宏观经济校准模型的拟合优度研究

发布时间：2018-01-04 08:03

本文关键词：宏观经济校准模型的拟合优度研究　出处：《南京大学》2014年硕士论文　论文类型：学位论文

【摘要】：自从龚刚和Willi Semmler(2003)首次在国内期刊上介绍RBC模型及校准方法之后,我国学者就开始将此方法运用于研究中国经济,并且基本侧重于进行关于中国经济特征事实的实证研究。而对于RBC模型中所使用的校准方法则研究较少,本文通过首次引入新的拟合优度测量工具对RBC模型的校准结果进行测量,试图弥补RBC模型校准结果检验不正规的缺陷、填补国内关于RBC校准结果的检验这方面的研究的空白。本文着重介绍了一种不同于常规的Kydland-Prenscott比率的拟合优度测量工具—一相对平均近似误差(RMSAE)。本文中并没有以假定模型准确模拟实际经济中数据为基础；相反,经济模型被视为是产生实际数据的随机过程的近似,而拟合优度则衡量了近似程度。在计算拟合优度时使用了随机误差这一标准工具。这些测量方法表明应在模型生成数据中引入多少随机误差使其自协方差等于观察数据的自协方差。随后,本文通过建立基本RBC模型来拟合中国经济现实,通过频谱分析、拟合优度测量工具RMSAE来与现实经济进行比较。结合中国实际经济特色以及前人的研究可知,基本RBC模型无法出色地拟合实际经济情况,引入了政府支出冲击之后的RBC模型表现必将更为理想。而模型运行的结果显示,引入了政府支出冲击的RBC模型的RMSAE确实比基本RBC模型小,这证明了RMSAE确实能够表现模型的拟合优度。另外,RMSAE的大小表明,本文模型对于就业的拟合仍然不够理想。这与模型的设定有关,未来的研究可将就业外部性、闲暇的跨期替代性引入RBC模型,以便更好地拟合就业指标。
[Abstract]:Since Gong Gang and Willi Semmlern introduced the RBC model and calibration method in the domestic journals for the first time, Chinese scholars have applied this method to the study of Chinese economy. And basically focus on the empirical study on the facts of China's economic characteristics. But the calibration methods used in the RBC model are less studied. In this paper, the calibration results of RBC model are measured for the first time by introducing a new goodness of fit measurement tool, in order to make up for the defect of RBC model calibration result test. To fill the gaps in the domestic research on the test of RBC calibration results, this paper focuses on the introduction of a good fit measurement tool, which is different from the conventional Kydland-Prenscott ratio. A relative average approximate error (. RMSAE. This paper is not based on the assumption that the model accurately simulates the data in the real economy. Instead, the economic model is seen as an approximation of the stochastic process that produces the actual data. The standard tool of random error is used to calculate the goodness of fit. These measurement methods show how many random errors should be introduced into the data generated by the model so that the self-covariance is equal to the observation. Look at the autocovariance of the data. In this paper, the basic RBC model is established to fit the economic reality of China, and the spectrum analysis is used. Combining with the characteristics of China's real economy and previous studies, the basic RBC model can not fit the actual economic situation well. The performance of RBC model after introducing government expenditure shock will be more ideal, and the result of model operation shows that the RMSAE of RBC model with government expenditure shock is indeed smaller than that of basic RBC model. In addition, the size of RMSAE shows that the model is still not good enough to fit employment, which is related to the setting of the model. The future research can introduce the intertemporal substitution of employment externality and leisure into RBC model in order to fit the employment index better.
【学位授予单位】：南京大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F124;F224

【共引文献】