I. Introduction
Sovereign defaults are a recurring feature of the global economy. Nearly every state has defaulted on its debt at least once; several have done so multiple times.Footnote 1 According to one measure, there have been over three hundred sovereign defaults since 1800.Footnote 2 According to another, over six hundred sovereign debt instruments have been subject to default or restructuring since 1950.Footnote 3 Sovereign defaults tend to happen in waves, with multiple countries entering into debt crises at the same time.Footnote 4
A new wave of sovereign defaults appears to be approaching.Footnote 5 In the wake of COVID-19, public spending needs have surged and economic activity has collapsed.Footnote 6 Over one hundred countries have petitioned the International Monetary Fund (IMF) for emergency assistance.Footnote 7 Several countries that were previously teetering on the edge—including Argentina, Ecuador, and Lebanon—have defaulted or begun the process of restructuring their debts.Footnote 8 As emergency measures enacted at the start of the pandemic “are fast becoming insufficient,”Footnote 9 several other countries are expected to follow close behind.Footnote 10
Like the hundreds of defaults that have preceded them, the next sovereign debt crises will occur in the absence of a comprehensive sovereign bankruptcy mechanism. They will instead be subject to a number of interdependent and uncoordinated institutions that have governed previous debt crises, including national contract law, restructuring agreements, holdout litigation, and lenders of last resort. Due to a series of recent jurisdictional decisions by investor-state arbitral tribunals, the next wave of debt crises is also likely to be governed by international investment law.
International investment law is a creature of treaties, mostly bilateral, entered into by states. There are nearly three thousand such bilateral investment treaties (BITs) in force today, creating a web of protections that cover an expansive set of economic assets and transactions.Footnote 11 Through investment treaties, states promise certain protections to investors of counterparty states, including protections against discrimination, uncompensated expropriation, unfair and inequitable treatment, and, in some cases, contractual breach. Importantly, investment treaties grant covered foreign investors the right to commence international arbitration directly against states and claim damages for treaty violations.Footnote 12
In 2006, Argentina became the first state to be subject to such claims for sovereign debt. It faced three separate investment arbitrations, brought by thousands of Italian creditors, for breaches of the Italy-Argentina BIT allegedly arising out of Argentina's 2001 default and subsequent restructuring.Footnote 13 An investment arbitration against Greece followed in 2013, alleging similar breaches of the Slovak-Greece and Cyprus-Greece treaties arising from Greece's 2012 bond restructuring.Footnote 14
At the time these claims were asserted, it was unclear whether international investment tribunals would accept jurisdiction over sovereign debt. Greece and Argentina each argued that sovereign bonds did not constitute protected “investments” as defined by the applicable investment treaties and the Convention on the Settlement of Investment Disputes Between States and Nationals of Other States (ICSID Convention).Footnote 15 They further argued that, even if sovereign bonds were investments, they were not made in the territory of the respondent states, and (in the case of Argentina) that mass claims were disallowed by ICSID.Footnote 16 These arguments were consistent with the weight of the commentary at the time.Footnote 17
Today, those arguments no longer appear persuasive. Three of the four tribunals presented with the Argentine and Greek sovereign debt disputes accepted jurisdiction, and the one tribunal that did not employed questionable reasoning and interpretative methods.Footnote 18 None of these decisions is binding in future cases, and some were issued with dissents.Footnote 19 They nevertheless signal that—whether directly through investor-state claims or indirectly in the shadow of potential claims—international investment law will continue to be a factor in sovereign defaults for the foreseeable future. Yet because all claims brought to date have been settled or dismissed before reaching a final decision on the merits, basic legal questions remain unresolved.
With a new wave of debt crises in view—and new treaty claims potentially on the horizon—this Article considers how the application of investment treaties to sovereign bonds is likely to impact states, creditors, and the institution of international investment law itself, and, moreover, how those impacts should inform the adjudication of sovereign debt disputes on the merits. The Article makes two principal claims.
The first claim is that while most investment treaties appear to cover sovereign debt, that coverage threatens to undermine the core economic purpose of international investment law. Investment treaties and arbitration enforce state compliance with investment commitments ex post, in order to reduce the cost of capital and thus promote investment ex ante.Footnote 20 But the predominant problem with respect to sovereign debt is over-, not under-compliance. Multiple studies have led to a consensus that states rarely default on their debts opportunistically; instead, they resist restructuring unsustainable debts even when it is efficient to do so.Footnote 21 BITs threaten to worsen this inefficient behavior by both impeding the restructuring of unsustainable debt and by exacerbating the market imperfections that lead states to over-comply in the first place.
Investment treaties are likely to create inefficiencies in other ways as well, including by exacerbating holdout problems, increasing litigation costs, and creating uncertainty. All of these inefficiencies increase the cost of borrowing for states or the cost of lending for creditors, and thus undermine, rather than advance, the core economic purpose of international investment law. Sovereign debt thus poses a “BITs and bonds dilemma” for investment tribunals, who face the choice between giving effect to BITs’ coverage of sovereign debt or the fundamental purpose of those treaties.
The Article's second claim is that this dilemma can be resolved through the equal treatment of like creditors. Because, as explained below, equal treatment would only allow claimants to recover damages that placed them in the same relative position as other like creditors, it would not increase the leverage of creditors in general and thus would not worsen the ex post over-compliance problem. Nor would it empower holdouts in a way that undermines necessary restructurings, creates uncertainty, or leads to wasteful proceedings. The equal treatment of like creditors would instead prevent states from discriminating against creditors, which, as this Article demonstrates, is the one positive role that international investment law can play in sovereign debt markets.
The equal treatment of like creditors can be achieved through treaty reform, but it need not be. Standard nondiscrimination provisions in investment treaties directly support the equal treatment of creditors. In addition, and as explained below, other provisions—such as prohibitions against uncompensated expropriations or contractual breaches—can also be reasonably interpreted to lead to a similar result.
Ultimately, whether or not such equal outcomes are achieved—and thus whether or not the BITs and bonds dilemma is resolved—depends upon four ongoing debates in international investment law, to which this Article contributes, including: What constitutes a sovereign act (and is thus compensable under major treaty norms)? How should fair market value compensation for sovereign debt be calculated? What is the nature of “umbrella” clause violations? And what is the proper measure of compensation for such violations?
More broadly, resolution of the BITs and bonds dilemma depends upon a debate over the basic nature of investment law itself, including whether it creates mandatory rules that cannot be contracted out of, and, relatedly, whether tribunals should prioritize the stability or flexibility of investment norms. As the case of sovereign debt illustrates, there may be serious perils to imposing a set of mandatory one-size-fits-all rules on a wide range of assets and transactions. As it further illustrates, if such mandatory rules are nevertheless imposed, flexibility in their application may be needed to avoid their most harmful effects.
The remainder of this Article proceeds as follows: Part II situates BITs and bonds in the context of the broader international economic order. Part III lays out the BITs and bonds dilemma. Part IV proposes an interpretive solution to that dilemma. Part V concludes with an analysis of recent treaty developments and recommendations for both states and creditors.
II. BITs and Bonds in the International Economic Order
A. The Sovereign Debt “Regime”
Sovereign debt—defined as debt issued or guaranteed by a central government—measures in the tens of trillions of dollars.Footnote 22 And, as noted above, sovereign defaults are reoccurring, if irregular, events.Footnote 23 Nevertheless, there is no international sovereign bankruptcy law that allows states to declare bankruptcy or discharge their debt.Footnote 24 Nor is there a central forum for managing creditor claims or rationally distributing sovereign assets.Footnote 25 Sovereign defaults are instead governed by a number of interdependent yet uncoordinated institutions.
1. National Contract Law and Courts
First among these institutions are national contract law and national courts. Most sovereign debt is issued in the form of bonds.Footnote 26 A significant portion of sovereign bonds, including the “lion's share” of bonds issued by advanced economies,Footnote 27 are domestic bonds, meaning that they are governed by the issuer country's contract law and domestic courts.Footnote 28 The remainder are foreign or external bonds, meaning that they are subject to foreign laws and jurisdictions, often New York or English law and often New York or London courts.Footnote 29 In the event of default, creditors can bring claims for breach of contract under these laws and in these forums.
