UCLA

While the impact of agent heterogeneity has long been recognized in the Economicliterature, the link between agent heterogeneity and international asset pricing is yet to be fully understood. In this dissertation I use an overlapping generations framework to study the impact that agent heterogeneity in risk aversion has on the real exchange rate determination. Chapter 1 presents and develops the theoretical model used to study the implications of agent heterogeneity in risk aversion on the real exchange rate. I introduce a twocountry model that features heterogeneous risk aversion profiles for agents, both within and between countries. Furthermore, it is shown that the model can explain the Cyclicality puzzle documented in Backus and Smith (1993), which highlights the empirical disconnect between the exchange rate and relative consumption growth. This chapter also presents the numerical outcome of the model. Chapter 2 explains the quantitative methodology used to code and find the numerical solution of the model presented in chapter 1. The model does not admit a closed form solution and thus the presented outcome relies on the application of Monte Carlo Methods, the Feynman-Kac Theorem and the Piccard Iteration Theorem. Finally, Chapter 3 presents recent empirical evidence on the Cyclicality puzzle between the US and 4 OECD countries: UK, France, Germany and Italy.