1: \begin{abstract}

2:   In this paper we extend the existing literature on the asymptotic behaviour of

3:   the partial sums and the sample covariances of long memory stochastic

4:   volatility models in the case of infinite variance. We also consider models

5:   with leverage, for which our results are entirely new in the infinite variance

6:   case. Depending on the interplay between the tail behaviour and the intensity

7:   of dependence, two types of convergence rates and limiting distributions can

8:   arise. In particular, we show that the asymptotic behaviour of partial sums is

9:   the same for both LMSV and models with leverage, whereas there is a crucial

10:   difference when sample covariances are considered.

11: \end{abstract}

12: