Asymptotic results for heavy-tailed Lévy processes and their exponential functionals (Q1983635)

The author describes the behavior of a heavy-tailed Lévy process \((\xi_t)_{t\geq 0}\) with a negative drift \(\mathbb{E}[\xi_1] = -a < 0\), conditioned to stay positive for a large time. Similarly as in the case of a sum of i.i.d.\ heavy-tailed random variables, the key idea is the \textit{big jump principle}: for large \(t\), conditional on staying positive up to time \(t\), the process \(\xi\) quickly makes a jump to a position larger than \(at\), then essentially follows a regular course. This allows the author, under quite general conditions, to obtain the following asymptotic result: \[ \mathbb{E}\bigg[F\Big(\int_0^te^{-\xi_s}\,ds\Big)\bigg] \sim_{t\to\infty} C_F\cdot\mathbb{P}(\xi_1 > at). \] Similarly as before, the key observation is that sample paths that contribute the most to this expression are those where the Lévy process remains large for a long time. Finally, as an application of this result, the asymptotics of the survival probabilities of a continuous-state branching process in a heavy-tailed Lévy random environment (CBRE) are described.