An empirical examination of maximum entropy in copula-based simultaneous equations model

Abstract

© 2017 by the Mathematical Association of Thailand. All rights reserved. In our previous work, copulas were used to connect different marginal distributions in a system of linear equations. And in this study, we aim to further improve the way to join the marginal distributions. We propose entropy copula-based simultaneous equations model, in which the copulas are employed to improve the modeling of joint distribution. Under this model, we do not have to assume a specific distribution for the margins; instead, they are derived from the entropy method. This study will provide a conceptual idea of bivariate entropy copula-based simultaneous equations model and an explanation of how this method works. Then, an empirical experiment will be conducted to explore the performance of our proposed model.