Developing a Predictive Model for S&P 500 Stock Price Forecasting Using the Prophet Algorithm

                @article{CRSIJ26000462,
                  author = {Vivian A. Kelechi Njoku, Bartholomew O. Eke, Martha O. Musa},
                  title = {Developing a Predictive Model for S&P 500 Stock Price Forecasting Using the Prophet Algorithm},
                  journal = {Cosmo Research and Science International Journal},
                  year = {2026},
                  volume = {2},
                  number = {2},
                  pages = {787-797},
                  issn = {3108-1584},
                  url = {https://cosmorsij.com/published/CRSIJ26000462.pdf},
                  abstract = {Stock market forecasting remains one of the most challenging problems in financial data science due to the inherent complexity, volatility, and non-linear behaviour of equity indices. This study presents a systematic investigation into the application of the Prophet algorithm, an open-source time-series forecasting framework developed by Meta (Facebook), for predicting the closing prices of the S&P 500 index. Using 3,623 historical daily records spanning January 2010 to December 2023, a structured data science pipeline was implemented encompassing pre-processing, exploratory analysis, model training, evaluation, and an interactive Stream lit-based deployment interface. The model was trained on 80% of the dataset and evaluated on the remaining 20%, achieving forecasting performance metrics including a Mean Absolute Error (MAE) of 68.53, Root Mean Square Error (RMSE) of 116.10, Mean Absolute Percentage Error (MAPE) of 2.18%, and an R² score of 0.9754. Directional accuracy was recorded at 52.7%, with trend classification yielding an F1 Score of 0.6963. Comparisons with baseline approaches, including the Seasonal Autoregressive Integrated Moving Average (SARIMA) model, demonstrate the competitiveness of Prophet for capturing long-term structural trends in financial time series. The results confirm that Prophet offers a transparent, interpretable, and computationally accessible alternative for stock price forecasting, with practical implications for data-driven investment decision-making.},
                  keywords = {Prophet Algorithm, S&P 500, Stock Price Forecasting, Time-Series Analysis, Machine Learning, Financial Prediction, SARIMA, Stream lit},
                  month = {September}
        }