Investigating the Nonlinear Dynamics of Inflation Rate in Iran

Given the importance of understanding inflation behavior for modeling, forecasting and policymaking, the main question is whether stationary tests for inflation rate in Iran are accurate and correct? An answer to this question depends on studying the inflation dynamics. Overlooking this fact might result in using wrong stationary tests and hence incorrect forecast about inflation movement. The purpose of this paper is to study inflation dynamics in Iran for the period 1991:2-2011:4. Most stationary tests in Iran are based on linear behavior of inflation rate and show that this variable is stationary. This means that the cost of fighting inflation is not high. This paper questions these results and examines the possibility of nonlinear behavior of inflation rate in Iran. The results suggest that inflation follows a nonlinear path and its movement is very well explained by an exponential smooth transition autoregressive (ESTAR) model. Given the nonlinearity of inflation rate in Iran, we apply the nonlinear unit root tests suggested by Kapetanios, Shin and Snell (KSS) to inflation rate. In contrast to most studies in Iran, we show that the inflation rate in Iran is a nonlinear non-stationary variable.  This finding casts doubt on previous researches that considered inflation rate to be a stationary variable. Finally, it seems that shocks might have permanent effects on inflation rate and it would be very costly for monetary authorities to implement anti-inflationary policies in Iran.