Soverign lending spreads

Abstract

This thesis studies the determinants of sovereign lending spreads. The objective of the first chapter is to identify and disentangle various risks embodied in foreign currency denominated sovereign bond spreads. Its empirical approach tries to attribute the explanatory power of country fundamentals in a spread equation to their predictive power for default and illiquidity risk. For this, I incorporate rational expectation predictions into the spreads and propose an IV estimation method. The over identification test offers a test whether the spread can be explained by predicted risk probabilities. Applying this approach to developing country bond data from 1975 to 1995, I find that the non-structural explanatory power of fundamentals can be completely attributed to their influence on predicted risk probabilities. The second chapter takes a broader view across all public sovereign lending. Data from the World Bank suggests that the average spread on all forms of borrowing by developing countries is smaller than for top-rated US corporate bonds. After documenting these facts (with particular care for resolving data problems), the analysis looks behind the averages. Once identifying various sub-types of borrowing, I find that official and other private lending (trade-related) are the main source of the low average spreads. Bond and commercial bank lending shows reasonable spreads. Unlike other and official, bond and bank lending move nearly one in one with world interest rates. All types of private lending significantly differ from each other in the way they incorporate country fundamentals. The third chapter offers a potential source of liquidity risk in bond markets: in a Diamond-Dybvig type model, where agents face a risk of becoming more risk-averse early consumers, changes in the speed of public learning about default risk may increase bond spreads, and decrease investor welfare. This effect operates through a link between future price volatility and current prices: increased expected future price volatility leads to lower prices today.