On 29 March, a large asset manager announced a strategic overhaul of portions of its active equity management approach, focusing on the use of quantitative modeling over “traditional” human active management. Click here for a copy of the press release.
Just a couple of weeks ago, our Pittsburgh office hosted its inaugural artificial intelligence program, The Artificial Intelligence Gateway For the Investment and Business Community that featured keynote speakers and panel discussions regarding the increasing awareness of artificial intelligence (AI) across all industries and the impact this new form of technology will have on business. It was fascinating to hear from our keynote speakers about how AI actually works, how AI is used in self-driving cars, and future use of AI in various industries, from manufacturing to financial services. In addition, one of the panels, focused on robo-advice and AI, where we discussed technological growth and how AI might be used in the investment management industry and some of the related regulatory and fiduciary issues.
A main point of the panel’s discussion was the industry trend away from active money management towards lower-cost, passive strategy funds (such as exchange-traded funds), and whether machines could potentially have more success than humans at outperforming the market over the long term through multiple economic and market cycles. Interestingly, the attendees at our seminar from the technology side were absolutely confident that machines that are able to process (and evaluate) vast amounts of market data in minutes or seconds could outpeform broader markets. Interestingly, this manager is only making the changes for funds and strategies that tend to closely track the performance of indices (such as US large cap stock portfolios), and are retaining the human touch for country-specific strategies or strategies that are more focused or concentrated (presumably because they believe those markets to be less efficient and they will have an edge).
It is unclear from the press release whether this manager is using true machine learning (AI), although I suspect that there is an element of that involved. Quantitative managers and quant strategies are not new, and it is hard to imagine a significant market player making these types of fundamental and philisophical changes just using the same quant-like modeling that has been around for years without some form of differentiating factor (e.g., deep learning capability).