B0185
Title: Sparse high-dimensional vector autoregressive bootstrap
Authors: Robert Adamek - Aarhus University (Denmark) [presenting]
Ines Wilms - Maastricht University (Netherlands)
Stephan Smeekes - Maastricht University (Netherlands)
Abstract: A high-dimensional multiplier bootstrap is introduced for time series data-based capturing dependence through a sparsely estimated vector autoregressive model. Its consistency is proven for inference on high-dimensional means under two different moment assumptions on the errors, namely sub-gaussian moments and a finite number of absolute moments. In establishing these results, a Gaussian approximation is derived for the maximum mean of a linear process, which may be of independent interest.