Skip to content
All library documents

Combining News Sentiment and Time-Series Models for S&P 500 Trading

Article arXiv papers · Author: Haojie Liu et al.
Read the original

Summary

The study describes a trading approach that combines sentiment extracted from financial news with technical indicators and time-series forecasting. It uses GPT-2 and FinBERT for sentiment analysis, pairs the resulting signals with momentum and trend measures, and evaluates strategies alongside ARIMA and ETS models and a buy-and-hold benchmark. Performance is assessed using asset values and returns.

The document reports that combining sentiment signals with traditional models improved trading performance in its evaluation, and frames the method as a way to respond to changing market conditions. It gives no sample dates, detailed strategy rules, numerical performance results, or statistical tests, so the strength and generality of the finding cannot be assessed from the description. The evidence is specific to the S&P 500; it does not establish that the approach transfers to other markets or remains effective after trading costs.

Key ideas

  • The approach combines financial-news sentiment with technical indicators and time-series models.
  • GPT-2 and FinBERT are used to derive sentiment signals from news.
  • The evaluation compares combined strategies with buy-and-hold and sentiment-based approaches.
  • The reported improvement lacks numerical results and detailed validation information in the description.

Tags

EquitiesSentimentMachine learningTechnical indicatorsBacktestingUS markets
Full text
# Enhancing Trading Performance Through Sentiment Analysis with Large Language Models: Evidence from the S&P 500


# Enhancing Trading Performance Through Sentiment Analysis with Large Language Models: Evidence from the S&P 500









This study integrates real-time sentiment analysis from financial news, GPT-2 and FinBERT, with technical indicators and time-series models like ARIMA and ETS to optimize S&P 500 trading strategies. By merging sentiment data with momentum and trend-based metrics, including a benchmark buy-and-hold and sentiment-based approach, is evaluated through assets values and returns. Results show that combining sentiment-driven insights with traditional models improves trading performance, offering a more dynamic approach to stock trading that adapts to market changes in volatile environments.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.