Econometrics Seminar Series. Sparse High-Dimensional Vector Autoregressive Bootstrap
Published: 23 October 2025
21 November 2025. Professor Stephan Smeekes, Maastricht University
Professor Stephan Smeekes, Maastricht University
"Sparse High-Dimensional Vector Autoregressive Bootstrap"
Friday, 21 November. 16:00 - 17:30
Online
Abstract
("Sparse High-Dimensional Vector Autoregressive Bootstrap" by Stephan Smeekes, Robert Adamek and Ines Wilms)
We introduce a high-dimensional multiplier bootstrap for time series data based on capturing dependence through a sparsely estimated vector autoregressive model. We prove its consistency 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, we derive a Gaussian approximation for the maximum mean of a linear process, which may be of independent interest. Our simulation studies demonstrate the potential of the method in finite samples, even if the data are not generated by a finite-order vector autoregression. We thus establish a general framework for applying accurate bootstrap methods to large systems of time series (or panel data) characterised by temporal and cross-sectional dependence.
Bio
Stephan is Professor of Econometrics in the Department of Quantitative Economics at the School of Business and Economics, Maastricht University. His research focuses on the statistical analysis of time series data - developing methods at the intersection of econometrics, statistics, and data science. He specializes in modeling complex dynamic systems involving multiple time series, with a strong emphasis on uncertainty quantification, often leveraging bootstrap techniques. He also applies these tools to explore causality, trends, forecasting, and risk measurement across a range of fields - from macroeconomics and finance to interdisciplinary research on climate, environmental and medical applications.
For further information, please contact business-seminar-series@glasgow.ac.uk
We foster a positive and productive environment for seminars through our Code of conduct.
First published: 23 October 2025