One-period joint forecasts of Polish inflation, unemployment and interest rate using Bayesian VEC-MSF models

The paper aims at comparing forecast ability of VAR/VEC models witha non-changing covariance matrix and two classes of Bayesian Vector ErrorCorrection – Stochastic Volatility (VEC-SV) models, which combine the VECrepresentation of a VAR structure with stochastic volatility, represented by theMultiplicative Stochastic Factor (MSF) process, the SBEKK form or the MSF-SBEKK specification.Based on macro-data coming from the Polish economy (time series ofunemployment, inflation and interest rates) we evaluate predictive densityfunctions employing of such measures as log predictive density score, continuousrank probability score, energy score, probability integral transform. Eachof them takes account of different feature of the obtained predictive densityfunctions