 (2012) 2057; Peter Kaye, The New Private International Law of the European Community (1993) 182.

45  [2015] EWHC 3419 (Comm), para 92.

46  See, eg, Haeger & Schmidt GmbH v Mutuelles du Mans Assurances IARD (MMA IARD), Case C-305/13, ECLI:EU:C:2014:2320, paras 43–51 (finding that when the first forwarding agent has been replaced by the second forwarding agent, the court has to apply the law of habitual residence of the first forwarding agent).

47  Megliani (n 1) 198–99.

48  Eva Lein, ‘Jurisdiction and Applicable Law in Cross-Border Mass Litigation’ in Fausto Pocar, Ilaria Viarengo, and Francesca Clara Villata (eds), Recasting Brussels I (2012) 169–70; Linda Silberman, ‘The Role of Choice of Law in National Class Actions’ (2007–08) 156 U Pa L Rev 2001, 2022–24, 2031; Astrid Stadler, ‘Conflicts of Laws in Multinational Collective Actions—A Judicial Nightmare?’ in Duncan Fairgrieve and Eva Lein (eds), Extraterritoriality and Collective Redress (2012) 207–08.

49  Stadler (n 48) 208, 212–13; Lein (n 48) 171.

50  Stadler (n 48) 208, 211.

51  André Jacquemont, L’Emission des emprunts euro-obligataires (1976) 176; Megliani (n 1) 196–97.

52  Smith v Weguelin (1869) LR 8 Eq 198, 213 (holding that ‘the same law should regulate the whole’ of loan obligations contracted by the sovereign); Goodwin v Robarts (1876) 1 App Cas 476, 495; PCIJ, Serbian Loans Case, Ser A, No 21 (1929), 42 (attempting to determine a single law to govern the network of relations between various bondholders and the sovereign debtor, implicitly rejecting the possibility of subjecting the bonds to diverse laws). See also the decision of the Swedish Supreme Court in Skandia Insurance Company Ltd v Swedish National Debt Office (1937) 18 BYBIL 215, 216.

53  [2015] EWHC 3419 (Comm), para 94.

54  Mount Albert Borough Council v Australasian Temperance and General Mutual Life Assurance Society [1938] AC 224, 239.

55  Borchard (n 18) 67; FA Mann, ‘The Proper Law of Contracts Concluded by International Persons’ in FA Mann, Studies in International Law (1973), 220; Jean-Flavien Lalive, ‘Recent Version: Abrogation or Alteration of an Economic Development Agreement between a State and a Private Foreign Party’ (1961–62) 17 Bus Law 434, 436, fn 7; Smith v Weguelin (1869) LR 8 Eq 198, 213; Skandia Insurance Company Ltd (n 52) 215, 216; Conseil d’Etat 28 novembre 1958, ‘Dame Langlois’ (1960) Clunet 444; Gouvernement Ottoman c. Comptoir d’escompte et consorts, Trib civ de la Seine, 3 mars 1875 (1877) Sirey 25.

56  US–Mexican General Claims Commission, George Cook v United Mexican States (1927) RSA IV 213, 215.

57  Ser A, No 21 (1929), 42, No 22 (1929), 121. See also Separate Opinion of Judge Sir Hersch Lauterpacht in the Case of Certain Norwegians Loans (France v Norway) (1957) ICJ Rep 37; Separate Opinion of Abdel Badawi in the Case of Certain Norwegians Loans (France v Norway) (1957) ICJ Rep 30; Saudi Arabia v Aramco (1963) 27 ILR 117, 167.

58  Dissenting Opinion in the Brazilian Loans Case, Ser A, No 21 (1929) 130.

59  Henry Batiffol, Les Conflits de lois en matière de contrats (1938) 200–02 (albeit for private contracts).

60  Ser A, No 21 (1929), 114. Provided, of course, that the doctrine has any role to play in debt contracts (William Bratton, ‘The Interpretation of Contracts Governing Corporate Debt Relationships’ (1983–84) 5 Cardozo L Rev 371, 380–81). Sovereign bonds are not strictly speaking contracts of adhesion, given that the underwriters would generally negotiate the terms and conditions of sovereign bonds.

61  [1951] AC 201.

