A generalized beta copula with applications in modeling multivariate long-tailed data

► We propose a new copula class that we call the MGB2 copula. ► It can adapt non-elliptical and asymmetric dependencies using a parsimonious number of parameters. ► It features positive upper tail dependence. ► It includes many useful copulas as special or limiting cases. ► Its practical usefulness is established through data analysis.