股指期权定价的实证分析

发布时间：2018-01-11 03:33

本文关键词：股指期权定价的实证分析　出处：《延安大学》2017年硕士论文　论文类型：学位论文

【摘要】：本文首先对上证50ETF期权进行实证分析.经过搜集、整理上市后的最新标的数据,借助Excel工具估计两种基本模型参数并对模型进行实证研究.通过均方误差,分析模型结果与期权实际价之间的差异性,得出:扩展B-S模型精度较高,拟合效果较好;而且该模型下的期权偏差规律与合约规则中的非线性涨跌幅度设定相契合,说明扩展B-S模型更适合上证50ETF期权的定价.其次,基于以上结论对沪深300股指期权展开实证分析.通过搜集、整理最新期标的仿真数据,利用Eviews、MCMC、WinBUGS、Gibbs等多种统计软件和方法进行模型选择、参数估计,并借助BUGS程序语言对模型进行求解.接着,对该期权做关于SV-T(随机波动模型)和B-S模型的实证研究,通过分析误差、套利空间?等因素,得出:SV-T模型对该期权的定价更为有效.最后,基于考虑交易成本和BP过程下的推广B-S模型,对沪深300股指期权展开实证分析.通过选取、处理最新标的数据,估计各参数,利用Matlab求公式解、二叉树和三叉树数值解,得出:泊松分布参数?对模型的结果影响较大;通过分析所得图形,发现三叉树法优于二叉树法,并且两种树图法的数值解和公式解之间的误差随着参数?的增大而增大.本文的结论给出一个较为完整的系统框架,一方面验证了各理论模型的有效性和实用性;另一方面对我国尽快发行沪深300股指期权以及推出各新型期权提供了有效的数据支撑;同时也为证券决策者提供了一定的参考信息.
[Abstract]:This article first carries on the empirical analysis to the Shanghai 50 ETF option. After collecting, collates the latest target data after listing. Using Excel tools to estimate the parameters of the two basic models and empirical research on the model. Through the mean square error, the difference between the model results and the actual option price is analyzed. It is concluded that the extended B-S model has higher precision and better fitting effect. Moreover, the law of option deviation under this model is consistent with the nonlinear fluctuation range in the contract rules, which shows that the extended B-S model is more suitable for the pricing of Shanghai 50ETF options. Secondly. Based on the above conclusions, this paper makes an empirical analysis of Shanghai and Shenzhen 300 stock index options. Through collecting and collating the simulation data of the latest target, we use EviewsMCMCU WinBUGS. Gibbs and other statistical software and methods for model selection, parameter estimation, and with the help of BUGS programming language to solve the model. This paper makes empirical research on SV-T (Stochastic volatility Model) and B-S model. By analyzing the error, arbitrage space? Finally, based on the consideration of transaction costs and BP process, the extended B-S model is used. Through selecting, processing the latest target data, estimating the parameters, using Matlab formula solution, binary tree and triple tree numerical solution. Obtained: Poisson distribution parameters? It is found that the tri-tree method is better than the binary tree method, and the error between the numerical solution and the formula solution of the two tree graph methods is with the parameter. The conclusion of this paper gives a relatively complete system framework, on the one hand, it verifies the validity and practicability of each theoretical model; On the other hand, it provides effective data support for China to issue Shanghai and Shenzhen 300 stock index options and launch new options as soon as possible. At the same time, it also provides certain reference information for the securities decision makers.
【学位授予单位】：延安大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F224;F724.5

【参考文献】