There is an interesting story behind the inception of State Street’s Quantextual Idea Lab, the winner of the 2017 AFTA for the best front-office initiative. The Lab was the brainchild of Stephen Lawrence, a former analyst who now heads up State Street’s Quantextual Idea Lab department.
Lawrence had been producing research for portfolio managers for many years and, as with a number of his colleagues, was frustrated by how little of his work was being consumed. But it was not until he spoke with a portfolio manager that he realized how unfeasible it was for them to use any of the research he was producing.
“A former colleague showed me his email inbox which contained multiple pages of research,” Lawrence explains. “He told me that he didn’t get the chance to really read them—he just looked at the email subject and decided if he was going to read it.”
Lawrence understood that portfolio managers were interested in reading his research, but they didn’t have a practical method of deciding which pieces were relevant and which were not. So he came up with Quantextual Idea Lab, which he developed with the assistance of State Street, the second oldest financial institution in the US, with something of a startup mentality.
The final product is a research aggregator, although its uniqueness lies in the fact that it blends machine-learning algorithms with human expertise. “I started thinking if there was a way to come up with a more intelligent curation process for research,” Lawrence says. “The lab consists of technology that automatically tags and categorizes research, and then human expertise can continue to refine, train and improve the ability to understand the changing subject matter like financial markets research.”
The platform uses various classification algorithms that scan documents and decide what topics, asset classes and regions they are referring to. These algorithms, according to Lawrence, identify a range of words, starting from simple phrases and progressing to complex financial terms. In addition, natural-language processing technology is included allowing users to query research in a national language, as well as collaborative capabilities that allow them to communicate and share research with one another.
Most importantly, the Quantextual Idea Lab uses actual people to provide quality services. Working behind the scenes, the Lab employs researchers to address questions pertaining to complex investment processes. These researchers provide human-curated conclusions rather than mechanical interpretations, allowing research summaries to maintain their nuanced academic or economic perspectives.
Bloomberg's Gerard Francis comes on the podcast to talk about how data fragmentation issues are becoming more challenging.Subscribe to Weekly Wrap emails