Research On The Volatility And Risk Measurement Of Copper Price On Shanghai Futures Market

As an important part of the capital market, futures markets were built under the needs of avoiding price risk in spot markets. The first futures market in China was built in 1990s, and copper futures traded on Shanghai Futures Market (SHFE) is one of the most active futures and which also has a large volume of transactions. Since 2003, there has been a greater volatility arising in international copper price. And as the spread of international financial crisis triggered by American subprime mortgage in 2008, the real economy was influenced and meanwhile the futures markets suffered from oscillation. In financial markets, uncertainty means risk. From the perspective of the whole framework of risk measurement, an accurate understanding of the inherent law of price volatility lays the foundation for identifying and measuring the risk, in futures marketsIn the paper, the selected closing price data of copper futures contracts on SHFE is from early 1990 to the end of 2010, with three different sampling frequencies. First, it reviews the works of domestic and foreign scholars in the field of chaotic time series and risk measurement of futures markets. Then, the relevant theories and the models needed in the analysis are introduced. Again, the inherent chaotic characteristics of the price volatility of copper futures contracts is analyzed with the method of correlation dimension, Hurst exponent, surrogate data, the largest Lyapunov exponent and Kolmogorov entropy. Finally, GARCH models are built to identify the risk of copper futures markets and VaR models are built to measure the risk. The VaR values are calculated with the parameters from GARCH models. Also, Kupiec failure frequency method is used to test the accuracy of VaR values.In the whole analysis, the market risk of SHFE is analyzed with the relevant theories and methods of nonlinear chaotic dynamics and econometrics. Besides, various methods of testing chaotic characteristics are first applied to copper futures. According to the empirical results, the price of copper futures has the feature of leptokurtic and heavy tail, and it is a chaotic system which is sensitive to its initial conditions. The effects of prediction of GARCH models built on the basis of the chosen samples depend on the sampling frequency, the concrete model and the distribution assumption. The VaR models, which are used to measure the risk of copper futures on SHFE, are all reasonable under the assumption of Generalized Error Distribution (GED). When daily return series is used as sample to built risk-measurement models, VaR-GARCH models are more suitable than VaR-EGARCH models for the copper futures market. While the weekly return series is used, the result is reverse.