Measuring persistence in industrial output: The Indian case

Monthly data on the index of industrial production (IIP) in India, is analysed to check for the long-run effect or persistence of an 'innovation' or a 'shock' on the level of the series. Using (i) an exact maximum likelihood estimation method and constructing the likelihood function through Kalman filtering recursions and (ii) Cochrane's variance ratio technique, we find little long-term persistence in IIP. Our results imply a one percent change now should not affect much the level of IIP over a long horizon.