A Markov Chain Modulated Short-Term Interest Rate Model: Inference on Central Bank Transparency
This paper analyzes a short-term interest rate model with mean and volatility driven by an unobserved Markov chain. There are two main innovations in the approach. Firstly, we allow the state generating process to be endogenous as opposed to most such studies which assume this is an exogenous process. This model captures the dynamics of the short-rate process successfully and conclusively performs better than a GARCH process. Secondly, we use heuristic arguments to determine the relationshi p between predictable policy, i.e. transparency and measures of openness and accountability on th e part of the Central Banks. In this sense this study is unusual in the literature on Markov switching applications to short rate, since we draw economic conclusions rather than simply analyzing econometric properties. Results suggest that the market generally anticipates rate changes one-period-ahead and that the move by Banks towards greater openness and accountability in policy may have contributed to this. Finally, a non-parametric statistic, the concordance measure, indicates the nature of the co- movement in regimes and finds evidence of similar levels and regimes of predictability across Anglophile countries.