Essays on Information and Financial Frictions in Macroeconomics

Abstract

This dissertation consists of three independent chapters analyzing the role that information and credit frictions play in goods and financial markets. Within these chapters, I develop dynamic stochastic general equilibrium (DSGE) models to study the implications of these frictions on the macroeconomy, both at the national and international level. In the first chapter, I provide a novel explanation for the observed large and persistent fluctuations in real exchange rates using a model with noisy, dispersed information among price-setting firms. Chapter two studies how entrepreneurs' attitudes towards risk affect business cycles in a framework with agency frictions between borrowers and lenders. Finally, chapter three introduces a liquidity channel in a business cycle model with agency frictions to rationalize the highly volatile behavior of default recovery rates observed in the data. Real exchange rates have been extremely volatile and persistent since the end of the Bretton Woods system. For many developed economies, real exchange rates are as volatile as nominal exchange rates, and their fluctuations exhibit a half-life in the range of three to five years. Traditional sticky-price models struggle to jointly account for these features under plausible nominal rigidities (Chari, Kehoe, and McGrattan, 2002). Is it possible to reconcile, in a single framework, the enormous short-term volatility of the real exchange rate with its extremely long half-life? The first chapter of this dissertation addresses this question within a framework in which information is noisy and heterogeneous among price-setting firms. In this context, the continuing uncertainty that firms face about the state of the economy and about the beliefs of their competitors, slows down the price adjustment in response to nominal shocks, generating large and long-lived real exchange rate movements. I estimate the model using real output and output deflator data from the US and the Euro Area and show, as an out-of-sample test, that the model successfully explains the observed volatility and persistence of the Euro/Dollar real exchange rate. In a Bayesian model comparison, I show that the data strongly favor the dispersed information model relative to a sticky-price model à la Calvo. The model also accounts for the persistent effects of monetary shocks on the real exchange rate that I document using a structural vector autoregression. The second chapter, joint with Mikhail Dmitriev, studies how entrepreneurs' attitudes towards risk affect business cycles in a model with agency frictions. Entrepreneurs are inevitably exposed to non-diversified risk, which likely affects their willingness to borrow and to invest in risky projects. Nevertheless, the financial friction literature has paid little attention to how entrepreneurs' desire to take on this risk affects their investment choices in a general-equilibrium setting. Indeed, business cycle models with credit market frictions that feature idiosyncratic risk assume, for tractability, that entrepreneurs are risk neutral (Bernanke, Gertler, and Gilchrist, 1999, BGG). In this chapter, we generalize the BGG framework to the case of entrepreneurs with constant-relative-risk-aversion preferences. In doing so, we overcome the aggregation challenges of this setup and maintain an analytically tractable, log-linear framework. Our main result is that higher risk aversion stabilizes business cycle fluctuations in response to financial shocks, such as wealth redistribution or risk shocks, without significantly affecting the dynamic responses to technology and monetary shocks. Our findings suggest that, within this class of models, the ability of financial shocks to account for a large portion of short-run output fluctuations found in previous work (e.g., Christiano, Motto, and Rostagno (2014)) crucially hinges on borrowers' risk neutrality. The third chapter, joint with Mikhail Dmitriev, examines the implications of the cyclical properties of default recovery rates for aggregate fluctuations. We document that recovery rates after default in the United States are highly volatile and strongly pro-cyclical. These facts are hard to reconcile with the existing financial friction literature. Indeed, models with limited enforceability à la Kiyotaki and Moore (1997) do not feature defaults and recovery rates in equilibrium, while agency costs models following Bernanke, Gertler, and Gilchrist (1999) underestimate the volatility of recovery rates by one order of magnitude. In this chapter, we extend the standard agency costs model allowing liquidation costs for creditors to depend on the tightness of the market for physical capital. Creditors do not have expertise in selling entrepreneurial assets, but when buyers are plentiful, this disadvantage is minimal. Instead when sellers are abundant, the disadvantage of being an outsider is higher. Following a negative shock, entrepreneurs sell capital and liquidation costs for creditors increase, driving down recovery rates. With higher liquidation costs, creditors cut lending and cause entrepreneurs to sell even more capital. This liquidity channel works independently from standard balance sheet effects, and amplifies the impact of financial shocks on output by up to 50 percent.