While investors around the world rallied in the aftermath of the COVID-19 market crash, Scott Juds’s artificial intelligence-driven ETF was languishing.
The WIZ Bull-Rider Bear-Fighter Index, which uses AI to track changes in markets that determine whether it should shift its portfolio of exchange-traded funds to skew either more aggressive or conservative, suddenly couldn’t make sense of the data after an aberration as large as COVID-19.
“You had the initial shock of things which was followed by a series of closings and openings,” said Mr. Juds, the co-founder of Merlyn.AI, which runs WIZ.
He said the constant back-and-forth threw off the signals that AI use. “When it does that in a period of three months or less, you can’t properly determine momentum.”
As a result, WIZ is up just 7 per cent since its inception in October, 2019. Compare that with the S&P 500, which is up 37 per cent over the same period. Other simple index-tracking ETFs have posted similarly positive returns.
The performance of Mr. Juds’s ETF, which has suffered consistent losses since 2020, has sent investors running. At its peak, WIZ and DUDE (another Merlyn.AI ETF) had assets-under-management values of roughly US$250-million. Today, it’s just US$50-million.
But with AI constantly learning and big tweaks being made to the software with help from advisers, Mr. Juds is optimistic and said there has been considerable new interest in his products as long as they remain steady in the near future.
While multiple AI-driven ETFs have so far failed to beat the market, people like Mr. Juds still believe AI will be able to look through cluttered data to make investment decisions and eventually extract the best gains. Others believe it’ll be revolutionary for the user experience by giving retail investors greater education and control in customizing their portfolio while guiding them through different risk profiles.
Artificial intelligence has been one of many tools that large investing firms have consulted for years, but the popularity of ChatGPT has brought discussions of how AI can be relevant to investors at the retail level.
Mr. Juds said AI’s ability to find opportunities is rooted in the signal-to-noise ratio in investing.
There are countless data points in the world of equities that are simply background noise: They don’t mean anything. But buried deep within are valuable signals that point to meaningful investment opportunities in specific sectors.
“A very small signal like a tenth of a per cent in one day trend could get a 25-per-cent gain over a year,” said Mr. Juds. “But a 0.1-per-cent change in the market, that can get lost.”
There are still people who remain skeptical about whether AI can actually bring greater returns simply by trying to make smarter or quicker trades.
Joel Blit, an associate professor at the University of Waterloo specializing in the economics of innovation, said there’s an old adage that fund managers are no better than monkeys when it comes to picking stocks.
“Why would we think that an AI system would be any better? If they can parse through large amounts of data and find the needle in the haystack, then presumably they could do better,” said Prof. Blit, but he said there are examples of AI stock pickers that have been unable to beat the market so far.
When it comes to faster trading times, Prof. Blit said that if every big hedge fund had ultrapowerful AI making quick trades to make the best gains, then everyone having similarly powerful programming could negate any real increase in returns.
On the flip side, if AI reads too deeply into certain signals and makes ill-conceived decisions, it could lead to more market volatility.
“If there’s some kind of signal that’s heavily correlated to past bear performance in the market and all of a sudden all these algorithms start selling at once, it could lead to a major market collapse,” said Prof. Blit.
That’s why a whole other world of AI experts see a different potential for the technology: to bring knowledge that has always been inaccessible and dense to the everyday investor in a way that’s personalized and meant to help guide investing strategy.
This form of AI could help someone figure out if their portfolio is too heavily weighted to a certain sector, or is too susceptible to changes in the credit market or an economic downturn.
Companies such as Global Predictions are already providing advice to thousands of investors with billions of dollars, with input from AI.
Led by Canadian chief executive officer and co-founder Alexander Harmsen and based in San Francisco, the company created an AI-driven platform to help people make future investing decisions.
The program, called PortfolioPilot, allows people to simply plug in the details of their financial life such as their debt, real estate and investment accounts to receive nuanced advice on whether the investments they’re making actually match the goals and risk appetite they have.
A new ChatGPT plug-in by the company allows people to have basic conversations with an AI that can make similar suggestions, simply by reading a copy and paste of your investing statements. If you have a couple extra thousand dollars you’re looking to invest, it can give you suggestions based on your existing portfolio about where to spend next.
The idea builds on the already-revolutionary effect that robo-advisers and simple investing products such as ETFs have had on making it easier to be a self-directed investor.
“The main value in AI is about personalization and being able to democratize access to this sort of expertise,” said Mr. Harmsen.
“We can reflect back to people that you’re not diversified enough or that your risk-adjusted return you’re taking isn’t high enough. We feel people need that extra step of, ‘Here’s a couple things you can do to improve your portfolio.’”
PortfolioPilot is free to use and has advised roughly 5,500 users with more than US$3.4-billion in assets. The company plans to eventually make money by releasing its own set of AI-guided ETFs.
