The ability to judge what will happen in the future is a full-time occupation for business managers as well as financial analysists and central bankers. Where to invest, which markets to enter and which products to phase out are common predicaments yet making those judgement calls has become a lot easier in some circles.

This week’s Economist, a newspaper, explains how the US Federal Reserve has recently published papers on how it uses data gleaned from messages posted on Twitter, now ‘X’, to act as predictors for forthcoming economic events such as share price movements, economic conditions such as unemployment as well as monetary policy decisions.

Sentiment analysis is nothing new and Twitter/X plays no role in the shaping of events, only its data is used, as far as we can tell. A 2020 paper presented to the SSERS 2nd International Conference on Human Capital, in Amsterdam by Buckinghamshire New University Business School academics applied sentiment analysis to Initial Public Offering, IPO, data for a selection of firms floated on the London Stock Exchange between 2018 and 2020.

Using a methodology that is very similar to the one now adopted by the Federal Reserve, and in no way proprietary to the authors, using pre-IPO announcements posted on Google. Messages were analysed to assess sentiment as to the future direction of travel for the share price over the next 3 months. Positive sentiments rated a long or ‘buy’ label whilst ‘negative’ and ‘neutral’ sentiments rated a ‘short’ label. Using a 24-month period to smooth out stock market and economic shocks the data revealed an accuracy that would have seen a 21% return on any original investment. An examination of fund managers performances during the same period ranged from +44% to -7%, whilst the FTSE 100 annual return for 2018 was -12% and 2019 +12%. The total return based on data from 1984 to 2022 presents an annualised return of 7.4%. (Remember investments can go up as well as down and your capital is at risk as with any investment, look at what happened in 2018!) So, does sentiment analysis offer scope for decision making in various scenarios? The answer appears to be it could and the Federal Reserve of the United States of America appears to agree!

In business it’s not finding the right data, it’s knowing how and where to use it that counts.