MEASURING MARKET RISK: THE CASE FOR MALAYSIA’S NON-FINANCIAL SECTORS

ABSTRACT

This paper compares two types of volatility models for returns from an empirical point of view in measuring market risk; namely GARCH(1,1) and EGARCH(1,1). The models are estimated on Malaysian non-financial sectors data in which seven sectors are chosen to estimate the relevant parameters. The models are used to obtain daily volatility forecasts and these volatilities are used to estimate the Value-at-Risk (VaR) for each sector based on the Monte Carlo Simulation (MCS) approach. To complement the estimates, several conservatism tests namely the Mean Relative Bias (MRB) and Root Mean Squared Relative Bias (RMSRB) are conducted to identify the riskiest position. Although, t-distribution theoretically is appropriate in handling any reasonable amount of fat tail or asymmetric biases, this research provides evidence that its appropriateness in VaR application depends on the type of volatility model which has been integrated together with MCS and also the nature of the non-financial sectors involved.

Keywords:       Value-at-Risk, confidence level, conservatism.