Essays on Sovereign Debt and Partial Default

My dissertation involves the study of sovereign debt flows with an emphasis on the determinants and characteristics of sovereign default in emerging markets, accounting for the partial nature of this default. It consists of three chapters: Firstly, Chapter One and Chapter Two empirically explore the determinants of sovereign default. Secondly, Chapter Three quantitatively investigates the macroeconomic implications of the partial nature of sovereign default. In Chapter One and Chapter Two, I empirically examine the monetary and default responses of sovereign countries to the fluctuations in world commodity prices in a panel of 21 emerging countries with annual observations from 1970 to 2013, constructing and using a country-specific commodity price index with time-varying weights. The selection of the sample countries is based on both the definition of an emerging market by the International Monetary Fund (IMF) and the available information of the external public and publicly-guaranteed (PPG) debt arrears in the World Development Indicators (WDI) database. It is commonly known that emerging markets are vulnerable to global shocks, such as the shocks to world commodity prices. Large fluctuations in world commodity prices over the business cycles can greatly affect their foreign revenues, causing excessive external imbalances in international payments. Countries with external imbalance, especially with excessive current account deficits, are most likely to experience limited spending due to constraints on the inter-temporal substitution in expenditure and are likely to default on their external debt denominated in foreign currencies. I choose world commodity prices as the main predictor for the monetary and default responses to investigate, because for many of the countries in the sample, commodities are a large proportion of their export and foreign revenues. Large fluctuations in commodity prices, causing external imbalance, greatly affect their ability to service the external debt, which is typically denominated in foreign currencies. Chapter One, "World commodity prices, money, and foreign exchange in emerging markets: New evidence", investigates countries' monetary responses (the change of broad money and the choice of exchange rate regimes) when they are encountering the foreign revenue reduction caused by the fluctuations in world commodity prices, before making the decision to default. The estimated results show that the declines in world commodity prices significantly, positively affect the ratio of broad money to GDP and that countries tend to have more flexible exchange rate regimes when world commodity prices are depressed for an extended period. The investigation of the response of the broad money supply is consistent with the open economy trilemma and the estimates of the exchange rate regime flexibility fills the gap of the literature on the determinants of the choice of exchange rate regimes. By capturing the type of global shock as well as the time-varying country characteristics, the effect of the price index (excluding the country fixed effect) well explains the time-series variation and country-specific variation of the exchange rate regimes. Chapter Two, "World commodity prices and partial default in emerging markets: An empirical analysis", mainly explores the effects of the fluctuation in world commodity prices on sovereign default. The results show that the decrease in the price index increases the default rate. The response of the default rate varies across countries and it generally increases with a country's dependence on the commodity exports and external indebtedness. This chapter provides the first economically-significant, quantitative estimates of the effect of world commodity prices on the default rate. A few unique features of my approach allow me to make this contribution: The first feature is the price index. In this chapter, the details of developing the novel country-specific commodity price index are given. The price index is constructed with a two-stage aggregation using time-varying weights based on commodities exported and is used as the main explanatory variable. By accounting for the changes in export structure, the time-varying weights allow me to use data for a longer period, from 1970 to 2013, for my analysis, covering the emerging markets debt crisis, currency crisis, and the recent contraction. This country-specific nature of the price index helps control for other common, global shocks in the estimation. The second feature is that I focus on the realized default risk and use the partial default rate, other than default events, country spreads, or credit ratings, as the proxy for that risk. Along with the price index, the default rate provides longer-period data and allows me to do the analysis starting from 1970. In Chapter Three, "Sovereign debt: A quantitative comparative investigation of the partial default mechanism", I build and quantitatively solve the partial default models of a small open economy, in both endowment and production environments, to investigate the responses of the borrowing, default, and pricing of sovereign debt to economic shocks and to examine how the partial default mechanism improves the predictions of the sovereign default models. The simulation results of the models can well predict the country spreads, default-related statistics, and other business cycle indicators. My models assume the non-exclusion from the international capital market after default. Thus, I can also examine the impulse responses of various macroeconomic variables to the shocks to better understand the underlying propagation mechanism of partial default. The un-realistic assumptions and the limited prediction performance of full default models are the two main reasons that motivate me to choose the partial default mechanism and to build the partial default models. Firstly, the standard theory of sovereign default assumes that countries always default on all of their debt and are excluded from the international capital market after default. However, the empirical regularities show that countries always default on only part of their debt and they continue to borrow while having debt arrears. Besides being inaccurate assumptions, the full default model has the limitations in terms of predicting some of the critical debt indicators, like the debt-to-output ratio and the default frequency, although it can predict that default happens in bad times and can predict counter-cyclical country spreads. The partial default models can improve the predictions of the debt level and the default frequency without losing the performance of matching other data moments. Moreover, it also can predict the partial default rate, which the full default model is not designed to and cannot predict. The partial default models in Chapter Three have three features: firstly, the partial default is endogenously-determined, which allows me to compute the default rate; secondly, there is a preemptive recovery payment of the default, which enables the price function of the short-term debt to have the feature that the price of the long-term debt has; and thirdly, there is no exclusion from the international capital market after default, so I can examine the impulse responses of various macroeconomic variables. Compared with the partial default model with endowment, the partial default model with production generates better predictions for the debt service and improves the over-predicted volatilities of consumption and interest spreads. Besides simultaneously matching the mean spread and the debt-to-output ratio, its simulated results can predict the pro-cyclical investment and closely match the relative volatility of investment.