Modelling Inflation Rate Volatility in Kenya Using Arch -Type Model Family

Abstract

This paper describe the empirical study based on financial time series modelling with special application to modelling inflation data for Kenya. Specifically  the theory of time series is modelled and  applied  to  the  inflation  data  spanning  from  January  1985  to  April 2016 obtained  from  the  Kenya National Bureau of Statistics.  Three  Autoregressive Conditional  Heteroscedastic  (ARCH)  family  type  models  (traditional  ARCH, Generalized  ARCH  (GARCH),  GJR GARCH and  the  Exponential  GARCH  (EGARCH))  models were fitted and forecast to the data. This was principally because the data were characterized by changing mean and variance. The outcome of the study revealed that the ARCH –family type models, particularly, the EGARCH (1, 1) with generalized error distribution (GED) was the best in modelling and forecasting Kenya’s monthly rates of inflation. The study recommends that governments, policy makers interested  in  modelling  and  forecasting  monthly  rates  of  inflation  should take into consideration  Heteroscedastic models since it  captures the volatilities in the monthly rates of inflation.

Keywords: Inflation, Volatility, GARCH