基于极值分布VaR模型的中国股指期货风险评估

发布时间：2018-07-10 10:32

本文选题：VaR模型 + 股指期货　；参考：《上海师范大学》2013年硕士论文

【摘要】：近二十多年来，世界经济的快速发展，全球一体化特别是经济全球一体化的趋势越来越明显，伴随着世界经济的快速发展，特别是以布雷顿森林体系的崩溃为标志，全球金融市场的稳定性，变得越来越脆弱，金融市场风险一天都不曾消失，如同“达摩克利斯之剑”一样悬在上空。金融风险将影响作为一个整体的世界经济格局，而不是一个单一的金融机构，它的影响，就可能会影响世界经济的稳定发展。金融风险度量已成为一个非常重要的课题，研究它具有十分重要的理论和现实意义。风险测度理论在风险管理实践过程中不断发展，VaR风险度量方法使金融风险的定量研究得到进一步的发展。我国金融市场目前尚不成熟，套期保值工具品种还较为缺乏，伴随着金融市场的快速发展，其所承担的风险也越来越大。股指期货作为基础性风险管理工具，具有价格发现、套期保值的功能，沪深300股指期货合约正式上市，标志着我国股指期货的诞生，它的推出填补了我国股市缺乏针对系统性风险的管理手段这一空白，将改变股票市场缺乏规避系统性风险工具的现状，本文用极值分布VaR模型对股指期货的数据进行了实证研究，分析VaR值对股指期货风险评价研究具有重要意义。 VaR值的准确度量不仅与资产收益率的分布有关，还与资产收益率的波动性有关，，准确估计资产的波动率对于VaR的度量显得十分必要。本文对VaR模型进行了深入研究，包括极值波动下的VaR风险度量及其高频数据波动的VaR风险估计。介绍了极值分布中阈值和特征参数的估算方法，同时用两种方法，即分为广义极值模型和广义Pareto模型，对风险价值VaR作了实证研究，由于正态分布与金融时间序列的分布不符，其收益率具有“尖峰厚尾”的特性，从分布来看，五分钟损失序列的分布显然偏离正态分布，它表现出一个粗大的尾巴，带有负偏态和相对较大的峰度，并呈现波动集聚性。其次，广义帕累托分布是最有用和最实用的极值理论模型的研究成果，在分析金融市场风险时，由于样本数据具有厚尾分布,厚尾分布具有显著的优势。实证研究结果表明，极值理论可以准确度量尾部分布的极端事件的风险价值。
[Abstract]:In the past twenty years, with the rapid development of the world economy, the trend of global integration, especially the global economic integration, has become more and more obvious, accompanied by the rapid development of the world economy, especially marked by the collapse of the Bretton Woods system. The stability of global financial markets has become increasingly fragile, and financial market risks have not disappeared for a day, hanging like the sword of Damocles. Financial risk will affect the world economic pattern as a whole, not a single financial institution. Its influence may affect the stable development of the world economy. The measurement of financial risk has become a very important subject, and it has very important theoretical and practical significance to study it. In the process of risk management practice, the theory of risk measurement develops VaR risk measurement method, which makes the quantitative study of financial risk further developed. The financial market of our country is not mature at present, the variety of hedging tools is still lack, with the rapid development of the financial market, the risk it bears is becoming more and more big. As a basic risk management tool, stock index futures have the functions of price discovery and hedging. The Shanghai and Shenzhen 300 stock index futures contracts are officially listed, which marks the birth of stock index futures in China. Its introduction fills the blank of the lack of systematic risk management means in China's stock market, and will change the present situation of stock market lacking tools to avoid systemic risk. This paper makes an empirical study on the data of stock index futures by using the VaR model of extreme value distribution, and analyzes the importance of VaR value to the risk evaluation of stock index futures. The accuracy of VaR value is not only related to the distribution of return rate of assets, but also to the risk evaluation of stock index futures. It is also related to the volatility of the return on assets. It is necessary to estimate the volatility of assets accurately for the measurement of VaR. In this paper, the VaR model is deeply studied, including the VaR risk measurement under extreme fluctuation and the VaR risk estimation of high frequency data fluctuation. This paper introduces the estimation methods of threshold and characteristic parameters in the extreme value distribution, and uses two methods, namely, the generalized extreme value model and the generalized Pareto model, to make an empirical study on VaR of risk value, because the normal distribution does not agree with the distribution of financial time series. From the distribution, the distribution of the five-minute loss series deviates from the normal distribution obviously. It shows a coarse tail with negative skewness and relatively large kurtosis, and shows a fluctuating agglomeration. Secondly, the generalized Pareto distribution is the most useful and practical research result of extreme value theory model. In the analysis of financial market risk, because of the thick tail distribution in the sample data, the thick tail distribution has a significant advantage. The empirical results show that the extreme value theory can accurately measure the risk value of extreme events with tail distribution.
【学位授予单位】：上海师范大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.51;F224

【参考文献】