63  Barry Eichengreen, Ricardo Hausmann, and Ugo Panizza, ‘Original Sin: The Pain, the Mystery, and the Road to Redemption’ (2002) Paper prepared for the conference Currency and Maturity Matchmaking: Redeeming Debt from Original Sin 15; BIS ‘The Development of Bond Markets in Emerging Economies’ (2001) BIS Paper No 11; IMF, ‘Local Securities and Derivatives Markets in Emerging Markets: Selected Policy Issues’ (2003) IMF Financial Markets Quarterly.

65  Re United Railways of Havana and Regla Warehouses Ltd [1961] AC 1007, 1060. For a contrary view, see Julius Weigert, The Abrogation of Gold-Clauses in International Loans and the Conflict of Laws (1940) 5–7.

66  On diverse factors affecting the composition of government debt, see Allan Drazen, ‘Towards a Political–Economic Theory of Domestic Debt’ (1997) NBER Working Paper 5890; Peter Diamond, ‘National Debt in a Neoclassical Growth Model’ (1965) 55(5) Am Econ Rev 1126; Eduardo Levy Yeyati, ‘Dollars, Debt, and International Financial Institutions: Dedollarizing Multilateral Lending’ (2007) 21(1) World Bank Economic Review 21.

67  For dicta that might be read in the same way, see Smith v Weguelin (1869) LR 8 Eq 198, 213; Goodwin v Robarts (1876) 1 AC 476, 494. Rex v International Trustee for the Protection of Bondholders [1937] AC 500, 566 rejected such an extensive reading of these cases.

68  [1937] AC 500.

69  Note, ‘Decisions of National Tribunals involving Points of Public or Private International Law’ (1937) 18 BYBIL 210, 218; Wilfred Jenks, ‘The Authority in English Courts of Decisions of the Permanent Court of International Justice’ (1939) 20 BYBIL 1, 6.

70  Jenks (n 69) 6.

71  Note, ‘Conflict of Laws and the Gold Clause in Foreign Government Bonds’ (1937) 46(5) Yale LJ 891, 895; Bruno Zanghirati, ‘Sovereign Indebtedness: The Complex Relations between Banks and States’ (1896–97) 7 IYBIL 133, 145.

72  Delaume (n 34) 101–02. See also Jean-Flavien Lalive, ‘Unilateral Alteration or Abrogation by either Party to a Contract between a State and a Foreign National’ in Symposium on the Rights and Duties of Foreigners in the Conduct of Industrial and Commercial Operations Abroad, Rights and Duties of Private Investors Abroad (1965) 271; Cresswell et al (n 36) F-305. Contra see Georges Sausser-Hall, ‘La clause-or dans les contrats publics et privés’ (1937) 60 Recueil des Cours 651, 754.

73  The Metamorphosis [1953] 1 WLR 543, 547.

74  George van Hecke, Problèmes juridiques des emprunts internationaux (1964) 75–76 (citing Swedish, Austrian, and Norwegian cases).

75  Schmitthoff (n 12) 192–93.

76  Haeger & Schmidt GmbH v Mutuelles du Mans Assurances IARD (MMA IARD), Case C-305/13, ECLI:EU:C:2014:2320, paras 43–51 (finding that the habitual residence of the second forwarding agent should be taken into account in locating the manifestly more closely connected law).

77  Delaume (n 34) 91. But the language of the contract might not carry great weight.

78  Davidson Sommers, Aaron Broches, and Georges Delaume, ‘Conflict Avoidance in International Loans and Monetary Agreements’ (1956) 21 L & Contemp Prob 463, 472.

79  Richard Fentiman, International Commercial Litigation (2015) 210–14.

80  Mount Albert Borough Council v Australasian Temperance and General Mutual Life Assurance Society [1938] AC 224, 239 (in respect of local authorities),

81  Apple Corps v Apple Computer [2004] EWHC 768, para 63.

82  Perry v United States, 294 US 330, 348 (1935).

83  See, eg, Art 115(1) of the Basic Law of the Federal Republic of Germany on which see in detail, Paul Kirchhof, ‘The Public Debt, Democratic Principles and the Rules of Law’ in Detlev Dicke (ed), Foreign Debts in the Present and a New International Economic Order (1986) 339.

84  See, eg, Treaty on Stability, Coordination and Governance in the Economic and Monetary Union, 2 March 2012.

85  Borchard (n 18) 108.

86  Cresswell et al (n 36) F-306.

87  Derek Bowett, ‘State Contracts with Aliens: Contemporary Developments on Compensation for Termination or Breach’ (1988) 59 BYBIL 49, 53.

