Total variation approximations and conditional limit theorems for multivariate regularly varying random walks conditioned on ruin (Q2448699)

The focus of the paper is the development of a precise asymptotic description of the distributions of multidimensional regularly varying random walks, conditioned on hitting a rare target set represented as the union of half spaces. In particular, tractable total variation approximations (in the sample path space), based on change-of-measure techniques, are developed for such conditional stochastic processes. Using these approximations, joint conditional limit theorems for such specific objects as the time until ruin, a Brownian approximation up to (just before) the time of ruin, the ``overshoot'', and the ``undershoot'' are obtained as a corollary. The general strategy is based on the use of a suitable change of measure and later on coupling arguments. The idea is to approximate the conditional distribution of the underlying process step-by-step, in a Markovian way, using a mixture of a large increment that makes the random walk hit the target set and an increment that follows the nominal (original) distribution. The approximation constructed in the paper is tractable in the sense that it is given by a Markovian description which is easy to describe and is explicit in terms of the increment distribution of the associated random walk. It is mentioned that the presented tractability of total variation approximations can be used to study the random walk conditioned on bankruptcy, via efficient Monte Carlo simulation.