Nonstationarity and nonlinearity in inflation rate: Some further evidence

In this study, we investigate the time series properties of the inflation rate in all countries in Africa. In order to gauge whether the inflation rate is nonstationary or stationary, we employ: (i) two linear unit root tests based on different null hypotheses and (ii) nonlinear ESTAR-type unit root tests based on symmetric and asymmetric adjustments toward PPP. In addition, we use two methods to search for common stochastic trends between each African country's inflation rate and the inflation rate of the United States or the United Kingdom. The results provide vast evidence of nonstationarity and cointegration. ► Inflation data for 35 African countries are nonstationary. ► Non-linear ESTAR-type unit root tests are used. ► We find both symmetric and asymmetric adjustments. ► Linear and nonlinear unit root tests are important to capture individual nuances. ► Evidence of cointegration is shown.