Essays in the econometrics of macroeconomic models
Title: Essays in the econometrics of macroeconomic models
Author: TRYPHONIDES, Andreas
Citation: Florence : European University Institute, 2016
Series/Number: EUI PhD theses; Department of Economics
The thesis has focused on issues related to the use of external information in the identification, estimation and evaluation of Dynamic Stochastic General Equilibrium (DSGE) models, and comprises three papers. The first paper, entitled Improving Inference for Dynamic Economies with Frictions - The role of Qualitative Survey data, proposes a new inferential methodology that is robust to misspecification of the mechanism generating frictions in a dynamic stochastic economy. I derive a characterization of the model economy that provides identifying restrictions on the solution of the model that are consistent with a variety of mechanisms. I show how qualitative survey data can be linked to the expectations of agents and how this link generates an additional informative set of identifying restrictions. Moreover, I show how the framework can be used to formally validate mechanisms that generate frictions. Finally, I apply the methodology to estimate the distortions in the Spanish economy due to financial frictions and derive an optimal robust Taylor rule. The second chapter, entitled Estimation and Inference for Incomplete Structural Models using Auxiliary Density Information considers an alternative method for estimating the parameters of an equilibrium model which does not require the equilibrium decision rules and produces an estimated probability model for the observables. This is done by introducing auxiliary information about the conditional density of the observables, and using density projections. I develop and assess frequentist inference in this framework. I provide the asymptotic theory for parameter estimates for a general set of conditional projection densities and simulation exercises. In the third chapter, entitled Monetary Policy Rules and External Information, I analyze how conclusions about monetary policy stance are altered when we explicitly acknowledge that model concepts like the output gap and inflation are non-observable and we utilize many proxies that are available in the data. I document the effects on Bayesian inference of introducing such proxy information.
Defence date: 30 September 2016; Examining Board: Professor Fabio Canova, EUI, Supervisor; Professor Peter Hansen, The University of North Carolina at Chapel Hill; Professor Giuseppe Ragusa, LUISS; Professor Frank Schorfheide, University of Pennsylvania
Type of Access: embargoedAccess