Crude oil Price forecasting: Leveraging machine learning for global economic stability

Abstract

The volatility of the energy market, particularly crude oil, significantly impacts macroeconomic indices, such as inflation, economic growth, currency exchange rates, and trade balances. Accurate crude oil price forecasting is crucial to risk management and global economic stability. This study examines various models, including GARCH (1,1), Vanilla LSTM, GARCH (1,1) LSTM, and GARCH (1,1) GRU, to predict Brent crude oil prices using different time frequencies and sample periods. The LSTM and GARCH (1,1)-GRU hybrid models showed superior performance, with LSTM slightly better in predictive accuracy and GARCH (1,1)-GRU in minimizing squared errors. These findings emphasize the importance of precise crude oil price forecasting for the global energy market and manufacturing sectors that rely on crude oil prices. Accurate forecasting helps ensure economic sustainability and stability and prevents disruptions to production and distribution chains in both developed and emerging economies. Policymakers may choose to implement energy security measures in response to the significant impact of crude oil price volatility on the macroeconomic indicators. These measures could include maintaining strategic reserves, diversifying energy sources, and decreasing the dependence on volatile oil markets. By doing so, a country's ability to handle oil price fluctuations and ensure a stable energy supply can be enhanced.