Enforcement of sovereign debt contracts through national courts is rarely successful, however.Footnote 30 The enforcement of domestic debt faces a “high[] probab[ility]” of failure because of changes in law or resistance by the state and its judiciary.Footnote 31 The enforcement of foreign debt—which is insulated from these risksFootnote 32—faces a different obstacle: the execution of foreign judgments.Footnote 33
The execution of foreign judgments in the debtor state faces the same obstacle as the enforcement of domestic debt: namely, changes to domestic law and reluctant local courts.Footnote 34 Execution outside of the debtor state, on the other hand, is impeded by the limited number of government assets located abroad,Footnote 35 and foreign sovereign immunity protections that shield many such assets from attachment.Footnote 36 It is, in fact, so difficult to recover damages in the case of default that bond prospectuses often expressly warn creditors of the challenges of doing so.Footnote 37
2. Restructuring Agreements
Due in part to the difficulty of enforcing sovereign debt contracts, restructuring plays a critical role in resolving sovereign debt crises. Sovereign debt restructurings, which typically take the form of an exchange of old bonds for new bonds, can include the deferral of interest or principal payments, or “haircuts” of the face value of the bond.Footnote 38 Depending on the applicable law and contract terms, restructurings can be voluntary—whereby only those creditors who agree to the exchange receive the new bonds—or they can be mandatory—whereby even the bonds of creditors who reject the offer are subject to the exchange.Footnote 39 Restructuring can occur prior to a default, or it can occur after a default has already occurred.Footnote 40
Creditors may agree to restructure sovereign debt not only because of the obstacles to enforcing sovereign debt instruments, but also because of the impact that unsustainable debt can have on states’ capacity to service future debt obligations. High debt loads can place a state in a debt “spiral,” in which burdensome debt payments crowd out other investments, lead to lower growth, less revenue, and the need to assume further debt, which in turn leads to lower growth, less revenue and so on.Footnote 41 As macroeconomic conditions deteriorate and debt-to-GDP levels rise, the cost of continued debt service can rise in a dynamic fashion.Footnote 42 As the IMF explains:
Allowing an unsustainable debt situation to fester is costly to the debtor, creditors and the international monetary system. For debtors, a situation of debt overhang depresses investment and growth and creates a sense of financial uncertainty that can raise the eventual magnitude of the debt problem. … Delays that magnify the scale of economic dislocations also tend to reduce the economic value of creditors’ claims. …Footnote 43
Where debt becomes unsustainable, agreeing to a collective restructuring, whereby each creditor agrees to accept less in future payments than they had contracted for, may make all creditors better off.Footnote 44
3. The Holdout Problem and Solutions
Even where creditors as a group recognize that debt restructuring is in their own interest, some creditors may nevertheless demand full payment on their bonds. These “holdout” creditors count on a restructuring agreement to reduce the state's overall debt burden to a sufficient degree to allow the state to pay the holdouts in full.Footnote 45 Holdout creditors, in other words, attempt to free-ride on the agreement of other creditors.Footnote 46
Where too many creditors hold out, however, the strategy may backfire, as the restructuring can fail, leaving no agreement on which to free-ride. Even the possibility that certain creditors may hold out can be enough to scuttle or delay a restructuring. Indeed, “[i]f creditors know that a ‘holdout’ can obtain full repayment conditional on a previous debt restructuring, everyone will want to be that holdout, and no one will want to restructure.”Footnote 47 Paying off holdouts or even signaling a willingness to pay holdouts can thus undermine necessary debt restructurings and make all creditors worse off.Footnote 48 At the same time, not paying holdouts can lead to protracted creditor litigation with costs to both states and creditors alike.Footnote 49
States and creditors have devised certain contractual solutions to address the holdout problem. For example, exit consents allow a qualified majority of creditors to amend the nonfinancial terms of bonds not subject to an exchange in order to make the prospect of holding out less favorable.Footnote 50 Collective action clauses, on the other hand, allow a supermajority of creditors to agree to a restructuring in a way that binds all creditors and thus renders litigation futile.Footnote 51
While collective action clauses are “now a well-established market practice” in foreign bondsFootnote 52 and have “played a useful role in achieving high creditor participation,”Footnote 53 they are not a complete solution to the holdout problem. They are not found in all foreign sovereign bonds, and are they relatively rare in domestic bonds.Footnote 54 Moreover, even where they are included, their functionality is limited where they only bind bondholders of the same debt series. In that case, “a creditor, or a group of creditors, can obtain a ‘blocking position’ in a particular series and effectively nullify the operation of CACs in that series.”Footnote 55
B. Introduction of International Investment Law to the Sovereign Debt Regime
Subjecting sovereign debt to investment treaties can be expected to both enhance creditors’ enforcement rights and, relatedly, exacerbate the holdout problem.
1. Strengthening Creditor Enforcement
Investment treaties strengthen enforcement rights in two ways. First, with respect to domestic debt, they protect against ex post changes to governing laws, which may violate the treaties’ provisions on expropriation or fair and equitable treatment. The investment arbitration claims against Greece—which retroactively changed the terms of its domestic debt—were brought for precisely this reason.Footnote 56 Moreover, investor-state arbitration—which allows parties to choose the arbitrators who hear the disputeFootnote 57—is likely to provide a friendlier forum than the courts of the defaulting sovereign, even in the absence of a change to governing law.Footnote 58
Second, with respect to both domestic and foreign debt, investor-state arbitral awards provide certain enforcement advantages over national court judgments. The claims brought against Argentina—which concerned debt subject to foreign law and jurisdictionFootnote 59—appear to have been motivated by these advantages.
While investor-state arbitral awards are subject to the same major limitations as foreign judgments—namely, sovereign immunity and the lack of attachable assetsFootnote 60—their enforcement is facilitated by two global “pro-enforcement” regimes.Footnote 61 Investor-state awards can be enforced under either the ICSID or New York Convention.Footnote 62 The ICSID Convention requires that each of its 153 contracting states enforce ICSID awards as “as if [they] were a final judgment of a court in that State.”Footnote 63 The New York Convention, likewise, requires its 161 members to recognize and enforce arbitral awards except in “narrow” and limited exceptions.Footnote 64 Foreign judgments, by contrast, do not benefit from any comparable global enforcement regime.Footnote 65 Their enforcement instead depends upon bilateral or less comprehensive multilateral treaties, or, absent that, local law, “which often makes it difficult, if not impossible, to effectively enforce them.”Footnote 66
2. Exacerbating the Holdout Problem
International investment law may also exacerbate the holdout problem. Creditors litigate where the expected net gains of doing so are greater than the expected net gains of agreeing to a restructuring.Footnote 67 Thus, to the extent investment treaties make enforcement less costly or more effective, they increase the expected net gains of litigating, and, all else being equal, increase the likelihood that creditors do so.Footnote 68
International investment law may also exacerbate the holdout problem by elevating the enforcement rights of some creditors above others. While there are almost three thousand investment treaties in force and investors can structure their investments to satisfy nationality requirements,Footnote 69 it is unlikely that all creditors—particularly unsophisticated or small domestic creditors who are less likely to employ such creative legal tactics—can avail themselves of these advantages. Investment treaties could thus be expected to confer a type of “secured” status on covered creditors, placing them in a superior position to others.Footnote 70 The resulting heterogeneity could make agreement among creditors more difficult, increase the marginal cost of avoiding holdout litigation,Footnote 71 or both, and thus lead to more such litigation.
Finally, investment treaties may exacerbate the holdout problem by providing an end-run around collective action clauses. As noted above, such clauses are unlikely to preclude all holdout litigation, both because of their incomplete coverage and because they leave open the possibility of holdouts gaining a blocking position in a particular debt series.Footnote 72 However, as Waibel notes, even where collective action clauses bind all creditors and preclude litigation for breach of contract in national courts, state actions relating to restructurings facilitated by collective action clauses may still give rise to investment treaty claims.Footnote 73 For example, creditors could plausibly claim that a repudiation or threat of nullification committed prior to such a restructuring constituted an expropriation or a violation of fair and equitable treatment, even if it does not give rise to a breach of contract.Footnote 74
3. Jurisdiction
While the potential for international investment law to enhance creditor enforcement rights and increase holdout litigation appears significant, only four arbitrations against two states and pursuant to two treaties have been brought to date. That heretofore limited use may be partly explained by the time and cost associated with investor-state arbitration.Footnote 75 Another likely cause, however, is the uncertainty that until recently surrounded threshold jurisdictional questions, including whether sovereign bonds were even covered under investment treaties and the ICSID Convention.
That jurisdictional uncertainty now appears to have been resolved largely in favor of creditors. As noted above, the weight of arbitral precedent has concluded that sovereign bonds are in fact protected investments. While not binding, those decisions are relevant because most BITs contain definitions of investments similar to the treaties at issue in those arbitrations, and encompass, as those treaties did, “every kind of asset” or “all assets,” including claims to money.Footnote 76 Indeed, it is difficult to argue that sovereign bonds do not constitute a “kind of asset” or an “investment” under the plain meaning of those terms.Footnote 77
Following these decisions, some states have negotiated new investment treaties that either expressly exclude sovereign debt or limit applicable substantive and procedural protections.Footnote 78 Those treaties remain a small minority, however, and many have not yet come into force.Footnote 79 Moreover, older treaties may continue to apply to previously issued sovereign debt under sunset provisions, which ensure terminated treaties apply to investments made while those treaties were still in force, often for a decade or more.Footnote 80 These new treaties nevertheless signal a movement toward differentiating sovereign bonds from other investments.
The next Part—which analyzes the impact of applying BITs to bonds from an efficiency perspective—helps to explain and evaluate those treaty reforms. It moreover provides the basis for a proposal for adjudicating sovereign debt disputes under the vast majority of treaties that still cover sovereign debt without distinction.
