Essays on Sovereign Default with Unobservable Physical Capital and Debt Relief

Abstract

This thesis provides three novel theoretical frameworks of sovereign default with unobservable physical capital and debt relief. The models are calibrated to the Argentina’s economy and simulate the effect of productivity shocks between 1980 Q1 and 2017 Q4. We show results in terms of bond prices and a borrower’s capital accumulation, debt, consumption, and debt relief, in addition to the default options at the steady-state and on the transition path with productivity shocks.

The first paper is concerned with a model of sovereign default with unobservable physical capital. To our knowledge, this is the first paper in the literature of sovereign default to model default probability with unobservable capital accumulation. It aims to solve the bond premium and perceived equilibrium default with incomplete information about the borrower’s capital accumulation. The model applies to developing countries with undetectable capital accumulation. We derive the bond premium as well as the optimal level of foreign assets. In comparison with the existing literatures, the steady-state of capital and foreign assets from our model is lower and closer to the actual data. This is because the sovereign borrowers have less incentive to accumulate capital for influencing the future bond price. The bond premium is provided from the perceived information only.

In the second paper, we introduce a partial default option with the debt reduction and the probability of re-entry as exogenous variables. The novelty of this paper is the illustration of how sovereign borrowers decide on a partial default. Previous literature only includes absolute default that is sovereign borrowers are not required future repayments. In this model we include an additional option to default that is partial repayment and a shorter stay in debt hangover. These two main variables are exogenously added into the default option. Thus, sovereign borrowers can decide among repayment, absolute default, and partial default. The outcome of the simulations provides the optimal level of debt relief and the length of stay in default for sovereign defaulters. The option of partial default leads to a higher utility for indebted countries. Under the calibration for the Argentina economy, the optimal debt relief in the simulation matches actual data. Results confirm the importance of using partial default as the utility of sovereign borrowers can be significantly higher than with absolute default. More importantly, the results show that sovereign defaulters who receive debt relief experience an increase in both consumption and investment. This outcome is in accordance with the statistics found by Fonchamnyo (2009) and Arslanalp and Henry (2005) that the countries receiving debt relief will spend more on both consumption and capital investment after re-entry into the credit market. On the other hand, our result contradicts the model of Romero-Barrutieta et al. (2015) where their results show only an increase in consumption but lower capital investment after receiving debt relief. Specifically, this is because Romero-Barrutieta et al. (2015) implements the fully observable capital accumulation and the exogenous bond price. Therefore, our model can show an importance of unobservable physical capital which should be considered in the bond price schedule because our results from simulation are in accordance with the stylised facts of developing countries over the past decades; consumption and investment are both increased after receiving debt relief.

Finally, the debt relief and timing of debt restructuring become endogenous variables in the third paper. The results show the dynamic choices of optimal debt relief under different levels of initial capital, foreign assets, and productivity. Countries facing a large negative productivity shock prefer an absolute default regardless of foreign assets. On the contrary, indebted countries with small or moderate productivity shocks optimise their own utilities over debt relief choices. The model simulations capture the behaviour of the Argentina’s economy whose defaults happened for different levels of capital, foreign assets and productivity. The main novelty of this model is the dynamic choice of debt relief and its impact on the sovereign borrowers’ decision. In comparison with the existing literature, our model provides a non-linear choice of debt relief over various level of initial capital, foreign assets, and productivity. Our theoretical framework is flexible enough to embody further development such as heterogenous agents, switching regime of default scheme, and partial default with mixture of bond maturities. This is left for future work.