Thesis advisor: James E. Anderson / Thesis advisor: Rosen Valchev / The effect of uncertainty on firms' behavior and on the macroeconomy is generally negative in the literature. Extensive research has also demonstrated that financial frictions limit the extent of firms' activities and growth prospects. In the first two chapters of this dissertation, I study both empirically and theoretically how a specific type of uncertainty, exchange rate uncertainty, interacts with financial frictions to affect the behavior of exporting firms. In line with the existing literature, I find in the first chapter that exports of manufacturing sectors in which firms are more financially constrained decrease by more in times of high uncertainty. Having more tangible capital, which can potentially be used as collateral, makes the effect of uncertainty less negative, especially im sectors where firms are large. Relying more on external financing, on the other hand, makes the effect more negative and affects sectors with small firms more. Current theoretical models have little to say about the effect of uncertainty on heterogeneous firms. To address this issue, I introduce in the second chapter a model of financially-constrained heterogeneous exporting firms in which credit conditions depend on the degree of exchange rate uncertainty. Firms in different sectors face different types of financial constraints, and are therefore differently affected by uncertainty. I use the calibrated model to evaluate potential policies that could be implemented to alleviate the negative effect of exchange rate uncertainty on exports. The uncovered interest parity puzzle is the empirical finding that countries with higher risk-free interest rates tend to see their currencies appreciate in the short run. Typical two-country macroeconomic models instead predict that high interest-rate currencies depreciate, with arbitrage opportunities eliminating profitable carry trade strategies. The international finance literature responded to this puzzle by providing several alternative theoretical models able to explain the puzzle. In the third chapter of this dissertation, I study how the predictions of two of these alternative models - the habit model of Verdelhan (2010) and the distorted belief model of Gourinchas and Tornell (2004) - are affected when re-cast in a standard dynamic stochastic general equilibrium framework. I investigate how the mechanisms rely on specific parameter values in order to find under which conditions, if any, they can explain the UIP puzzle. In addition, I obtain business cycle moments from model simulations and compare them to the moments obtained from a standard two-country DSGE model and from the data. My results show that for the first model, the habit model, the UIP results disappear under realistic calibrations. For the second model, the distorted beliefs model, UIP properties remain under some calibrations. In addition, business cycle predictions remain close to empirical evidence. / Thesis (PhD) — Boston College, 2019. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Economics.