State Street combines machine learning with human insight to tackle research overload
22 September 2017 | 3362 views | 0
Source: State Street
Investment professionals today are inundated with research reports from the sell side, the buy side and academia.
Top financial research teams can produce enough content to consume 24 reams or 12,000 sheets of paper per day. To help address this information overload, State Street (NYSE:STT) has launched QuantextualSM Idea Lab, which combines the power of machine learning with the knowledge of human experts in order to help investment professionals efficiently read and interpret lengthy research reports, and apply relevant findings to their investment strategies.
Part of the next generation of intelligent client solutions developed by State Street Global Exchange℠, State Street’s data and analytics business, Quantextual Idea Lab uses machine learning algorithms to consume complex research reports, tag them by investment themes and assets as well as suggest new, relevant materials based on the end-user’s specific needs, preferences and observed reading behavior. The solution then carefully overlays the experience and market knowledge of Quantextual’s own research team to continuously improve the algorithms’ performance and accuracy.
“Today, the speed, sophistication and volume of information far outpaces our ability to consume it,” said Stephen Lawrence, head of Quantextual Research, State Street Global Exchange. “Efficient intake of research is essential for identifying investment opportunities and managing risk. With limited time investors may overlook or delete important research insights. Those who can make sense of large amounts of information quickly will front run peers.”
Quantextual Idea Lab includes the following features:
• Automated document classification that can accelerate the speed at which research is acquired. By having access to automatically tagged and summarized research, the investment professionals can focus more time on developing strategies and making informed decisions.
• Human-curated content and conclusions created from the research which preserves their nuanced academic or economic tone.
• Enhanced search capabilities including tag and keyword search as well as natural language search with an intuitive question-and-answer function capable of instantaneously responding to investment-related questions, returning the key research and insights connected to the inquiry
• Social collaboration tools allowing users to annotate research reports and share them with colleagues, which can help users steer the most important ideas to the right decision-makers.
“Through Quantextual, we’ve designed a workflow that empowers the transformation of carefully curated ideas into decisive and strategic action,” continued Lawrence. “As more investment firms shift to a model of covering research costs out of their own P&L’s, an additional benefit of Quantextual is that it allows them to systematically review research consumption, to help them optimize their research spend.”