New algorithms are searching social media for clues as to why bitcoin has experienced a bullish renewal in 2019 and to predict future price movements.
Social Media to Predict Crypto Prices
According to a report by Reuters, hedge funds and asset managers are attempting to gain an edge on the market by mining social media for data. These companies are increasingly focused on creating computer algorithms that use social media metrics to reliably predict the next market movement. Thus far, the models are looking beyond Facebook to find “signals” buried in the mountain of content on sites like Reddit, WeChat and Twitter.
Bin Ren, CEO of Elwood Asset Management, explained that implementing accurate social media algos is both complex and expensive, but provides a serious advantage.
He told Reuters in an interview,
“It’s an arms race for money managers. Very few players are able to implement and deliver it, but I believe it is highly profitable.”
While the massive amount of content produced daily on platforms like Reddit and Twitter may give the appearance of finding a needle in a haystack, Ren believes computers are capable of detecting profitable trends.
“The information propagates not randomly, but through a very well-defined structure - it’s like a tree. It’s very similar to modeling the spreading of a virus.”
Many of the programs rely upon “sentiment analysis,” which involves computers mining social media messages to gauge the overall investor mood towards crypto-assets.
These companies are banking on the belief that the crypto markets are more susceptible to emotionally driven behavior. Bitcoin investors lack the same authoritative sources of information provided to the traditional markets, with even less reliable data like economic indicators and financial statements.
Sentiment analysis also highlights the abundance of retail investors in crypto, who may not react rationally to market behavior in the same way as institutions.
Despite the promise of finding usable data on social media, algorithm creators are running into a new problem: fake news.
While machines can be taught to filter posts for certain information, it becomes more difficult for the algorithms to discern biased or inaccurate statements.
Cedric Jeanson, CEO of BitSpread, explained the challenges involved,
“It’s a matter of gathering all the info, trying to understand who is trading where, what kind of liquidation can appear...The sentiment itself, what we see on Twitter, can be really geared toward fake news. We are always very cautious about what we’re reading in the news because, most of the time, we’ve seen that there’s a bias.”
Hedge funds looking for an edge on the market is nothing new, but the rise of social media algorithms adds a twist to cryptocurrency investors. While social media provides a plethora of data to be mined, it will likely take time for algorithms to become sophisticated enough to turn the abundance of noise into reliable information.