Quantile cointegration analysis of the Fisher hypothesis

This paper intends to provide possible explanations for the empirical failure of the Fisher hypothesis in terms of economic shocks by employing the quantile cointegration methodology recently proposed by Xiao (2009). Our empirical results for six OECD countries suggest that though the nominal interest rate and inflation move together in the long run, the cointegrating coefficients between the two variables display an asymmetric pattern depending on the sign and size of the shocks, in sharp contrast to the counterparts with the conventional cointegration methods. In details, in the lower quantiles, the nominal rate is low, and would rise less proportionally than the inflation, leading to the so-called Fisher effect puzzle; by contrast, in the upper quantiles where the level of the nominal rate is high, the former would adjust on a one-to-one basis to changes in the latter, and therefore, support the Fisher hypothesis. Asymmetric monetary policies may be responsible for the findings. Finally, a further checking shows that our findings are robust to the changes of econometric modeling and data frequency.