非对称金融资产波动预测与建模影响研究（塞拉利昂标准普尔500指数及汇率实证研究）

发布时间：2018-11-12 11:34

【摘要】：建模金融资产收益和回报并进行预测对于资本市场的决策者和参与者追求利润最大化的目标和做出精准判决是至关重要的。尽管在相关的文献中已经出现了对金融资产收益率动态波动的相关的学术文章,但是建模和预测金融资产收益的动态波动性在金融数学,金融计量经济学,金融经济学,应用统计与经济增长中所能够起到的关键作用时至至今,仍然是一个悬而未决的问题。因此,本论文详细阐述并探讨了关于标准普尔500股指日收益率和塞拉利昂月汇率收益的动态波动模型的建模和预测,这其中包括自回归滑动平均模型,广义自回归条件异方差模型(GARCH), Taylor,Schwert广义自回归条件异方差模型(GARCH), Glosten, Jagannathan, Runkle GARCH模型及条件正态分布,t分布的非对称ARCH模型。于此同时,此篇文章也详细论述了由Dickey-Fuller, Phillip-Perron提出的单位根检验,误差纠正模型。同时证明了在简单序约束条件下,平稳APARCH(p)模型的一致收敛性,强相合性和渐进正态性。同时也推导了APARCH模型解析得分表达式。结合极大似然估计推导了GARCH模型的相关性质。将2002年1月1日至2012年至12月31日期间的日标准普尔500指数回报数据,1991年1月1日至2013年12月31日,塞拉利昂月汇率收益数据对上述提及的模型进行拟合,结果表明使用重尾t分布的APARCH模型是最有效和最成功预测模型,能够成功预测日股指回报序列。在重尾t分布下的GJR-GARCH模型能够最有效的预测塞拉利昂月汇率收益率序。误差修正的结果显示,如果长期的均衡出现偏离,则可通过对日股票指数和月度汇率的不平衡性的调整,来达到恢复平衡的效果。最后,研究表明,文中给出的模型能够成功建模日500股票指数和塞拉利昂月汇率波动。同时由于杠杆效应的存在,导致了非对称GARCH模型显示出了非对称的收益率序列。在非常态分布下,非对称(GARCH)和GARCH模型能够更好的拟合并提升整体的对条件方差的估计。鉴于股指和汇率波动的实际含义,此研究将对资本市场的决策者,投资人及行业从业研究人员在经济行业实践中,所涌现出的风险管控和外汇市场波动等领域起到至关重要的作用。
[Abstract]:Modeling and forecasting the returns and returns of financial assets is essential for capital market decision makers and participants to pursue the goal of maximizing profits and making accurate decisions. Although there have been relevant academic articles on the dynamic volatility of financial asset returns in the relevant literature, modeling and predicting the dynamic volatility of financial asset returns in financial mathematics, financial econometrics, financial economics, Application of statistics and economic growth can play a key role until now, is still an outstanding issue. Therefore, this paper discusses the modeling and forecasting of the dynamic volatility model of the daily yield of the S & P 500 stock index and the Sierra Leone monthly exchange rate return, which includes the autoregressive moving average model. Generalized autoregressive conditional heteroscedasticity model (GARCH), Taylor,Schwert generalized autoregressive conditional heteroscedasticity model (GARCH), Glosten, Jagannathan, Runkle GARCH model and asymmetric ARCH model with conditional normal distribution and t distribution. At the same time, this paper also discusses the unit root test and error correction model proposed by Dickey-Fuller, Phillip-Perron in detail. At the same time, the uniform convergence, strong consistency and asymptotic normality of stationary APARCH (p) model are proved under the condition of simple order constraint. At the same time, the analytic score expression of APARCH model is derived. The related properties of GARCH model are derived by using maximum likelihood estimation. Using the daily S & P 500 return data for the period from 1 January 2002 to 31 December 2012 and the Sierra Leone monthly exchange rate gains data from 1 January 1991 to 31 December 2013 to fit the model mentioned above, The results show that the APARCH model with heavy-tailed t distribution is the most effective and successful prediction model and can successfully predict the daily stock index return series. The GJR-GARCH model under the heavy-tailed t distribution can best predict the monthly exchange rate return order in Sierra Leone. The result of error correction shows that if the long-term equilibrium deviates, the effect of restoring equilibrium can be achieved by adjusting the imbalance between the Japanese stock index and the monthly exchange rate. Finally, it is shown that the proposed model can successfully model the daily 500 stock index and the Sierra Leone monthly exchange rate fluctuation. At the same time, because of the existence of leverage effect, the asymmetric GARCH model shows an asymmetric return sequence. Under the abnormal distribution, asymmetric (GARCH) and GARCH models can better estimate the conditional variance of the whole quasi combined lifting system. In view of the actual implications of stock index and exchange rate fluctuations, the study will apply to capital market decision makers, investors and industry researchers in the practice of the economic industry. Emerging risk management and foreign exchange market volatility and other areas play a vital role.
【学位授予单位】：大连理工大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：F830;F224

【相似文献】