Core measures of WPI inflation and growth in IIP in India: trimmed mean approach

Abstract

This report aims to measure the core inflation in India using trimmed mean approach proposed by Bryan, Cecchetti and Wiggins II (1997) and to find out whether there is any increase in efficiency using this statistical approach compared to the more popular exclusion based approach Non-food Manufacturing WPI. Measuring core inflation is crucial to Central Bank to conduct monetary policy. The core inflation represents the persistent source of inflation whereas Headline inflation, owing to short term fluctuations, supply side shocks and sticky price adjustments, contains noise which needs to be removed. There is no globally accepted definition of core inflation. Smoothing techniques like Centered moving averages, Filtering techniques like Kalman Filter and Exclusion based techniques like ex-food and ex-fuel price averages have been used as such measures historically. This report compares trimmed mean WPI to Non-food manufacturing WPI. The concept of trimmed mean has not been applied to growth in Index of Industrial Production (IIP) so far. This report extends the study to identify a core measure of IIP. The rest of the report is divided into six sections. The next section explains the concept of trimmed means. The third section discusses cases of Turkey and Colombia where trimmed mean approach has been applied. The fourth section describes the statistical characteristic of the price distribution in India. The fifth section discusses various tests to determine optimal trim. Optimal trim has been identified in the sixth section for WPI. In seventh section, the study has been extended for growth in IIP and the last section puts the concluding remarks.