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.

Abstract: In the first essay, 'Financial Frictions, Intersectoral Adjustment Costs, and News-Driven Business Cycles', I show that an RBC model with financial frictions and intersectoral adjustment costs can generate sizable boom-bust cycles and plausible responses of stock prices in response to a news shock. Booms in the labor market, which make it possible for both consumption and investment to increase in response to positive news, are caused through two channels: the increases in value of marginal product of labor and the increases in value of collateral. Both of these channels enable firms to hire more workers. Intersectoral adjustment costs contribute to both channels by increasing the relative price of output and capital during expansions. Financial frictions enter in the forms of collateral constraints on firms, which influence the latter channel, and the financial accelerator mechanism driven by agency costs, which amplifies all the key variables. My model differs from previous studies in its ability to generate boom-bust cycles without restricting the functional form of consumption in household preferences and without requiring investment adjustment costs, variable capital utilization, or any nominal rigidities. In the second essay, 'Financial and Real Frictions as Sources of Business Fluctuations', I show that a negative shock to a financial or real friction in an economy can generate quantitatively significant and persistent recessions, even without a decrease in exogenous aggregate total factor productivity in a heterogeneous agents DSGE model. The increase in uncertainty that a firm is facing when it makes capital adjustment, however, is found to have a limited or dubious influence on economic activities. The roles of collateral constaints as a financial friction and nonconvex capital adjustment costs as a real friction in aggregate fluctuations are examined in this propagation mechanism. When these frictions become strengthened, the degree of capital misallocation is intensified, which leads to a drop of endogenous aggregate total factor productivity. As agents expect that the return to investment and endogenous TFP decrease, they reduce aggregate investment sharply, which also leads to a drop in employment. Interruption of efficient resource allocation coming from these two frictions is found out to be enough to generate a large and persistent aggregate flucutations even without introducing heterogeneity in firm-level productivity.

In this thesis I have investigated three aspects of market frictions. Chapter 1 is about financial frictions, i.e. frictional forces prevailing in the financial lending markets and how monitoring and legal fines imposed on banks affect financial fragility. Chapter 2 explores the frictional labor market, i.e. frictional forces that prevent the smooth matching process between employees and employers in labor markets. In this chapter I investigate the sources of fluctuations in labor market volatility. Chapter 3 investigates the asymmetrical information in lending markets and how bankruptcy law could potentially affect this asymmetrical information between a borrower and its lenders. In Chapter 1, I have investigated the implications of legal fines and partial monitoring in a macro-finance model. This primary motivation of this work was the unprecedented level of fines banks faced in recent years. The research in this field is very sparse and this work is one of the few to fill in the void. I have tried investigating the implications of fines and partial monitoring in static and dynamic frameworks. There is partial monitoring in the sense that dubious behavior of intermediaries is not always observed with certainty. Moreover intermediaries can pay some litigation fees to mitigate the punishment for their conduct should they get caught. Several insights can be drawn from introducing such concepts in static and dynamic frameworks. Partial monitoring and legal fines make the incentive constraint of intermediaries more relaxed, in the sense that bankers are required to pledge less collateral to raise fund. This decrease in the asset pledgeability pushes the corporate spread down. In a dynamic set-up due to changes in asset qualities caused by such possibilities, recovery in output and credit become sluggish in response to an adverse financial shock. The dynamic implications of the model for the post-crisis period are investigated. This paper calls for further research to broaden our understandings in how legal settlements interact with banks' behaviors. In Chapter 2 (joint with Elisa Guglielminetti) I have investigated the time-varying property of job creation in the United States. Despite extensive documentation of the US labor market dynamics, evidence on its time-varying volatility is very hard to find. In this work I contribute to the literature by structurally investigating the time-varying volatility of the U.S. labor market. I address this issue through a time-varying parameter VAR (TVP-VAR) with stochastic volatility by identifying four structural shocks through imposing robust restrictions based on a New Keynesian DSGE model with frictional labor markets and a large set of shocks. The main findings are as follows. First, at business cycle frequencies, the lion share of the variance of job creation is explained by cost-push and demand shocks, thus challenging the conventional practice of addressing the labor market volatility puzzle à la Shimer under the assumption that technology shocks are the main driver of fluctuations in hiring. Second, technology shocks had a negative impact on job creation until the beginning of the '90s. This result is reminiscent of the "hours puzzle" à la Gali. In Chapter 3 (joint with Garence Staraci) I provide an additional rationale why creditors include covenants in their contracts. The central claim is that covenants are not only included as a means of shifting the governance from debtors to creditors, but also to potentially address the concerns creditors might have about how the bankruptcy law is practiced. To investigate this claim, I take advantage of the fact that covenants are nullified inside bankruptcy. This fact permits us to show that any change to the bankruptcy law affects the spread through changes that it brings to the contractual structure...

The first chapter explores liquidity risks and analyzes the relationship between funding liquidity and market liquidity theoretically and numerically. Funding liquidity is measured by borrowing constraints of secured loans. Market liquidity is measured by the trading frequency and the ease of traders' negotiations. Borrowing constraints affect the fundamental value of assets traded on the market as well as traders' negotiations. On the other hand, trading efficiency changes borrowing constraints. The above dynamic interactions cause the economy respond persistently to liquidity shocks. Pushing further, the simulated liquidity moments move together and present business cycle property. Moreover, money has an essential role as the medium of exchange in the exchange process that impose a non-trivial monetary policy implication. The second chapter studies the efficiency and default risk of long-term non-recourse loans in bank lending, where limited commitments present. With predetermined terms of the loan, the borrower has incentive to terminate the loan earlier either by prepayment or default. The incentive-compatible loan contract in favor of a lower default risk encourages earlier prepayments, and vice versa. If the loan market is frictional that it takes time for the bank to find another borrower, the interest loss on prepayment becomes severe. Under this condition, allowing default increases the efficiency of lending. Moreover, the bank would allow a higher default rate if the initial market interest rate is lower or the size of the loan is larger. The third chapter studies the effect of recourse law on homeowners' behavior during the residential mortgage foreclosure process, and shows evidence from 7 counties of the Illinois state. We construct the dataset from a loan-level foreclosure and land lien database to capture the individual-level heterogeneity. Although the percentage of deficiency judgment granted is low, we find that the fear of banks' recourse right affects homeowner's bankruptcy and private sale decision during foreclosure. The estimation results show that, if there is no deficiency threat, the bankruptcy chapter 7 claims would be lower by 3%, the bankruptcy chapter 13 claims would go up by 10% and the probability of public sale would be increased by 4%.