Sovereign Debt Restructurings

Abstract

Sovereign debt crises involve debt restructurings characterized by a mix of face-value haircuts and maturity extensions. The prevalence of maturity extensions has been hard to reconcile with economic theory. We develop a model of endogenous debt restructuring that captures key facts of sovereign debt and restructuring episodes. While debt dilution pushes for negative maturity extensions, three factors are important in overcoming the effects of dilution and generating maturity extensions upon restructurings: income recovery after default, credit exclusion after restructuring, and regulatory costs of book-value haircuts. We employ dynamic discrete choice methods that allow for smoother decision rules, rendering the problem tractable.

Improving Sovereign Debt Restructurings

“…The probability of continuing to have a "high risk aversion" conditional on being in that state is 0.42. As Dvorkin, Sánchez, Sapriza and Yurdagul (2019) show, this captures well the periods of systemic emerging market risk in the last three decades.…”

“…Following Dvorkin, Sánchez, Sapriza and Yurdagul (2019), we set the Markov transition probability of lenders continuing to have a "low risk aversion" (a = 0), conditional on being in that state, to be 0.88 per year. The probability of continuing to have a "high risk aversion" conditional on being in that state is 0.42.…”

“…In addition, the value of the probability of remaining excluded after restructuring is set at 75 percent, δ = 0.75, also following Dvorkin, Sánchez, Sapriza and Yurdagul (2019).…”

“…Passadore and Xu (2018) introduces bonds with long maturity and debt restructuring in a sovereign default model, but focuses on illiquidity in the secondary markets of sovereign debt and considers both the maturity and the outcomes of debt restructurings to be exogenous. The closest to our framework is the work of Dvorkin, Sánchez, Sapriza and Yurdagul (2019), which explains why debt maturity extensions are usually the output of debt restructurings.…”

“…While the value of the one-period risk free bond seems high at first sight, the low probability of staying in this regime implies that longer-term yields, like for a 10-year risk free bond, will be much lower, and in line with the data for U.S. yields.9 To solve the model numerically, we employ the method inDvorkin, Sánchez, Sapriza and Yurdagul (2019), which entails introducing small shocks into the values of each discrete choice in the model. We follow this paper in terms of the distribution of these shocks and the parameters governing this distribution.…”