The Inflation Forecast and the Loss Function


Lars E.O Svensson
Princeton University,

First draft: November 2001
Version 1.2: January 2003

In Paul Mizen, ed. (2003), Central Banking, Monetary
Theory and Practice: Essays in Honour of Charles Goodhart, Volume One, Edward Elgar,


This paper argues that inflation-targeting central banks should announce explicit loss function with numerical relative weights on output-gap stabilization and use and announce optimal time-varying instrument-rate paths and corresponding inflation and output-gap forecasts. Simple voting procedures for forming the Monetary Policy Committee’s
aggregate loss function and time-varying interest-rate paths are suggested. Announcing an explicit loss function improves the transparency of inflation targeting and eliminates some misunderstandings of the meaning of “flexible” inflation targeting. Using time-varying instrument-rate paths avoids a number of inconsistencies and other problems inherently associated with constant-interest-rate forecasts.

JEL Classification: E42, E52, E58

Keywords: Inflation targeting, interest rate paths.