Working Paper No. 30
By M A S Joyce
This paper estimates UK quarterly inflation uncertainty over 1950-94,conditional on a univariate specification of mean inflation, using a variety of ARCH-related volatility models. To discriminate between these models, we employ the partially non-parametric methodology of Engle and Ng (1993), which focuses on measuring the 'news impact curve'. Our results reject the symmetry restriction imposed in standard ARCH and GARCH models, suggesting that inflation uncertainty is much more sensitive to 'bad news' than 'good news'. Our preferred estimates of the conditional variance of inflation are found to be positively associated with the level of inflation.