III. The BITs and Bonds Dilemma
The impact of international investment law on sovereign debt has been the focus of only limited commentary. Waibel, who has conducted the most comprehensive legal analysis,Footnote 81 has observed that investment treaties have the potential to increase holdout litigation and undermine collective action clauses.Footnote 82 Ostřanský has observed that investment law is likely to introduce uncertainty into sovereign debt pricing and restructurings.Footnote 83
This Part builds on those observations to examine the impact of international investment law, not just on the restructuring process, but on the social (or aggregate) cost of sovereign debt. To date, no empirical studies have directly addressed this issue.Footnote 84 However, as explained below, existing empirical studies, together with economic models, strongly suggest that that investment treaties are likely to increase, rather than decrease, the social cost of sovereign debt and thus undermine, rather than advance, the core economic function of investment law.
A. The Efficiency Function of Investment Law
The promotion of economic efficiency is a—if not the—core function of international investment law.Footnote 85 Investment treaties provide investors substantive and procedural protections against state action for the purpose of reducing risk and thus promoting foreign investment.Footnote 86 From a social welfare perspective, investment treaties should protect and promote investment only to the point where the marginal benefits of doing so are greater than the marginal costs. They should, in other words, incentivize states to act efficiently with respect to investment.
1. Credible Commitments
According to the classic contract theory of investment law, investment treaties promote efficiency by enabling states to make enforceable commitments to investors.Footnote 87 States and investors, like two private commercial parties, may desire to enter into a cooperative relationship that is mutually beneficial. But because states face an incentive to renege on their commitments after the investment is made,Footnote 88 investors may refuse to invest in the first place, leaving both investors and states worse off.Footnote 89
The domestic legal system may not be sufficiently credible to assure foreign investors. States can change legislation and nullify contracts, and their judicial systems may not provide a fair hearing.Footnote 90 Investment treaties, on the other hand, allow states to make international commitments that cannot be unilaterally altered or nullified, and whose enforcement does not depend upon domestic courts. By strengthening enforcement, investment treaties reduce the risk that states will renege on their commitments ex post, and thus enable mutually beneficial investment transactions ex ante.Footnote 91
2. An Imperfect Market Theory of BITs
The contract theory of investment law, however, overlooks that states are quintessential repeat players. As such, states have an incentive to abide by their commitments, since reneging could jeopardize their ability to enter into other beneficial transactions in the future.Footnote 92 As Sykes observes, under perfect market conditions, states would internalize the cost of any adverse investment action in the form of increased capital costs.Footnote 93 Where investors have perfect information and capital is perfectly elastic, any adverse investment action would lead investors to demand a higher rate of return or withdraw from the market. Where states seek to maximize overall long-term welfare, they therefore have “a unilateral incentive to steer clear of policies” whose marginal cost (in terms of increased capital costs or lost investment) exceeds their benefit. In other words, under perfect market conditions, states would act efficiently toward foreign investment and there would be “no problem [for investment treaties] to solve.”Footnote 94
Investment law has an economic role to play only because perfect market conditions do not exist. Among other things, high discount rates may lead states to take actions that increase short-term gains even where they lead to greater long-term costs.Footnote 95 Agency problems may lead officials to maximize their own self-interest at the expense of the state.Footnote 96 And information asymmetries may allow states to take adverse actions without internalizing the costs, and thus keep investors from allocating capital efficiently.Footnote 97
As such, investors cannot rely on rational self-interest to restrain states from breaching their investment commitments. By offering creditors compensation in the event of such actions, investment treaties enforce compliance ex post, thereby reducing the cost of capital and promoting investment ex ante.Footnote 98
B. The Effect of BITs on States’ Costs of Sovereign Borrowing
Applying BITs to bonds, however, threatens to undermine rather than advance those objectives. That is because, as explained below, the predominant problem with respect to sovereign debt is inefficient over-, not under-, compliance. In fact, the same market imperfections that lead states to inefficiently breach other investment commitments likely lead states to inefficiently over-comply with sovereign debt. In this context, investment treaties threaten to increase the cost of sovereign debt in two ways. First, they threaten to do so directly, by impeding the restructuring process and adding to the deadweight losses associated with unsustainable debt. Second, they threaten to do so indirectly, by exacerbating the market imperfections that lead states to inefficiently resist restructuring unsustainable debt in the first place.
1. The Restructuring Process
As discussed above, unsustainable debt can lead to deadweight losses that harm states’ economic growth, reduce their capacity to pay creditors, and cause economic costs for society as a whole.Footnote 99 In such situations, restructuring debt in a timely and sufficient manner can benefit both states and their creditors.Footnote 100
Yet multiple studies have found that states resist restructuring debts even when it is efficient to do so. For example, an IMF review of nine sovereign debt restructurings from 2000 to 2013 found that “debt restructurings often took place a considerable period after Fund staff had assessed that the [country's] debt was no longer sustainable” and that “[w]hen debt restructurings do occur, they often do not restore debt sustainability.”Footnote 101 Other studies have come to similar conclusions,Footnote 102 including finding that states resist restructuring unsustainable debts “even if that implies running down reserves, shortening the maturity of the debt, and ceding part of their economic policy sovereignty to multilateral institutions.”Footnote 103
Conversely, the empirical evidence shows that it is “rare”Footnote 104 or “extremely rare”Footnote 105 for states to default when they have the capacity to pay their debts. As such, a consensus has emerged that the predominant public policy problem with respect to sovereign debt is not inefficient under-compliance, but rather inefficient over-compliance.Footnote 106 Indeed, according to the IMF—hardly an apologist for debtor states—the problem is not that states renege on their sovereign debt commitments too often or too much, but that they often do “too little and too late.”Footnote 107
BITs threaten to further increase the deadweight losses associated with unsustainable sovereign debt by impeding the restructuring process. By exacerbating the holdout problem, BITs threaten to undermine or delay the completion of restructurings, thus making them even later than they already would have been.Footnote 108 By strengthening creditor enforcement, BITs force states to choose between assuming additional litigation costs and the execution of additional assets, or increasing exchange offers to avoid litigation in the first place—either one of which would reduce the effective relief provided by restructurings, thus making them even littler than they already would have been.
The additional deadweight losses from these littler and later restructurings would fall directly on states and thus effectively raise their cost of borrowing. Notably—in contrast to any additional payments to creditors themselves—these additional deadweight losses would confer no benefit on creditors and thus no ex ante benefit on the cost of capital. To the contrary, these losses would indirectly harm creditors by reducing states’ capacity to pay.Footnote 109
2. State Behavior Ex Post
In addition to impeding the restructuring process, BITs threaten to increase costs by incentivizing states to resist restructuring unsustainable debt even more than they already do. This is supported by empirical studies, discussed below, that suggest that the same market imperfections that lead states to breach their other investment commitments are driving states’ over-compliance with sovereign debt. Instead of correcting those market imperfections, as one would expect with respect to other investments, BITs threaten to exacerbate them in the context of sovereign debt.Footnote 110
a. High discount rates and the economic costs of default
With respect to investments in general, we expect states to inefficiently breach their investment commitments where they overvalue the short-term benefits of the breach, while discounting its long-term costs in terms of lost investment. Even in the case of sovereign debt, we might expect high discount rates to lead to excessive defaults to the extent such actions produced immediate economic gains (in terms of reduced debt servicing) despite greater long-term costs (in terms of higher borrowing costs).
Yet not only is excessive default rarely observed, empirical studies suggest that breaches of sovereign debt commitments reverse the timeline on gains and losses expected of typical breaches. Instead of short-term gains and long-term costs, sovereign debt restructurings produce short-term costs and long-term gains. Indeed, multiple studies have shown that breaches of sovereign debt commitments are associated with immediate economic costs, including declines in economic output,Footnote 111 foreign direct investment,Footnote 112 other private capital flows,Footnote 113 and international trade;Footnote 114 risks of banking and currency crises;Footnote 115 the exclusion from international capital markets;Footnote 116 and, upon return to those markets, increased borrowing costs.Footnote 117 State compliance with sovereign debt is widely attributed to these costs in the economic literature.Footnote 118
At the same time, many of these costs have been found to be short-lived. A review of empirical studies shows that defaulting states “have regained access to international capital markets fairly quickly,”Footnote 119 and that there is no detectable effect on growth or risk spreads after one or two years.Footnote 120 Moreover, while the costs of default diminish over time, gains increase as the benefits of reduced debt payments begin to accumulate.Footnote 121
The timeline of short-term costs and long-term gains suggests that—in contrast with other investments—high discount rates may be contributing to over-compliance with sovereign debt. It moreover suggests that BITs, which are likely to increase the short-term costs of default—by increasing litigation costs, extending the duration of restructuring and its associated disruptions, and increasing payouts to creditors—are likely to exacerbate that over-compliance rather than correct it.Footnote 122
b. Agency problems and the political costs of default
The same appears to be true of agency problems, which, in contrast to how they affect other investments, appear to lead to over-, not under-, compliance with sovereign debt commitments.
In the case of investments generally, we expect agents to cause the state to inefficiently breach investment commitments where agents benefit personally or politically from breaches while avoiding responsibility for the costs. In the case of sovereign debt, we might expect political agents to be too willing to default where doing so produces political gains (whether that be through a boost to national pride or economic benefits) that outweigh political costs (which might not materialize or might fall on future agents).
