How we are building the best prediction trading engine for the crypto markets
Prediction trading is about analyzing historical data in order to predict future price trends. Market sentiment has been shown to be an important prediction factor in this equation, which is why we are adding this capability to Enigma.
One of the biggest and most active public forums these days is Twitter, where more than a million users post over 140 million tweets every day. The posts are very concise thanks to being limited to only 280 characters and use hashtags for tagging content and @ for addressing other users. This makes Twitter an ideal information source for sentiment analysis. It also provides for easy tracking of very large and highly influential accounts, with the power to move the markets.
An example of the latter would be John McAfee’s Twitter account @officialmcafee, which has over 845,000 loyal followers who can create substantial sentiment signals. For this reason many traders are tracking his account and executing trades based on the projects mentioned in his tweets.
While some may benefit from this strategy trading smaller market caps coins and tokens, it is unhealthy for the market stability and needs to be considered with caution.
The research
Research has shown that historical price movements are not the only source of information for prediction trading. It can be improved upon by using additional indicators such as public sentiment, which can be measured though social media.
Ruiz et al. have used time-constrained graphs to study the problem of correlating the Twitter micro-blogging activity with changes in stock prices and trading volumes.
Bordino et al. have shown that trading volumes of stocks traded in NASDAQ-100 are correlated with their query volumes (i.e., the number of users requests submitted to search engines on the Internet).
Gilbert and Karahalios have found out that increases in expressions of anxiety, worry and fear in weblogs predict downward pressure on the S&P 500 index.
Bollen showed that public mood analyzed through twitter feeds is well correlated with Dow Jones Industrial Average (DJIA).
The techniques involved
Raw data collected from Twitter via their API is first processed to remove unnecessary data like pictures, emoticons and URLs and then filtered for specific keywords of interest. The resulting data can then be analyzed using various techniques which assign each post a positive, negative or neutral rating.
Sentiment analysis is very much field specific, therefore a lot of research is required to adapt the protocols to each specific field, in our case cryptocurrencies.
By combining market sentiment with other indicators such as statistical analysis of trading data on exchanges, search engine trends, developer activity, corporate announcements and other relevant news, our Enigma prediction trading engine will be able to optimize the trading strategies for higher returns, while at the same time minimizing any risks.
We are seeing encouraging results in our testing and are confident that once the development is complete, we will be able to very accurately and efficiently collect and interpret relevant information and use it to gain advantage in the market.
If you wish to stay in touch with us, please be sure to follow us on our social media channels.
