‘AI 2.0 is about finding some of those new applications as well as some players that might be doing things more efficiently,’ says Chhad Aul
The AI investment boom that drove markets in 2023 and 2024 is now being put to the test, as investors who once tolerated massive spending on artificial intelligence infrastructure are insisting on proof of returns.
But Chhad Aul sees AI’s return on investment as a central market driver rather than a theme that gets “resolved” in a single year. He expects sentiment to swing between phases when investors happily tolerate stretched AI valuations and periods, especially around earnings seasons, when they scrutinize which companies are genuinely turning that spend into profits.
“There's going to be periods where there's a concern or focus on who are the relative winners versus the relative losers. You already began to see that last year. There’s lots of things going on with these companies because they are fairly diversified,” noted Aul, CIO and head of multi-asset solutions at Sun Life Global Investments (SLGI), pointing to last year’s divergence among big tech names as evidence.
“But if you look at even just Alphabet and Meta, two of the Mag 7 that were amongst the biggest, in terms of making capital investments in AI and data centers, they had very different price action. Alphabet was up 60 per cent last year while Meta was essentially flat,” said Aul, emphasizing that gap reflects the market already distinguishing “profitable investments versus those that were a little bit less clear to see where that was going to drive the bottom line,” he added.
That kind of discrimination, he argues, is only going to intensify. After two or three years of pilots and rollouts, investors and analysts will press harder as they will begin to want concrete examples of productivity gains and better customer experiences, rather than just vague ambitions. Aul sees this as the start of a multi‑year “super cycle,” where AI continues to influence market action across sectors and through multiple phases of the broader economic cycle, rather than a short‑lived fad, he said.
He splits AI’s impact on returns into two main categories. First are the hard, quantifiable metrics that flow straight into the numbers. AI should lower labour costs and trim operating expenses by cutting errors and streamlining processes, which in turn should lift margins and show up clearly in earnings. There is also the top-line effect, where AI helps drive higher sales or better conversion rates - outcomes that can be measured and directly linked to specific deployments.
Then there are the softer, harder-to-measure gains. Better customer experiences, deeper engagement and stronger brands don’t lend themselves to a neat ROI calculation, but he still views them as central to the value case for AI.
When taken together, these hard and soft impacts are set to become a standing item on earnings calls, as investors press companies to spell out how their AI investments are actually delivering, he noted.
Aul sees institutional investors and pension funds facing the same calculus as any other business when it comes to AI adoption. For his own investment team, the opportunity lies in boosting productivity across multiple functions like reallocating resources to higher-value work as well as automation and data transformation of tasks that were previously done manually.
While those gains can be measured in concrete terms, some benefits are harder to quantify, noted Aul. Delivering clearer analysis to investors, improving the quality and speed of reporting, and strengthening client communication all matter when it comes to fulfilling fiduciary responsibilities, even if they don't translate neatly into a single ROI figure.
According to Aul, the framework is the same whether you're running a pension plan or a corporate operation, underscoring that institutional investors need to weigh the upfront costs against a mix of hard metrics and softer, longer-term improvements in how you serve your stakeholders.
Aul argues investors should resist the urge to make an all-or-nothing call on AI as he expects continued strong performance punctuated by bouts of valuation anxiety, which makes extremes risky in either direction.
"You don't want to be taking a strategic perspective or even medium-term perspective or an extreme position on whether this is a bubble. This is not a bubble. It's not a black and white scenario that's going to kind of play out in a single year," he said.
He also pushes back on the idea that AI is the only game worth playing. While it dominated headlines in 2023 and 2024, last year told a different story as Canadian metals stocks and European defence names both outperformed AI.
“A lot of disruption and change is unlocking other areas of the market that have been unloved for years with better valuation. The broad answer is maintain some diversification globally and have some exposure to AI, but don't make that your only play,” he added.
That’s why he cautions against trying to define a single “good” ROI number for all AI initiatives, suggesting it’s more useful to think in terms of payback periods.
Based on what he’s seen so far, many AI projects take longer to earn back their costs than traditional IT investments, often in the range of two to three years. In that context, he views something around a 30 per cent return as a reasonable benchmark for AI, not as a rigid rule but as a practical starting yardstick.
The key, however, is whether these projects can realistically recoup their investment over that first two‑to‑three‑year window, recognizing that the answer will vary by use case and sector.
Aul expects stretched AI valuations to outlast most investors' patience. He takes some comfort in the fact that doubt still exists around current prices and the path to profitability.
"Usually, once you're into a bubble and approaching the popping of that bubble is where all sort of skepticism and disbelief has been dispelled. Everyone's kind of bought into the story so it's hard to find a skeptic," he said.
He contrasts today's environment with the dot-com era, when investors abandoned traditional valuation methods in favour of metrics like clicks and eyeballs. While that kind of rationalisation isn't happening yet, he believes it's coming as transformative technologies tend to attract excess capital.
"It's just kind of how capitalism works,” he said. “When there's something new and so transformative, like AI, there's going to be an excess of investment. And then there's gonna be some winners and losers and some consolidation. Then, there will be a bit of a mess and it'll be cleaned up after the fact," he said.
Ultimately, Aul believes the next phase of AI investing may look different from the chatbot-dominated narrative of recent years as Chinese competitors like DeepSeek have demonstrated that AI capabilities can be achieved more efficiently, while robotics applications represent what Aul calls "a physical manifestation of AI."
For active managers willing to look beyond the obvious names, "AI 2.0 is about finding some of those new applications as well as some players that might be doing things more efficiently," he said.
