Essays in international finance

This dissertation is composed of three essays which address a range of topics in international finance. In the first, we confront three distinct sets of stylized facts which appear anomalous to a world of high financial integration. In spite of the considerable lowering of legal and institutional barriers to cross-border investment of the last twenty years, investors continue to hold disproportionately large claims to domestic output, fund domestic investment mostly out of domestic savings, and consume at very different rates than agents residing abroad. We investigate investment behavior in a world in which agents have superior information regarding domestic returns than those overseas. We show that such a setting provides a unified explanation for these anomalies. In particular, agents prefer to hold domestic equities, yet willingly trade abroad. Their cross-border trading serves to amplify, rather than dampen, cross-border consumption differences and domestic savings-investment correlations. We conclude that information heterogeneity may play an important role in our understanding of international capital markets.;In the second essay, we address the strong partiality towards domestic securities in international investment portfolios. We argue that the preference for local investments is not a uniquely international phenomenon at all. We measure the degree of preference for locally headquartered firms exhibited in domestic portfolios of U.S. investment managers. Distance turns out to be as important as borders in separating and thereby discouraging investors from potential investments. We demonstrate that small, highly levered firms are most likely to be locally held. We claim that investor partiality towards geographically proximate firms accounts for a substantial component of the home bias phenomenon, and may also provide an important clue for cross-sectional asset pricing puzzles.;In the final essay, we derive an optimal tax management strategy for financing of subsidiaries by multinational corporations that takes into account exploitation of tax-loss credits. We develop a formal multi-period dynamic model to characterize the optimal dividend repatriation policy and the optimal choice of debt-equity mix. The model generates several testable empirical implications that are consistent with available evidence and several others that have not been either discussed or empirically tested in the literature.