Course Catalogue

The lecture will focus on multivariate time series models. After reviewing some fundamental theoretical time series concepts, we will first deal with stable VAR models and their use for forecasting, Granger causality and impulse response analysis. To this end, we will also discuss important issues on asymptotic- and bootstrap-based inference. Afterwards, we discuss integrated multivariate processes, i.e. will we deal with unit root econometrics as well as cointegration. If time permits, we may consider factor models or high-dimensional VARs. The course both addresses asymptotic analyses as well as implementation issues. Accordingly, tutorial sessions are also devoted to coding and empirical problems besides addressing theoretical problems. In the last part of the course, participants introduce or discuss in more details (further) model classes by giving presentations and writing a paper. We may cover e.g. Bayesian VARs, structural VARs, factor-augmented VARs, VARMA models, etc.. This course is complementary to the course Structural Vector Autoregressive Analysis offered by Matthias Meier. While the latter course focus on structural modelling approaches from an applied macro perspective, we take an econometric approach and deal with multivariate I(1) approaches, VARs, VARMA models, etc.. This course is complementary to the course Structural Vector Autoregressive Analysis offered by Matthias Meier. While the latter course focus on structural modelling approaches from an applied macro perspective, we take an econometric approach and deal with multivariate I(1) approaches, VECM and VARMA models in more detail.