By Allan Lane, PhD, Founding Partner, Algo-Chain.
As the ETF industry launches into ‘Passive Investing 3.0’ in the year of the 50th Anniversary of Stanley Kubrick’s science fiction masterpiece 2001: A Space Odyssey, it seems fitting that we can now talk about new ETFs that fall into the Artificial Intelligence & Machine Learning (AIML) category. The trend for themed ETFs continues apace, so what exactly constitutes an AIML based ETF and do they provide a good investment opportunity?
Beyond 2018 – An ETF Odyssey
Searching for AIML ETFs one finds a couple of US listed ETFs which deliver innovate ideas in entirely different ways.
The first example, with ticker BOTZ US, is the Global X Robotics & Artificial Intelligence ETF, launched by the boutique provider Global X in September 2016. To date this fund has gathered $2.3bn in AUM, comes with a TER of 0.68%, and invests in companies that potentially stand to benefit from an increased adoption and utilization of robotics and AI, including those involved with industrial robotics and automation, non-industrial robots, and autonomous vehicles. The ETF’s current holding of 39 stocks sees a high concentration of Japanese companies (43%), followed by the US (34%), comprising well-known names such as Nvidia and Mitsubishi Electric.
The second category of ETFs within the AIML genre are the ones that track a benchmark index that looks to use sophisticated statistical models to select the weight of the assets when constructing the index. One such example is the Rogers AI Global Macro ETF, launched in June this year by the ETF Managers Group with ticker BIKR US. With a TER of 1.18% this fund stands out from the crowd, but perhaps not in the way intended.
The portfolio is built using an asset allocation model that utilizes AI techniques to analyse macroeconomic data on a monthly basis to identify likely changes in market directions among a list of 40 countries. The model can take an overweight position in an ETF with exposure to short-dated US Treasury Bonds if that looks like the better option.
It’s way too early to comment on either the performance of the fund or indeed the assets under management, but the lack of facts in the factsheet leave a lot to be desired. While boasting that this ETF is not a “Blackbox,” the statement that the allocation of each single-country ETF in the Index is based on a proprietary artificial intelligence-driven algorithm suggests otherwise.
High fees and a lack of transparency are not the characteristics of a successful ETF product. If the fund delivers good returns, then it will be a useful investment tool, but it might be a mis-step to make so much noise about the association with Jim Rogers. That type of marketing proposition is more akin to the cult of the star ‘Stock Picker’ which is the antithesis of what made the ETF industry the success that it is today.
As a theme, AIML ETFs will most likely increase in popularity, and I have no doubt the best AI-centric algos will in time deliver on their promise of ‘faster, better, cheaper’ products, but what’s in a name. Nobody doubts that the tech sector has been the powerhouse that has driven the markets to record highs, but more than ever when investing in an AIML themed ETF, it seems the due diligence process will need to play a bigger role than usual.