How does social media sentiment provide investors with an additional tool with which to gain insights in financial markets, cryptocurrencies and inflation?
- Social media sentiment can be mined by investors to provide them with insights on stocks, cryptocurrencies, and commodities they may be unable to find through more conventional sources.
- Refinitiv MarketPsych Analytics sentiment scores collect data from thousands of news and social media sites from around the globe. From this data, a net sentiment score is calculated by comparing positive sentiment and negative sentiment over time about a particular stock.
- The research found that stocks that received positive sentiment scores in the analysis outperformed those that got negative sentiment scores.
For more data-driven insights in your Inbox, subscribe to the Refinitiv Perspectives weekly newsletter.
With the proliferation of content on social media, an increasing number of investment professionals are using this medium to track, monitor and gauge interest in particular stocks, using this as an additional channel to monitor sentiment in financial markets.
At its most basic level, the technology quantifies the sentiment expressed about specific stocks. In a series of research studies we found that this sentiment forecasts future stock prices.
“As there are so many people who cannot wait to follow the prevailing trend of opinion, I am not surprised that a small group becomes an army.”
~ Josef de la Vega, 1688, Confusion de las Confusiones
De la Vega observed that the transmission and spread of opinion drove price trends on the Amsterdam Exchange in the first book written on financial markets.
Account holders at Robinhood – an online brokerage where investment ideas are socially transmitted – on average outperform professional investors1.. However, when Robinhood investors jump into specific stocks at the same time on the back of positive price movement (”herding events”), those investors tend to underperform.2.
Using text analysis of media, we tested de la Vega’s observation of over 300 years ago on the trends of opinion in modern social media.
Retail traders are moving markets in stocks and cryptocurrencies, and the forums where they gather can be systematically mined for useful investment insights.
Measuring social media sentiment
Our research team performed a series of studies on stock sentiment, using monthly aggregations of media sentiment data. They used the Refinitiv MarketPsych Analytics (RMA) sentiment scores.
RMA social sentiment scores represent aggregate scores from thousands of news and social media sites including Twitter, Reddit, and thousands of others.
The sentiment score is a time series of the net positive versus negative references to a given company. Every minute, the business and financial news and social media is scanned for new references to tens of thousands of assets.
The sentence “Tesla is performing great” is scored as a positive sentiment value (+1) for Tesla (TSLA.O). When thousands of such references in millions of daily online new articles are aggregated, a net sentiment score (positive vs negative balance) can be derived and plotted over time, giving an overview of the media tone expressed about a specific firm.
Figure 1 is a heatmap of U.S. stock sentiment from 1 August 2021 to 1 September 2021. Note that Fisker and Salesforce appear to have the most positive media sentiment (brightest green in the heatmap).
According to de la Vega’s observations, investors should follow the positive trend of opinion and buy Fisker (FSR) and Salesforce (CRM) stock.
Following the trend of opinion
MarketPsych’s research team ranked more than 30,000 global stocks each month using their average RMA sentiment scores over the prior month and binned them into deciles (approximately 2,500 stocks in each). They tracked the performance of each decile portfolio over the following 90+ days.
The most positive stocks by media sentiment decile (green lines) significantly outperform the most negative stocks (redder lines) over the following 90 days, as seen in Figure 2.
There is a significant, steady and consistent divergence in performance between the most positive and most negative groups. The spread averages 4 percent annualised between the top and bottom sentiment deciles from 2006-2020.
When broken down by region, the outperformance of high sentiment stocks is evident in every country and region we tested, from Brazil to the U.S., and it was true of both large and small cap stocks.
Watch – Refinitiv MarketPsych Indices: understand the mental state of the market
Finding new investment ideas
There is a complex interaction between sentiment, fundamentals and asset prices.
Some researchers find that “filtering” sentiment – removing the effects of recent price action on retail sentiment – leads to an even stronger predictive signal.
The goal of this analysis is to demonstrate how online media sentiment can provide value to investors. And as we found, a simple sentiment momentum strategy shows promise in periods from monthly to yearly windows in every country studied.
2. Barber, Brad M. and Huang, Xing and Odean, Terrance and Schwarz, Christopher, Attention Induced Trading and Returns: Evidence from Robinhood Users (July 27, 2021). Available at SSRN: https://ssrn.com/abstract=3715077 or http://dx.doi.org/10.2139/ssrn.3715077