论文部分内容阅读
基于我国2002年到2012年的季度数据,本文使用广义矩估计方法检验了耐用品长期风险模型。模型参数依赖的状态变量不可观测,通过推导把状态变量表示成可以观测的指标价格红利比率,从而可以对模型参数进行估计。研究表明,投资者的风险厌恶系数和跨期替代弹性相互分离,模型能够有效刻画投资者的消费和投资行为特征。此外,模型预测成分对于红利增长率、市场收益率和股权溢价有较强的预测能力,而对消费增长率的预测能力有限。
Based on the quarterly data from 2002 to 2012 in our country, we use the GMM method to test the long-term risk model of durable goods. Model parameters depend on the state variables can not be observed, through the derivation of the state variables can be expressed as observable index price dividend ratio, so that the model parameters can be estimated. The research shows that investors ’risk aversion coefficient and intertemporal substitution elasticity are separated from each other, and the model can effectively characterize the characteristics of investors’ consumption and investment behavior. In addition, the model predictive component has strong predictive power for the dividend growth rate, market return rate and equity premium, while the predictive power for the consumption growth rate is limited.