Essays on Frictional Financial Markets

This dissertation consists of three essays that uncover the origins of market frictions and their implications for the functioning of the global foreign exchange (FX) market. The first research paper speaks to the hegemony of the US dollar in FX trading. Over 85% of all FX transactions involve the US dollar, despite the United States accounting for less than one quarter of global economic activity. I show both theoretically and empirically that the US dollar dominates FX volumes because FX market participants are strategic about their trading costs. Hence, they avoid directly transacting in non-dollar currency pairs if the expected trading cost is too large. Instead, market participants exchange non-dollar pairs indirectly by using the US dollar as a vehicle currency. That is, market participants first exchange a non-dollar currency into US dollars, and then trade those US dollars for their target currency. I derive a set of theoretical conditions for currency dominance in FX trading volume. To validate these conditions empirically, I use a granular and globally representative FX trade data set. My empirical findings are consistent with the predictions of my theoretical framework and corroborate the importance of strategic behaviour as a novel determinant of currency dominance. Using a novel identification strategy, I show that up to 36-40% of the daily volume in the most liquid dollar currency pairs are due to vehicle currency trading. The second paper studies the information content of trades in the FX market. Specifically, we analyse a novel, comprehensive order flow data set, distinguishing among different groups of market participants and covering a large cross-section of currency pairs. We find compelling evidence that global FX order flows convey superior information heterogeneously across agents, time, and currency pairs. These findings are consistent with theories of asymmetric information and over-the-counter market fragmentation. A trading strategy based on exposure to asymmetric information risk generates high returns even after accounting for risk, transaction cost, and other common risk factors shown in the FX literature. Finally, the third paper analyses the cross-sectional asset pricing implications of liquidity risk in the FX market. Precisely because of its sheer size and despite its decentralised nature, the FX market is commonly known as one of the most liquid and resilient trading venues. However, a clear understanding of whether FX liquidity matters for asset prices is still missing. This paper aims to fill this gap by providing the first systematic study of the pricing implications of FX liquidity risk. We show that, even in this market, exposure to liquidity risk commands a non-trivial risk premium of up to 4% percent per annum. In particular, systematic (marketwide) and idiosyncratic liquidity risk are not subsumed by existing FX risk factors and successfully price the cross-section of currency returns. However, we also find that liquidity and carry trade premia are significantly correlated. The carry trade is a simple trading strategy that aims to profit from the interest rate differential between high- and low-yielding currencies. The correlation between liquidity and carry trade premia lends support to a liquidity-based explanation of the infamous carry trade risk premium. To illustrate this point, we decompose carry trade returns and show that the commonality with liquidity risk stems from periods of high market stress and is confined to the static but not the dynamic carry trade.

This dissertation consists of three essays that uncover the origins of market frictions and their implications for the functioning of the global foreign exchange (FX) market. The first research paper speaks to the hegemony of the US dollar in FX trading. Over 85% of all FX transactions involve the US dollar, despite the United States accounting for less than one quarter of global economic activity. I show both theoretically and empirically that the US dollar dominates FX volumes because FX market participants are strategic about their trading costs. Hence, they avoid directly transacting in non-dollar currency pairs if the expected trading cost is too large. Instead, market participants exchange non-dollar pairs indirectly by using the US dollar as a vehicle currency. That is, market participants first exchange a non-dollar currency into US dollars, and then trade those US dollars for their target currency. I derive a set of theoretical conditions for currency dominance in FX trading volume. To validate these conditions empirically, I use a granular and globally representative FX trade data set. My empirical findings are consistent with the predictions of my theoretical framework and corroborate the importance of strategic behaviour as a novel determinant of currency dominance. Using a novel identification strategy, I show that up to 36-40% of the daily volume in the most liquid dollar currency pairs are due to vehicle currency trading. The second paper studies the information content of trades in the FX market. Specifically, we analyse a novel, comprehensive order flow data set, distinguishing among different groups of market participants and covering a large cross-section of currency pairs. We find compelling evidence that global FX order flows convey superior information heterogeneously across agents, time, and currency pairs. These findings are consistent with theories of asymmetric information and over-the-counter market fragmentation. A trading strategy based on exposure to asymmetric information risk generates high returns even after accounting for risk, transaction cost, and other common risk factors shown in the FX literature. Finally, the third paper analyses the cross-sectional asset pricing implications of liquidity risk in the FX market. Precisely because of its sheer size and despite its decentralised nature, the FX market is commonly known as one of the most liquid and resilient trading venues. However, a clear understanding of whether FX liquidity matters for asset prices is still missing. This paper aims to fill this gap by providing the first systematic study of the pricing implications of FX liquidity risk. We show that, even in this market, exposure to liquidity risk commands a non-trivial risk premium of up to 4% percent per annum. In particular, systematic (marketwide) and idiosyncratic liquidity risk are not subsumed by existing FX risk factors and successfully price the cross-section of currency returns. However, we also find that liquidity and carry trade premia are significantly correlated. The carry trade is a simple trading strategy that aims to profit from the interest rate differential between high- and low-yielding currencies. The correlation between liquidity and carry trade premia lends support to a liquidity-based explanation of the infamous carry trade risk premium. To illustrate this point, we decompose carry trade returns and show that the commonality with liquidity risk stems from periods of high market stress and is confined to the static but not the dynamic carry trade.