But again, not only are opportunistic defaults rare, empirical studies suggest that sovereign defaults are associated with negative, not positive, political consequences, and therefore that agency problems may be contributing to the more significant problem of over-compliance. For example, Borensztein and Panizza have found that following sovereign defaults, political parties and politicians in power are more likely to suffer electoral losses,Footnote 123 and leaders are more likely to be removed from office.Footnote 124 These findings are consistent with previous studies showing the negative impact of currency crises on political fortunes.Footnote 125 They are further supported by the anecdotal experiences of Greece, which saw its prime minister resign and one of its two major political parties collapse, and Argentina, which cycled through five presidents in as many weeks during their last restructurings.Footnote 126
These political costs make sense in light of the economic and political costs of defaults. Legal and economic scholars have recognized that such costs may lead “self-interested agents” and “myopic policymakers” to delay restructuring as long as possible to avoid the costs of default, and then restructure too little in order to hasten the return to normal.Footnote 127 The IMF, likewise, has attributed states’ reluctance to restructure unsustainable debt to “fear[] [of] the economic, financial and political fallout of a restructuring.”Footnote 128
In this context, instead of correcting for agency problems, international investment law is likely to exacerbate them.Footnote 129 For if political agents inefficiently delay and minimize restructurings in order to avoid the immediate economic—and thus political—costs that result, then a mechanism that increases those costs and lengthens their duration would likely incentivize further inefficient delays.
At the same time, agency problems also point to a potential positive role for investment treaties. While agency problems may incentivize compliance with sovereign debt obligations in general, they do not necessarily incentivize states to impose the burdens of any eventual default equitably among creditors. To the contrary, agency problems may incentivize imposing such costs on a select group of creditors—such as foreign creditors or other creditors without political clout. In fact, while the evidence shows that states are generally reluctant to default,Footnote 130 “some defaults appear to have successfully discriminated between foreign and domestic debtholders.”Footnote 131 International investment law may thus still have a role in protecting creditors from discriminatory treatment. With respect to ex post compliance in general, however, investment law threatens to exacerbate the effects of agency problems, not correct them.
c. Information asymmetry and access to courts
Information asymmetries may also contribute to states’ ex post over-compliance. In other contexts, we expect information asymmetries to lead to inefficient under-compliance because they allow states to surreptitiously take investment-adverse actions without facing the consequences of future lost investment. But not only do we rarely observe inefficient under-compliance, such information asymmetries are unlikely to exist in the context of sovereign defaults. Sovereign debt defaults are well-publicized events, and the market reacts quickly to defaults by decreasing demand for debt from the defaulting state.Footnote 132
Instead, information asymmetries are likely to contribute to inefficiency in another way. Specifically, where creditors are not able to independently verify the sustainability of a state's debt stock or their capacity to pay, states may face an incentive to “postpone default to ensure that there is broad market consensus that the decision is unavoidable and not strategic,”Footnote 133 and thus obtain creditor support for a restructuring arrangement.
While international investment law is unlikely to worsen such information asymmetries, it not likely to effectively ameliorate them either. The adjudicative process can play a transparency-enhancing role with respect to sovereign debt,Footnote 134 but investor-state arbitration is poorly suited to play that role, at least where national courts exist as alternative forums.Footnote 135 This is so for several reasons.
First, disclosure in investor-state arbitrations is controlled by the tribunal and generally more limited than party-controlled discovery in national courts, at least in common law jurisdictions such as New York and London.Footnote 136 Second, investor-state proceedings are typically subject to a higher level of confidentiality than national litigation.Footnote 137 And third, because of the uncertainty and inconsistency associated with investment-arbitration, the ability of investor-state tribunals to make legal determinations that are accepted by the market and serve as an objective evaluation of state behavior is relatively weak.Footnote 138 Indeed, the current ad hoc system, which has no appeals mechanism and is not constrained by stare decisis, allows for two creditors holding the exact same debt and subject to the exact same treaty to be treated incongruously.Footnote 139
At the same time, investor-state arbitration might play some role as an information intermediary where no other forum is available for creditors to pursue their claims. For example, in the event a state nullifies its domestic debt or otherwise forecloses access to its national courts, investor-state arbitration may still have some utility as an information intermediary (though that utility would have to be weighed against the costs of providing holdouts another forum to litigate). Outside of those circumstances, however, investment arbitrations are likely to be redundant and ineffective.
* * * *
In sum, in addition to impeding and increasing the costs of the restructuring process, international investment law threatens to exacerbate the market imperfections that lead states to inefficiently resist restructuring unsustainable debt in the first place. These effects could be expected to compound each other. By impeding and increasing the costs of restructurings, investment treaties would incentivize states and their agents to further delay and diminish the restructuring of unsustainable debt. Those same treaties would then impede the restructuring process when states finally do attempt to restructure. The end result would be even littler and later restructurings, more deadweight losses, and an increase in the effective cost of sovereign borrowing for states.
3. State Behavior Ex Ante
If states over-comply with sovereign debt obligations, something else must explain the recurring nature of sovereign defaults. Certainly, some of the factors that lead states to default are outside of their control and impervious investment law's effects. For example, individual debt crises are often instigated by broader global economic shocks.Footnote 140 This explains why defaults tend to happen in waves, with multiple countries entering into debt crises at the same time.Footnote 141 To cite one particularly grave example, nearly half of all countries defaulted during the years following the Great Depression and World War II.Footnote 142 Needless to say, international investment law would hardly be effective at reducing such risks.
Another possible explanation, however, is that states inefficiently overborrow ex ante.Footnote 143 There are, in fact, some indications that ex ante overborrowing is a driver of sovereign debt crises. First, borrowing by crisis-prone countries tends to be pro-cyclical, meaning that they tend to borrow more during periods of economic expansion than contraction,Footnote 144 which is the opposite one would expect of borrowers seeking to smooth consumption during economic downturns.Footnote 145 Second, there is wide variation in the debt stock of similarly situated nations, suggesting that borrowing is not driven by economic fundamentals, but rather by policy choices.Footnote 146 Finally, the IMF has found that most emerging countries’ debt stocks are higher than those countries could repay by maintaining their respective average primary balances.Footnote 147
Assuming that inefficient overborrowing ex ante contributes to debt crises, it is likely driven by the same market imperfections that cause states to inefficiently over-comply with sovereign debt obligations ex post. High discount rates and agency problems could be expected to lead states and their agents to overborrow in the near-term while discounting the costs of doing so in the long-term. Politicians, for example, could have an incentive to overborrow to boost the economy while they are in office, knowing (or hoping) that the consequences of doing so will fall on future governments. Additionally, opacity as to fiscal conditions may keep lenders from accurately assessing a state's capacity to pay.
Increasing ex post costs could be expected to have some impact on ex ante overborrowing. Rational states would reduce overborrowing on the front end to avoid the increased costs of default on the back end. This could ameliorate the negative effects of investment law on ex post over-compliance. However, any positive impact on overborrowing is likely to be limited by the market imperfections that drive that overborrowing in the first place. For example, the ex ante agency problem that leads to overborrowing is unlikely to be influenced by ex post enforcement, since that agency problem exists in spite of the costs that accrue ex post. High discount rates likewise, by definition, would lead states to discount any future additional ex post costs that international investment law may impose. And to the extent over-lending is a function of ex ante opacity, investor-state arbitration would be largely useless, as it is only available ex post, and in any event, as described above, is of limited utility as an information intermediary. Thus, while encouraging further inefficient state behavior ex post, international investment law is unlikely to have any significant impact on inefficient state behavior ex ante.Footnote 148
C. The Effect of BITs on Creditors’ Costs of Lending
For all the above reasons, international investment law is likely to increase the deadweight losses associated with sovereign debt restructurings, by both impeding restructurings and by encouraging the inefficient behavior of states. The resulting losses can be expected to accrue to states, thus effectively increasing their cost of borrowing, without conferring any benefit on creditors. In fact, at least some of the deadweight losses can be expected to fall indirectly upon creditors by reducing states’ overall capacity to pay.Footnote 149
In addition to these indirect costs, investment law is also likely to impose other costs directly on creditors by creating incentives for inefficient creditor behavior. In particular, by incentivizing holdout arbitration in all the ways discussed above, investment treaties may directly increase creditors’ costs in two ways.
First, holdout arbitration may increase litigation costs by leading to parallel filings in both national courts and investor-state arbitration.Footnote 150 A race to attach a limited number of assets could lead creditors to file arbitration claims to either gain an enforcement advantage or preserve a position relative to other creditors. Investment arbitration may thus place creditors in a prisoners’ dilemma, whereby each faces the incentive to file a treaty claim, regardless of what other parties do, and regardless of whether it leaves all creditors worse off.Footnote 151
Second, holdout arbitration may introduce uncertainty into the distribution of payments to creditors. In particular, BITs could be expected to shift payments to holdout creditors, who insist on full payment, from non-holdout creditors who do not. Relatedly, BITs could be expected to shift payments to creditors of certain nationalities who have access to investment-treaty arbitration, from creditors of other nationalities who do not. And finally, BITs may shift payments to domestic bonds from foreign bonds. As noted above, domestic bonds are likely to benefit more from investment-treaty protection than foreign bonds. In addition to the enforcement advantages that accrue to both domestic and foreign bonds, domestic bonds stand to benefit from protections against changes to governing law and from the provision of a more neutral adjudicative forum (both of which foreign bonds already enjoy). While sovereigns are less likely to default on domestic bonds than foreign bonds,Footnote 152 these additional protections may make domestic bond defaults even less likely, and thus, all else equal, make foreign bond defaults more likely.
