1: \begin{abstract}

2: We continue our investigations on the stable-regenerative double-stable model. This is a stationary regularly varying stochastic process, and here we focus on the case where the memory parameter $\beta\in(0,1/2)$, the tail parameter $\alpha\in(0,1)$, and also the process is non-negative. At the microscopic level, the local behavior of extremes of this process has been fully understood, as the convergence to the tail process has been established recently. At the macroscopic level, our first result enhances the recent extremal limit theorem on the running maxima of the model to a convergence of point processes for the clusters of extremes. Most interestingly, our second result reveals that the model exhibits a phase transition at the mesoscopic level in the following sense: if  the block size $d_n$ satisfies $d_n\to\infty, d_n = o(n)$ and $\limn d_n/\log n = \rho\in[0,1]$, then

3: \[

4: \proba\pp{\frac1{b_n}\max_{k=1,\dots,d_n}X_j>y} \sim \vartheta_\rho d_n\proba\pp{\frac {X_1}{b_n}>y} \qmwith \vartheta_\rho = (1-2\rho\beta)\mathfrak q_{F,2},

5: \]

6: where $b_n$ is such that $\proba(X_1>b_n)\sim n\inv$, and  $\mathfrak q_{F,2}$ an explicit constant. The phase transition refers to the fact that the constant $\vartheta_\rho$ now depends on the rate $d_n\uparrow \infty$, a feature that is prohibited by the anticlustering condition that is satisfied by most examples of regularly varying stochastic processes in the literature. In particular,

7: the model has recently been shown to have a distinct candidate extremal index and extremal index, two key parameters in the studies of extremes for regularly varying processes. Here, the constant $\vartheta_\rho$ interpolates between the candidate extremal index (with value $\vartheta_0$) and the extremal index (with value $\vartheta_1$). We are unaware of any other examples of regularly varying stochastic processes with such a feature.

8: \end{abstract}

9: