Ination Dynamics and Time-Varying Persistence: The Importance of the Uncertainty Channel

In this article, we employ a time-varying GARCH-type speci?cation to model in?ation and in- vestigate the behaviour of its persistence. Speci?cally, by modelling the in?ation series as AR(1)- APARCH(1,1)-in-mean-level process with breaks, we show that persistence is transmitted from the conditional variance to the conditional mean. Hence, by studying the conditional mean/variance independently, one will obtain a biased estimate of the true degree of persistence. Accordingly, we propose a new measure of time-varying persistence, which not only distinguishes between changes in the dynamics of in?ation and its volatility but also allows for feedback between the two variables. Analysing the in?ation series for a number of countries, we ?nd evidence that in?ation uncertainty plays an important role in shaping expectations, and a higher level of uncertainty increases in?ation persistence. We also consider a number of unit root tests and present the results of a Monte Carlo experiment to investigate the size and power properties of these tests in the presence of breaks in the mean and the variance equation of an AR(1)-APARCH(1,1)-in-mean-level data generating process. The Monte Carlo experiment reveals that if the model is misspeci?ed, then commonly used unit root tests will misclassify in?ation as a nonstationary, rather than a stationary process.