基于分位数决策理论的资产定价研究

发布时间：2018-01-12 17:30

本文关键词：基于分位数决策理论的资产定价研究　出处：《复旦大学》2013年硕士论文　论文类型：学位论文

【摘要】：资产定价的核心是要把未来不确定或者确定的收益通过随机折现因子(SDF)将其折现为现在的价格,基于消费的资产定价(C-CAPM)就是用消费的效用构造出折现因子,而在标准的效用函数的设定下,C-CAPM不能解释有关资产定价的“股权收益率溢价之谜”、“无风险收益率之谜”等问题。因此,一些学者从效用的设定上入手,假设人们对消费的减少更为敏感,且敏感程度超过了边际效用递减带来的程度,从而采用非对称偏好的效用函数,使定价模型与市场数据较为吻合。本文也采用非对称偏好的效用函数,从一个简单的方向上认为人们直接关注消费的下行风险,即消费者关注其效用的分布的某一个分位数,则该分位数值可直接用于衡量消费者对风险的厌恶程度,进而得出分位数效用决策下的资产定价模型。本文第二章提出并讨论了分位数效用函数下的资产定价模型,并对模型的有效性与标准效用模型进行了对比分析。本文第三章分别利用美国消费数据和资本市场数据,采用与分位数回归相结合的广义矩估计方法,对模型的参数进行了估算,进而得到了较为合理的风险厌恶参数和跨期替代弹性参数值。由于中国的市场数据的时期较短,因此本文在增加假设条件的基础上对模型进行调整后,应用中国的消费数据和资本市场数据,同样采用分位数回归和广义矩估计相结合的方法进行了参数估计。并发现国内市场上,风险厌恶参数在不同的时间段有所变化,而EIS却变化不大,以及不同类型的股权投资表现出了不同的风险厌恶参数和EIS参数,对此,本文做了一定的分析解释。本文第四章将模型应用于动态调整中,从而在分位数效用决策框架下对风险溢价的逆周期性、无风险收益率的顺周期性和风险溢价的可预测性做出解释。
[Abstract]:The core of asset pricing is to discount future uncertain or determined returns into the current price through a random discount factor (SDF). C-CAPM-based asset pricing is to construct a discounted factor with the utility of consumption, and under the setting of standard utility function. C-CAPM can not explain the "riddle of equity return premium" and "riddle of risk-free rate of return" about asset pricing. Therefore, some scholars begin with the setting of utility. Assuming that people are more sensitive to the reduction of consumption and the sensitivity is greater than that brought about by diminishing marginal utility, the utility function of asymmetric preference is used to make the pricing model more consistent with the market data. This paper also uses the utility function of asymmetric preference, from a simple direction that people pay attention to the downward risk of consumption directly, that is, consumers pay attention to a certain quantile of its utility distribution. The quantile value can be directly used to measure the risk aversion of consumers, and then the asset pricing model under the quantile utility decision can be obtained. In the second chapter, the asset pricing model under the quantile utility function is proposed and discussed, and the validity of the model is compared with the standard utility model. In the third chapter, the parameters of the model are estimated by using the generalized moment estimation method combined with quantile regression, using the American consumption data and the capital market data, respectively. Then we get more reasonable risk aversion parameter and intertemporal substitution elasticity parameter. Because the period of market data in China is relatively short, this paper adjusts the model on the basis of adding hypothetical conditions. Using Chinese consumption data and capital market data, the method of quantile regression and generalized moment estimation is used to estimate the parameters, and it is found that in the domestic market. Risk aversion parameters change in different time periods, but EIS does not change much, and different types of equity investment show different risk aversion parameters and EIS parameters. This article has made the certain analysis explanation. In chapter 4th, the model is applied to dynamic adjustment to explain the counter-periodicity of risk premium, the procyclicality of risk-free rate of return and the predictability of risk premium under the framework of quantile utility decision.
【学位授予单位】：复旦大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F224;F830.9

【参考文献】