Researchers have developed an innovative approach to predict future prices of the S&P 500 index using a combination of deep learning models and sentiment analysis. The experimental flow involved data collection, preprocessing, deriving technical indicators, generating sentiment scores from ESG-related news data, and using deep learning models to forecast future prices.
The S&P 500 index, representing 500 major U.S. companies, is a key indicator of the overall stock market trends. By integrating ESG information and the S&P 500 data, researchers aimed to enhance stock price prediction models and highlight the significance of sustainability information to investors.
Data collection involved gathering news articles related to ESG and historical data on the S&P 500 index from January 1, 2016, to July 31, 2023. Technical indicators such as opening price, closing price, RSI, SMA, EMA, MACD, and others were derived to analyze stock price movements.
Sentiment analysis using FinBERT, a specialized model for financial data, was conducted to assess the sentiment of news articles. The sentiment scores were then standardized and used as input features for the deep learning models.
Multiple datasets were generated with different window sizes to predict future stock prices. Deep learning models such as Bidirectional recurrent neural networks (Bi-RNN) and bidirectional long short-term memory networks (Bi-LSTM) were employed to capture patterns in the data and forecast stock prices accurately.
Overall, the experimental procedure illustrated a comprehensive approach to predicting S&P 500 index prices, combining technical indicators, sentiment analysis, and deep learning models for enhanced accuracy and performance. The findings of this study could have significant implications for investors and researchers in the financial markets.
