Functional convergence of linear processes with heavy-tailed innovations (Q300283)

The authors study the partial sum process of a linear process, especially in the case with heavy-tailed innovations. They give a necessary and sufficient condition for the finite-dimensional convergence of the partial sum process, where the limit is an alpha-stable Lévy motion. To obtain functional convergence, they use the space of càdlàg functions equipped with the \(S\) topology, which is weaker than Skorokhods \(M_1\) or \(J_1\) topology. After discussing convergence in distribution with respect to the \(S\) topology, they state their main result: if the coefficients of a linear process are summable and the distribution of the innovations fulfill some regularity conditions, then the convergence in distribution of the partial sum process with respect to the \(S\) topology follows from the finite-dimensional convergence.