What happens when humans stop picking stocks and give the job to AI? Not only was this the subject of one of the best-attended panels at this year’s INN8 Invest Summit, but it’s a question that both retail investors and asset managers are grappling with now.
Will it make the average retail trader an investment superstar? Will it usher in the demise of the asset management industry as we know it? Not quite. Or at least, not yet.
“There’s an enormous amount of hype in the press at the moment about AI. And I think, with years of market cynicism, you’ve got to look at some of what’s going on as talking your book,” Giles Heeger, head of Stanlib South Africa, tells the FM.
Consider that the two most prominent US AI firms — Anthropic and OpenAI — are set for potentially record-breaking IPOs this year and next, notwithstanding mounting fears that they cannot control the technology they have unleashed through their large language models, Claude and ChatGPT.
Heeger, who began his career in markets in 1996, started the year by tinkering in his personal capacity with AI tools to get a feeling for the tech. As for his staff, he says: “We’ve got some investment teams that have quite a big investment in systematic approaches, so they go through a lot of very big data sets. And what used to be a painful, slow process is now incredibly fast.”
As a general executive, on the other hand, it’s a bit like having a few MBA grads on tap. An AI tool “can do a whole lot of stuff that an MBA grad would do for you, but it’s just as likely to be wrong, and it does need the same explicit instructions as an MBA grad”, he says.
Critically, says Heeger, AI can help humans do less of the frustrating, mundane, routine admin. But humans still have to apply their minds. “If you don’t know what to ask it, and you don’t know how to brief it, and you don’t know how to spot a hallucination, then it’s actually not helpful at all. And you only get that from having a lot of experience,” Heeger says.
A judgment cul-de-sac
There’s another point he says shouldn’t be overlooked in the rush to AI everything. “In our industry, I think it’ll be a mistake to use its usefulness to replace trainees because in seven or 10 years’ time, you won’t have the humans who’ve got the experience to use it. And then without them, you don’t have that judgment, and you’ve walked yourself into a cul-de-sac,” he believes.
Where AI can be transformational is in dealing with clients; now that bots have got that much better, in basic transactions AI “can give you an absolutely human kind of experience, which will improve service levels for the customer. That’s a big opportunity for an industry like ours,” Heeger argues.
But if all asset management firms, presumably, are looking at the potential of AI in much the same way, what is there to differentiate them in the industry?
For starters, says Heeger, this isn’t exactly new if you consider how the internet, in the past 25 years, “massively democratised data”.
“Those digital capabilities brought hundreds of thousands of day-traders into the market and removed a ton of moats. I would argue that’s why a lot of asset managers are battling to beat the index, because there is now so much information available,” he says. Latest figures from Morningstar bear this out: its research found that only 13% of active US large-cap managers beat passive rivals over the 10 years to June 2026.
Getting the culture right
This is why there’s a twofold imperative, he says, for asset managers: know what you want to achieve, and build the right team to do it.
“What is the way that you manage money? What is your philosophy? What is your process? It takes huge, consistent efforts to actually do that. That gives you a very small group of people, I think, who are competitive. And the second piece is culture.”
He recalls how, when he joined Stanlib nine years ago — leaving Libfin, the specialist risk and investment division he helped Liberty set up in 2008 — he asked one of South Africa’s top asset consultants for advice.
“They said: ‘You know, Giles, this game is super easy. All you’ve got to do is get a group of exceptional people together, get them working as a team and keep them together for a long time.’”
Which is of course anything but easy — especially when you’re wrangling a group of A-type individuals who’ll all inevitably hold strong views about their subject.
“Building an environment where these guys are comfortable challenging each other; where they get to compound their intellectual capital over a long period — that is the magic,” he says. “And the glue that makes it all possible is culture. And I don’t think AI can replace culture.”
Heeger’s first “real” job was at Standard Bank.
“I’d managed to scrape up some funding to do an MSc in finance in London, and while I was over there the rand blew out. So suddenly my rand financing didn’t look so strong,” he recalls.
This was in 1996, two years after South Africa’s first democratic elections, when initial panic over the health of the then president Nelson Mandela was compounded by political anxiety and rumours that the Reserve Bank was set to scrap exchange controls. It prompted a speculative attack on the currency, which fell about 20%, forcing the then Bank governor Chris Stals to maintain interest rates higher than 19%.
Libfin’s ‘aha’ moment
Heeger managed to “talk his way” into a part-time job at Standard Bank in London, returning to the investment bank after his accounting articles, and staying until he moved over to help establish Libfin.
It was then that Heeger had his first brush with the technology that has defined the AI boom of the past four years: graphics processing units (GPUs) and the firm behind them, Nvidia.
“We had a bunch of youngsters who were experimenting at the time with GPUs to train Standard Bank trading models, because they could compute so much faster,” he says. The aim was to run the far more complex life insurance market valuations on the GPUs.
“When I started at Libfin, the world didn’t have the computing power. And the guys on the frontier of it were trying to do it by renting downtime on other computers, when office parks were quiet. And we had a bunch of young actuaries who said: ‘Hey, I wonder if you could take these things out of the Sony PlayStation, because they’re really fast graphics, and if we could programme the spreadsheet maths into that, maybe we could run them faster.’”
Impressively, they got it to work. “For a very short period of time, like about eight weeks, we were leaders at this stuff,” laughs Heeger.
Fully invested
Stanlib has signed two transformative deals in the past two years to sharpen up its own tech: implementing US asset giant BlackRock’s Aladdin software to power its front-end investment processes, and working with JPMorgan Asset Management to provide back-office services.
“It is so hard to generate good investment performance. You’ve got to be, all the time, thinking about how to stack the odds in our favour. And one of the ways that you stack the odds is you reduce the amount of time that you spend on other stuff,” says Heeger.
“Our portfolio managers are fond of saying to me: ‘Giles, we need to be spending 90% of our time on investment decision-making, and Giles, your job is to make that possible.’”
As for his own personal investment philosophy, Heeger’s views were heavily influenced by the intersection between his time as a trader and working for a life insurer.
“I started in a world which was very trading-dominated. It was all very short-term, very risk-focused. And then from there, I moved across to a life insurance company where only five of us had a trading background and the rest were all actuaries who had a very long-term thought process, and they approached everything with a power-of-compounding mindset,” he says.
“The biggest takeout for me in this industry in so many respects is time: whether it’s building a good team or building a good business. It sounds trite but the most valuable thing is time, because it hides a lot of mistakes.”
In other words, if you have a long time to invest, don’t worry about the mistakes you might make on the way.
“It doesn’t matter, because time smooths [those] out. And if I’ve got the time, then I try to be as fully invested as possible.”
