The Life-Cycle-Permanent-Income Model: A Reinterpretation and Supporting Evidence

The consumption path associated with the life-cycle-optimising version of the permanent- income model is commonly agreed to be a random walk with drift. The persisting failure of the latter to conform to data could, however, raise questions about the suitability of the life- cycle-permanent-income framework within which the random walk model is developed. We propose an alternative interpretation of the permanent-income revision rule which implies consumption follows an ARIMA(1,1,0) with drift. We show that this path can also be derived as a solution to a life-cycle optimising problem with habit formation and precautionary saving motives. U.S. data for 1929-2001 strongly supports the model.