Mr. Harmsen doesn’t see AI bringing human wealth managers to extinction. But he does hope it could push wealth managers to step up their game by increasing interactions with clients, lowering fees and moving away from simple risk-profile categories for investors to choose from, and instead allow for more personalized approaches to portfolios.
Already, he’s seen interest from portfolio managers who’d like to use his program.
Edward Kholodenko, CEO of Questrade, said he sees client experience as the lowest hanging fruit where AI can make a difference. That includes helping clients understand how to make trades or how to complete a certain function on the company’s website.
He said Questrade is also developing and using AI internally, but providing advice to the public could be tricky because there are no regulations around the technology or its use of user data.
“It’s an area you have to be very, very careful … in terms of mining and using the data. We’re examining how to use the data to help our customers become more successful and financially secure,” said Mr. Kholodenko.
James Rockwood, founder and CEO of the fintech company CapIntel, said another obstacle is ensuring that AI remains compliant with rules if it were to directly provide financial advice, a service that is heavily regulated.
“People talk about how ChatGPT is so confident in everything it says but it doesn’t yet know if what it says is correct,” said Mr. Rockwood.
“You could run into issues where an AI could say something like, ‘This guarantees 100-per-cent returns,’ and a person can land in hot water.”
The Globe and Mail reached out to the Office of the Superintendent of Financial Institutions and the Canadian Securities Administrators, but neither regulatory body provided comment on whether national regulations for AI usage in Canada are coming.
A study commissioned by the Autorité des marchés financiers, Quebec’s financial regulatory body, and undertaken by the University of Montreal and Polytechnique Montréal, recommended the government create a framework for the use of AI that would identify unacceptable practices and data regulations for the use of this technology.
Depending on self-thinking robots for investment decisions certainly doesn’t come without risks. Mr. Kholodenko said investors should have a sober approach to AI, since people could create bots that push people to buy products that are not in their best interest.
One thing that experts agree on is that AI technology is in its infancy. As the technology develops and the amount of historical data that AI is able to access continues to grow, Mr. Rockwood said any prediction about where AI will prove to be most valuable in the financial world is simply that: a prediction.
“I don’t think anybody is talking about what AI is today when they’re having these discussions,” he said. “They’re talking about what could happen in the future, and that future has such a wide set of potential outcomes.”
TORONTO – Canada’s main stock index was up more than 100 points in late-morning trading, helped by strength in base metal and utility stocks, while U.S. stock markets were mixed.
The S&P/TSX composite index was up 103.40 points at 24,542.48.
In New York, the Dow Jones industrial average was up 192.31 points at 42,932.73. The S&P 500 index was up 7.14 points at 5,822.40, while the Nasdaq composite was down 9.03 points at 18,306.56.
The Canadian dollar traded for 72.61 cents US compared with 72.44 cents US on Tuesday.
The November crude oil contract was down 71 cents at US$69.87 per barrel and the November natural gas contract was down eight cents at US$2.42 per mmBTU.
The December gold contract was up US$7.20 at US$2,686.10 an ounce and the December copper contract was up a penny at US$4.35 a pound.
This report by The Canadian Press was first published Oct. 16, 2024.
TORONTO – Canada’s main stock index was up more than 200 points in late-morning trading, while U.S. stock markets were also headed higher.
The S&P/TSX composite index was up 205.86 points at 24,508.12.
In New York, the Dow Jones industrial average was up 336.62 points at 42,790.74. The S&P 500 index was up 34.19 points at 5,814.24, while the Nasdaq composite was up 60.27 points at 18.342.32.
The Canadian dollar traded for 72.61 cents US compared with 72.71 cents US on Thursday.
The November crude oil contract was down 15 cents at US$75.70 per barrel and the November natural gas contract was down two cents at US$2.65 per mmBTU.
The December gold contract was down US$29.60 at US$2,668.90 an ounce and the December copper contract was up four cents at US$4.47 a pound.
This report by The Canadian Press was first published Oct. 11, 2024.
TORONTO – Canada’s main stock index was little changed in late-morning trading as the financial sector fell, but energy and base metal stocks moved higher.
The S&P/TSX composite index was up 0.05 of a point at 24,224.95.
In New York, the Dow Jones industrial average was down 94.31 points at 42,417.69. The S&P 500 index was down 10.91 points at 5,781.13, while the Nasdaq composite was down 29.59 points at 18,262.03.
The Canadian dollar traded for 72.71 cents US compared with 73.05 cents US on Wednesday.
The November crude oil contract was up US$1.69 at US$74.93 per barrel and the November natural gas contract was up a penny at US$2.67 per mmBTU.
The December gold contract was up US$14.70 at US$2,640.70 an ounce and the December copper contract was up two cents at US$4.42 a pound.
This report by The Canadian Press was first published Oct. 10, 2024.
