The Predictability of GARCH-Type Models on the Returns Volatility of Primary Indonesian Exported Agricultural Commodities

Abstract

Agricultural sector plays an important role in Indonesia’s economy; especially for the plantation sub-sector contributing high revenues to Indonesia’sexporting sectors. The primary agricultural commodities in Indonesian export discussed in this study would be Crude Palm Oil (CPO), Natural Rubber TSR20, Arabica Coffee, Robusta Coffee, Cocoa, White Pepper and Black Pepper. Meanwhile, the returns volatility nature of agricultural commodity is famous. The volatility refers to heteroscedasticity nature of the returns which can be modeled by GARCH-type models. The returns volatility can be describe by the residual of the mean equation and volatility of error variances in the previous periods. The aims of this study are to examine the predictability of GARCH-type models on the returns volatility of those seven agricultural commodities and to determine the best GARCH-type models for each commodity based on the traditional symmetric evaluation statistics. The results find that the predictability of ARCH, GARCH, GARCH-M, EGACRH and TGARCH, as type of GARCH models used in this study, are different for each commodity.