用仿真动差函数法来检验长期风险模型

发布时间：2018-03-27 17:14

本文选题：长期风险模型　切入点：股权溢价之谜　出处：《厦门大学》2014年硕士论文

【摘要】：三十多年来在金融经济领域中取得的辉煌成就之一便是基于消费的资本资产定价模型(CCAPM)。但是被寄予厚望的CCAPM却无法解释金融市场中经典的三大未解之谜,如“股权溢价之谜”、“无风险利率之谜”和“股权波动性之谜”。长期风险模型(BansalYaron (2004))是基于要解决三大未解之谜而提出的,而模型的实证结果可以合理、完美地诠释三大谜。但是长期风险模型是否真的可以正确合理地模拟现实世界,至今仍存在很大的争议。本文尝试用仿真动差函数法(EMSM)来检验长期风险模型是否合理。长期风险模型最为核心的模型设定是假设存在一个不可观测的很小的关于长期消费的风险因子xt,并假设它分别与消费增长、分红增长的关系是存在含残差项的线性方程。故本文在假设不可观测变量xt存在的基础上,用EMSM来检验它与消费的增长、分红的增长关系是否显著,并进一步检验长期风险模型的关键方程的内部设定是否合理。
[Abstract]:One of the brilliant achievements in the financial and economic field over the past three decades has been the consumption-based capital asset pricing model (CCAPM). But the CCAPM, which has high hopes, cannot explain the three classic unsolved mysteries in financial markets. For example, "the riddle of equity premium", "the riddle of risk-free interest rate" and "the mystery of equity volatility". The long-term risk model Bansal Yaron / 2004 is based on solving three unsolved riddles, and the empirical results of the model are reasonable. A perfect interpretation of the three mysteries. But can long-term risk models really simulate the real world correctly and reasonably? There is still a lot of controversy. This paper attempts to use the EMSM method to test whether the long-term risk model is reasonable. The core of the long-term risk model is to assume that there is an unobservable and small model. The risk factor for long-term consumption, xt, and assuming that it is separately associated with consumption growth, The relationship between dividend increase and consumption growth is linear equation with residuals. Therefore, on the basis of supposing the existence of non-observable variable xt, this paper uses EMSM to test the relationship between the increase of dividend and consumption, and whether the increase of dividend is significant or not. Furthermore, the internal setting of the key equation of long-term risk model is proved to be reasonable.
【学位授予单位】：厦门大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F224;F830.91

【参考文献】