In this paper we conduct a seasonal adjustment (SA) of the 2011-12 base series of Index of Industrial Production (IIP). We use the x-13 ARIMA-SEATS iterative process and follow an indirect approach of first identifying seasonality at the product level and then recompile the manufacturing index with seasonally adjusted series. The SA process shows identifiable seasonality in 206/405 (50%) items spread within broad NIC groups of food, beverages, textiles, leather & apparels. Seasonally adjusted levels also provide a smooth and low fluctuation series that can be used for extrapolation in the advance and provisional estimate stage of GDP estimates. However, the SA process reveals several data quality issues of inexplicable outliers, growth rates and changes in pattern of individual items. While seasonal adjustment has advantages, the process pre-supposes pristine data quality and given the trends shown by item level data, both the SA and actual IIP are inadequate in explaining the growth performance of the manufacturing sector.

 

Keywords: Seasonal Adjustment, X-13 ARIMA-SEATS, Index of Industrial Production, Fluctuations, India

 

JEL Classification: C43, C50, P44

 

Acknowledgments

 

We are thankful to Pronab Sen and GC Manna for helpful comments and suggestions.