Enhancing high-dimensional dynamic conditional angular correlation model based on GARCH family models: Comparative performance analysis for portfolio optimization

In this paper, we present a novel extension of the dynamic conditional angular correlation framework through several influential GARCH family models. This extension aims to enhance the precision in capturing volatility dynamics and broaden their applicability across diverse market conditions. Furthermore, the application of stock portfolio optimization based on the real financial data is conducted to evaluate and compare the estimation performance of dynamic correlation matrices produced by the different extended models. These experiments reveal the significant superiority of the dynamic conditional angular correlation with fractionally integrated GARCH model in markets exhibiting the long-term memory characteristics, effectively capturing persistent volatility.