Essays on contagion of financial crises

This dissertation covers selected issues on contagion of financial crises. Chapter 1 “Why do financial crises tend to occur together? Some stylized facts on Contagion” presents a set of empirical facts regarding financial contagion analyzing the financial crises of the 1990s: the Mexican crisis, the Asian crisis and the Russian crisis. In particular, the perception that countries may suffer financial crises without having any problems with their fundamentals is empirically assessed. Applying a methodology that uses consistently a definition of contagion to a sample of twelve Latin American countries and eight Asian countries, the following results emerge. First, contagion cases are the exception and not the norm. Real world cases of contagion seem to be of the trigger kind. That is, countries that seem to have experienced contagion had fundamental weaknesses that were exacerbated by crises in other countries. Second in both Mexican and Asian crises contagion had a very strong regional component. In the Mexican crisis only Latin American countries suffered contagion while in the Asian crisis only Asian countries did. The Russian crisis breaks the norm having two Latin American countries, Brazil and Colombia, infected. Third, most countries mentioned as victims of contagion shared at least some common characteristics with the countries where the crises originated. Brazil and Argentina shared with Mexico the fact that they were engaged in exchange rate based stabilizations with structural reforms and that these countries had banking sector problems before the crises started. Brazil and Colombia shared with Russia serious fiscal difficulties. Chapter 2 “Large Investors, Market Power and Contagion” inquires on the implications of having large investors such as mutual funds operating in emerging markets for contagion of financial crisis in the context of an economic model. An important conclusion reached is that large investors allocate more funds where the effect on the prices of the emerging market assets of changing their positions is smaller ceteris paribus, that is to say where their market power effect is smaller in magnitude. Associating positively the magnitude of their market power effect with the degree of illiquidity of emerging markets, the previous statement implies that countries with comparable fundamentals but with more liquid financial markets are the ones more likely to suffer contagion during an international financial crisis episode. More generally, if countries' fundamentals are allowed to differ, countries with weaker fundamentals but with relatively more illiquid markets can be partially “shielded” from contagion. The baseline model is also used to shed some light on the apparent hostility of politicians towards large investors and the substantial capital flow volatility during the 1990s.