Skip to content

Long Only Monthly Portfolio with Twitter Sentiment

Context Analytics (CA) leads the way in leveraging social sentiment data into actionable financial signals. A variety of metrics in the S-Factor feed allows users to combine and aggregate data to create custom factors. In this blog we showcase how users can create factors to deliver a portfolio that outperforms the market.

Monthly Weighted Sentiment per Tweet

The signal is created to evaluate which security has the most positive sentiment accounting for the popularity of the security.

  1. Core Factors
    1. Raw-S : Aggregate of Sentiment from all Tweets on a stock in the past 24 hours
    2. S-Volume : Number for Tweets on a stock in the past 24 hours
Monthly Aggregation
    1. Applied an exponential weight (λ = 0.9) to Raw-S to give more weight to Tweets from more recent days, creating Weighted Sentiment
    2. Monthly summation of S-Volume and Weighted Sentiment across each month and security
Monthly Strategy

Using a universe of only S&P 500 securities, below is a Monthly Long Only strategy with Monthly Weighted Sentiment per Tweet.

At the end of each month, the 50 securities in the S&P 500 with the highest Monthly Weighted Sentiment per Tweet are selected. The portfolio is rebalanced one market day later at market close of the first market day each month. The 50 securities are equally weighted each month.

Performance

 

Screenshot 2025-07-10 at 8.55.26 AM

 

 

Screenshot 2025-07-10 at 8.56.47 AM

The strategy has consistently outperformed the equally weighted S&P 500 index (RSP) since the beginning of 2014. In over 10 years, the strategy has increased annualized returns by nearly 1% compared to RSP with reduced volatility. This is while only selecting 50 securities each month, which is 10% of the benchmark’s holdings.

When looking at a more recent timeframe, since the beginning 2022, the strategy outperforms the benchmark by over 2% annually with similar volatility. This is an outperformance of 10.08% cumulatively over a 3.5 year time period.

Unlock Alpha with Context Analytics Data

The portfolio above is one demonstration on how using social sentiment metrics can enhance stock selection leading to higher returns and reduced risk. Context Analytics has numerous sentiment metrics from a variety of sources that can be combined and aggregated to create unique factors. For more information on Context Analytics, please visit contextanalytics-ai.com or email us at contactus@contextanalytics-ai.com.