非对称随机波动建模及其在金融风险管理中的应用研究

项目名称： 非对称随机波动建模及其在金融风险管理中的应用研究

项目编号： No.71471173

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 管理科学

项目作者： 张波

作者单位： 中国人民大学

项目金额： 60万元

中文摘要： 本项目研究在非对称随机波动条件下的金融风险管理与资产定价。在连续时间框架下研究：（1）非对称随机波动中杠杆参数的精确估计，充分利用股市高低频数据，将杠杆参数、RSV和跳作为解释项，建立波动率预测模型；（2）解决模型离散化偏差以及小样本下参数估计导致的期权定价偏差问题。在离散时间框架下研究：（3）非对称MA-SV、MS-SV和时变SV建模，给出快速有效的分块MCMC算法；（4）基于非参贝叶斯方法的SV建模及VaR计算问题；（5）决策理论框架下，基于贝叶斯方法的SV模型的假设检验。本项目的研究考虑到了多资产配置以及金融时序的变结构问题，能分析处理我国人民币外汇数据、股票高频交易数据以及Shibor隔夜拆借利率等重要金融数据，有助于深刻理解金融市场风险形成规律及其传导机制，合理规避金融风险，为投资者的资产配置与风险管理以及监管部门的宏观监控提供决策支持。

中文关键词： 非对称随机波动；金融高频数据；统计学；金融计量；非参贝叶斯

英文摘要： We study the financial risk management and asset pricing under the condition of asymmetric stochastic volatility. In the framework of continuous time models to study: (1) the accurate estimation of leverage parameter in asymmetric stochastic volatility models; make the most of high and low frequency data of stock market and take the leverage parameter、RSV and jump as explanations to build a volatility forecasting model; (2) the deviation of model discretization and asset pricing due to parameter estimation under small samples. In the framework of discrete time models to study: (3) the modeling of asymmetric MA-SV、MS-SV and time-varying SV models; (4) the modeling of SV model and the calculation of VaR based on nonparametric Bayesian method; (5) the hypotheses testing of SV model based on Bayesian method in the framework of decision theory. The research of this project takes the asset allocation and the structural change of financial time series into consideration, which can be used to analysis and process the important financial data of China foreign exchange rates、high frequency stock market trading data and Shibor overnight rates. Therefore, this proposal is to deeply comprehend the formation rule and transmission mechanism of the financial market risk, which can be reasonably avoided, and provide decision support for the asset allocation and risk management of investors and the macroscopic supervision of regulators.

英文关键词： Asymmetric Stochastic Volatility;High Frequency Financial Data;Statistics;Financial Econometrics;Bayesian nonparametric analysis