June 2025

The article at a glance: Market expectations about sovereign default can be crucial determinants of political decision making and can even become self-fulfilling in triggering sovereign default. I provide novel causal evidence that an increase in the economic cost of default can credibly commit a government to repay and can curb market expectations for the propensity of a default event by eliminating belief-driven debt crises equilibria. However, such a shift comes at the cost of increasing the severity of fundamental crises.

Can Bond Vigilantes be Reassured by a Higher Economic Cost of default?

By Lennart Niermann, CERF/CCFin Scholar, Faculty of Economics

Market expectations about sovereign default can be crucial constraints for political decision making and debt sustainability.

Bond vigilantes and the associated market expectations about sovereign default can play a crucial role in shaping policy, as seen when bond-vigilantes swiftly ended Liz Truss’ 2022 UK premiership after just 44 days, or when they substantially contributed to a roll-back of Donald Trump’s 2025 reciprocal tariffs after a 50 basis point hike in the 10-year bond yield following their announcement. Beyond forcing governments to change policy, a sudden loss in investor confidence can even become self-fulfilling by preventing the sovereign from rolling over its debt and forcing default as was the case for the Mexican debt crisis of 1994-95 (Cole and Kehoe, 2000). I provide novel evidence that such market expectations are not only affected by the fiscal rule a government is committed to but can rather also be influenced by the economic cost a government would incur in case of default.

Court Rulings on Debt Litigation Cases Provide an Exogenous Source of Variation

For sovereigns who raise their external debt under foreign jurisdiction, creditor lawsuits have become an increasingly important component of the cost of default in recent decades (Schumacher et al., 2021). Namely, creditor lawsuits can exert an economic penalty by delaying or blocking debt restructurings, which prolongs exclusion from financial markets or via the attachment of export revenues, which can effectively function as a trade embargo. After sovereign default in the Republic of Congo for example, several US-based debt funds launched attachment attempts on Congo's crude oil exports and successfully blocked Congo from receiving royalties or export revenues - Congo's most important source of foreign exchange - for multiple years. In Argentina’s case, a 15-year legal dispute with holdout creditor NML Capital resulted in blocked payments to restructured bondholders, as well as high-profile asset enforcement actions, including the seizure of a naval vessel. Crucially though, the degree to which creditors can litigate is governed by the jurisdiction under which external debt is raised. As approximately 70 percent of international sovereign debt in recent decades is raised under New York law which emphasizes case law and legal precedent, my empirical identification strategy leverages U.S. legal rulings on sovereign debt litigation cases. These provide a natural experiment to identify the impact of shifts in the perceived economic cost of default on the default probability.

An Increase in Litigation Induced Default Costs Curbs Default Expectations

Empirically, I analyse the response of USD-denominated bond markets subject to U.S. jurisdiction in Brazil, Peru, Chile, Costa Rica, and Panama in response to a series of 23 U.S. legal rulings on litigation cases against Argentina. I follow Vrugt (2011) and Merrick (2001) in simultaneously estimating each country's daily recovery rate - the share of the bond principal creditors expect to recoup - as well as the default probability from a daily cross-section of their USD-denominated government bonds. I find that an increase in the expected recovery rate by one percentage point – induced by a legal news shock that raises the economic cost of default - reduces the expected likelihood of a default event over the next year by 0.46 percentage points. Notably, governments can endogenously affect the economic cost they would face after default by the design of their debt portfolio: They can choose the legal framework under which they borrow, they can choose whether or not to include features such as collective action clauses in the legal design of their bond contracts, and they can choose which creditors to borrow from which can  determine their exposure to heavily litigating plaintiffs. My finding indicates that a shift to more punitive debt portfolios can curb markets’ default expectations independent of the fiscal rule.

Default Penalties Affect Expectations by Eliminating Belief-Driven Debt Crises Equilibria

I further develop these findings in a stochastic general equilibrium model framework where a government faces a classical commitment problem as first formulated in Cole and Kehoe (2000). I extend their model by introducing creditor heterogeneity which determines the default penalty: in my setup, governments can borrow from two types of creditors ‘painful’ and ‘safe’ creditors via notionally identical bonds. A high reliance on ‘painful’ creditors will imply a high default penalty. The government cannot selectively default on only one creditor. In line with my empirical findings, I observe that a higher reliance on ‘painful’ creditors, which implies a larger default penalty, can reduce the probability of default by eliminating belief-driven debt crisis equilibria. The mechanism works as follows. In the standard framework, a belief-driven debt crisis can occur for intermediate debt levels at which the government would prefer to default if it can’t roll over its debt, while it would prefer to not default if it can roll over its debt. At that state, there exist two market beliefs that are supported by a distinct equilibrium each – if creditors belief a crisis will occur, they stop lending, and it indeed becomes optimal for the government to default; if creditors belief no crisis will occur, they keep lending, and the government finds it optimal to repay. In my model extension, an increase in the default penalty via higher reliance on ‘painful’ creditors can eliminate this equilibria multiplicity at some debt levels, as default now becomes so painful, that even if creditors would stop lending, the government would rather repay than default. The increase in the economic cost of default thus functions as a commitment device and can eliminate belief-driven debt crisis expectations. However, governments still face a trade-off. On the one hand, relying more on painful creditors allows governments in my model setting to sustain higher levels of government debt without facing the risk of a belief-driven debt crisis; on the other hand, it is more punishing if a crisis does materialize for fundamental reasons. I find that as long as the government perceives a debt crisis as an off-equilibrium threat, when hit by an MIT shock to output - an unexpected shock that hits an economy at its steady state, leading to a transition path back towards the economy’s steady state – it will always find it optimal to increase debt levels and shift to a more punitive creditor base instead of cutting government spending. If hit by a series of adverse MIT shocks, a shortsighted government not anticipating future shocks to output will converge to a point of high debt and high economic cost of default which makes it very vulnerable to fundamental debt crises.