Variance Bounds on the Permanent and Transitory Components of Stochastic Discount Factors

Author(s):  
Gurdip S. Bakshi ◽  
Fousseni Chabi-Yo
2019 ◽  
Vol 65 (8) ◽  
pp. 3541-3558 ◽  
Author(s):  
Fousseni Chabi-Yo ◽  
Riccardo Colacito

We propose a new entropy-based correlation measure (coentropy) to evaluate the performance of international asset pricing models. Coentropy captures the codependence of two random variables beyond normality. We document that the coentropy of international stochastic discount factors (SDFs) can be decomposed into a series of entropy-based correlations of permanent and transitory components of the SDFs. We employ the cross section of G-10 countries to obtain model-free estimates of all the components of coentropy at various horizons and we show that the generalization of the long-run risk model featuring two predictable components of consumption growth rates, global disasters, and recursive preferences can account for the composition of codependence at all horizons. This paper was accepted by Tomasz Piskorski, finance.


2021 ◽  
Vol 123 ◽  
pp. 106018
Author(s):  
Nicole Branger ◽  
Michael Herold ◽  
Matthias Muck

2011 ◽  
Vol 62 (1) ◽  
Author(s):  
Benjamin R. Auer

SummaryThis paper analyses whether the consumption based capital asset pricing model is consistent with asset return data from Denmark, Italy, Norway and Austria. The performance of the CCAPM is evaluated by applying the nonparametric methodology of Hansen and Jagannathan (1991) and adopting five alternative specifications of utility. In addition to standard power utility the recursive preferences model proposed by Epstein and Zin (1989) is adopted. Both internal and external habit formation (persistence) using the models proposed by Constantinides (1990), Abel (1990) and Campbell and Cochrane (1999) are also considered. The findings are evaluated using the test of Burnside (1994) and the pricing error measure of Hansen and Jagannathan (1997). It is found that the majority of models produce stochastic discount factors consistent with the data (at mostly low degrees of risk aversion). As far as the pricing errors of the models are concerned the model incorporating recursive utility can be considered best.


Sign in / Sign up

Export Citation Format

Share Document