In this paper a new Bayesian approach is proposed to test a point null hypothesis based on the deviance in a decision-theoretical framework. The proposed test statistic may be regarded as the Bayesian version of the likelihood ratio test and appeals in practical applications with three desirable properties. First, it is immune to Jeffreys’ concern about the use of improper priors. Second, it avoids Jeffreys–Lindley’s paradox, Third, it is easy to compute and its threshold value is easily derived, facilitating the implementation in practice. The method is illustrated using some real examples in economics and finance. It is found that the leverage effect is insignificant in an exchange time series and that the Fama–French three-factor model is rejected.
Bayes factor, Decision theory, EM algorithm, Deviance, Markov chain Monte Carlo, Latent variable models
Journal of Econometrics
LI, Yong; ZENG, Tao; and YU, Jun.
A New Approach to Bayesian Hypothesis Testing. (2014). Journal of Econometrics. 178, (3), 602-612. Research Collection School Of Economics.
Available at: http://ink.library.smu.edu.sg/soe_research/1550
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.