Artificial intelligence continues to move into the ETF space. On October 18, 2017, the Artificial Intelligence run ETF powered by IBM’s Watson launched on the NYSE. The fund, AIEQ, is managed by EquBot and brought to market by ETF Managers Group.
It’s hard to deny the lure of such an elegant concept behind an ETF in what is now undeniably a technological revolution within the asset management space. The supercomputing machine, Watson, has been learning Finance applications for some time and is now set to manage a portfolio of stocks based on what it has learned. It’s important to point out the machine learning component of this. There are numerous applications of AI hitting various parts of the marketplace and how the machines are taught and learn is important to the outcome. The analogy was of two chefs who have the same ingredient but different skills. The end meal will come out different. However there’s no denying the application of this type of approach in investment management space.
The ETF itself is targeted toward active managers. The value proposition enhanced by the AI application is that it allows the computer system to analyze information faster and more efficient than a group of analysts. Efficient market hypothesis postulates that a stock’s market price is reflective of all publicly available data. With the increase in volume but availability of information and subjectivity of the integrity and reliability of the information, this has created efficiencies and challenges to the art and science of investment analysis. A system that can more effectively filter through this information conceptually seems to be a value proposition, for sure. The technology behind the ETF is also designed to create trade points based on the metrics.
The objective is a more efficient approach toward active management, designed to accelerate current methods of data analysis, potentially improving outcomes under the Efficient Market Hypothesis. The parameters and constraints on the allocation will shape the outcome. Watson won’t be allowed to allocate to anything it wants. Additionally, what is has learned and what it will learn will effect the outcome and its ability to outperform other active managers and indexed strategies, whether that be traditional passive or smart / strategic beta.
What does this all mean for investors? The success of this ETF will ultimately depend on its outcome. The ETF itself will find its way into the same screening processes and comparisons that other active managers and new strategic beta funds, meant to target traditional active management, find themselves. One of the challenges that has faced recent strategic and factor models has been their ability to perform in line with ex-ante versus ex-post metrics based on their indexes. Without an index behind the method, there will be an observation period for AEIQ for investors to gain a level of comfort with the approach. They need to know what to expect and if it can perform in line.
As an asset manager whose job it is to evaluate ETFs, this will be part of our process too. I think this marks an exciting time in the investment management space, for investors and advisors alike. This will be the first of many.
2017-191
All information contained herein is for informational purposes only. This is not a solicitation to offer investment advice or services in any state where to do so would be unlawful. Analysis and research are provided for informational purposes only, not for trading or investing purposes. All opinions expressed are as of the date of publication and subject to change. Astor and its affiliates are not liable for the accuracy, usefulness or availability of any such information or liable for any trading or investing based on such information.