文章基本信息

标题：An application of Regular Vine copula in portfolio risk forecasting: evidence from Istanbul stock exchange
本地全文：下载
作者：Novan Nandiwilastio ; Tien R.Muchtadi ; Nugraha E.Suyatma 等
期刊名称：Quantitative Finance and Economics
电子版ISSN：2573-0134
出版年度：2021
卷号：5
期号：3
页码：452-470
DOI：10.3934/QFE.2021020
语种：English
出版社：AIMS Press
摘要：In times of financial turbulence, it is a well-documented fact that the co-movement of financial returns tends to increase leading to unexpected portfolio losses. The magnitude of the losses can severely be underestimated when the characteristics of univariate return series and the dependence structure between the returns are not represented well by a risk forecasting model. From a growing literature on the available multivariate modelling tools, this paper aims to investigate daily portfolio Value at Risk and Expected Shortfall forecasting performance of elliptical as well as Regular Vine copulas. For this purpose, return series of twelve stocks that are listed in Istanbul Stock Exchange are obtained for the period of June 2010 to December 2018. The series are modelled with univariate Generalized Auto-Regressive Conditional Heteroskedasticity models with Normal and Student's t innovations. Equally weighted portfolio returns are forecasted depending on the univariate GARCH marginals and their multivariate dependence structure. Estimated daily portfolio Value at Risk and Expected Shortfall values with varying levels are compared with the traditional Variance-Covariance and Dynamic Conditional Correlation Multivariate GARCH model estimates. While the models performed well at the Value at Risk backtests, according to the applied ES backtests R-vine copula GARCH found better at yielding more accurate Expected Shortfall forecasts.
关键词：copula theory;Regular Vine copula;value at risk;expected shortfall;risk management