Title: Central bank interest rate policy: Bayesian analysis using a cross nested AOP model
Authors: Armin Seibert - Augsburg University (Germany) [presenting]
Gernot Mueller - Augsburg University (Germany)
Andrei Sirchenko - National Research University Higher School of Economics (Russia)
Abstract: The decisions of central banks (e.g. ECB) to lower or raise the key interest rate have a high impact on macroeconomic conditions like other interest rates, asset prices and employment. We investigate a model for these interest rates consisting of three (cross nested) autoregressive ordered probit (CNAOP) models considering different economic covariates. The hierarchy of those probit models is intended to mimic different levels of the decision. It can handle the high number of zeros (i.e. most times the interest rate is unchanged) very well. The presence of latent variables prevents it from a maximum likelihood analysis. Therefore we develop a Bayesian estimation algorithm and test it in a simulation study. Finally, we apply the model on real world data.