已实现NGARCH模型及应用研究

发布时间：2018-03-11 08:24

本文选题：高频金融数据　切入点：波动率　出处：《重庆理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：近年来,随着电子化交易在金融市场的广泛应用以及信息技术的迅猛发展,金融市场的波动性也日趋激烈。如何更精确地预测资产的收益风险引起了人们的高度重视。估计资产收益的波动率是预测收益风险的关键问题之一,而波动率的估计精度又与模型的假设及数据的采集频率密切相关。一般而言,模型的波动率估计值越精确以及所使用的数据频率越高,波动率的估计精度就越好。因此,以高频金融数据为研究对象,如何建立一个具有统计优良性的波动率模型是本文的主要研究目标。本文的主要研究内容及创新点如下:1.基于NGARCH模型刻画了波动率的杠杆效应特征,本文在已实现GARCH模型的波动率方程中引入参数的扰动,提出了已实现NGARCH模型。在新模型中,引入的参数与误差项序列成负相关关系,使得新息既在大小上对当前收益作出扰动,又在方向上对当前收益作出扰动。2.鉴于模型的参数估计精度直接影响风险预测的准确性,本文采用蒙特卡罗方法对提出的已实现NGARCH模型的参数估计的稳健性进行检验。随机模拟结果显示,在%5的显著性水平下,所有参数估计值的均方误差均显著。同时,当设定模拟次数为500次时,随着样本量的增大,所有参数的估计值依然显著。模拟结果表明,本文提出的已实现NGARCH模型的波动率估计方法有较好的稳健性。3.基于文中提出的已实现NGARCH模型对上证50指数和上证380指数5min频率的高频数据进行了实证分析,并对其风险预测结果进行了比率检验。其次,对已实现NGARCH模型和已实现GARCH模型的风险预测结果进行了比较,结果表明,已实现GARCH模型比已实现NGARCH模型高估了市场风险。本文提出的已实现NGARCH模型,为金融风险管理提供了新的方法,在一定程度上丰富了金融风险管理的理论。
[Abstract]:In recent years, with the wide application of electronic transactions in financial markets and the rapid development of information technology, The volatility of financial markets is becoming more and more intense. How to predict the return risk of assets more accurately has attracted great attention. Estimating the volatility of asset returns is one of the key problems in predicting income risk. The accuracy of volatility estimation is closely related to the assumptions of the model and the frequency of data collection. In general, the more accurate the volatility estimate is and the higher the frequency of the data used, the better the accuracy of volatility estimation. Taking high-frequency financial data as the research object, How to establish a volatility model with statistical excellence is the main research objective of this paper. The main contents and innovations of this paper are as follows: 1. Based on the NGARCH model, the characteristics of volatility leverage are described. In this paper, the parameter perturbation is introduced into the volatility equation of the realized GARCH model, and the realized NGARCH model is proposed. In the new model, the introduced parameters have a negative correlation with the series of error terms, which makes the innovation not only disturb the current income in the magnitude, but also the new model. In view of the fact that the accuracy of parameter estimation of the model directly affects the accuracy of risk prediction, In this paper, Monte-Carlo method is used to test the robustness of the proposed parameter estimation of NGARCH model. The results of random simulation show that the mean square error of all parameter estimates is significant at the significant level of 5. At the same time, When the number of simulations is set to 500 times, with the increase of sample size, the estimated values of all parameters are still significant. The simulation results show that, The volatility estimation method of realized NGARCH model proposed in this paper has good robustness. 3. Based on the realized NGARCH model proposed in this paper, the high frequency data of Shanghai 50 index and Shanghai 380 index 5min frequency are empirically analyzed. The results of risk prediction are compared between the realized NGARCH model and the realized GARCH model, and the results show that, The realized GARCH model overestimates the market risk compared with the realized NGARCH model. The realized NGARCH model proposed in this paper provides a new method for financial risk management and enriches the theory of financial risk management to a certain extent.
【学位授予单位】：重庆理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F224;F832.51

【参考文献】