OUR INVESTMENT PROCESS
The data for our innovative trading strategies is generated through natural language processing in German, English and Chinese via individual crawling of a unique set of hundreds of sources from traditional content to user generated content on blogs, forums and social media to reflect “who says what to which security”..
More than 2 million messages from news, articles, research reports, blog posts, forum entries and tweets are collected per day while monitoring more than 45,000 securities in real-time. As an innovative German FinTech company we know that unique access to alternative data sources in combination with artificial intelligence (A.I.) and the know-how how to use it is the key competitive advantage in the quantitative asset management industry going forward.
Data collection from hundreds of data sources across the web, ranging from traditional editorial content to user-generated content from blog, forum and social media posts (2+ million posts per day).
Automated language processing in three languages (German, English and Chinese) and weighing: “Who says what – to which security” by observing more than 45,000 securities in real time.
Making predictions based on machine learning and aggregated Big Data values and security prices. The AI system is able to systematically change its trading style and behavior.
Execution of trading strategy with high focus on risk management. Our focused trading is fully automated until the final order confirmation.
The Big Data and A.I. based investment approach transforms data in predictive trading signals. Our fully automated long / short equity strategy uses a unique combination of stock selection, market timing, smart exposure management and tailored risk management with the goal to generate above market returns.
While the whole process from data crawling to trading execution is fully automated, certain parameters are still predefined, such as the equity universe, maximum exposures, regulatory requirements, etc. The A.I. system is capable of changing its trading style and behavior in a systematic way. Once a trading signal is generated, a portfolio manager only confirms each order with one final click.