Investigation of multivariate pairs trading under copula approach with mixture distribution

Pairs trading is typically implemented using two assets. The copula approach can allow us to consider the dependency among multiple assets and use multivariate pairs in this strategy. The goal of this article is to investigate this strategy under the copula approach for a group of assets that have mixture distributions. Increasing the consideration of multivariate pairs, especially in the trivariate case, enhances the amount of dependent information. In fact, the results show that multivariate pairs increase trading opportunities. Computational pieces of evidence are brought forward to support the proposed algorithm of this work.