The test: A Nordic AI-powered fund versus a Swedish Index Fund

Jacob Bergdahl, a full-stack developer and AI expert, examines how a fund that used AI to pick stocks compared with an index fund during the acid test of a pandemic. (Medium)

Humans do a poor job at beating the index, long term

It’s been proved time and again that humans picking stocks are generally unable to beat the performance of an index fund over an extended period.

Warren Buffet famously won his million-dollar bet against Protégé Partners LLC that including fees, costs and expenses, an S&P 500 index fund would outperform a hand-picked portfolio of hedge funds over 10 years.

Take mutual funds. According to Tony Robbins, 96% of all mutual funds in the United States fail to match or beat the market over any extended period. (Tony Robbins, Money: Master The Game, 2016, p. 96, ISBN: 978–1–4711–4861–3).

Bergdahl pivots to AI, and the shorter term

The question in Bergdahl’s mind: Does AI give better results versus an index fund, particularly in the shorter-term, and during a black swan event such as the coronavirus?

To find out, he compared a Nordic fund that uses artificial intelligence to select investments based on the object’s economic data – the FIM Artificiell Intelligens A. Interestingly, over half the fund’s corpus is invested in U.S. companies. The fund commenced using AI for stock selection on November 20, 2019.

The FIM Artificiell Intelligens A fund was pitted in the contest against Avanza Global. Avanza Global is a Swedish index fund that contains over 1,500 assets and closely follows the Amundi Index MSCI World Index.

The time frame for comparison was from November 20, 2020, when FIM started using AI, to June 19, 2020, a period that included the impact from COVID-19.

The result:

• FIM Artificiell Intelligens A: (-) 8.91%
• Avanza Global: (-) 4.85%

Clearly, AI was unable to match the power of an index fund during the virus crisis. If you take fees into account the gap would widen further – fees for FIM are 1.64% compared to Avanza’s 0.10%.

“If the AI is unable to handle volatile, unexpected markets, then what’s the point?” asks Bergdahl.

Across the pond: The AI-Powered Equity ETF (NYSEARCA: AIEQ)

Separately from Bergdahl’s exercise let’s take a look at the AI-Powered Equity ETF (NYSEARCA: AIEQ). Here are the salient features of the ETF:

• First and only actively managed ETF to fully utilize AI (IBM Watson) as a method for stock selection
• The system seeks to represent a team of 1,000 research analysts working 24/7 analyzing millions of data points daily

The multi-cap equity ETF’s inception date is October 17, 2017. Its expense ratio is 0.77%.

Here is how the AI-powered fund stacked up against the S&P500.

Since inception:

February to June 2020:

It’s observed that over the longer term AIEQ (+8.86%) has generally under-performed the S&P500 (+18.56%).

However, during pandemic times, AIEQ (-6.18%) performed slightly better than the S&P500 (-8.25%).

Related Story:   Can AI Predict a Child’s Future? No, it Can’t.