An Analysis on Stationarity of Consumer Price Index (CPI) Time Series in Iran and Using ARIMA Model to Forecast For a Future Time Period

This study aims to determine stationarity or non-stationarity of time series data of consumer price index (CPI) in 1980-2012 (i.e. 1359-1391 based on Iranian calendar) time period where 2004 (i.e. 1383 Iranian calendar) is the base year. applying the appropriate methods to detect the trend of time series and determine the autoregressive, moving average and autoregressive moving average of CPI time series autoregressive, moving average and autoregressive moving average of CPI time series, we perform unit root test that results in non-stationarity of CPI time series. using the Augmented Dicky-Fuller test, it is concluded that the time series has only one unit root. therefore, an ARIMA model is developed for the time series data. using Portman-Teau test, the adequacy of the model and its pure randomness is proved. Finally, we use the proposed ARIMA model to forecast future data from 2013 to 2022 (i.e. 1392 to 1401 Iranian calendar) in 80 and 95 percent levels of confidence.