If creditors could accurately anticipate these distributional effects ex ante,Footnote 153 they could price them into bond prices efficiently. But the unpredictable nature of investor-state arbitration is likely to impede creditors from doing so.Footnote 154 These distributional impacts are thus likely to impose uncertainty costs on creditors.Footnote 155
* * * *
In theory, although suboptimal, the costs that accrue directly to creditors—as well as some of the costs that accrue directly to states—could be justified by other benefits that investment treaties produce. For example, empowering holdout creditors could discipline states that under-comply with their sovereign debt commitments.Footnote 156 The additional uncertainty and transaction costs that investment law creates could also be outweighed by the benefits created by marginally more efficient state behavior. Indeed, that is precisely the justification for applying international investment law—despite its associated transaction costs—in other contexts.
In reality, however, where states already inefficiently over-comply with sovereign debt commitments, and where investment law threatens to worsen that inefficient state behavior, the additional costs that investment treaties impose will only further increase the cost of sovereign debt and further undermine the economic purpose of international investment law.
IV. Resolving the BITs and Bonds Dilemma
The above analysis helps to explain recent reforms undertaken by some states to exclude sovereign debt from new investment treaties or impose procedural and substantive limitations on its protection.Footnote 157 However, the vast majority of treaties continue to define “investment” broadly in a way that includes sovereign debt. How should tribunals proceed when faced with claims under those treaties? Several imperfect alternative responses present themselves.Footnote 158
One alternative is for tribunals to accept jurisdiction and proceed as if sovereign debt were any other contract. This approach would recognize that the plain language of the ICSID Convention and most BITs covers sovereign debt, but ignore that such coverage undermines international investment law's core goals as described above. That may have been the road the three tribunals hearing claims against Argentina would have taken had those arbitrations not settled or been discontinued before reaching a decision on the merits.Footnote 159
A second alternative is to ignore the plain language of BITs and the ICSID Convention and reject jurisdiction. That is the path urged by commentators and by the tribunal seized of the Greek debt dispute. While having the virtue of avoiding all of the problems of subjecting BITs to bonds, this alternative relies an implausible interpretation of the term “investment,” and moreover, strips BITs of their capacity to protect creditors from discrimination.
This Part proposes a third alternative. It argues that international investment law's fundamental norms can be advanced rather than undermined—and states and creditors made better not worse off—by guaranteeing the equal treatment of like creditors, but going no further. Moreover, it shows how the reasonable interpretation of current treaties can lead to that result.
A. Equal Treatment as the Solution to the BITs and Bonds Dilemma
The equal treatment of like creditors solves the BITs and bonds dilemma in two ways. First, it prevents creditors from using BITs in a way that introduces inefficiencies and thus undermines the economic function of those treaties. As discussed above, by expanding creditor enforcement rights and exacerbating the holdout problem, BITs threaten to increase inefficiencies in four ways: (1) incentivizing states to further delay and diminish necessary restructurings; (2) impeding necessary restructurings once undertaken; (3) increasing litigation costs; and (4) introducing uncertainty. The equal treatment of like creditors would mitigate each of these effects.
If investor-state arbitration only put creditor-claimants in the same position as other similarly situated creditors—such that holdout creditors were compensated as much as, but no more than, non-holdouts—it would not increase the aggregate leverage of creditors. It would not require the disbursement of additional resources, but rather the equitable distribution of whatever resources states do disburse. Moreover, it would effectively neutralize any benefit to holding out, except in the case of discriminatory treatment. By doing so, it would (1) eliminate the additional incentives for states to further delay and diminish restructurings, (2) prevent creditors from undermining restructurings, (3) reduce the proliferation of additional proceedings and the attendant litigation costs, and (4) minimize uncertain distributional shifts.
Second, the equal treatment of like creditors would preserve international investment law's capacity to play one of the key roles that the sovereign debt regime leaves for it to play: ensuring the nondiscriminatory treatment of creditors. As discussed above, the market imperfections that incentivize states to renege on commitments with respect to other types of investments—high discount rates, agency problems, and information asymmetries—actually incentivize compliance with sovereign debt obligations in general.Footnote 160 Nevertheless, agency problems may still incentivize states to spread the costs of default inequitably among creditors.Footnote 161 Guaranteeing the equal treatment of like creditors would protect against such discrimination.
In its most effective form, the equal treatment solution would keep the restructuring of sovereign debt out of investor-state arbitration entirely. States, knowing that discriminating among creditors would trigger liability—and having the capacity to avoid such liability—would steer clear of such actions. Creditors, knowing they could not recover more through investor-state arbitration than through a restructuring agreement, would be disincentivized from bringing such claims. Both states and creditors would thus have the incentive to resolve debt crises outside of investor-state arbitration, without the costs that it entails.
In that sense, the equal treatment solution would position investor-state arbitration where (at least ideally) it sits with respect to other investments: “in reserve,” as a tool to ensure “processes function as intended,”Footnote 162 and whose use is obviated by its success in incentivizing efficient behavior. That contrasts with how investor-state arbitration would otherwise be used with respect to sovereign debt: as a rent-seeking tool that creditors could be expected to use regardless of whether states acted efficiently or not, and which, through that use, would undermine rather than advance the core function of international investment law.
B. A Path Toward Equal Treatment Through Four Debates
The equal treatment of like sovereign bond creditors could most obviously be achieved by reforming BITs to limit their application to nondiscrimination protections. Specifically, treaties could be amended to guarantee sovereign creditor-claimants nondiscriminatory treatment—via national treatment, most favored nation treatment, and stand-alone nondiscrimination clauses—but to go no further. Such clauses would allow creditor-claimants to obtain damages only where they are offered less favorable restructuring terms than other like creditors, i.e., creditors holding the same or similar debt.Footnote 163 This should lead to the equal treatment of like creditors, because damages for discrimination would, at a minimum, undo the effects of discrimination,Footnote 164 or because states would themselves treat creditors equally in order to avoid such claims.
But BITs need not be reformed to achieve this outcome. Reasonable interpretations of current treaties can achieve it as well. Of course, nondiscrimination norms are not the only norms of the international investment law regime. Most BITs also contain absolute standards of treatment that do not depend upon the relative treatment of other creditors. For example, most BITs contain clauses requiring compensation for expropriation, the fair and equitable treatment of investors, and, in some cases, umbrella clauses that prohibit contractual breach.
As explained in this Section, however, these absolute standards of treatment do not preclude the equal treatment of both creditor and non-creditor claimants. To the contrary, reasonable interpretations of such standards may, counterintuitively, lead to relative equality in outcomes, even when applied to the benefit of only some creditors. As demonstrated below, the path toward equal treatment—and the resolution of the BITs and bonds dilemma—runs through four ongoing and fundamental debates in international investment law.
1. What Amounts to a Compensable Sovereign Act?
The first fundamental debate upon which the resolution of the BITs and bonds dilemma depends is over what amounts to a compensable sovereign act. As Waibel has observed, “[i]nternational courts and tribunals have repeatedly stressed that property rights and investment protection are no insurance against ordinary commercial risks.”Footnote 165 International tribunals have likewise drawn a line between breaches of contract, which do not create treaty claims, and sovereign interference with a state's contractual obligations, which do.Footnote 166 In other words, a breach of contract does not amount to a breach of treaty “unless it [is] proved that the State or its emanation has gone beyond its role as a mere party to the contract, and has exercised the specific functions of a sovereign authority.”Footnote 167
Pursuant to this principle, as Waibel shows, there is “ample authority” supporting that a mere default on debt—which amounts to a breach of contract—is a noncompensable commercial act.Footnote 168 The tribunal in SGS v. Philippines, for example, found that “[a] mere refusal to pay a debt is not an expropriation of property, at least where remedies exist in respect of such a refusal.”Footnote 169 Likewise, with respect to a default on Polish bonds, the Foreign Claims Settlement Commission of the United States held that “[m]ere nonpayment of an obligation of this nature does not constitute a nationalization or other taking of property.”Footnote 170
The principle that sovereign default is a commercial, not sovereign, act is moreover consistent with U.S. Supreme Court jurisprudence on the Foreign Sovereign Immunities Act of 1976. In Republic of Argentina v. Weltover, the Court found that the issuance and restructuring of sovereign debt were covered by the Act's commercial activity exception.Footnote 171 In reaching that conclusion, the Court held that the distinction between commercial and sovereign actions depends upon “whether the particular actions that the foreign state performs (whatever the motive behind them) are the type of actions by which a private party engages in ‘trade and traffic or commerce.’”Footnote 172 Because the payment and restructuring of debt are “the type of actions by which a private party engages in ‘trade and traffic or commerce,” they were found to constitute commercial activities under the Act.Footnote 173
The principle that default is not a sovereign act is a fundamental milestone on the path toward equal treatment. It should prevent holdout creditors from claiming that a mere default breaches the major absolute protections of BITs, including expropriation and fair and equitable treatment. However, simply because default does not amount to a sovereign act does not eliminate the possibility of state liability. Tribunals and commentators have identified at least two other specific acts with respect to sovereign debt that could lead to state responsibility under expropriation or fair and equitable treatment clauses. First, nullifying a debt contract, by rendering it legally unenforceable, may amount to a compensable sovereign act. For example, the holding in SGS v. Philippines that “a mere refusal to pay a debt is not an expropriation of property” was dependent upon the existence of “remedies [] in respect of such a refusal.” As the tribunal explained, there was no expropriation because “[w]hatever debt the Philippines may owe to SGS still exists; whatever right to interest for late payment SGS had it still has.”Footnote 174 As a practical matter, states may only nullify domestic debt subject to their own laws, as states cannot alter foreign laws to render external debt unenforceable.
