Forecasting Inflation by Using the Sub-Groups of both CPI and WPI: Evidence from Auto Regression (AR) and ARIMA Models

The undertaken study is conducted to forecast the inflation of Pakistan for the financial year FY2018-19 using two different time series techniques. In this research, we used consumer price index (CPI) and wholesale price index (WPI) with their sub-groups as inflation indicators for Pakistan. The undertaken research analyzes the proficiency of two important econometrics time series approaches such as Autoregressive (AR) with seasonal dummies, and Autoregressive integrated moving averaged (ARIMA) models by using root mean square (RMSE) criteria. In any economy, inflation and its forecasting are an imperative factor for the fiscal and monetary policies. The study is pertinent, as the forecasted figures of inflation start before the FY2018-19, which helps the policy makers to set the inflation target for FY2018-19. The month-to-month data has been considered for this study for the period from July 2008 to June 2018, and this research is concentrated on forecasting for the year 2018-19. In order to forecast CPI, we use 12 sub-groups and for WPI we use 5 sub-groups in both baskets for the 2007-08 base year. The result of this study reveals that the forecasted value of period average of CPI for the period FY2018-19 is 6.23 percent, however, for WPI is 8.96 percent.