Introduction
Traditional U.S. stock markets run from 9:30 AM to 4:00 PM Eastern Time. Overnight trading occurs between market close to the opening of the next day’s trading session. This period is often volatile due to factors such as international market movements, geopolitical events, and company-specific news, which can cause significant price movements and result in gaps up or down at market open.
Holding positions overnight exposes investors to the unpredictability of price movements, as many price adjustments occur during this period, presenting substantial risk. This research identifies signals heading into market close that can provide more information and help manage the risk associated with holding overnight positions. Conversely, these signals can also help identify securities that may experience positive jumps overnight.
Context Analytics and StockTwits
Context Analytics (CA) excels in processing and structuring textual data for sentiment analysis. CA gathers data from sources like StockTwits, a platform where traders and investors share their thoughts on various securities.
CA assigns a sentiment score to each StockTwits message, ranging from -1.000 (extremely negative) to 1.000 (extremely positive). These scores are aggregated over 24 hours and compared to a 20-day historical baseline to generate S-Factors. The S-Factor feed, one of CA's flagship products, includes the S-Score, which evaluates securities based on sentiment. The S-Score offers a cross-sectional view of social sentiment, contextualized against its historical baseline.
The SV-Score, another important metric, is calculated similarly to the S-Score but focuses on the volume of messages. The SV-Score provides standardized volume score, comparing the number of messages about a security on StockTwits to its historical baseline. By measuring the relative volume of discussion, the SV-Score helps identify unusual levels of activity and potential interest in a security.
Overnight Signal with S-Score and SV-Score
In this research, securities were grouped into five quintiles based on their S-Score at 15:40 ET, 20 minutes before market close. The analysis focused on securities with positive SV-Scores, indicating higher-than-usual volume on StockTwits before market close. Quintile 5 contained the top 20% of S-Scores, while Quintile 1 contained the lowest 20%.
Daily Close-to-Open returns for each security were calculated, and returns within each quintile were equally weighted to create daily quintile returns, cumulated since 2019. This analysis was conducted on two universes: securities priced over $5 and S&P 500 constituents.
The results showed a clear correlation between positive sentiment on StockTwits and overnight returns. Securities in the top quintiles (Quintile 5 and 4) outperformed those in the lower quintiles (Quintile 1 and 2), highlighting the strong connection between sentiment and overnight price movements. This pattern was consistently observed in both the broader universe of securities priced over $5 and the S&P 500 constituents. Additionally, securities with higher message volumes on StockTwits and an extremely low S-Score were clear indicators of overnight underperformance. Conversely, an extremely high S-Score could signal an overnight jump.
Conclusion
Insights from Context Analytics' S-Score and SV-Score are invaluable for managing the risk of overnight positions and identifying potential positive overnight movements. By leveraging sentiment data from platforms like StockTwits, investors can make more informed decisions, potentially mitigating losses and capitalizing on positive sentiment trends. For more information, visit www.contextanalytics-ai.com or click the button below.