Dependence analysis of exchange rate and international trade of Thailand: Application of vine copulas

• Main Authors: Chakorn Praprom, Songsak Sriboonchitta
• Format: Book Series
• Published: 2018
• Subjects: Computer Science; Engineering
• Online Access: https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84897834670&origin=inward; http://cmuir.cmu.ac.th/jspui/handle/6653943832/53426
• Institution: Chiang Mai University

This paper aims to investigate the correlation of multivariate dependences between the international trade of Thailand and the USD/THB exchange rate using vine copulas, including canonical (C-vine) and drawable (D-vine) vine copulas which are very flexible dependency structures. Another advantage is that thesemethods overcome limitations and complex dependencymodels. Before we built the paircopula constructions of the vine models, ARMA(1,1)-GARCH(1,1) was adopted to remove time dependence in each of the marginal time series. Furthermore, we got the various standardized residuals to transform into appropriate uniform margins [0,1]. The results can be seen for C-vine case, Gaussian, Rotated Joe, and BB1 which are suitable bivariate copula families for each pair-copula construction. On the other hand, D-vine case, Gaussian, and Rotated Joe are appropriate copula families for the pair-copula construction. In addition, the sequential log-likelihood is quite close to the one obtained by joint maximization; it means that both the vine models are appropriate-fit models. In order to confirm that it is not possible to distinguish between the two models, we employed the Vuong and Clarke tests to verify the suitability of the non-nested model. These tests confirm that the C-vine and Dvine copulas are not distinguishable. It can be concluded that our pair constructions of the time-varying Gaussian copula could be appropriate fits, better than those of the static copula. This study will help policy makers take action to combat the exchange rate volatility. © Springer International Publishing Switzerland 2014.