Modelling Colombo consumer price index: a vector autoregressive approach

Abstract

The consumer price index measures changes in the price level of consumer goods and services purchased by households, which reflect inflation. Various studies have been conducted to modelling and forecasting Consumer Price Index (CPI) by developed countries . However, such studies have not been reported in Sri Lanka. This paper is an attempt to modelling the Colombo Consumer Price Index (CCPI) by using monthly CCPI data from January 2003 to May 2011. For this purpose, Stepwise Regression, Principal Component Analysis and Vector Autoregressive (VAR) approach were used. The VAR model with the first principal component of selected CCPI components was identified the best fitted model for the CCPI series. The model was also tested to an independent data set using CCPI from February 2010 to May 2011.