Applications of GARCH Models in Forecasting Financial Market Volatility: Insights from Leading Global Stock Indexes

Abstract

This study investigates the volatility dynamics of major global stock indexes, including the FTSE 100, Hang Seng Index, NIKKEI 225, and S&P 500, using a range of Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models. The analysis spans a comprehensive 20-year period from January 1, 2004, to December 31, 2023, encompassing diverse market conditions such as bull and bear markets, the 2008 financial crisis, and the COVID-19 pandemic. The methodology includes preprocessing steps such as calculating daily log returns, performing descriptive statistics, and conducting stationarity and ARCH effect tests to ensure data suitability for volatility modelling. The study evaluates several GARCH models, including GARCH, EGARCH, NGARCH, APARCH, GJR-GARCH, and TGARCH, to forecast volatility and address both symmetric and asymmetric effects. The TGARCH model exhibits strong performance in capturing leverage effects and asymmetries, particularly for the FTSE 100 and Hang Seng Index. The APARCH model performs well for the S&P 500, demonstrating sensitivity to past shocks. Overall, the findings underscore the importance of advanced GARCH models in accurately predicting volatility in global financial markets, highlighting the TGARCH model's effectiveness in addressing asymmetries and providing insights into selecting appropriate models for enhanced financial analysis and risk management.