Investors have traditionally looked to price and volume for direction. But increasingly alternative data is providing signals.
Some of the data we are most excited about at Bloomberg are derived from news articles and social media posts.
Specifically we have created historical data sets for the volume of news and social media posts and algorithms to calculate sentiment based on the content.
Traders are realizing spikes in Twitter volume can indicate engagement as effectively as trading volumes. Big increases in news coverage can predict trading orders or validate market moves.
Paul Besson, head of quantitative search at Kepler Cheuvreaux in Paris, analyzed Bloomberg data. For companies with the biggest increases in news he found:
–High news flow was correlated to higher trading volumes and client trading activity in those securities.
–High news flow correlated to absolute outperformance for stocks and played a similar role as outsized volume in “validating” past trends in share prices by mitigating future reversions to the mean.
These charts can be visually compelling. The volume of news on Tencent versus Huawei tells a great story. The sentiment of articles written about Danske Bank tells another.