A Data Driven Hybrid Approach for Stock Market Forecasting Using Arima and Asymmetric Garch Models

Abstract

Due to the highly unpredictable character of financial markets, precise forecasting of stock markets is crucial for traders, analysts, and policymakers. The NASDAQ Composite Indicator is challenging to predict due to its numerous geopolitical and economic factors. The GJR-GARCH model addresses asymmetric volatility effects, particularly the pronounced impact of negative market fluctuations on instability, whereas the ARIMA model identifies linear patterns within the time series. This study forecasts the NASDAQ Composite Index by analyzing historical price trends and fluctuations in the market through the ARIMA and GJR-GARCH methods. The model demonstrates strong predicted accuracy when this strategy is utilized on daily closing prices from 2010 to 2024. The outcomes aim to assist financial professionals and investors in enhancing decision-making and effectively managing risks.