Quantitative investing, one of the latest paths available to hedge fund managers and incorporating computer analytics in innovative new ways in order to make precision investment decisions, may have a new challenger emerging from within its own ranks. According to a report in January by Eurekahedge, the quickly-changing landscape of alternative investing strategies has seen a sudden rise in the prominence of artificial intelligence-based (AI) funds, and that many of these funds are vastly outperforming so-called "traditional quants", as well as human-led management teams.

AI Funds Won the Past Several Years

According to a report by ValueWalk, Eurekahedge's AI/Machine Learning Hedge Fund Index, monitoring performance of 23 hedge funds utilizing this investment strategy, has managed to outperform generalized hedge funds as well as traditional quant funds decisively since 2010. In fact, AI funds have netted annual returns of 8.44% for the past 6 years. This is dramatically higher than the other indices that Eurekahedge uses, including the CTA/Managed Futures index, with 2.62% returns for the same period, and the trend following index, which saw a mere 1.62% return level at the same time. At a time in which most hedge funds across the industry are facing significant struggles, AI funds are doing better than their peers. How is this happening?

Faults With Traditional Quant

ValueWalk argues that AI's dominance in recent years can be attributed in part to two failures of traditional quant-based approaches. First, as traditional quants have become more and more popular (with more than $40 billion of inflows in just the past two years), the area has become overcrowded. This problem doesn't show signs of slowing down, either: the first two quarters of 2016 alone saw inflows of $10.8 billion, a record. As overcrowding continues to develop, returns across the industry are depressed.

The second issue is the traditional research model that many of these funds utilize. According to the Eurekahedge report, "trading models built using back-tests on historical data have often failed to deliver good returns in real time."

How do AI funds work to beat these odds? According to Yoshinori Nomura, a director of Simplex Asset Management, algorithms should be as simple as possible, monitoring the environment and adapting its strategy as needed. While some quant funds do not follow the environment for price persistence, he sees this as a key to AI's flexibility. Modifying the investment strategy in accordance with these findings is also crucially important. Beyond that, an innovative back-testing strategy which is more focused than blind collections of data and which weighs out recency relevance against weight analysis. What many AI fund managers are still working on, however, is understanding the causation of the market fluctuations that their funds are becoming better able to predict.