论文部分内容阅读
本文利用1996年1月5日至2008年9月17日的我国银行间隔日同业拆借利率序列,通过GARCH模型对收益数据中的自相关和异方差现象进行了实证研究,采用MLE方法估计模型参数,再利用所得参数分别计算了不同收益率分布假设下的不同置信水平的VaR值;在此基础上,进行回测检验,比较了各种模型估计效果,并进一步分析了我国同业拆借利率市场的系统性风险历史波动趋势;最后提出了相关结论与政策建议。
In this paper, we use the GARCH model to analyze the autocorrelation and heteroscedasticity of income data from January 5, 1996 to September 17, 2008 in China. The MLE method is used to estimate the model parameters , And then calculate the VaR of different confidence levels under the different return distribution hypothesis by using the obtained parameters respectively. On the basis of this, the back-testing test is carried out to compare the estimated results of various models, and the further analysis of the inter-bank interest rate market Systemic risk history trend; Finally, the relevant conclusions and policy recommendations.