ABSTRACT
This chapter discusses the three approaches, parametric estimation based on asymmetry generalized autoregressive conditional heteroskedasticity (GARCH) type model and semi-parametric approach based on moments and nonparametric approach based on Kernel function are used to measure the risk in China's futures market so that the risk could be regulated by the Chinese government. Moreover, the empirical results show that for Downside Value at Risk (VaR) and Upside VaR, nonparametric approach exhibits better than the other two approaches at all the three confidence levels. Generally, financial markets consider the downside risk. Short selling mechanism in futures market urges the investors and supervisors concerned not only price fall but also price rise. This section introduces 'Upside VaR', and estimates the Downside VaR and Upside VaR with the above three approaches. The core of parametric VaR estimation is to estimate the volatility of the financial market. Generally speaking, higher volatility implies higher risk.
