Big Data in the Investment Industry March 2020
Subscribe to Insights in Brief to be notified about new Featured Content as it becomes available!
"AI: The Next Phase for Financial Data" in Explorer's February 2017 Big Data Viewpoints identified that data-driven AI systems were starting to automate the decisions that fund managers typically made. Such AI systems were focused on long-term investments and not on the already-common high-frequency trading algorithms that concentrate on speed. During the past two years, the role of data-driven AI has continued to evolve—though the path has not always been smooth. For example, Sentient Technologies closed its AI investment fund in 2018.
Today, many financial firms appear more interested in augmenting the work of fund managers with AI and big data than in using the technology to automate their work. A recent Financial Times article—"Stockpickers turn to big data to arrest decline" from 10 February 2020—claims that "quantamental" investment advice—combining (traditional) fundamental analysis with data-driven quantitative analysis—is now "one of the most powerful trends in asset management." In another example, AI investment firm Alpaca, now markets itself as a human and AI collaboration platform for investment markets.
In addition to combining human and machine analysis, financial firms continue to look for opportunities to add new big-data sources into their analysis (often termed "alternative data"). For example, JPMorgan Asset Management has created a natural-language-processing system that analyzes text from multiple data sources, including investment-bank research reports, social media, earnings-call transcripts, job advertisements, news stories, and regulatory findings. Instead of using the analysis to automate investment decisions, the system presents the analysis in dashboard format to fund managers.
Despite the continued success of high-speed trading algorithms, software that truly automates the role of fund managers is still rare. Equbot, which makes use of IBM Watson software, claims to operate the only actively managed exchange-traded investment fund that utilizes artificial intelligence for stock selection. Other examples include AI-driven hedge fund Rebellion Research and Walnut.ai, which operates the AI-driven Singularity Fund. Most often, however, AI and big-data analysis only advise human investors. This "hybrid" approach reflects a broader trend in data-driven AI: For high-stakes decisions, current technology limits mean that humans need to be in the loop.
For the big-data industry, the role of alternative data in finance is particularly significant, though many firms seem to be struggling to exploit alternative data to the fullest. Tammer Kamel—CEO of financial, economic, and alternative-data firm Quandl—says that "there are few dozen firms that are having real and sustainable success with alternative data. But I don't think the club is expanding very quickly." Companies report problems with data sets that are expensive (such as satellite or geolocation data), are hard to turn into actionable insights, or that come with privacy issues that concern some investors. As with the big-data field in general, data clearly can yield profitable insights; however, the practical reality of turning data into action is harder than many firms expect.