For years, AI companies have been training their machines using copious amount of data. The result? AI that can provide seemingly insightful and sometimes humanesque answers.
In recent months since the release of ChatGPT, Artificial Intelligence (AI) has captured the world’s attention. It can hold a conversation, come up with party ideas, write logical essays and even pass US law entrance exams and the US medical licensing exam!
So, what if your fund manager is an AI?
Can it help you capture greater returns in the markets and grow your wealth more efficiently?
AI-powered ETFs have been around since 2017, the first being powered by IBM’s Watson. This means we have real market data to determine if an AI fund manager can actually outperform the market (and their fellow human counterparts).
If you’re also wondering whether AI can help us invest better, read on.
What are AI powered ETFs?
AI powered ETFs are exchange-traded funds that leverage Artificial Intelligence technology to automate portfolio management decisions. By utilizing machine learning algorithms, these AI-powered ETFs are said to provide higher returns than traditional investing methods.
Why use AI powered ETFs?
No matter how experienced a human fund manager is, there is no way they can monitor every available data point, backtest every possible strategy nor get rid of biasness completely.
The draw of having an AI is that with sufficient computing power, it can do all of the above.
But AI powered ETFs are not built the same…
There are several AI powered ETFs available in the market now. What makes each of them unique boils down to how their AI machine were trained and how the ETFs are being managed.
Let’s take a look at the popular AI powered ETF available and how they stack up:
Best AI powered ETFs
| AI Powered Equity ETF (AIEQ) | WisdomTree U.S. AI Enhanced Value Fund (AIVL) | SPDR S&P Kensho New Economies Composite ETF (KOMP) | Qraft AI-enhanced US Large Cap ETF (QRFT) | AdvisorShares Let Bob AI Powered Momentum ETF (LETB) | |
| Inception Date | 17 Oct 2017 | 18 Jan 2022 | 22 Oct 2018 | 20 May 2019 | 8 Feb 2022 |
| Investment Objective | Uses AI to build predictive financial models that improves performance | Seeks alpha using AI trained on smart beta value concepts | Invests in innovative companies driving the fourth industrial revolution | Provide capital gains by investing in factors that are outperforming in current market | Select for outperformance based on a mix of fundamentals and price momentum |
| Expense ratio | 0.75% | 0.38% | 0.2% | 0.75% | 1.07% |
| AUM | US$104M | US$0.42M | US$1.7B | US$4.7M | US$26M |
| Performance (Since Inception) | 30% | 0.5% | 50% | 45% | -9% |
| Performance (YTD) | 14.12% | 3.57% | 12.56% | 7.17% | 0.17% |
AI Powered Equity ETF (AIEQ)
AIEQ is powered by an AI that relies on IBM’s Watson. It claims to “equal a team of 1,000 research analysts, traders and quants working around the clock”, minus the human bias and errors.
At the point of writing, it is up 30% since inception, contains 142 stocks, of which its top 3 holdings include Las Vegas Sands Corporation, RH and Constellation Energy.
- Inception Date: 17 Oct 2017
- Expense ratio: 0.75%
- Exchange: NYSE ACRA
- AUM: US$104M
AIEQ’s Investment Process
AIEQ’s AI builds predictive models on 6000 listed US companies using millions of data points that include financial data, news, market sentiments, macro economic data and many more.
Did AIEQ beat the market?
If you had bought AIEQ since its inception, you would be up ~30%. Unfortunately, you would be underperforming the market (S&P500) which returned 58% in the same period.
AIEQ claims that with time and more market data, its artificial intelligence and machine learning capabilities would allow its investment system to improve. Only time will tell.
WisdomTree U.S. AI Enhanced Value Fund (AIVL)
AIVL aims to deliver outsized returns by using their quantitative AI model to pick value stocks.
At the point of writing, it is up 86% since inception, contains 561 stocks, of which its top 3 holdings include Bunker Corporation, Teledyne Technologies and Leidos Holdings.
- Inception Date: 16 Jun 2006 (Updated fund objective on 18 Jan 2022)
- Expense ratio: 0.38%
- Exchange: NYSE ACRA
- AUM: US$0.42M
AIVL’s Investment Process
AIVL allows human and AI to link hands in order to chase for greater returns. It relies on human fund managers to collect and clean data points which are used to train the AI. The AI then selects the best stocks using today’s data points. The human fund managers uses the selection provided to build the portfolio, taking risk management and timing trades for maximum return potential.
Did AIVL beat the market?
Original known as the WisdomTree US Dividend ex-Financials Fund, AIVL updated its fund objective on 18 Jan 2022.
