مذاکره دوباره، بند اقدام جمعی و بازارهای بدهی های مستقل

Collective action clauses (CACs) are provisions specifying that a supermajority of bondholders can change the terms of a bond. We study how CACs determine governments' fiscal incentives, sovereign bond prices, and default probabilities in environments with and without contingent debt and IMF presence. We claim that CACs are likely to be an irrelevant dimension of debt contracts in current sovereign debt markets because of the variety of instruments utilized by sovereigns and the implicit IMF guarantee. Nonetheless, under a new international bankruptcy regime like that recently proposed by the IMF, CACs can increase significantly the cost of borrowing for sovereigns, contrary to what is suggested in previous empirical literature.

Collective action clauses (CACs) are provisions in debt contracts specifying that the terms of the contract regarding principal, interest, and maturity can change if there is consent of a predetermined supermajority of bondholders. This paper studies how CACs determine governments' fiscal incentives, bond yields, and default probabilities. Understanding these interactions is essential for the design of the so-called “Sovereign Debt Restructuring Mechanism” (SDRM) proposed by the IMF and currently under discussion.1 CACs introduce flexibility in situations of financial distress by facilitating renegotiation.2 In their absence, bondholders have no incentives to enter into the renegotiation process since, individually, they are unable to affect the probability of repayment (as long as the debt is not held by a large lender). CACs solve the problem of free riding among creditors within a legal jurisdiction because a supermajority of bondholders can make the outcome of the renegotiation mandatory for all. But the existence of CACs does not always imply a friendly restructuring process. Sovereigns tend to issue debt in different jurisdictions, and while CACs coordinate creditors within each one, the free riding problem between jurisdictions remains. This is a feature of the 1990s not present in the 1980s, when few banks concentrated most of the sovereign bonds. To attend to this problem, the idea of an international bankruptcy procedure (or an SDRM), to coordinate creditors in different jurisdictions, has been put forward.3 It has been argued that facilitating renegotiation can have both positive and negative consequences. Because renegotiation relieves countries from debt overhang, governments might run reckless fiscal policies that increase the likelihood of financial crisis. Since lenders anticipate this behavior, the cost of the lack of commitment to run responsible fiscal policies is borne by the country itself. In the end, the severity of the moral hazard problem determines whether facilitating renegotiation, by creating an SDRM, make countries worse or better off. The debate about the value of an SDRM lies precisely on this trade off.4 We setup up a model to understand the determinants of this tension. We focus on environments where countries can strategically issue debt with and without CACs, and in different legal jurisdictions, both in the presence and absence of the IMF. We show that an SDRM is never a good idea when debt contracts are state contingent. Under uncontingent debt payments, we derive a series of implications that we believe are both new and relevant for the discussion of an international bankruptcy procedure. Furthermore, we point at some empirical evidence to question conclusions from previous empirical results. First, our analysis sheds light on the discussion of the role of CACs and an SDRM in affecting the trade off between ex post restructuring cost and ex ante moral hazard. Recent work by Eichengreen and Mody (2000) shows that yields on primary sovereign debt markets (initial auctions) are higher when bonds have CACs, especially for low rated borrowers.5Becker et al. (2003) and Gugiatti and Richards (2003) argue that bond prices are not affected very much by the implicit (legal jurisdiction) or explicit inclusion of these types of clauses when looking at yields in secondary markets. Hence, they conjecture that either financial markets are not really aware of the role of those clauses, or the moral hazard problem that these clauses bring to international credit markets does not outweigh the ex post inefficiencies of no renegotiation. Therefore, switching to an SDRM would not increase the yields paid by sovereigns.6 We argue that these empirical exercises suffer from the Lucas' critique. The reason is that bond yields are estimated under the current regime, characterized by no renegotiation due to a “compositional effect” and the presence of the IMF. This compositional effect, which is missing in the literature, comes from the free riding problem among creditors of different jurisdictions. We claim that these clauses are likely to be irrelevant in sovereign debt markets, and hence, spreads of yields of bonds with and without CACs are uninformative about moral hazard problems. Nonetheless, our framework suggests that these yields and the moral hazard problem could worsen in a regime with an SDRM and CACs (under full coordination among creditors). Quantitatively, these compositional effects are relevant. By 2002, 59% off all international borrowing occurred under US jurisdiction, 10% under German jurisdiction, and 6% under Japanese law, all with no collective action provisions, while 24% resided in the UK, where the opposite is true.7 It is then reasonable to expect no major difference between yields of bonds with and without CACs. Once the country is financially distress, holders of bonds with friendly restructuring provisions might not forgive because they posses a minority of the total outstanding debt, and they can only marginally affect the probability of repayment. In particular, we show the compositional effect was present in the case of Argentina 2001. Furthermore, we show that yields of bonds with and without CACs where not only similar before but also during the crisis. This is evidence against the argument that the ex ante moral hazard problem is balanced by the ex post gains from renegotiation. Instead, this evidence favors our story claiming that compositional effects make CACs irrelevant, explaining why the sovereign debt markets do not really care about these clauses. We also show that compositional effects are likely to have been present in other cases of default. We also show that the presence of the IMF affects the international allocation of capital and hence default probabilities and yields. Furthermore, it can also affect the decision of governments to include CACs in bonds or not. When the IMF has a strong interest in the destiny of a country, it will be inclined to launch bailouts to avoid a financial crisis. This implies that lenders would always get paid, either by the country or by the IMF, and hence, yields would never reflect default probabilities. In this case, yields are uninformative with respect to moral hazard (which is maximized under full bailout). For that reason, we should carefully understand the interactions of governments, lenders, and the IMF in sovereign debt markets before drawing conclusions about an SDRM. The theory suggests that sometimes there will be conflict between governments and the IMF regarding the creation of an SDRM. Given that the IMF will intervene, governments sometimes prefer not to have an SDRM to then enjoy the subsidy of the implicit IMF guaranty. But the IMF would prefer to have an SDRM in place so that lenders internalize the costs of lending to reckless governments. The theory also suggests that conflict does not necessarily exist between lenders and the IMF. When moral hazard issues are important, countries and the IMF would prefer not to have an SDRM to induce fiscal responsibility as a commitment device not to renegotiate. Lastly, most of the literature works under the assumption that an SDRM always induces some moral hazard. On the contrary, our framework illustrates the possibility that an SDRM could actually induce better incentives. To see this point, first suppose that the IMF would prefer not to intervene if lenders were to renegotiate under an SDRM but would prefer to launch a bailout in the absence of an SDRM. If the moral hazard problem is important, countries might prefer to be punished by the outcome of the renegotiation with lenders than by having a generous IMF bailout. In other words, countries would be better off under an SDRM precisely because it can provide greater incentives for governments to avoid financial crises. The rest of the paper is divided in three sections. In Section 2, we present the model and analyze different contractual environments without the IMF. In Section 3, we introduce the IMF and derive its implications for sovereign debt markets. Finally, in Section 4, we explain the implications of our theory regarding sovereign yields and discuss previous literature.