Bias in Federal Reserve inflation forecasts: Is the Federal Reserve irrational or just cautious?

Inflation forecasts of the Federal Reserve seem to have systematically under-predicted inflation from the fourth quarter of 1968 until Volckerʹs appointment as Chairman, and to systematically over-predict it afterwards until the second quarter of 1998. Furthermore, under quadratic loss, commercial forecasts seem to have information not contained in those forecasts. To investigate the cause of this apparent irrationality, this paper recovers the loss function implied by Federal Reserveʹs inflation forecasts. The results suggest that the cost of having inflation above an implicit time-varying target was larger than the cost of having inflation below it for the period since Volcker, and that the opposite was true for the pre-Volcker era. Once these asymmetries are taken into account, the Federal Reserveʹs inflation forecasts are found to be rational.