Eaton–Gersovitz Sovereign Debt Model

The Eaton–Gersovitz sovereign debt model is the founding analytical framework of the modern literature on international lending to governments. Jonathan Eaton and Mark Gersovitz presented it in the 1981 Review of Economic Studies article "Debt with Potential Repudiation: Theoretical and Empirical Analysis". The article starts from a problem that distinguishes sovereign from private borrowing: a government cannot be forced by foreign courts to repay, so any loan is repaid only if the borrower prefers repaying to defaulting. The model treats debt as a voluntary contract in which the borrower's willingness to pay, rather than its ability to pay, is the binding constraint. It belongs to the broader family of enforcement-constrained contracts in business and economics theory and sits close to problems of moral hazard and information asymmetry.

A small open economy receives a stochastic endowment each period and has concave preferences over consumption. It can smooth consumption by borrowing from risk-neutral foreign lenders who have access to a competitive supply of funds at the world interest rate r. Debt is non-contingent: a one-period bond promises a fixed repayment, and the borrower decides each period whether to honor it. The crucial assumption is that no court can seize the borrower's assets, so the only penalty for repudiation is an endogenous loss of reputation: lenders will not extend new loans to a country that has defaulted.

The government compares two continuation values. If it repays, it keeps access to credit markets and can continue smoothing shocks. If it defaults, it avoids the repayment but is excluded from borrowing, and must consume its endowment in each future period. Because an endowment that fluctuates is costly for a risk-averse agent, the value of market access is positive, and this value is what supports repayment. In the original paper the exclusion is permanent; later work adds probabilistic re-entry.

A central result is that the lender faces a quantity constraint, not only a price. For a given debt level, there is a threshold endowment below which the government defaults. A higher promised repayment raises the default threshold and thus the probability of default. Competitive lenders offer a loan whose face value compensates them for that default risk, so the interest rate rises with the loan size. Beyond some debt level, a larger promised repayment lowers the expected repayment because default becomes so likely that the return on the loan falls. The expected-repayment schedule is therefore hump-shaped, and lenders will not lend beyond its peak. This is a credit ceiling: the country is rationed even though it would accept a higher interest rate, a feature reminiscent of the equilibrium rationing studied by Stiglitz and Weiss under adverse selection.

The ceiling depends on the country's volatility, its patience, and the world interest rate. A country with a more variable endowment values access to credit more and can therefore sustain more debt, while an impatient government discounts the future cost of exclusion and has a lower ceiling. Eaton and Gersovitz also noted that default occurs in equilibrium only in states of the world in which the endowment is low, so defaults cluster in bad times, which is consistent with the historical record of sovereign crises during global downturns.

The punishment is a credible equilibrium outcome: if each lender expects every other lender to refuse to lend to a defaulter, it is individually rational to refuse too. In the language of game theory, the exclusion is a trigger strategy supported as a subgame-perfect Nash equilibrium of the repeated lending game. Importantly, the borrower's reputation is not about its type, since the model has no hidden information. It is about the shared expectation of a market-wide boycott. The mechanism resembles the repeated-game cooperation models behind collusion and relational contracting, and it implies that the debt contract is self-enforcing.

The framework yields several empirical predictions that Eaton and Gersovitz examined with data from developing countries: that debt levels should be limited by a ceiling rather than by the demand for loans, that default risk should increase with the debt-to-income ratio, and that borrowers with higher income volatility might hold more debt. These predictions have been tested repeatedly with mixed results, and the broader finding that sovereign spreads depend on the cycle has been robust.

For about two decades the model remained a stylized theoretical benchmark. Two papers turned it into a quantitative tool for studying emerging-market business cycles.

• Aguiar and Gopinath (2006) embedded the Eaton–Gersovitz structure in a dynamic stochastic model with a finite exclusion period, so that the country regains access after default. They show that the model can generate default at a plausible debt ratio and that, with shocks to trend growth, the model reproduces the high volatility of consumption relative to output in emerging markets.
• Arellano (2008) used a one-period, non-contingent bond, endogenous default, and an output cost of default (a loss of income on the default date) to generate default episodes in which spreads rise sharply before default and consumption falls with output. The model produced a spread that is strongly countercyclical, and a default policy in which the government defaults when income is low and debt is high. Her calibration to Argentina reproduced the cyclical behavior of the trade balance and spread.

Subsequent work added long-maturity debt, which introduces dilution and debt-overhang-like inefficiencies, renegotiation after default, and self-fulfilling rollover crises. These extensions show that when debt is long-term, a government that issues new bonds reduces the value of old claims, so equilibrium spreads reflect expectations of future borrowing.

Critics note that permanent exclusion is not observed: defaulting countries typically regain market access within a few years. Furthermore, a government that is cut off from lending can still save, and a government could in principle use savings to self-insure. This objection leads directly to the argument of the Bulow–Rogoff sovereign debt model, which shows that reputation of this type cannot support lending if the defaulter can buy insurance-like contracts on a cash-in-advance basis. The empirical literature also finds that the costs of default, such as trade sanctions and disruption of banking systems, matter beyond the loss of borrowing. The Eaton–Gersovitz framework is still the standard workhorse because it delivers a tractable equilibrium default problem: the government chooses debt and default to maximize utility, lenders price default risk, and the equilibrium is a fixed point between the two. It also connects to the broader question of designing contracts under limited commitment, a topic taken up in mechanism design.

1. Eaton, J., & Gersovitz, M. (1981). Debt with Potential Repudiation: Theoretical and Empirical Analysis. The Review of Economic Studies, 48(2), 289–309. https://doi.org/10.2307/2296886
2. Arellano, C. (2008). Default Risk and Income Fluctuations in Emerging Economies. American Economic Review, 98(3), 690–712. https://doi.org/10.1257/aer.98.3.690
3. Aguiar, M., & Gopinath, G. (2006). Defaultable Debt, Interest Rates and the Current Account. Journal of International Economics, 69(1), 64–83. https://doi.org/10.1016/j.jinteco.2005.05.005
4. Jeremy I. Bulow, Kenneth Rogoff (1986). "A Constant Recontracting Model of Sovereign Debt." Journal of Political Economy. https://doi.org/10.1086/261596