Hedge funds and asset managers have turned to algorithms to learn more from social media about cryptocurrency markets, Reuters writes. They are working on algos that would be used to analyse social media to detect triggers that can affect cryptocurrencies.

A number of investors are already using computer models to be able to spot price differences across cryptocurrency exchanges.

“Sentiment analysis” can be used to analyse social media moods, and is already applied to traditional markets, such as equities and foreign exchange. The analysis is used to determine consumer sentiments towards a certain company or asset. With few authoritative sources of information, little reliable data, and a large number of individual investors, sentiment analysis could come in handy when analysing cryptocurrency markets, Reuters writes.

However, analysing all social media, such as Reddit, Twitter and Telegram, could be pricey.

According to president of investment company Bluesky Capital Andrea Leccese, you need to pay between $500,000 to $1 million to get a robot capable of reading just Twitter in English. Most of the funds go towards hiring skilled developers, Leccese said. Considering the number of relevant social media channels, the required investment is substantial.

The algorithm would also need to be able to recognise bias and inaccurate information.