Essays on Capital Flows, Financial Frictions, and Exchange Rates

This dissertation contains three essays on the relationships between exchange rates, capital flows and intermediary frictions, from both empirical and theoretical perspectives. In Chapters 1 and 2, I aim to contribute to the intermediary-based exchange rate literature following the work of Gabaix and Maggiori (2015) by exploring the role of market making entities - and the frictions faced by them - in determining capital flows' exchange rate effects. Despite their essential role as the main intermediaries in real-world currency markets, market makers have thus far been absent in macro-financial exchange rate models. The main justification stems from the notion that these agents bear little end-of-day currency risk resulting from capital flows, making microstructure effects negligible from a macro perspective. In Chapter 1, I utilize novel supervisory datasets from the Bank of Thailand to provide new evidence that greatly conflicts with this established notion. In addition, I also show that the exchange rate effects of international capital flows in relation to these entities are indeed large, persistent, and attributable to both inventory and information microstructure frictions. Motivated by these empirical findings, I propose in Chapter 2 a new macro-microstructure exchange rate model centered on market making frictions, and use the model to examine the heterogeneous effects of different types of capital flows, as well as the merits of policy tools used to address such flows. Finally, Chapter 3 attempts to quantify the magnitude of the original Gabaix and Maggiori (2015)'s FX intermediary friction across a range of small-open economies. To do so, I propose a new estimation strategy for identifying the size of the intermediary friction. I find the estimates of the friction to be economically sizable for several countries, and to co-move with several observable FX market characteristics in an economically sensible manner. These results thus confirm the empirical relevance of the intermediary-based mechanism that has become an important modelling device in both academic and policy circles, but which has thus far received relatively limited empirical verification