Date of Award
2025
Document Type
Thesis
Degree Name
Master of Business Administration
Department
Business Analytics and Information Systems
First Advisor
Michael D Herley
Abstract
This thesis investigates whether integrating alternative data sources with conventional financial indicators enhances S&P 500 forecasting accuracy in modern financial markets characterized by increased complexity and rapid information flows. The study analyzes daily S&P 500 returns using twelve predictor variables comprising conventional indicators (VIX volatility index, WTI crude oil, gold futures, AAA bond yields, Federal Funds Rate, Treasury yield spreads) and alternative data sources (Bitcoin returns, Daily News Sentiment Index, Indeed job postings, Economic Policy Uncertainty Index) across 1,341 observations from February 2020 to June 2025. The methodology employs ordinary least squares regression and ARIMAX(5,5) modeling with rigorous preprocessing including Augmented Dickey-Fuller stationarity testing and logarithmic transformations, estimated via maximum likelihood procedures in EViews with 30-day out-of-sample validation. Results reveal alternative data contributes a 249% improvement in explanatory power over conventional indicators alone (adjusted R-squared: 0.213 vs. 0.061), with ARIMAX models achieving superior performance (0.666 adjusted R-squared). These findings offer practical applications for portfolio management, risk assessment, and trading strategies while demonstrating the value of alternative data integration in modern financial forecasting.
Recommended Citation
Scarborough, Patrick M., "PREDICTING PRICE MOVEMENTS OF THE S&P 500 USING CONVENTIONAL AND ALTERNATIVE MARKET INDICATORS" (2025). Theses and Dissertations. 1011.
https://scholar.stjohns.edu/theses_dissertations/1011