The research presented in this dissertation has been motivated by the Great Recession that has shown us once again how the financial system can amplify and propagate relatively mild economic shocks into larger scale recessions. In the past decade, we witnessed what many may call the worst crisis since the Great Depression that appears to have resulted from a perverse interaction between real estate markets and borrowers' balance sheets. The three chapters of this dissertation are an attempt to shed light on the mechanisms that may have allowed for such perverse effects to arise. I believe that to understand crisis episodes such as the recent one, it is imperative to study why some economic agents become overly exposed to risk, why financial markets fail to function at times, and what policy makers can do to ameliorate the incidence and repercussions of such adverse events.
In Chapter 1, "A Theory of Balance Sheet Recessions with Informational and Trading Frictions," I propose a novel theory to rationalize the limited risk-sharing that drives balance sheet recessions as a result of informational and trading frictions in financial markets. I show that borrowers and creditors will find it costly to share macroeconomic risk in environments where creditors value the liquidity of financial claims but where information about the future states of the economy is dispersed and the secondary markets for financial claims feature search frictions. As a result, borrowers will optimally choose to retain disproportionate exposures to macroeconomic risk on their balance sheets, and adverse shocks will be amplified through the balance sheet channel. I show that the magnitude of this amplification becomes closely linked to the level of information dispersion and the severity of search frictions in financial markets. In this setting, I study the implications of the theory for macro-prudential regulation and find that subsidizing contingent write-downs of borrowers' liabilities can be welfare improving.
In Chapter 2, "Informed Intermediation over the Cycle," a joint work with Victoria Vanasco, we construct a dynamic model of financial intermediation in which changes in the information held by financial intermediaries generate asymmetric credit cycles as the ones documented by Reinhart and Reinhart (2010). We model financial intermediaries as "expert'' agents who have a unique ability to acquire information about firm fundamentals. While the level of "expertise'' in the economy grows in tandem with information that the "experts'' possess, the gains from intermediation are hindered by informational asymmetries. We find the optimal financial contracts and show that the economy inherits not only the dynamic nature of information flow, but also the interaction of information with the contractual setting. We introduce a cyclical component to information by supposing that the fundamentals about which experts acquire information are stochastic. While persistence of fundamentals is essential for information to be valuable, their randomness acts as an opposing force and diminishes the value of expert learning. Our setting then features economic fluctuations due to waves of "confidence'' in the intermediaries' ability to allocate funds profitably.
In Chapter 3, "Credit Crises, Liquidity Traps, and Demand Externalities," I extend the work of Eggertsson and Krugman (2012) to study welfare implications of households' consumption-saving decisions in New Keynesian economies with incomplete asset markets. My contribution is to show that due to aggregate demand externalities the amount of debt pre-contracted in such economies is generally excessive, and that the amount of "over-borrowing" is increasing in the Central Banker's inflation-aversion. This externality arises because an individual household does not internalize its contribution to the overall fragility as the latter is only a function of aggregate indebtedness of all borrowers. These findings suggest that macro-prudential policies geared towards limiting household leverage can indeed be welfare improving.
