Time Series Modeling and Forecasting of CPI of Bangladesh

Time series data is very important in the case of financial development of any economy. But it is very difficult to deal with time series data. This study is consists of time series modeling and forecasting of some economic variables of Bangladesh. For this study we have considered CPI as selected economic variables of Bangladesh. The variable was collected from the “Monthly Economic Trends” published by the Bangladesh Bank. For modeling purpose we have used Box-Jenkins Methodology which is very popularly known in the sector of time series modeling. On the study at first we have checked the stationarity. For checking stationarity we have used both graphical method and Dickey Fuller test. Then to find the appropriate model for CPI we have used the ACF, PACF curves. After finding the appropriate models we have performed the diagnostic checking for each of the fitted model, which was good. Then by using the best fitted model we have find some forecasted values for each of the variables.