Analysis of Relationship Among V4 Countries and Germany by Their Gross Domestic Products and Copula Models

Abstract

We analyzed quarterly seasonally-filtered time series (OECD link) of GDP in EURO per capita for the V4 countries (Visegrad treaty - Czech republic, Hungary, Poland, Slovakia) and Germany in the period 1996/Q1 – 2018/Q1. First, ARIMA models were used to clear the temporal dependence, then marginal distribution functions were utilized to standardize the data such that it is U(0,1) distributed. Dependence among the 5 random variables was analyzed in terms of correlation strength and joint distribution modeled by elliptical, vine and factor copulas, both in the whole 22 years period and 6-year rolling windows with 3-year overlaps. The choice of such different model classes allows us study the nature of underlying dependence structure from different standpoints.