The realm of news serves as a vital tool for investors, traders, and risk assessors, enabling them to monitor companies effectively. Maintaining awareness of all released information that could influence stock prices is important. At Context Analytics, our Quantitative News Feed diligently monitors thousands of securities daily, drawing from hundreds of diverse sources to assess sentiment surrounding these companies. Changes in news sentiment can wield significant impact, prompting readers to act. In this blog, we delve into the analysis of news fluctuations for companies and their repercussions on price movements.
Our approach was to create a metric called the Delta S-Score, which serves as a key indicator of changes in news sentiment prior to market open. By quantifying these changes, we aimed to discern their influence on subsequent stock returns.
To develop this metric, we adopted a methodology similar to previous research: establishing a standardized sentiment score for each security and subsequently calculating the change (delta) from the previous score. In this sample, we applied a threshold of 5 for the maximum number of tickers identified in an article. This means any article with more than 5 tickers mentioned were excluded from the analysis as they do not reliably reflect any company's importance in the article. The following outlines the methodology behind our score:
This score captures the sentiment shift from the previous news day, offering a gauge of the extent of sentiment change in comparison to its historical fluctuations. The resulting metric demonstrates a normal distribution, ensuring a consistent and reliable metric for trading purposes.
To evaluate the predictive power of the Delta S-Score, we segmented stocks into quintiles based on their respective Delta S-Scores prior to market open. Each quintile represents a different level of sentiment change, ranging from highly positive to highly negative.
We then implemented a trading strategy wherein we held positions in each quintile from market open to close, compounding returns over time. Our analysis covered a substantial historical period, from August 2021 to the present day.
Our findings revealed a compelling relationship between the Delta S-Score and subsequent open-to-close returns. Notably, we observed a monotonic spread across quintiles, wherein:
The observed relationship between the Delta S-Score and stock returns underscores its potential utility as a predictive tool. This research represents just one of the myriad ways to harness our Quantitative News Feeds for alpha generation. For further information and inquiries, click the button below or reach out to us via email at ContactUs@contextanalytics-ai.com .