AI In Asset Management

By: Joel Schlesinger

INVESTMENT industry veteran Peter Simmie has long been on the hunt for what makes a company a good investment.

But after 35 years, he’s largely given up on finding them.

That is, without the help of a robot.

OK, it’s not quite a “robot.”

Rather, it’s advanced hardware married with even more advanced software to produce an artificial intelligence-driven investment process to find stocks with the best chances of future high-dividend growth.

“In our business, you’re trying to get rid of personal preferences and judgment,” says Simmie, chief investment officer with Bristol Gate Capital Partners in Toronto, one of the pioneers in the industry when it comes to implementing artificial intelligence (AI).

And using a computer program that can pore over vast amounts of information much faster than a team of analysts could, which then also comes up with near-flawless analysis, does just that: it takes the human-error factor out of investing.

“I’m not sure AI, in the way we’re using it, could do anything that we couldn’t do ourselves intellectually, but it does it a whole lot faster and with much more consistency,” he says.

AI, machine learning and data analytics are the buzzwords when talking about tech today. Indeed, they represent significant innovations in computing that have truly become game-changers for just about everything — from how we drive to how we bank to our health care.

But Simmie says the investment world has been lagging behind when it comes to implementing AI and related technologies.

“A lot of the people in the investment management business have been trained to be in the investment management business.”

In other words, they’re often steeped in the orthodoxies of investing. But they’re not computer geeks. And while AI has been slowly adopted by financial institutions for other reasons, like risk analysis of loans, using a computer brain to pick investments hasn’t been a priority.

But that’s changing fast, driven by firms like Bristol Gate, says Simmie, an alumnus of the University of Winnipeg. In fact, nobody paid much attention to the company and its strategy of successfully finding stocks with high-dividend growth until it started using AI to do it.

“It’s such a hot topic that we now get asked to make presentations in the States.”

That’s hardly surprising to longtime investment industry insider John Guttormson, who has worked with some of Canada’s largest firms on implementing innovation.

“The traditional active management participants in the industry are in a bit of a tizzy,” the Winnipeg-based investment product consultant says.

Another industry innovation, exchange-traded funds (ETFs) that generally track and mirror a benchmark index, like the S&P 500, charge extremely low management fees relative to mutual funds and even personalized portfolio management.

“It’s always been so difficult to outperform the benchmark, and it used to be if you got a benchmark rate of return, clients were happy,” he says.

Now everyone can buy the benchmark performance for pennies on the dollar paid to have a professional struggle year in and year out to do the same — thanks to ETFs.

Guttormson says AI is the industry’s “white hat to the rescue” in this respect.

AI allows investment managers to do what they do, but presumably much better, he says. And what is it investment managers aim to do? Why, that’s to outperform their benchmarks. (Otherwise, why do you pay them extra money, right?) With AI, the theory is that they should. And Simmie says Bristol Gate’s AI-driven portfolio has done just that.

Central to AI’s promise — what makes it stand out from past computerdriven strategies — is its ability to “get smarter over time.”

Like us, it learns as it goes… only faster and with fewer errors.

This is potentially good news for investors of all kinds — though for now AI is most likely to benefit very wealthy and large investors like pension plans.

Yet AI will reshape how we all invest, Guttormson says.

“There is a professor at Stanford who says that AI is the new electricity, and we all know how pervasive electricity is,” he adds, referring to Andrew Ng, the chief scientist at Baidu Research, one of the world’s leading R&D firms in AI, and who also teaches at Stanford University in California.

Already, the intrepid little investor can access AI technology to boostreturns — presumably — in his or her portfolio.

Case in point is Canadian ETF provider Horizons, which has offered its MIND Active AI Global Equity ETF — one of the world’s first AI-run funds — since late 2017.

Anyone with an online brokerage account can buy and sell it.

“The idea of a product like MIND is it’s learning to learn as it goes along,” says Hans Albrecht, vice-president with Horizons ETF.

“With these products, it’s quant, turbocharged.”

Albrecht is referencing quantitative analysis, which has been around for a few decades, and involves using advanced mathematics to create computer algorithms to find needles in haystacks of data. But the old (non-AI) quant programs are limited. They do the same thing over and over, whereas AI learns from success and, more importantly, mistakes to get better as time passes.

So it should come as no big surprise that MIND’s performance since inception is hardly earth-moving. If you had invested $10,000 at its founding, you’d still have $10,000 today.

That uninspiring performance is to be expected.

MIND is powered by a “neural network” built by a Korean company. And the best way to think of it is like a child learning rapidly over a short time.

So the theory goes: MIND’s software should get better at picking outperforming markets around the globe over time.

Whether that proves true or not, expect more products like MIND to come in the near future.

Albrecht argues AI is reshaping the industry just like other innovations have in recent years (i.e. the internet and smartphones).

“Fifteen years ago, it was like, ‘The internet is really cool,’” but not essential like it is today, he says. Google, Facebook and Amazon were not yet market drivers (and, today, big investors in AI).

Apple was just developing the iPhone when few people could have predicted that in a few short years, “everybody would have a powerful computer in their pocket.”

Thanks to these technological advances, we’ve seen an explosion in data and its sharing.

In fact, 90 per cent of the world’s data in all of human history has been created in the past couple of years, Albrecht adds.

“That has changed everything.”

So given the promise of AI, just imagine what the world — and your investment portfolio — might look like 15 years from now.



Learn about News, Events and how we are using Data Science to eliminate human bias.