Title: Refined measures of dynamic connectedness based on time-varying parameter vector autoregressions
Authors: David Gabauer - Johannes Kepler University (Austria) [presenting]
Abstract: The dynamic connectedness measures are combined with a time-varying parameter vector autoregressive model (TVP-VAR) with a time-varying variance-covariance structure. This framework allows us to capture possible changes in the underlying structure of the data in a very flexible and robust manner. Specifically, there is neither need to arbitrarily set the rolling-window size nor a loss of observations in the calculation of the dynamic measures of connectedness as no rolling-window analysis is involved. Since this TVP-VAR-based connectedness framework rests on multivariate Kalman filters it is less outlier-sensitive than the originally proposed rolling-window VAR approach. Those merits are illustrated by conducting various Monte Carlo simulations. Moreover, we are investigating the dynamic transmission mechanism of the four most-traded foreign exchange rates by comparing the results of the TVP-VAR model with three different rolling-window VAR models. Finally, we introduce confidence intervals for TVP-VAR-based and rolling-window VAR-based dynamic connectedness measures.