Elimination of long term memory in volatility in portfolio optimization: A study via some bivariate copulas (frank and BB6 copula) Available to Purchase

Copulas are useful tools for modeling the dependence structure between two or more variables especially in financial risk modeling. Copulas are becoming a quite flexible tool in modeling dependence among the components of bivariate and mul-tivariate vector, in particular to predict losses in insurance and finance. The study of long term memory elimination via copula in the context of portfolio opimization in not new in literature, for details, see [1] and the references cited therein. However, the search for an appropriate bivariate copula to address this issue has not been explored in terms of several desirable properties, such as, asymptotically independent, Tail-increasing and decreasing property; Markov property that are very important in analyzing fi-nancial data that are quite often spatial-temporal in nature. In this paper, we explore the applicability of two bivariate copula—one from the Archimedean class, namely the Frank copula; and the other from the Gumbel family of copula, namely the BB6 copula for modeling temporal dependence and dependence structure, using a real data from financial domain. We establish all such neces-sary desirable properties (in the appendix) of the two selected copula in connection of eliminating long term memory in portfolio optimization. On the aspect of portfolio optimization and elimination of long term memory and by utilizing a well-known dataset this article heavily draws on the paper by [1] but the major contribution and focus of this article is to provide a framework/layout for selecting an appropriate copula in terms of identifying certain desirable properties.