Second, commentators and tribunals have observed that governments are sovereign rather than commercial actors where they repudiate debt by signaling either explicitly or implicitly their intention to never service their debt obligations.Footnote 175 Because a state's capacity to reject its obligations is not dependent on the applicable governing law, states may repudiate both domestic and foreign debt. For example, the tribunal in Abaclat found that the passage of Argentina's Padlock Law, which precluded the government from making payments to foreign holdout creditors, thus “entitling it not to perform part of its obligations,” amounted to a sovereign act.Footnote 176
While the most critical principle for ensuring the equal treatment of creditors is that default does not amount to a compensable sovereign act, whether tribunals apply the nullification or repudiation standard may also affect whether BITs play a positive role in the sovereign debt regime. In particular, holding that repudiation is a sovereign act may constrain states from discouraging holdouts. As discussed above, efficient and successful restructuring depends on states’ capacity to clearly and credibly signal to creditors that holdouts will not be paid.Footnote 177 Prohibiting the repudiation of debt would impede states from doing exactly that.
The policy implications of prohibiting nullification are mixed, although on balance they are more consistent with a positive role for BITs than a prohibition of default or repudiation. On the one hand, prohibiting nullification may lead to additional domestic litigation and increase the costs of restructuring domestic debt. On the other hand, it would not foreclose states’ ability to send clear signals to holdouts and thus encourage participation in restructurings. Moreover, it would preserve creditors’ access to courts and the information intermediary role courts can play.
The legal case that repudiation is a compensable sovereign act is likewise significantly weaker than the legal case for nullification. Holding that repudiation is a sovereign act is inconsistent with the principle that states act commercially when they engage in the “type of actions by which a private party engages in ‘trade and traffic or commerce.’”Footnote 178 Clearly, private parties can signal an intent to never perform a contract or reject the contract's validity outright. They may do so to gain negotiating leverage and in good or bad faith. Regardless, a repudiation of one's contractual obligations is squarely commercial as that term has been defined.Footnote 179
Arguments to the contrary confuse the type of act with its source. For example, the tribunal in Abaclat reasoned that the Argentine law “entitling it not to perform part of its obligations” was a sovereign act on the basis that it was a legislative act that “derives from Argentina's exercise of sovereign power … it is neither based on nor does it derive from any contractual argument or mechanism.”Footnote 180 But that standard would define any legislative or executive act as sovereign, regardless of whether it is the type of action that a commercial actor could take. It would, taken to its logical end, define the issuance of and default on sovereign debt—which require the approval of some sovereign authorityFootnote 181—as sovereign acts, contradicting “ample authority” to the contrary.Footnote 182
Nullification of debt, on the other hand, is plainly not the type “type of action[] by which a private party engages in ‘trade and traffic or commerce.’”Footnote 183 While private parties may breach contracts, they do not have the power to render those contracts legally unenforceable. The better answer, therefore, both as a matter of law and policy, is that nullification of sovereign debt constitutes a compensable sovereign act, but that neither default nor repudiation does.
2. How Is Fair Market Value Calculated?
The second fundamental question upon which the resolution of the BITs and bonds dilemma depends is the calculation of compensation for sovereign acts. Compensation for expropriation is typically determined by reference to the fair market value (FMV) of the expropriated investment.Footnote 184 Calculating FMV requires assessing what a buyer would be willing to pay for the asset in an arm's length transaction, accounting for the commercial risk associated with that asset.Footnote 185 All else equal, where the commercial risk associated with the asset has increased since its original purchase, FMV will be lower than the asset's sticker price.Footnote 186
The principle that a mere default is a commercial act becomes highly significant in this context. First, it establishes that a creditor holding a defaulted bond would not have a claim for damages on account of the default itself. The FMV of defaulted bonds would simply be the price at which those bonds trade on the market (which incorporates the effect of the default), or, alternatively, the price of any new bonds that the creditor could have obtained through an exchange.Footnote 187 In other words, even if the creditor did have a cognizable claim for default, it could only obtain damages equivalent to what it could obtain on the market or in a bond exchange, and would thus have no incentive to hold out and bring a claim in a first place.
Second, treating default as a commercial act establishes that, where a state acts in its sovereign capacity—for example by nullifying its debt—creditors should be compensated only for losses due to the sovereign act, not any losses due to default. In other words, it establishes that the proper compensation for a sovereign act should not be the face value of the bond, but instead its fair market value, incorporating losses due to default. This should further reduce creditors’ leverage and the holdout problem by reducing the damages available for cognizable claims and minimizing any difference between what is offered through an exchange and what is available through investor-state arbitration.
In practice, determining the FMV of an asset subject to a sovereign act such as an expropriation presents challenges, because both the act and the expected remedy can affect the asset's market price. Where less than complete compensation is expected, an expropriation or impending expropriation will reduce what buyers are willing to pay for the asset and thus its market value.Footnote 188 Where complete or above-market price compensation is expected, an expropriation or impending expropriation may actually increase the market price of the asset, creating a situation of moral hazard.Footnote 189 To minimize such distortions, international investment law often calculates the FMV of an asset by looking to the market value of the asset before the expropriation took place or was anticipated. For example, the World Bank Guidelines on the Treatment of Foreign Direct Investment calls for fair market compensation to be “determined immediately before the time at which the taking occurred or the decision to take the asset became publicly known.”Footnote 190
In the case of sovereign debt, calculating the FMV of nullified debt is likely to be particularly challenging, because the (compensable) risks of nullification and the (noncompensable) risks of default are almost certainly correlated. As such, the risk of nullification can be expected to rise and fall with the risk of default, making it difficult to disentangle the two.
As a practical matter, however, FMV in this context could be determined by using a number of benchmarks. First, consistent with World Bank Guidelines, and assuming there is a delay between a default and a nullification, tribunals could determine FMV by way of the price of the bond after default but before the nullification occurs or becomes known. However, this method may misprice the FMV of the bonds, because, as discussed above, the price after default likely already incorporates a higher risk of nullification. Second, tribunals may look to the face value of new bonds that are offered in an exchange. This approach, too, may misprice the bonds if it takes place following a nullification or in the shadow of an expected nullification. On the other hand, because of the incentives faced by states and their agents discussed above, there is good reason to believe that the face value of exchange bonds would meet or exceed the state's capacity to pay, and thus effectively compensate for any nullification effects.Footnote 191 Third, tribunals could look to the change in market value of comparable debt instruments that have been subject to restructuring or default but not nullification, and compare that with the change in market value of bonds subject to default and nullification. This approach could more cleanly separate the price impact of nullification from that of default, although it would require that there be similar non-nullified debt to compare.Footnote 192
None of these benchmarks is perfect, but together they provide practical ways to disentangle commercial from sovereign risk and preclude compensation for the former. By doing so, they minimize any differences in compensation between holdouts and non-holdouts and thus more fully achieve the equal treatment of like creditors.
3. Are Umbrella Clause Claims Allowed Notwithstanding an Exclusive Forum Selection Clause?
In addition to claims for expropriation or fair and equitable treatment, creditors may also bring claims for contractual breach under so-called umbrella clauses. Umbrella clauses are not contained in all treaties,Footnote 193 are by no means uniform,Footnote 194 and are highly controversial.Footnote 195 But at least some tribunals have found that umbrella clauses allow investors to bring claims predicated on a contractual breach.Footnote 196 Where such a clause applies and is so interpreted, a creditor-claimant could establish liability by simply showing that the state breached its debt contract by defaulting.
Whether umbrella clauses end up playing a role in the arbitration of sovereign debt disputes hinges on a contested threshold question: do forum selection clauses that select a national court as the exclusive forum for contractual breach foreclose umbrella clause claims in investor-state arbitration? This question is relevant because external bonds often select New York or London as the designated forum.Footnote 197 If exclusive versions of such clauses effectively foreclosed contractual claims for default, then it will not be possible to assert umbrella clause claims for at least some sovereign debt. If, on the other hand, creditors can bring umbrella clause claims notwithstanding exclusive forum selection clauses, then the adjudication of such claims could be determinative as to whether international investment law plays a positive or negative role in the sovereign debt regime.
