Lack of Data Inhibiting AI for Asset Managers? 
Interesting commentary in FundFire this week discussing the challenges of utilizing AI within the portfolio management process. While AI can be very effective where there is complete data availability, in the endeavor of fundamental investing, which is a predominant investment strategy on the buy side, this may not always be the case. Firstly, with fundamental investing, each manager typically has a limited number of different data characteristics that are screened for. So not a lot of data in relative terms. Secondly, there can often be a lack of availability of historical data, such as an incomplete data set. Therefore, applying an AI approach like deep learning would be largely ineffective in detecting patterns, since these types of approaches are typically only good if there is a lot of data and complete the data set.
Does this negate the current and potential future value of AI within the portfolio management process? Absolutely not. As with any software tool, the overall goal is to perform a complex process or workflow as rapidly and efficiently as possible and AI tools can be leveraged to great effect in this respect. At INDATA, we are focused on the practical applications of AI that we use within our Architect AI portfolio modeling and rebalancing solution for buy-side portfolio managers. Instead of focusing on less tangible AI techniques like deep learning, we fundamentally treat AI as a supplemental tool that can act as a digital assistant to the portfolio manager, allowing them to do things faster and more efficiently. Learn More.
