Predicting stock market index prices using Facebook Prophet and XGBoost: evidence from the Saudi stock market

Khledj, Meryem; Mansouri, Hadjmoussa; Hela, Borgi

المجلة الدولية للاداء الاقتصادي
Volume 8, Numéro 2, Pages 293-307
2025-12-20

Predicting Stock Market Index Prices Using Facebook Prophet And Xgboost: Evidence From The Saudi Stock Market

Authors : Khledj Meryem . Mansouri Hadjmoussa . Hela Borgi .

Abstract

This study investigates the effectiveness of two machine learning models—Facebook Prophet and XGBoost—in predicting the Saudi Stock Market Index (TASI). The study relied on a daily series of prices during the post-COVID-19 recovery period. The results reveal that there is a difference in prediction accuracy, as the XGBoost model outperformed the Facebook Prophet model in various accuracy criteria (MSE, RMSE, and MAPE). The study concludes that integrating the strengths of both models can enhance stock index forecasting accuracy, providing valuable insights for financial analysts and policymakers.