Small deviations of general Lévy processes (Q1035870)

Let \(X=(X_t)_{t\in[0,1]}\) be a real--valued Lévy process and let \((\nu,\sigma^2,b)\) be the corresponding Lévy triplet where \(b\in \mathbb R\), \(\sigma^2\geq 0\) and \(\nu\) denotes the generating Lévy measure. The authors describe the behavior of the small deviation function \[ -\log\mathbb P(\sup_{0\leq t\leq 1}|X_t|\leq \varepsilon) \] as \(\varepsilon\to 0\) in terms of quantities derived from \((\nu,\sigma^2,b)\). This is a very surprising result because, to our knowledge, this is the first example where one is able to describe the small deviation behavior for processes of a whole class. The paper contains several examples where the above mentioned quantities are calculated explicitly. Among them are Gamma processes, compound Poisson processes or the class of Lévy processes with non-zero Gaussian component. In the latter case the small deviation behavior coincides with that of the Brownian motion.