Exponential GARCH Modelling of the Inflation-Inflation Uncertainty Relationship for Ghana

Abstract

This study examines the asymmetric effects of inflation on inflation uncertainty in Ghana for the period 1963:4 to 2014:2. Exponential Generalized Autoregressive Heteroscedasticity (EGARCH) model is employed on monthly inflation rates to estimate inflation uncertainty. Two complementary approaches are used to determine the empirical relationship between inflation and its uncertainty. In the first approach, inflation dummy is included in the variance equation and in the second, we employ the two-step procedure in which Granger causality test is performed on the monthly inflation rates and the conditional variance generated from the EGARCH model. We find strong support for both Friedman-Ball and Cukierman-Meltzer hypotheses for the full sample as well as the inflation targeting period. Given the current build-up in inflationary pressures in Ghana, our results warn of possible costs of not keeping inflation in check. The major policy implication that follows from this study is that the Bank of Ghana should strive to minimize the gap between actual and target inflation levels so the public will have consistent belief in all announced policy targets.