Abstract
A previsão da volatilidade dos retornos de ativos financeiros é uma abordagem desafiadora e tem atraído a atenção de participantes do mercado, reguladores e acadêmicos nos anos recentes. Este artigo propõe um modelo GJR-GARCH nebuloso para a previsão da volatilidade dos índices S&P 500 e Ibovespa. O modelo combina os conceitos de sistemas de inferência nebulosos e a abordagem GJR-GARCH para considerar os príncipios de variância condicional, efeitos de alavancagem e agrupamentos de volatilidade, em que as flutuações são associadas por similaridade. Além disso, um algoritmo de evolução diferencial (ED) é sugerido para solucionar o problema de estimação dos parâmetros do modelo GJR-GARCH nebuloso. Os resultados indicaram que o método proposto permitiu uma melhor capacidade de previsão da volatilidade em comparação com modelos da família GARCH e com um modelo GARCH nebuloso recentemente proposto na literatura. Além disso, o algoritmo de ED considerado obteve soluções ótimas com uma rápida taxa de convergência.
Highlights
Measuring and forecasting stock market volatility plays a crucial role for asset and derivative pricing, hedge strategies, portfolio allocation and risk management
The author showed that the forecasting performance is significantly improved if the leverage effect of clustering is considered along with the use of fuzzy systems and Generalized ARCH (GARCH) approaches
The Fuzzy GJR-GARCH model in (9) is a nonlinear time-varying equation to model the behavior of complex dynamic systems as stock market volatility processes, including mechanisms to ensure the description of volatility stylized facts, such as volatility dependence, leverage effects and volatility clustering
Summary
Measuring and forecasting stock market volatility plays a crucial role for asset and derivative pricing, hedge strategies, portfolio allocation and risk management. The Fuzzy GJR-GARCH includes clustering estimation using subtractive clustering algorithm to provide a more autonomous model, based on data-knowledge, different from the approach proposed by Hung (2011a), which considers expert knowledge to determine the number of stock return clusters. Empirical results considers comparisons with GARCH-type models and with a recent Fuzzy-GARCH approach proposed by Hung (2011a), estimated with particle swarm optimization as in its original version and with the differential evolution algorithm suggested. After this introduction, the reminder of this paper is organized as follows.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
