Essays on imperfect common knowledge in macroeconomics
Carregando...
Arquivos
Data
2018-05-22
Autores
Orientador(res)
Guimarães, Bernardo de Vasconcellos
Métricas
Título da Revista
ISSN da Revista
Título de Volume
Resumo
This dissertation study the implications of strategic uncertainty induced by imperfect common knowledge for macroeconomic models and economic policy. In the first chapter, I evaluate whether central bank’s transparency enhances the effectiveness of monetary policy. I study this question using a New Keynesian model in which firms do not observe the time-varying inflation target or monetary policy shocks. Two informational assumptions are considered: (i) firms observe the interest rate decisions only (standard assumption) and (ii) firms observe the interest rate and an idiosyncratic signal about the inflation target. Under the standard assumption, agents infer output and inflation fluctuations by realizing that other agents are acting exactly like them. That ceases to be true when agents face strategic uncertainty induced by the idiosyncratic signal. One key implication is that, in the case of a monetary contraction, greater transparency improves the inflation-output trade-off only under the second assumption. In the second chapter, for a general class of DSGE models, I show that whenever agents extract information from endogenous variables that depends directly on the underlying unobserved shock, there is a qualitative difference in the signal extraction from those variables under imperfect information and imperfect common knowledge. This difference in learning about unobserved shocks does not vanish even in the limiting case when the variance of the private signal goes to infinity. Intuitively, strategic uncertainty prevents agents from knowing other agents’ decision, despite that those actions are the same in equilibrium. This discontinuity challenges this benchmark assumption by showing the implicitly substantial knowledge about endogenous variables assumed available to agents under imperfect information. The third chapter develops a novel solution method for a general class of DSGE models with imperfect common knowledge. The main contribution is that the method allows for the inclusion of endogenous state variables into the system of linear rational expectations equations under imperfect common knowledge. One key implication is that the endogenous persistence of state variables is the same under full information and imperfect common knowledge. A primer empirical evaluation of the informational frictions suggests that the model under imperfect common knowledge explains better the expectation data but is relatively worse at explaining the macroeconomic aggregates.