on the use of time-varying vine copulas in multivariate time series analysis

Analyzing and forecasting time-varying multivariate time series is a challenging task because of their long/short-term patterns, and hence, adopting time-varying or dynamic approaches has been proposed in the literature. many of the traditional methods in the literature assume normality and hence a linear relationship for the data because of its simplicity in calculation and understanding, but despite the popularity, it is well known that this assumption is valid only within the gaussian framework. in this context, although the classical copula approach has been widely used, it lacks flexibility in modeling complex high-dimensional dependence, especially in multivariate time series analysis. the better performance of vine copula as opposed to the classical copula in modeling the dependence of high-dimensional data has been pointed out by many authors. in this work we apply vine copulas in modelling multivariate time series data. in this regards, we handle time series data, using a change point detection technique. we illustrate our approach using a simulation study as well as a real data set analysis.