We propose a Kronecker product structure for large covariance or correlation matrices. One feature of this model is that it scales logarithmically with dimension in the sense that the number of free parameters increases logarithmically with the dimension of the matrix. We propose an estimation...

This paper returns to de Mesnard's paper of 2000, which has exposed the so-called method of bicausative matrices. This method was created to analyze the structural change between two matrices, as an improvement of the causative method of Jackson and al. (1990). In its 2000 paper, de Mesnard has...