Deep Learning the Dynamics of Financial Contagion: LSTM Networks for Spillover Detection in Four GCC Stock Markets
Abstract
This paper presents Long Short-Term Memory (LSTM) networks for detecting and predicting financial spillovers in the Gulf Cooperation Council (GCC) stock markets. Although widely used, traditional Vector Autoregression (VAR) models do not capture the nonlinear and asymmetric dynamics common in regional financial contagion, especially during crises. We apply LSTM networks to the daily returns of four major GCC stock indices: Saudi Arabia, Oman, Dubai, and Qatar. The sample spans 2,273 trading days from March 2012 to December 2024.
Fields of Study & Subjects
Financial Risk and Volatility Modeling
Stock Market Forecasting Methods
Market Dynamics and Volatility
Publication details
- Journal / source
- International Research Journal of Innovations in Engineering and Technology
- Volume
- 10
- Issue
- 01
- Pages
- 74–86
- Publication date
- 1 January 2026