eToro’s AI Copy Portfolio Trades on Sentiment
Lucas Cacioli Oct 16, 2019 03:46
The eToro trading platform launched its latest user copy portfolio on Tuesday, an artificial intelligence (AI) tool that trades on the sentiments of 850,000,000 daily tweets.
The eToro trading platform launched its latest user copy portfolio on Tuesday, an artificial intelligence (AI) tool that trades on the sentiments of 850,000,000 daily tweets.
The eToro/Tie copy portfolio is the creation of a partnership between The Tie data analysis firm and Social Market Analytics (SMA) that leverages proprietary machine learning and natural processing technology to assess and quantify traders feeling towards digital assets. All eToro traders have to do is click copy on The TIE portfolio, with a minimum buy in of $2000, and the AI will diversify their crypto portfolio among the most positively discussed assets.
Crypto Market Driven By Crowd
According to an introductory presentation released by The Tie on its partnership with eToro, there is a strong correlation between twitter sentiment and crypto market movements.
From the presentation, “Unlike traditional assets, cryptocurrencies do not have fundamentals. Bitcoin has no revenue, dividends or debt. The value of digital assets are driven by supply and demand, and movement is determined by the wisdom of the crowd.” In this day and age, Twitter is where the crowd holds the discussion.
While the analytics tool may be new to consumers, Joshua Frank, CEO, The Tie explained in a YouTube video, that hedge fund managers and investment firms have been leveraging the same SMA data stream the eToro copy tool now uses. Frank said, "eToro is a social trading exchange in itself, making it the perfect partner for The Tie's technology."
Bots and Overlaps
Every day, The Tie’s AI scours through the 850,000,000 tweets and analyses, sorts and interprets social sentiment toward cryptocurrencies. Filters are used to assess the relevance and remove spam and bots, and sophisticated algorithms study the language of the tweets and assign ratings to different investors. Through the evaluation process, only 10% of tweets on any particular currency usually remain in the stream.
The presentation highlights how each tweet is scanned individually and extracted to a relevant category, but there are close to 80 overlaps between crypto-asset abbreviations and non-assets. “Using patented technology SMA is able to combat problems related to overlapping tickers across asset classes. For example, identifying whether an ETH tweet is on Ethereum or Ethan Allen, the publicly traded furniture company.”
Image via ShutterstockImage source: Shutterstock
