Essays on the microstructure of foreign exchange markets

The essays presented in this dissertation lie in the intersection of international economics and market microstructure theory. The first essay finds that previous market microstructure theory yields unsatisfactory predictions in the context of foreign exchange markets. Furthermore, the essay reveals a puzzle in the literature. The second essay presents a unified model that nests previous models, shows where they are misspecified, and presents new estimates robust to the critique in the first essay. The overriding contribution of the dissertation is that it presents a unified approach to thinking about dealer/specialist pricing that can accommodate the institutional features of each market, and robustly bear out theoretical predictions.; In the first essay, the empirical viability of a class of partial equilibrium microstructure models of market maker price setting behavior is addressed. While predicted asymmetric information effects are generally borne out in the literature, support for predicted inventory effects is less forthcoming. This essay revisits the one study that has found predicted effects. Evidence of structural breaks and model misspecification is presented. Although signed volume and inventory data are shown to have explanatory power, the model does not fit the data. For this class of models empirical results generally do not support model predictions, begging the question of whether there are wider misspecification problems present.; In the second essay, a new model of intraday market making is presented in the context of foreign exchange markets. Previous work suggests that market makers set prices to induce inventory imbalance—compensating trades, but this equates to intentionally buying high and selling low. This paper argues that market makers have multiple instruments with which to manage inventory and asymmetric information. In particular, foreign exchange dealers can call out to others in the over-the-counter foreign exchange market to unload inventory imbalances. This change has implications on both predicted inventory and asymmetric information effects. The model proposed solves the empirical puzzle of rejecting theoretically—predicted market-maker-level inventory effects that is pervasive in both the equity and foreign exchange microstructure literature. It nests previous theoretical work and is supported by the data.