Monetary policy rules: Some evidence from Markov switching models with time varying parameters

The purpose of my dissertation is to show some evidence of changes in monetary policy during the post-Korean war period. To capture this behavioral evolution, I estimated Taylor-type reaction functions taking into account the possibility that the parameters of that policy function can vary over time. Additionally, the proposed function distinguishes periods in which the “policy errors” show high and low volatility. This may imply discretionary monetary policy or lack of control over the monetary instrument.; My procedure provides a “better fit” to the data than “fixed-coefficient” estimations used in previous studies. My results demonstrate that the period of high volatility of the policy errors includes two sub periods: the aftermath of the break in Bretons Woods agreement and the Volcker disinflation. The former could be related to discretionary policies and the latter with a lack of control in the federal funds rate. Also, the non-linear reaction function captures periods of sharp upward reaction of the FED and downward response, as well, that could not be detected by previous studies based on “fixed-coefficient” reaction functions. My proposed way of estimate monetary policy rules might help to a better understanding of recent monetary history.