Two arbitrations involving the same claimant and a similar set of facts illustrate the divide on this question. In SGS v. Paraguay, the tribunal found that a contract with an exclusive forum selection clause did not foreclose investor-state arbitration of contract claims on the ground that the treaty's umbrella clause effectively converted any contract claims into treaty claims. Because the forum selection clause in the contract applied only to contract claims, it thus had no bearing on the adjudication of umbrella clause claims even if those claims were premised on a contractual breach.Footnote 198
The tribunal in SGS v. Philippines, on the other hand, came to the opposite result. It reasoned that while umbrella clauses may “make it a breach of the BIT for the host State to fail to observe binding commitments, including contractual commitments,” they do “not convert the issue of the extent or content of such obligations into an issue of international law.”Footnote 199 Because the “extent” and “content” of those obligations remained inherently contractual, the tribunal found that the exclusive forum selection clause applied, and it stayed the arbitration.Footnote 200
There are good commercial and economic reasons to prefer the SGS v. Philippines approach to the approach taken in SGS v. Paraguay. Forum selection clauses are part of a contractual bargain and can impact the price of the contract. Limiting the parties’ ability to contract out of investment arbitration thus limits party autonomy and may introduce inefficiency into the contracting process.Footnote 201 While it may be sensible to make access to investment arbitration a default rule under some circumstances—such as where the negotiation of access entails significant transaction costs and most parties are likely to agree to arbitration—those circumstances are unlikely to be present with respect to sovereign debt, where transaction costs are unlikely to be high (particularly in contracts that contain negotiated forum selection clauses).Footnote 202
Moreover, any procedural economy concerns that could favor the SGS v. Paraguay approach are unlikely to materialize. In theory, accepting jurisdiction over umbrella clause claims could prevent parallel proceedings, in which parties litigate other treaty claims in investment arbitration and contract claims in national court. In reality, however, there is unlikely to be an “identity of the parties’ object and cause of action” in the different proceedings, as contract claims will be based in national law and investor-state claims will be based on treaty law.Footnote 203 As such, neither res judicata, lis pendens, nor fork-in-the road provisions would prevent parallel proceedings and procedural waste.Footnote 204 In this context, the better answer is that exclusive forum selection clauses should foreclose the investor-state arbitration of umbrella clause claims predicated on contractual breach.
4. Do Contract or Fair Market Value Damages Apply to Umbrella Clause Claims?
Assuming, however, that umbrella clause claims for sovereign default are admissible, their impact on sovereigns and their creditors will turn on the calculation of damages. Two obvious choices present themselves: FMV and contract damages. While tribunals “rarely properly distinguish[]” between the two,Footnote 205 which of these is applied will have a significant impact on the role that international investment law plays in the sovereign debt regime.
As noted above, the standard metric for the calculation of damages for treaty claims is FMV, and the FMV of sovereign bonds will drop below the face value of the bonds in the event of a default.Footnote 206 If FMV was also used to calculate damages for umbrella clause claims, the compensation claimants received under such claims would mirror the compensation for other claims. As with respect to other claims, such compensation would effectively lead to the equal treatment of like creditors.Footnote 207
On the other hand, if normal contract damages were awarded for umbrella clause claims, creditor-claimants would be entitled to damages specified by the debt contract itself or by the applicable national law on contracts. Because contract damages do not account for changes in risk or fluctuations in the market value of the contract,Footnote 208 contract damages would be significantly higher than FMV in times of default.
The divergent approaches of SGS v. Paraguay and SGS v. Philippines provide a way forward. According to the tribunal in SGS v. Paraguay, because umbrella clauses effectively convert contract claims into treaty claims,Footnote 209 exclusive forum selection clauses do not preclude the adjudication of umbrella clause disputes. According to the tribunal in SGS v. Philippines, umbrella clauses do not convert contract claims in treaty claims, and thus exclusive forum selection clauses preclude the investment arbitration of claims arising from a breach of contract. Following the logic of this split, to the extent that umbrella clause claims are admitted, such claims should be considered treaty claims, not contract claims, and therefore treaty damages, i.e., FMV, not contract damages, should apply. To the extent such umbrella clause claims are found to be contractual in nature, the exclusive forum selection clauses found in at least some bond instruments should preclude their arbitration, and questions regarding damages should not arise.
* * * *
The path toward equal treatment, and the resolution of the BITs and bonds dilemma, runs through the four debates reviewed above. In brief, if (1) defaults are commercial and not sovereign acts, and (2) the appropriate measure of compensation for treaty violations is FMV, then the damages for any treaty claim related to sovereign debt should not be the face value of the bond, but rather the FMV of the bond, accounting for the fact that it is in default. If (3) umbrella clause claims are admitted on the basis that such claims are treaty and not contract claims, then (4) the appropriate measure of compensation for umbrella clause claims should not be contract damages, but instead the FMV of the bond, again accounting for the default.
Because the FMV of a defaulted bond can be established by reference to the trading price of such bonds, the value of a bond exchange in a restructuring agreement or the value of comparable instruments, claimant-creditors should only be able to recover damages equivalent or similar to what is available to other like creditors.Footnote 210 This should prevent states from discriminating against creditors. Moreover, because it limits any increases in creditor leverage in the aggregate and the potential for claimants to fare better in arbitration than in restructuring, this interpretation would mitigate the holdout and over-compliance problems and the costs they create.
C. Objections and Responses
Before concluding, two potential objections to the above proposal merit responses.
1. The Rare Case Versus the Typical Case
The first objection could be put as follows: The analysis, conclusions, and proposal in this Article are based upon data and observations regarding the average or typical default. While most states may wait “too late” to restructure and then do so “too little,” that does not preclude other states from defaulting too often or too much. Why should investment law be built around the typical or prevailing case, instead of protecting creditors against the rare case of opportunistic default?
There are three responses to that objection. First, unless investment law distinguishes between the rare case of opportunistic default and other more common defaults, building investment law around the rare opportunistic default will have all of the negative effects in other more common cases. With respect to those other cases, BITs will threaten to increase inefficiencies and deadweight losses, and thus undermine the core purpose of those treaties.
Second, to the extent international investment law attempts to distinguish between the rare case of opportunistic default and other more common defaults, it is likely to end up on a slippery slope toward the negative impacts described above. While some defaults may be obviously opportunistic, others are likely to be closer calls. Determining which is opportunistic would require assessing a state's ability to pay, something even the IMF—an institution equipped with teams of economists dedicated to doing precisely that—struggles to do effectively.Footnote 211 The notion that ad hoc international investment tribunals would be able to do any better is highly unlikely. Instead, a fragmented investment arbitration regime would almost certainly lead to conflicting, contradictory, and simply incorrect assessments.Footnote 212
Finally, while it is true that the above proposal does not offer additional remedies to creditors in the event of rare opportunistic defaults, it supports other processes that could provide relief in such cases.Footnote 213 For example, the proposal protects creditors’ other legal remedies, such as access to national courts, and contributes to the smooth functioning of restructurings by protecting against discrimination. It moreover leaves undisturbed other effective market-based tools for creditors to insure against risk, including through the purchase of credit default swaps.Footnote 214 In other words, the proposal supports, rather than supplants, existing functions.Footnote 215
2. Mandatory and Flexible Rules Versus Optional and Stable Rules
The second objection is broader and could be put as follows: The analysis underlying this proposal depends on a set of costs, limitations, and redundancies created by other institutions, including, in particular, the costs of default, foreign sovereign immunity, and national courts. If any of these costs or institutions were to change, the optimal role for investment law could change as well. Why should tribunals tailor their interpretations to specific disputes and to the external institutions and dynamics that surround the dispute rather than apply them indifferently to all manner of disputes regardless of the consequences? In other words, why should tribunals prioritize the flexibility rather than the stability of investment treaty norms?
National courts, at least in the case of external debt, certainly prioritize the latter. As Buchheit, Gulati, and Tirado observe: “Nothing can realistically be done to keep a holdout from obtaining a judgment in a foreign court on a foreign law-governed debt instrument. Attempting to unseat basic tenets of contract law in countries like the United Kingdom … will meet fierce resistance.”Footnote 216 In taking this approach, national courts have sometimes played a destructive role in sovereign default cases.Footnote 217
At the same time, by establishing clear rules that recognize sovereign debt instruments as enforceable contracts, national courts have slowly but surely encouraged creditors and states to agree to private solutions to sovereign debt problems. As discussed above, more and more sovereign debt contracts now contain collective action clauses that can avoid the worst consequences of the rigid enforcement of sovereign debt contracts. While these changes have been painstakingly slow, and have yet to catch up with some problematic provisions, including the pari passu clause,Footnote 218 the steady application of contract law by national courts has given states and creditors clear rules to negotiate around.
