In 1993 Engle and Kozicki proposed the notion of common features of which one example is a serial correlation common feature. We say that stationary, non-innovation processes exhibit common serial correlation when there exists at least one linear combination of them which is an innovation. Later on in 1993 Vahid and Engle combined the notions of cointegration among I(1) processes with common serial correlation within their first differences. It is commonly known that cointegrated time series have vector error correction (VEC) representation. The existence of common serial correlation leads to an additional reduced rank restriction imposed on the VEC model’s parameters. This type of restriction was later termed a strong form (SF) reduced rank structure, as opposed to a weak one introduced in 2006 by Hecq, Palm and Urbain.
The main aim of the present paper is to construct the Bayesian vector error correction model with these additional strong form restrictions.
The empirical validity of investigating both the short- and long-run co-movements between macroeconomic time series will be illustrated by the analysis of the price-wage nexus in the Polish economy.
In this paper we present the Bayesian model selection procedure within the class of cointegrated processes. In order to make inference about the cointegration space we use the class of Matrix Angular Central Gaussian distributions. To carry out posterior simulations we use an alorithm based on the collapsed Gibbs sampler. The presented methods are applied to the analysis of the price – wage mechanism in the Polish economy.
The concept of cointegration that enables the proper statistical analysis of long-run comovements between unit root processes has been of great interest to numerous economic investigators since it was introduced. However, investigation of short-run comovement between economic time series seems equally important, especially for economic decision-makers. The concept of common features and based on it the idea of two additional reduced rank structure forms in a VEC model (the strong and the weak one) may be of some help. The strong form reduced rank structure (SF) takes place when at least one linear combination of the first differences of the variables exists, which is white noise. However, when this assumption seems too strong, the weaker case can be considered. The weak form appears when the linear combination of first differences adjusted for long-run efects exists, which is white noise.
The main focus of this paper is a Bayesian analysis of the VEC models involving the weak form of reduced rank restrictions.
After the introduction and discussion of the said Bayesian model, the presented methods will be illustrated by an empirical investigation of the price – wage spiral in the Polish economy.
The main goal of the paper is the Bayesian analysis of weak form polynomial serial correlation common features together with cointegration. In the VEC model the serial correlation common feature leads to an additional reduced rank restriction imposed on the model parameters.
After the introduction and discussion of the model, the methods will be illustrated with an empirical investigation of the price-wage nexus in the Polish economy.
Additionally, consequences of imposing such additional short-run restrictions for permanent-transitory decomposition will be discussed.
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.
This paper investigates the relative importance of cost, demand, financialand monetary shocks in driving real exchange rates in four CEE countries over2000–2018. A two-country New Keynesian open economy model is used as atheoretical framework. In the empirical part, a Bayesian SVAR model withMarkov switching heteroscedasticity is employed. The structural shocks areidentified on the basis of volatility changes and named with reference to the signrestrictions derived from the economic model. Main findings are fourfold. First,real and financial shocks have similar contributions to real exchange variability,whereas that of monetary shocks is small. Second, financial shocks amplifyexchange rate fluctuations stemming from real shocks. Third, even though theexchange rate gaps change over time, they remain quite similar across CEEcountries except for Slovakia. Fourth, Slovakia introduced the euro at the timeof a relatively large real overvaluation, which subsided after a lengthy adjustmentprocess.
en
pl
