Title: Comparing the measures of core inflation in India: trimmed mean and structural vector auto-regression approach
Authors: Arundhati Mallick; Narayan Sethi
Addresses: Department of Humanities and Social Sciences, National Institute of Technology (NIT), Rourkela, Odisha 769008, India ' Department of Humanities and Social Sciences, National Institute of Technology (NIT), Rourkela, Odisha 769008, India
Abstract: This paper estimates core inflation through trimmed mean method and structural vector auto-regression (SVAR) method by using monthly data of wholesale price index (WPI) and index of industrial production (IIP) from April 2005 to March 2014 for India. In this paper, we have compared 25% trimmed mean and SVAR method and it was found that SVAR measure provides better results than trimmed mean method. In SVAR method, we found the exact movements of core and non-core shocks in impulse response functions and variance decomposition. It is based on the definitions of core inflation but trimmed mean method excludes the outliers in the price index, whereas SVAR method is difficult to estimate.
Keywords: core inflation; trimmed mean; SVAR; structural vector auto-regression; core shock; non-core shock; wholesale price index; WPI; index of industrial production; IIP; India; impulse response; variance decomposition.
International Journal of Monetary Economics and Finance, 2014 Vol.7 No.4, pp.288 - 301
Received: 30 Aug 2014
Accepted: 21 Nov 2014
Published online: 25 Feb 2015 *