BEKK models are known for being computationally demanding, but it is also complicated to find a way to estimate them in R. We found packages that allow to estimate BEKK variance parameters, but they do not offer any possibility to change the conditional mean parameters. Even more, it is not possible to simply look at which mean model is estimated by default. However, in almost any finance article employing BEKK models, an estimation of the conditional mean of the returns is made, therefore, we feel really confused at this respect.

Thanks to everyone.