The literature investigates the relation between savings and interest rate mainly for household sector, but in recent decades households ceased to be the main source of savings in the economy. We try to identify how the savings of different sectors respond to the interest rate change using the SVAR methodology. We focus on Poland and generalize the results for other European economies. We find that although the household savings rate tends to rise after an increase of interest rate, the corporate savings simultaneously fall, inducing a negative conditional correlation between them. The responses of savings rate of general government and foreign savings are diverse (although the former usually declines after an interest rate increase) and does not seem to be related to factors like the membership in the currency union or the level of public debt. We also check the existence of the ‘crowding-out’ effects and conclude it only applies in the case of government savings crowding out household savings.
The paper aims at comparing forecast ability of VAR/VEC models with a non-changing covariance matrix and two classes of Bayesian Vector Error Correction – Stochastic Volatility (VEC-SV) models, which combine the VEC representation of a VAR structure with stochastic volatility, represented by the Multiplicative Stochastic Factor (MSF) process, the SBEKK form or the MSF-SBEKK specification.
Based on macro-data coming from the Polish economy (time series of unemployment, inflation and interest rates) we evaluate predictive density functions employing of such measures as log predictive density score, continuous rank probability score, energy score, probability integral transform. Each of them takes account of different feature of the obtained predictive density functions.
The aim of the study is to formally compare the explanatory power of Copula-GARCH and MGARCH models. The models are estimated for logarithmic daily rates of return of two exchange rates: EUR/PLN, USD/PLN and stock market indices: SP500, BUX. The analysis is performed within the Bayesian framework. The posterior model probabilities point to AR(1)-tSBEKK(1,1) for the exchange rates and VAR(1)-tCopula-GARCH(1,1) for the stock market indices, as the superior specifications. If the marginal sampling distributions are different in terms of tail thickness, the Copula-GARCH models have higher explanatory power than the MGARCH models.