Extremes of Lévy driven mixed MA processes with convolution equivalent distributions (Q626294)

Mixed moving average processes of the form \[ Y(t)=\int_{R_{+}\times R}f(r,t-s)d\Lambda(r,s) \] are considered, where \(f\) is a kernel function, \(\Lambda\) is an infinitely divisible, independently scattered random measure with finite dimensional distributions belonging to the class of convolution equivalent distributions and lying in the maximum domain of attraction of a Gumbel distribution. The tail behaviour of the stationary distribution of \(Y(t)\) and \(\sup_{0\leq t\leq h} Y(t)\) are analysed. Extreme behaviour of \(Y\) is described in terms of a marked point process based on maxima of \(Y\) in random intervals.