The intention of this paper is to determine the most efficient risk model which can be implemented in diverse business sectors of an economy. The methodology involved using Value-at-Risk (VaR) technique with the integration of GARCH-based representation on three selected non-financial sectors in Malaysia. Using time-series data from 1993 until 2010, the efficiency test namely the Mean Relative Scaled Bias (MRSB) is then conducted. The evidence showed that the VaR forecast integrated with t-distribution GARCH has better capabilities to track movements in true risk exposures thus suggesting it as the most efficient model within specific assumptions and constraints. Keywords Value-at-risk, efficiency test, mean relative scaled biased (two more keywords)..

4 hits