Ruling out one explanation of the unhappy industrial production data

by Rudrani Bhattacharya.

Today's data release showed y-o-y growth in the Index of Industrial Production: this was at +2.7% in November when compared with the previous value of 11.3%. Some analysts have conjectured that this was driven by seasonal fluctuations including the placement of Diwali, holidays related to Diwali and the reduced number of working days in November. In the jargon of seasonal adjustment, Diwali is termed a `moving holiday': it's a holiday which shows up in different months in different years.

Our work on seasonal adjustment includes treatment of the overall IIP and some of its components while controlling for the Diwali effect. Even after this adjustment, the seasonally adjusted annualised growth rate for the month is negative (-1.97%) while the average growth over three months including the current month is 0.18%.

IIP consumer goods fared particularly badly (a value of -6.48% for the 3-month average of the annualised point-on-point change of the seasonally adjusted level). Another weak performer was IIP manufacturing (0.017%). However, IIP capital goods shows an encouraging picture with a 3-month average of seasonally adjusted annualised rate of 41.10%.

The year-on-year change is the sum of 12 shocks. Ordinarily it has a lot of inertia. It is, indeed, surprising that the y-o-y change flipped from +11.3% to +2.7% for two consecutive months. There may well be major difficulties in the statistical system which are leading to this. We can be pretty confident that the odd behaviour is not caused by seasonal effects.