Connect with us
AI Investment Funds AI Investment Funds

FinTech

The AI ETF: Frankenstein and the Swan

 3 min read / 

The Artificial Intelligence Equity ETF, that started trading last year, has the potential to revolutionise the investment industry, but what happens if it goes wrong?

“The development of full artificial intelligence could spell the end of the human race” – Stephen Hawking (Cellan-Jones, 2014)

On the 18th of October 2017, ‘EquBot’ launched a new fund that looks to outperform the S&P 500 benchmark, by investing in a diversified holding of 40-70 companies. The company claim that the fund manager (a man named Watson), has the capability to process millions of pieces of information and company announcements in seconds, and what’s more, he does it all on his own. Unfortunately, institutional investors looking to book lunch with Mr Watson may struggle as he’s a hard man to reach.

Officially, Watson goes by the name IBM Watson and is ‘a cognitive computer platform’ that connects large amounts of structured and unstructured data (EquBot, 2018). The AI Equity ETF partners with IBM Watson in order to trawl through information on thousands of US companies. It finds unique investment opportunities, based on a number of pre-programmed algorithms, but also uses the data to conduct the process of machine learning, where Watson ‘learns’ about market trends. Watson’s initial learning curve was obviously a steep one, with the AIEQ losing 4% in the first twenty days of trading, only to miraculously recover and generate a return of 6.53% (11% since the twenty day post-launch low) at the time of writing, compared with the S&P 500 return of 6.57% (since the 18th of October).

The AIEQ clearly offers an incredible way of mining extremely large datasets in order to find investment opportunities, but one of the key questions has to be how the AI would react in a market crash or black swan event. One of the theories behind the unexpected market volatility experienced towards the end of January/beginning of February was an ETF sell-off, with stop losses in these passive instruments forcing the sale of additional securities, causing the market to overcompensate. Could this problem be multiplied in a severe correction? Ultimately, one needs to question the ability of AI to deal with such events and under the assumption that the software learns from experience, does it have the capability to react appropriately in the first ‘black swan’ instance? Having said this, humans themselves have not proven themselves to necessarily be capable of dealing with such a situation themselves.

AIEQ combines big data, mining, machine learning and AI into an instrument that in theory should prove an incredible level of insight into the world of excess returns. The older AIEQ gets, the wiser Watson becomes, as it experiences a greater amount of market movement and generates a larger sample upon which to base its decision. From a downside perspective, however, AIEQ must be continually monitored like the infant it is, in order to avoid any teething troubles, such as Frankenstein-esque bugs that the owner has no control over.

The potential for AIEQ is enormous and as long as there is a manual downside plan implemented, the AI ETF may begin to break the bounds between active and passive investing. AIEQ’s development will almost certainly prove extremely interesting. 

Have your say. Sign up now to become an Author!

Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Send this to a friend