88  Delaume (n 34) 86–88, 91, 100.

89  Kronke (n 20) 349–50.

90  Henry Batiffol, Les Conflits de lois en matière de contrats (1938) 106–07 (arguing that the unity of the State’s debt calls for the application of its own law).

91  Michael McParland, The Rome I Regulation on the Law Applicable to Contractual Obligations (2015) 415.

92  See Chapter 1.

93  [1951] AC 201, 222.

94  Writers have since long noted the artificiality of the distinction (Anna Gelpern, ‘Domestic Bonds, Credit Derivatives, and the Next Transformation of Sovereign Debt’ (2008) 83 Chi-Kent L Rev 101, 133; UNCTAD (n 32) 5).

95  See Chapter 6, section II.

96  Banco Santander Totta Sa v Companhia Carris De Ferro De Lisboa Sa [2016] EWCA Civ 1267, paras 65–67.

97  Dexia Crediop SPA v Comune di Prato [2017] EWCA Civ 428, paras 130–34.

98  Jonathan Harris, ‘Mandatory Rules and Public Policy under the Rome I Regulation’ in Franco Ferrari and ‎Stefan Leible (eds), The Rome I Regulation (2009) 315.

99  Lord Collins of Mapesbury et al (n 44), vol 2, para 32-096.

100  Andrea Bonomi, ‘Article 9’ in Ulrich Magnus and Peter Mankowski, Rome I Regulation: European Commentaries on Private International Law (2017) 647.

101  Kareda v Benkö, Case C-249/16, ECLI:EU:C:2017:472.

102  Bonomi, ‘Article 9’ in Magnus and Mankowski (n 100) 645–46.

103  Definitely Maybe Ltd v Lieberberg GmbH [2001] 1 WLR 1745, para 4.

104  De Beéche v South American Stores (Gath & Chaves) Ltd [1935] AC 148 (bills of exchange handed over in Chile where the controls were adopted).

105  Central Hanover Bank and Trust Company v Siemens& Halske A.G., 15 F Supp 927 (SDNY 1937); Martin Domke, International Loans and the Conflict of Laws: A Comparative Survey of Recent Cases (1937) 11–20.

106  [1997] All ER (D) 124.

108  [1929] 1 KB 470.

109  Trevor Hartley, International Commercial Litigation: Text, Cases and Materials on Private International Law (2015) 672.

110  [2015] EWHC 2377 (Comm), paras 38–40.

111  Republik Griechenland v Grigorios Nikiforidis, Case C-135/15 [2016] ECR EU:C:2016:774.

112  [1929] 1 KB 470.

113  [1958] AC 301.

114  Ispahani v Bank Melli Iran [1997] All ER (D) 124. See also Republic Griechenland v Grigorios Nikiforidis, Case C‑135/15, ECLI:EU:C:2016:774.

115  Weber, ‘The Law Applicable to Bonds’ in van Houtte (n 35) 39.

116  Ispahani v Bank Melli Iran [1997] All ER (D) 124; Wilson, Smithett & Cope Ltd v Terruzzi [1976] QB 683. In support of this position, see, eg, Lester Nurick, ‘The IMF Articles of Agreement’ in David Sassoon and Daniel Bradlow (eds), Judicial Enforcement of International Debt Obligations (1987) 107.

117  Libra Bank Ltd v Banco Nacional de Costa Rica, 570 F Supp 870, 897 (SDNY 1983); J. Zeevi & Sons, Ltd v Grindlays Bank (Uganda) Ltd, 37 NY 2d 220 (NY 1975).

118  Weston Banking v Turkiye Garanti Bankasi, 57 NY 2d 315, 326 (1982).

119  George Weisz, Nancy Schwarzkopf, and Mimi Panitch, ‘Selected Issues in Sovereign Debt Litigation’ 12(1) U Pa J Intl Bus L 1 (1991), 1–2. In support of this position, see Dominique Carreau, ‘Rapport du directeur d’études de la section française du centre’ in Dominique Carreau and Malcolm Shaw (eds), The External Debt (1995) 22. But see Armin von Bogdandy and Mathias Goldmann, ‘Sovereign Debt Restructurings as Exercises of International Public Authority: Towards a Decentralized Sovereign Insolvency Law’ in Espósito Carlos, Li Yuefen, and Bohoslavsky Juan Pablo (eds), Sovereign Financing and International Law: The UNCTAD Principles on Responsible Sovereign Lending and Borrowing (2013) 63 (German courts do not accept the expansive interpretation of Art VIII).