ARTIFICIAL INTELLIGENCE has been a common trope in Hollywood for decades now. The idea that machines could one day
turn against humankind isn’t that ridiculous a premise for most people. Despite those concerns, AI, Big Data and analytics are tools of the present, not the distant future. According to Statista, worldwide revenue from Big Data and business analytics is
expected to top $US200 billion by 2020.
In asset management, firms are now getting behind this technology in a big way. One company throwing its weight behind artificial intelligence is Horizons ETFs Management. The firm launched Canada’s first AI-focused ETF in November, the Horizons Active AI Global Equity ETF (MIND), which uses artificial intelligence for all of its securities selections. Using a machine learning process known as Deep Neural Network Learning, the fund offers a glimpse into what the future of the industry might look like, explains Horizons co-CEO Steven Hawkins.
“We used 10 years of data points for the system to learn from; from that, it was able to generate a predictive portfolio for better
investing,” Hawkins says. “It uses what it has learned from running through millions upon millions of scenarios based on that data, and it has learned to become a better investor.”
Hawkins and his team have placed a lot of faith in MIND, to the extent that neither they nor subadvisor Mirae Asset Global
Investments will interfere in investment decisions. The fund is rebalanced every 30 days; AI dictates which underlying assets are retained or replaced.
“We are not going to override the trade decisions it makes; we will simply execute those decisions and see what happens,” Hawkins says. “We believe that an AI system lacking bias will make better decisions than a human portfolio manager.”
Security selection is one function AI brings to the table in asset management, but there are many more. Fred Tavan, global head of Sun Life Financial’s Innovation Lab, explains how his company is using AI right now.
“We are using it in corporate marketing, and what corporate marketing is developing will be rolled out across all our 18 business
units,” Tavan says. “Each one of those units will use analytics to find out where clients think we are doing well and where they think we are not doing well by analyzing the unstructured commentary they have fed us through surveys.”
Natural language processing and sentiment analysis are the next evolution of market research, and they’re tools Sun Life will increasingly use to get a general view of their customers.
“This is something that wasn’t used before in terms of data – usually it would be multiple-choice answers,” Tavan says. “For a human being to analyze a lot of unstructured text and come up with common themes, when you are looking at thousands of survey responses, the process is very difficult. For a machine, it’s much easier.”
While advancements in technology are undoubtedly a positive, there is a negative side to this progress. The likelihood of a
Terminator-style war against the machines is still pretty fanciful, but robots taking the place of humans in the workplace certainly isn’t. Much like automation in manufacturing has led to mass job losses for decades, will AI deal a similar blow to the financial services sector?
In a recent article, Gaurav Chakravorty, co-founder of asset management firm Qplum, predicted that about 90% of discretionary
traders will lose their jobs, being replaced by machine-learning engineers and data scientists as AI takes hold. This new era isn’t decades away; rather, we are already in the formative stages, as Hawkins points out.
“One of our subadvisors, Guardian Capital, is actively using AI every day to go through all of the news stories around the world affecting their universe of individual companies,” he says. “They use AI to interpret that data and predict growth, dividend growth, dividend cuts.”
While AI will clearly eliminate the need for certain jobs, it likely won’t lead to empty offices at asset management firms. When it comes to investing, there are certain times when the personal touch is necessary.
“There is always a human element that’s needed,” Hawkins says. “People want to know why decisions are being made, and because the AI system will never be able to tell us why it is making decisions ... I don’t think that will ever go away.”