本科毕业论文 我国期货市场的风险分析.docVIP

本科毕业论文  摘要 本文主要研究期货投资中的风险,通过合理的假设, 根据期货交易的原则建立数学模型． 风险价值VaR是度量市场风险的一种普遍使用的工具，因此本文就用VaR的值来评价期货市场的风险．而计算VaR值的方法很多，经过分析比较，本文采用极值理论来进行求解． 在计算位置参数，刻度参数和尾部指数时，考虑到非参数估计方法不需要假设服从某种确切的分布而直接估计尾部指数，而且偏误和均方误较小．于是我们就用Hill法来估计，用已学的C++程序实现；而用极大似然估计法来估计和，用matlab软件实现． 在对风险区间进行估计的问题中，我们采用了一种双标准风险系数判别法，并建立了有关期货市场风险的数学模型．然后我们利用这个模型对期货市场的风险分多种情况进行了讨论． 最后我们根据期货交易的历史数据，用极值理论计算出种期货产品的VaR值，选出其中值相对较小的5种，即为最理想的五种期货产品． 而对于期货价格的预测采用了灰色模型和神经网络模型． ` 关键词：风险价值VaR；广义极值分布（GEV）；极值理论； R/S分析；双标准差别法；灰色模型 Abstract This paper mainly research the risk in futures investment, and establish the model by reasonable assumptions and futures trade of principles. VaR is a tool that is used valuing market risk. In this paper we also use VaR value to evaluate the risk of futures market. Though there are a lot of methods how to calculate VaR, by analyzing and comparing to others, we use pole value theories to solve VaR value in this paper. When calculating position parameter, engraving degree parameterand tail parameter , we consider that the non- parameter estimate method doesnt need to obey a certain distribute accurately, and directly estimate tail parameter, which has less error margin. So we can use the method of Hill to estimate the tail parameter and get the result by c++ program we have learned， and use the most value method to estimate the tail parameterand, and get value of tail parameterand by using Matlab software. In the problem of estimating the risk of the interval, we adopted a kind of pair standard risk coefficients discretion method, and built up the model. This paper used the model of the future market to discuss at the risk of a variety. Finally, according to the history data of futures trades, calculating the VaR of futures product with the Pole value theories, and We elect the five species which is relatively small value, so these is the best of the five futures products. This text has used the pessimistic model and the neural network model to regard the forward price forecast Key