If you had bought AIVL since it updated its objective and let the AI take over, you would be down 0.5%. But you would have outperformed the S&P500’s -10% return in the same period.
SPDR S&P Kensho New Economies Composite ETF (KOMP)
KOMP focuses on disruptive and innovative companies listed in the US, with a focus on New Economies companies like those in emerging technologies, robotics, automation, artificial intelligence and more.
At the point of writing, it is up 50% since inception, contains 561 stocks, of which its top 3 holdings include Bunker Corporation, Teledyne Technologies and Leidos Holdings.
- Inception Date: 22 Oct 2018
- Expense ratio: 0.2%
- Exchange: NYSE ACRA
- AUM: US$1.7B
KOMP’s Investment Process
KOMP uses AI and a quantitative weighting methodology to track the S&P Kensho New Economies Composite Index which is also built using AI and Natural Language Processing (NLP). Using NLP, the AI looks for companies driving the new economy sector (aka the Forth Industrial Revolution) across more than 20 sub-indexes.
Did KOMP beat the market?
If you had bought KOMP at inception, you would be up 50%, outperforming the S&P500’s 49% in the same period.
Qraft AI-enhanced US Large Cap ETF (QRFT)
QRFT aims to deliver returns over the long term. Its AI is built to seek alpha based on 5 proven factors; quality, size, value, momentum and low volatility.
At the point of writing, it is up 45% since inception, contains 351 stocks, of which its top 3 holdings include Apple, Microsoft and Berkshire Hathaway.
- Inception Date: 20 May 2019
- Expense ratio: 0.75%
- Exchange: NYSE ACRA
- AUM: US$4.7M
QRFT ‘s Investment Process
QRDT uses AI to “continuously learn the correlation of factor returns with various macroeconomic and valuation conditions”. This allows the AI to shift its investment system dynamically towards factors that are outperforming based on the current markets.
This would be comfortable for investors who are familiar with factor investing, and believe that QRFT can identify patterns from outperforming factors that human fund managers may miss.
Did QRFT beat the market?
If you had bought QRFT since its inception you would be up by ~45%, outperforming the S&P500 which returned ~37% during the same period.
This is pretty encouraging. That said, only time will tell if QRFT’s AI can continue to outperform the markets over several market cycles.
AdvisorShares Let Bob AI Powered Momentum ETF (LETB)
LETB relies on a mix of fundamental and technical analysis. Its AI identifies listed companies with high price momentum.
At the point of writing, it is down 9% since inception and is fully in cash.
- Inception Date: 8 Feb 2022
- Expense ratio: 1.07%
- Exchange: NYSE ACRA
- AUM: US$26M
LETB’s Investment Process
LETB uses AI select for outperformers based on a mix of fundamental sentiment analysis and price momentum. As a risk management measure, it also takes macro signals into consideration.
Did LETB beat the market?
LETB is a relatively young AI powered ETF. If you had bought it at inception, you would be down 9%. Comparatively, the S&P 500 was down 15% in the same period.
LETB’s strategy looks similar to what we use in our Quality Momentum Trading strategy. During bear markets, there could be periods where the portfolio is fully in cash to avoid trading in a general downtrend.
Risk of AI Powered ETFs
While artificial intelligence provides an interesting take to investing, you should take note of the key risks of relying on AI powered ETFs.
1) Black Box Strategies
Although some funds provide detailed information about how their AI works, most of them remain black box strategies. This is understandable as most fund managers wouldn’t want to share their algorithm in entirety to prevent competitors from stealing their IP.
Hence, you may often need to understand the objective of the ETF and trust that the AI employed works as described.
2) Short Track Record
AI powered ETFs are fairly new in the market. You can only rely on the short history that we have thus far to make an investment decision. Only time will tell which AI can perform consistently through multiple market cycles.
3) Reliance on Training Data
As we often say in investing “past performance is not indicative of future returns”. The underlying principle of AI powered ETFs is that the AI can pick up patterns based on an insane amount historical data.
The biggest risk of this principle is that the AI may not be able to handle black swan event. That said, most human fund managers can’t either.
Conclusion
AI powered ETFs are providing investors with unprecedented access to automated, quantitative investing strategies. By removing human bias and error, these ETFs could have the potential to outperform the markets over long periods of time.
When it comes to picking the best AI powered ETFs, on top of reading about the Fund’s objective, you’ll need to understand how the AI is trained, how it is used in the investment process.
AI powered ETFs are still relatively new in the market and should only be a small part of your entire portfolio. Do keep the risks of investing in these ETFs in mind!
If you’re interested to invest in AI as a technology instead, check out our pick of the Best AI stocks.