Analysing volatility of Colombo consumer price index using GARCH models

Abstract

The objective of this paper is to analyse and modelling the volatility of Colombo Consumer Price Index (CCPI) in Sri Lanka using monthly data from January 2008 to April 2014. Three types of GARCH models (GARCH, TGARCH and EGARCH) were used for this study. Using various specifications for mean equation, study estimated GARCH (1, 1), TGARCH (1, 1) and EGARCH (1, 1) for CCPI. The estimation results reveal that ARMA (1, 0) - EGARCH (1, 1) comes out to be most appropriate specification for modelling CCPI volatility. The study finds that, no evidence of symmetry in the response of CCPI volatility to negative and positive shocks.