What Are Day Traders Thinking? S&P's Twitter Sentiment Index Has Answer

(Bloomberg) -- A year after meme mania broke out, Wall Street’s still obsessing over what day-traders are up to. 

S&P Dow Jones Indices said Thursday it is launching gauges that will measure daily sentiment on Twitter in a bid to get a grip on what “everyone” is thinking. Two new indices will be based on analyses of tweets that contain so-called “$cashtags,” which Twitter users deploy to reference stock symbols on the social-media platform.

“Factor-based indices are a popular strategy for passive investors, so we’re excited to work with Twitter to bring this unique tilt to the S&P 500,” Peter Roffman, global head of innovation and strategy at S&P Dow Jones Indices, said in a statement.

The index provider is debuting the S&P 500 Twitter Sentiment Index, which will measure the performance of 200 constituents with the highest sentiment scores. That gauge will be float-adjusted. The second, the S&P 500 Twitter Sentiment Select Equal Weight Index, will measure performance of 50 stocks on a more equal basis. The indexes are going to rebalance at the beginning of each month. 

Wall Street professionals have all year been trying to deduce what day-traders are thinking about and what moves they might make next after an en-masse effort at the start of 2021 showed they could affect the movement of stock prices. Many traders scroll through Reddit blogs to suss out the next big thing, while others obsessively track Stocktwits citations. There are even firms hiring WallStreetBets veterans to get inside the heads of the people who call themselves apes with diamond hands.

UBS Group AG recently also, amid elevated at-home trader participation, came out with a model that flags retail signals for stock returns. 

Stock discussions taking place on Twitter are having an increasingly significant impact on markets, S&P Dow Jones said. Conversations around finance were up more than 26% last year versus the year before, according to the social-media platform’s internal data cited by the company.

©2021 Bloomberg L.P.