Title: A mixed-frequency Bayesian vector autoregression with a steady-state prior
Authors: Sebastian Ankargren - Uppsala University (Sweden) [presenting]
Yukai Yang - Uppsala University (Sweden)
Mans Unosson - University of Warwick (Sweden)
Abstract: A Bayesian vector autoregressive (VAR) model is proposed which allows for the explicit specification of a prior for the variables' ``steady states''(unconditional means) for data measured at different frequencies, without the need to aggregate data to the lowest common frequency. Using a normal prior for the steady state and a normal-inverse Wishart prior for the dynamics and error covariance, a Gibbs sampler is proposed to sample from the posterior distribution. Moreover, a numerical algorithm for computing the marginal data density is suggested, which can be used to find appropriate values for the necessary hyperparameters. The proposed model is evaluated in an empirical forecasting situation in which Swedish GDP growth is being forecasted. The results generally indicate that the inclusion of monthly data is helpful and improves the accuracy of quarterly forecasts, in particular for shorter horizons.