The Futures Market Risk Research Based On TaR Models

The most commonly method being used is to measure the market risk by VaR value, VaR analysis the risk at the view of loss. While the essay makes the introduction of TaR, which is a new concept. At a certain confidence level, it is a possible interval that the asset or portfolio suffered loss threshold in the coming period. TaR analysis the market risk from the view of time, based on the return interval.The research on the return interval focuses on its probability density function nowadays. This paper studies the return interval in two aspects. The first aspect is the probability density function of the return interval. Existing research mainly concentrated in this part. This paper do empirical analysis on the Shanghai and Shenzhen300stock index futures, the Shanghai and Shenzhen300Index, Zheng cotton index, threaded index, gold index and some stocks’ return interval of logarithm returns. We find that the probability distributions obey power law distribution. Its parameters are not the same, but they have subjection at different thresholds.Based on the return interval, by computing the probability of loss on the Shanghai and Shenzhen300stock index futures, the Shanghai and Shenzhen300Index, Zheng cotton index, the thread index of market, comparative analysis of market risk, we can find that at different thresholds in different futures markets have different risk characteristics, the loss rate of the thread index is higher than Shanghai and Shenzhen300stock index futures market, at a low threshold probability of loss Shanghai and Shenzhen300stock index futures is greater than Zheng cotton index; the second aspect is to establish time series model of the return interval, establishing ARIMA model about the return interval of the Shanghai and Shenzhen300stock index futures, gold index and Zheng cotton index logarithmic yields, and estimate the return interval within the sample, the result shows good. This paper also found that it is with the subject of the return time interval at different thresholds which meet the same type of time series model. The results apply to the Shanghai and Shenzhen300stock index futures, gold index and Zheng cotton index of commodity futures.Finally, using three methods to calculate TaR value for analyzing market risk, methods are:1. the historical data method;2. the probability density distribution of the return interval;3. time series models of the return interval. TaR values of different thresholds in different measure comparing the futures market can analyze risk characteristics of each futures market, TaR value of the Shanghai and Shenzhen300stock index futures greater than Zheng cotton and thread, indicating a lower risk compared to the Shanghai and Shenzhen300stock index futures. At the same paper also compares the advantages and disadvantages of the three methods, the historical data distribution method does not need to consider the distribution issue, and the second method requires a consideration of the distribution, the need for distribution test, but the above two methods are not dynamic, and the third method is dynamic, with good results in the model estimation sample, but the sample forecasting poor results.