Abstract
This paper computes several statistical measures for core inflation for India and provides methodology of construction of these core measures. Some of these have been computed for the first time for India, such as: persistence weighted, variations of ‘Neo-Edgeworthian Index’, asymmetric trimmed mean, and month-by‐month exclusion (dynamic trimmed mean) core measures. For computing these core measures, the study uses both aggregate WPI and a detailed breakdown of the WPI. It covers the period April 1994–April 2009, with 1993–1994 as the base year. Subsequently, a comparison of these estimated core measures based on the criteria of usefulness of a measure of core inflation from a monetary policy point of view is carried out. The study finds only some representative measures of core inflation to be useful.
Notes
1. We also analyzed the standard deviation of monthly inflation of seasonally adjusted 69 individual series. From one can see that the most volatile components are the same as the ones that are excluded from the above core indexes.
2. This result is drawn from Bicchal and Sharma (2011). Details regarding computation and results are given in this paper.
3. For calculating standard deviation, we have taken all of the information available for the reference period of the study.
4. The computed weighting schemes of all the reweighted core inflation measures are available on request.
5. We have estimated the model for a large number of different values of g 0 and examined the explanatory power (see equation (8), section 5.3 below) of the resulting core inflation measure in terms of predictability of future inflation. The results suggest that the value of equation (8) increases with smaller g 0 parameter (i.e., with smother series).
6. The unit test results are available on request.
7. The choice of HP filter as a reference trend measure in the RMSE approach is motivated through Vega and Wynne (2003). They argue that the assessment of core inflation in terms of detecting changes in trend inflation should be assessed in real time.
8. These horizons are most relevant from the point of view of monetary policy, which is motivated by usual knowledge about the lags in the monetary policy transmission process in India.