Prediction of Oil and Foreign Currency Prices Using MGARCH and AI Models

Abstract

The research seeks to improve forecasting precision in oil and foreign exchange markets by applying MGARCH-BEKK and LSTM models. These methodologies, recognized for their capacity to predict nonlinear dynamics, examine the intricate volatility interrelations involving oil prices and currency exchange rates, utilizing data from January 2018 to June 2024. The anticipated results suggest that a hybrid MGARCH-LSTM model demonstrates superior predictive performance by effectively capturing the inherent interdependencies in these volatile financial markets. The LSTM model is anticipated to excel at processing sequential data and providing accurate volatility predictions for oil and currency markets. This method highlights the efficacy of hybrid models in tackling the intricacies of modern financial systems. The anticipated results correspond with current research, underscoring the revolutionary capacity of merging classic econometric methods with AI-driven approaches to improve market stability and predictability. Enhanced forecast accuracy underscores the strategic benefits of utilizing MGARCH and LSTM models for informed financial decision-making. Implementing these sophisticated technologies will empower stakeholders to navigate uncertainty in global markets more effectively.