Even on the most liquid stocks, Context Analytics’ Social Sentiment provides an edge. Using only S&P 500 constituents, this Monthly Social Sentiment Index outperforms the market by over 2.5% annually over the past 5+ years.
In this blog we demonstrate one way to incorporate Context Analytics Sentiment metrics into an actionable portfolio which can outperform the market. For this analysis we are using the S&P 500 as the benchmark and its constituents as the universe of securities available for this portfolio.
This is a Monthly Long Only portfolio. Trades are entered and exited on the Market on Close (MOC) price of the fourth market day of each month. Securities are selected the last day of the month prior based on their Monthly Weighted Sentiment Score. The Top 200 securities based on the highest Monthly Weighted Sentiment are selected into the index and weighted by market cap each month. This sentiment metric is derived at the end of each month from Context Analytics’ S-Factor Feed.
We use one of our sentiment factors `Raw-S` from the S-Factor feed to calculate the Monthly Weighted Sentiment. `Raw-S` is the aggregation of Sentiment from Tweets in the past 24 hours. We exponentially weight this factor (lambda = 0.9) so Sentiment from more recent days have a stronger weight than Sentiment from earlier in the month. This exponentially weighted factor is summed over the entire month by security to calculates the security’s Monthly Weighted Sentiment.
Since 2018, the CA Index has outperformed the market benchmark. The annualized return of the Index beat the benchmark by 2.61% while reducing its risk too. The Sharpe and Sortino ratios on the Monthly Sentiment Index are higher than the S&P 500 over this 5+ year window. Selecting a large percentage of the S&P, 200 stocks, means there isn’t much turnover in the portfolio month over month.
The Monthly Weighted Sentiment factor used in this analysis can be derived from the S-Factor feed, which is Context Analytics’ standard social sentiment feed. The S-Factor feed looks at 24-hours’ worth of Twitter sentiment/activity and compares it to the security’s historical baseline. Many other factors can be derived by aggregating data over longer periods to fit your desired holding period or by using a combination of metrics in the S-Factor feed. For more information about Context Analytics, click the link below or email us at ContactUS@ContextAnalytics-AI.com.