📝 SummarXiv

Abstract

The time-varying Vine Copula model has become a new direction in the Vine Copula class of models due to its time-varying structural parameters. We have observed that the Vine structures of the time-varying Vine Copula model currently used in economics and business research are C-Vine and D-Vine. These two structures are simpler than the R-Vine structure in modeling but will lose more details. Although truncation and simplification of the Vine structure are necessary when the number of variables is large, the number of variables in economics and business research is often small. Therefore, the R-Vine structure is definitely more suitable for constructing time-varying Vine Copula for economic research. Therefore, this paper uses the GAS (Generalized Autoregressive Score) model to dynamically parameterize the R-Vine structure to construct a time-varying Vine Copula model. The application of this model to the study of liquidity risks between China and Southeast Asian countries, including during the pandemic period, reveals that the time-varying model based on the R-Vine structure not only achieves better statistical test results but also better reflects economic and even political realities compared to the other two structural time-varying models.