Trading companies rely on social media algorithms for crypto trends
The world of cryptocurrencies can be daunting, with most of the data on the market’s sentiments not being readily available. Despite the market being worth hundreds of billions of dollars, most are private investors. This has led many trading firms to think outside the box when it comes to collecting data on the market. For some, social media holds all the answers.
Crypto hedge funds have been engaged in an arms race for algorithms that can scour the social media arena and predict the market’s movements. According to a report by Reuters, these algorithms have a widespread presence in all the major platforms, from Twitter and Reddit to more regional platforms such as Line and WeChat.
Known as sentiment analysis, this method of data collection is used in traditional markets such as foreign exchange markets. However, in the crypto market, its significance is more pronounced given the opaque nature of the market. Furthermore, cryptos are more reliant on online sentiments than many other assets. After all, Bitcoin was first spread online via chatrooms and forums.
Social media data collection is however quite a challenge for the firms that use it. One of the challenges that it faces is the high costs associated with purchasing a bot that can analyze all the information availed. Speaking to Reuters, Bin Ren, CEO of Elwood Asset Management revealed that while it’s profitable, only a few deep-pocketed players can afford to install and maintain such a system.
Andrea Leccese, the president of New York based investment firm Bluesky Capital told the outlet that the cost for a bot that could scour Twitter in English ranged between $500,000 and $1 million. Most of this money is spent on hiring developers with the required skills. This was what put off Bluesky from using this approach, Leccese said.
With such high costs, it’s easy to see why social media sentiment analysis isn’t widespread yet. Social media platforms are available in hundreds of languages apart from English. Moreover, there are tons of platforms aside from Twitter, from Reddit and Telegram to Line and WeChat in Asia to Odnoklassniki in Russia.
Piling on to the challenges is the risk of fake news which has continued to rise in popularity. As multiple reports have revealed in the past, some social media influencers and some websites as well accept money to post fake positive reviews on some cryptos. Algorithms could have a much harder time distinguishing the real from the fake.
Cedric Jeanson, the CEO of London-based crypto asset manager BitSpread admitted that the risk of fake news is one that’s ever present. However, despite this, the method works. He stated:
It’s a matter of gathering all the info, trying to understand who is trading where, what kind of liquidation can appear. It’s a strategy that makes sense. 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.
To receive the latest CoinGeek.com news, special discounts on CoinGeek Conferences and other inside information direct to your inbox, please sign up for our mailing list.