Why should international investment tribunals not do the same? As an initial matter, unlike national contract law, the Vienna Convention on the Law of Treaties requires that BITs be interpreted in their “context and in the light of [their] object and purpose.”Footnote 219 The object and purpose of investment treaties—investment promotion—compels consideration of the real world effects that BIT interpretations create, which in turn requires considering how those interpretations will interact with other factors and institutions. Ignoring the context in which investment treaties are applied and the effects of that application is incompatible with the law on treaty interpretation.
Just as importantly, the putatively mandatory nature of investment law, the breadth of its application, and its substance and procedure all favor flexibility over stability. The weight of arbitral precedent has treated investment treaties as creating a set of mandatory rules that cannot be contracted out of.Footnote 220 The mandatory nature of international investment law neutralizes one of the principle benefits to clear and stable rules, namely that they facilitate party autonomy and efficiency by providing a clear basis for negotiations.Footnote 221 Instead, mandatory rules, particularly inflexible ones, threaten to constrain party autonomy rather than facilitate it, and to undermine rather than advance efficiency.Footnote 222 This is particularly so because investment treaties establish substantive and procedural rights that in any other context would amount to important price terms subject to negotiation.Footnote 223
The BITs and bonds dilemma is an extreme illustration of the inefficiencies that mandatory rules can create. But mandatory rules can also create inefficiencies in more subtle ways, including by forcing substantive and procedural protections onto states and investors that they may otherwise have chosen to forgo. For example, investment treaties create an additional enforcement mechanism that insulates investors from the risk of sovereign acts. But that mechanism comes at a cost to states that we would expect to be priced into contracts ex ante.Footnote 224 Where investors do not desire or are not willing to pay that cost, investment treaties may actually impede rather than promote investment.Footnote 225
The potential for such inefficiencies is particularly significant in light of investment law's expansive reach. Most BITs purport to impose a single set of rules over nearly every aspect of the economy, including “all assets” or “every kind of asset”—from real property to concession contracts, equity shares, debt, and claims to performance and money.Footnote 226 This extraordinary breadth increases the risk that any one mandatory rule will be inefficient with respect to at least some transactions. As Arato has observed, applying the same set of rules designed for “Blackstonian property” to contracts, debt, and equity has distorting effects.Footnote 227 Imposing mandatory one-size-fits-all procedural rules can likewise cause inefficiencies.Footnote 228
In this context, the substantive and procedural flexibility of investment law makes sense. BIT standards are notoriously vague and capacious. Fair and equitable treatment,Footnote 229 national treatment, and even expropriationFootnote 230 standards leave ample room for a multitude of interpretations that can vary depending on the circumstances and the transaction at issue. The enforcement mechanism for BIT obligations is likewise optimized for flexibility. The ad hoc nature of investor-state arbitration, the lack of any meaningful appeals mechanism, and the lack of binding precedent allow tribunals the flexibility to tailor their interpretations to the transaction before them without constraining other tribunals with respect to future transactions.
This flexibility makes further sense in light of the genesis of BITs. BITs are negotiated and established among states, leaving no opportunity for negotiation with investor beneficiaries. Moreover, because their standards are meant to cover a wide range of investments, they are not tailored to any particular one. In that sense, BITs can be conceived as incomplete contracts whose interpretation and application calls for considering what optimizes the particular transaction at issue.
Flexibility, of course, has its own costs. Particularly where it leads to inconsistencies with respect to the same transactions and the same factual circumstances, the flexibility of investment law substance and procedure can create uncertainty that impedes states and investors from pricing treaty rules into their transactions ex ante. Ideally, these costs would lead to a reexamination of the mandatory nature of international investment law, at least with respect to state contracts in which investors are in a position to bargain—and pay—for the protections that they desire. In addition, or alternatively, states could specify more precise standards of treatment for different transactions, as some states are beginning to do with respect to sovereign debt.
Yet as long as the vast majority of investment agreements impose a set of mandatory rules that cover an extraordinarily wide range of assets and transactions without distinction, flexibility as opposed to stability may be the next best option. That is certainly the case with respect to sovereign debt.
V. Conclusion
Sovereign debt poses a dilemma to international investment tribunals. While most investment treaties cover sovereign debt, that coverage threatens to undermine the core purpose of those treaties. Instead of encouraging investment by correcting inefficiencies and thus reducing the cost of capital, applying investment treaties to sovereign debt threatens to do the opposite. Investment treaties threaten to worsen the holdout problem, make states even more reluctant to restructure unsustainable sovereign debt, and to increase litigation and uncertainty costs. All of these effects increase the cost of lending for creditors and the cost of borrowing for states, leaving both groups worse off than they otherwise would be in investment law's absence.
This Article identifies that dilemma and proposes a solution to it. It demonstrates how guaranteeing the equal treatment of like creditors, but going no further, can resolve all of the inefficiencies that investment treaties would otherwise create. It further demonstrates how guaranteeing the equal treatment of like creditors would allow treaties to play a positive role in the sovereign debt regime by preventing states from unjustifiably discriminating between creditors. Finally, it shows how current investment treaties can be reasonably interpreted to achieve that outcome.
Ultimately, however, what makes this proposal possible may also undercut its efficacy. That is because the flexibility of investment treaty norms that allows for an interpretive solution to the BITs and bonds dilemma also makes it less likely that tribunals will adopt it in a uniform manner on which states can rely. It is for this reason that states should take steps to limit the costs that investor-state arbitration may impose on sovereign debtors and creditors.
Some states have already been moving in this direction. A few newly negotiated investment treaties, such as the recently adopted (but not yet in force) Argentina-Japan BIT, exclude sovereign debt all together.Footnote 231 Other agreements, like the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP), take a more elaborate approach that only allows for discrimination claims in the event of a negotiated restructuring, but permits other claims in the absence of one.Footnote 232 The analysis in this Article reveals the strengths and weaknesses of each of these approaches.
On the one hand, the Argentina-Japan approach of excluding sovereign debt entirely solves the BITs and bonds dilemma by preventing BITs from increasing deadweight losses and encouraging inefficient state behavior. However, like arbitral decisions that reject jurisdiction, it deprives investment law of the opportunity to ensure the nondiscriminatory treatment of creditors.
On the other hand, the CPTPP, which allows for national treatment and most favored nation treatment claims in the event of a negotiated restructuring,Footnote 233 maintains a positive supportive role for international investment law. However, it does not resolve all of the inefficiencies of subjecting bonds to the strictures of that law. That is because the CPTPP defines “negotiated restructuring” as a restructuring agreed pursuant to the collective action clause of a particular instrument, or agreed by holders of at least 75 percent of the aggregate principal of that instrument.Footnote 234 Like other such clauses, therefore, the CPTPP allows holdout creditors to gain a blocking position in a particular debt series and bring claims for full paymentFootnote 235 (in the case of the CPTPP, by way of an expropriation claimFootnote 236).
If one were more concerned about opportunistic defaults than about inefficient over-compliance, then the CPTPP approach could make sense. Relative to no differentiation of sovereign debt at all, it balances some progress against the holdout problem with an increase in creditor leverage that could act as a check on opportunistic defaults. Alternatively, if one were concerned about inefficient over-compliance but not about creditor discrimination, then the Argentina-Japan approach of excluding sovereign debt would make sense. But if one accepts the consensus of empirical studies that states are too reluctant, not too willing, to restructure their debts and one recognizes that state agents may nevertheless have an incentive to unjustifiably discriminate among creditors, the better approach is to restrict tribunals’ jurisdiction over sovereign bonds to claims for discrimination, even in the absence of a negotiated restructuring.
In addition to future treaty reforms, states should also consider more immediate steps. They might, for example, issue interpretive statements that support the interpretations described in this Article. While tribunals have accorded differing weight to such statements, at least some have accepted them as binding or at least persuasive.Footnote 237 States and creditors could also attempt to limit the scope of international investment arbitration through modifying sovereign debt instruments, notwithstanding investment law's putatively mandatory nature. Such attempts are unlikely to be successful, although at least some tribunals have signaled that they would give effect to explicit contractual terms that purport to impose such limits.Footnote 238
All of these proposals—treaty amendments, interpretive statements, and revisions of bond covenants—face their own obstacles, however. In addition to the transaction costs of their negotiation and the time they will take to come into force, each may also face a coordination problem. Even though applying international investment law to sovereign debt may be costly for creditors and states in the aggregate, like holdout litigation more generally, it could still produce benefits for those creditors that employ it. And because other creditors of other states may be able to obtain such advantages through other BITs or through other debt instruments, creditors and creditor states may resist unilaterally disclaiming them.Footnote 239 Sovereign bonds, therefore, may pose a different version of the coordination problem that commentators have typically ascribed to investment treaties, in which capital importing states sign on to BITs because they are stuck in a prisoners’ dilemma.Footnote 240 With respect to sovereign debt, capital exporting states and their creditors may face such a dilemma instead.
All of these obstacles, including the coordination problem, are surmountable. But they do make it more difficult for states to implement a social-welfare maximizing solution to the BITs and bonds dilemma on their own, particularly in time to address the next wave of sovereign defaults. Which makes it all the more important that tribunals do so instead.
