At Context Analytics, we provide social sentiment data for a wide range of asset classes, including global equities, futures, FX, ETFs, and cryptocurrencies. One of our flagship products is the S-Factor Feed, which captures sentiment and volume metrics from Twitter. In this blog, we dive into an analysis of the Futures market, leveraging sentiment insights from our feed to uncover actionable trading strategies.
Futures Market Focus
For this study, we analyzed sentiment metrics across 45 commodities with sufficient Twitter activity, including notable names like crude oil, natural gas, silver, gold, copper, and soft commodities like cocoa, cotton, sugar, coffee, as well as currencies such as Bitcoin and British Pound among others. Our goal was to explore the relationship between Twitter activity (volume) and price movements in these futures markets, specifically looking at commodities with abnormally high or low Twitter volume change.
Twitter Volume & Sentiment Metrics
Twitter Volume is represented by our Standardized Volume Score (SV-Score). This score gives a standard normal distribution of Twitter activity surrounding each commodity for the previous 24 hours, compared to its 20-day baseline. By standardizing the volume, we can quickly identify which commodities are experiencing unusual social attention.
To take it one step further, we created a Delta SV-Score, which captures the standardized difference in Twitter volume over the past 24 hours relative to the previous day. A Delta SV-Score of 1, for example, indicates a one standard deviation increase in Twitter volume from the previous day. This score allows us to track real-time shifts in social media attention and leverage those signals for trading decisions.
Building a Trading Strategy
Using the Delta SV-Score, we implemented a simple strategy:
- If Delta SV-Score > 1 → Go Long (buy the commodity)
- If Delta SV-Score < -1 → Go Short (sell the commodity)
This strategy is based on the idea that "no publicity is bad publicity." For commodities, any increase in social attention, as reflected by a spike in Twitter volume, suggests growing interest or relevance, making it a signal to go long. Conversely, a significant drop in Twitter volume indicates waning attention or fading relevance, signaling a potential opportunity to short as the commodity is losing momentum.
We trade daily at the 9:30 AM ET timestamp, using Twitter data collected at 9:10 AM ET. Each day positions are rebucketed and adjusted based on the new Delta SV-Score. Our strategy is strict, meaning there are days where no long or short positions are taken if the thresholds are not met.
Performance Since 2021
Since 2021, we’ve seen some significant results:
- Our long-side strategy has returned 25% annually, with a Sharpe ratio over 1 and a Sortino ratio over 2, signaling robust performance and a good risk-adjusted return.
- The short-side has posted 10% annual losses, which holds promise as a counterbalance in our long short portfolio.
- Combined, the long-short strategy forms a balanced portfolio, delivering consistent and robust returns. We typically receive around 10 signals daily across both portfolios, with a roughly even split amongst long and short opportunities.
Conclusion: Twitter Volume as a Trading Signal
Our findings demonstrate that volume changes on Twitter have a significant impact on daily commodity price movements. This is just one of the many ways Context Analytics Twitter sentiment data can be manipulated to generate excess returns in the commodities market. Our standardized sentiment metrics provide a transparent and actionable insight into market trends, and this futures trading strategy showcases the potential of social sentiment to drive alpha in commodity trading.
For more information on commodities data or Context Analytics in general, visit www.contextanalytics-ai.com .