Abstract: Despite the discrete nature of monetary policy interest rates and the popularity of regime-switching approach in macroeconomic modeling, voluminous studies typically estimate policy rules using a regression model for a continuous dependent variable such as the Taylor-type rule. This paper demonstrates that a discrete-choice regime-switching version of the Taylor rule significantly outperforms the conventional Taylor-type rules (the classical and extended ones) and existing discrete-choice models both in and out of sample, successfully handles the zero lower bound period by a prolonged switch to a loose policy regime with no-change to the target rate, and delivers markedly different inference. Switching among three policy regimes interpreted as loose, neutral and tight policy stances elegantly addresses such stylized facts as asymmetric policy responses and preponderance of no-change decision by allowing them to be generated in three different regimes. Further, the paper shows that the endogeneity of regressors does matter in modelling monetary policy and can substantially distort the inference.
Link to work