A quantitative model of sovereign debt, bailouts and conditionality

Abstract

International Financial Institutions provide temporary balance-of-payment support contingent on the implementation of specific macroeconomic policies. While several emerging markets repeatedly used conditional assistance, sovereign defaults occurred. This paper develops a dynamic stochastic model of a small open economy with endogenous default risk and endogenous participation rates in bailout programs. Conditionality enters as a constraint on fiscal policy. In a quantitative application to Argentina the model mimics the empirical duration and frequency of bailout programs. In equilibrium, conditional bailouts generate high and volatile interest spreads. A Laffer-curve in conditionality reflects the trade-off between fostering fiscal reform and creating incentives for non-compliance.

Sovereign borrowing, financial assistance, and debt repudiation

“…In the Cole and Kehoe model there exists an interval of debt levels, the crisis zone, for which the government finds it optimal to default only in case it cannot issue new debt because of a run on the sovereign debt market. Our paper is also related to recent work by Boz (2011), Fink andScholl (2011), and Roch and Uhlig (2012) who study bailouts in a model of optimal default. 3 Boz (2011) includes bailout loans supplied by a third party (the IMF) along with market debt held by foreign private investors.…”

“…Total debt of the government increases by 16 percentage points (which is an increase of 50 percent). A rise of equilibrium debt levels is also present in the model by Fink and Scholl (2011), while Boz (2011) finds the opposite effect. In contrast to our results both studies find that the inclusion of financial assistance increases the default probability strongly.…”

“…Boz (2011) and Fink and Scholl (2011) also include financial assistance into a model of sovereign default. However, both studies have a focus different to ours and do not consider the full set of channels through which financial assistance affects the probability of default in our model.…”

“…In contrast to our results both studies find that the inclusion of financial assistance increases the default probability strongly. In Boz (2011) the number of defaults rises drastically from 5.8 to 64.6 per 10,000 quarters and in Fink and Scholl (2011) the default probability increases from 2.88 to 5.00 percent. One important reason for this difference is the presence of defaults due to 24 For all considered parametrizations the lending facility generates positive profits.…”

“…She assumes that IMF credits are repaid for sure and shows that her model is able to reproduce the countercyclical use of IMF loans while market debt is used procyclically. 4 Fink and Scholl (2011) model bailout loans as a grant that is constantly available to the government and associated with a restriction on government spending. Their model is able to mimic the empirical duration and frequency of bailout programs.…”