4 ways to play the long game in AI
March 25, 2026
-
5 min
The consensus view is that AI is the trade of the decade. The more interesting question is what happens when different stock-picking machines refuse to trade a theme and yet end up invested in companies that matter for AI’s future.
Taking four investment managers from the Natixis Investment Managers multi-affiliate family – Loomis Sayles, Harris | Oakmark, WCM Investment Management, and Vaughan Nelson – it’s interesting to note that none of them buys ‘artificial intelligence’ per se. There is no thematic sleeve, no top-down overlay, no model portfolio optimised around semiconductor capex cycles.
Instead, each starts with individual businesses and arrives at AI exposure as a consequence of its process. The difference in how they get there is what make the comparison instructive.
Loomis Sayles: Quality compounds
The Growth Equity Strategies (GES) team from Boston-based Loomis Sayles invested in Nvidia in 2019, years before the AI narrative dominated markets. The team’s conviction lay in Nvidia’s unique visual computing strengths, positioning it to lead secular AI growth, not simply gaming, over the team’s long investment horizon.
Since 2006, the GES team’s proprietary Quality-Growth-Valuation process has guided investment selection: quality and growth define what it wants to own; valuation determines when to own it. It diversifies its holdings by business driver, ensuring no single one dominates the mix.
‘AI GPU spending’ powers Nvidia’s growth and is the portfolio’s sole holding with AI as its primary long-term business driver. Meanwhile, Alphabet continues to thrive, with online advertising accounting for roughly 75% of its revenue – that’s the engine behind its secular growth. It leverages AI to enhance its large and profitable advertising business in ways that sustain and extend its competitive advantages.
Aziz Hamzaogullari, Founder & CIO of the GES team, says: “It's important to note that, like the disruption in internet-based computing, though the winners will be many in terms of end users and productivity boost, direct winners, those companies that directly sell into the value chain, will be very few because it requires a platform company and it requires a tremendous amount of research and development and CapEx dollars.”
The result is a concentrated set of high-conviction positions in businesses well diversified across distinct and lowly correlated business drivers to more effectively realize the benefits diversification.1
Harris | Oakmark: Hidden value revealed
Chicago-centred Harris analysts get excited when a company’s market price give them something for free. In their framework, AI is one more tool that can unlock such hidden value.
Take BNP Paribas, which reported €635 million in AI-driven value creation for 2025 through better customer engagement, lower data processing costs, and stronger risk management. Or IQVIA’s AI agents, that allow pharma companies to sift through clinical trial data faster and with better outcomes.
For Harris, AI is not the theme, rather it is a kaleidoscope of business-level improvements seeping through every sector, widening the gap between companies that integrate it well and those that do not.
Bill Nygren, Partner, Portfolio Manager and CIO for the US, says: “History shows the biggest winners in tech revolutions aren’t infrastructure companies but those using the technology to improve how they do business. Like Walmart’s adoption of computerisation drove out Sears and Kmart, and in the Internet era Amazon and Google dominated, not AOL or Cisco.” 2
WCM: Moat trajectory accelerated
California-situated WCM’s entire investment philosophy rests on the trajectory of a company’s competitive advantage: whether its moat is widening or narrowing. AI becomes relevant when it helps to widen the gap between a company and its competitors.
App Lovin’s AXON 2 AI engine, which dominates mobile ad mediation with roughly 80% market share, is a textbook example of a company with a widening moat. Celestica’s transformation from contract manufacturer to AI infrastructure architect positions it at the centre of hyperscale data centre buildouts. Siemens Energy captures the electrification wave as data centre power demand is projected to more than triple by 2035.
WCM monitors AI-related exposures across the portfolio, but always through the prism of moat trajectory and corporate culture and never as a standalone bet.
“Everybody has the same information, everybody can build a spreadsheet… it's the people that drive the competitive advantage to grow," says Paul Black, Portfolio Manager and CEO. “We have four analysts that just focus on business cultures. They don't try to assess the moat; they don't do any valuation work. They just try to assess the people side of the equation in every company that we invest in. And where we find strong alignment between culture and strategy we've got a long-term winner that's going do really well.” 3
Vaughan Nelson: Second-order opportunities
Houston-headquartered Vaughan Nelson employs a Capital Allocation Framework to develop insights into capital allocation trends and identify key inflection points that may shape future performance. Using this perspective, it believes that the most durable gains from AI are likely to emerge only if capital is redirected toward second-order changes – data platforms, production environments, and workforce retraining – and alignment with regulatory and energy considerations to enable scalable experimentation
The crucial question becomes whether a company’s capital allocation supports comprehensive reengineering around AI-enabled capabilities. Those who invest heavily in infrastructure without corresponding organisational redesign risk falling short of profitability and lasting impact. Conversely, those who pair infrastructure with disciplined second-order investment stand a stronger chance of achieving durable leadership and meaningful productivity gains.
Investment managers at Vaughan Nelson might suggest that if you must choose between ‘invest in AI today’ versus ‘invest in the systems that enable AI’, the mindful choice is to prioritise a restructured enterprise that can learn and adapt alongside AI, not just a stack of automation.
Adam Rich, Deputy CIO, comments: “Second-order innovation is not accidental, it is financed. The next era of market leadership will unlikely be defined by early adoption alone, but by disciplined capital allocation aligned with organisational redesign.” 4
Different styles, common threads
All four of these investment managers didn’t specifically seek out AI investments, they went looking for the companies they believed were best placed to outperform the market. What they recognised is that AI’s value emerges when it materially enhances a company’s economics rather than simply being an on-trend buzzword.
Rather than treating AI as a standalone theme, it is integrated into broader business models and corporate cultures. Exposure to AI arises from a combination of rigorous, bottom-up analysis of quality with gains rooted in genuine, scalable improvements to business economics rather than speculative AI bets.
These managers avoid the crowded trades and concentrate capital where the economics actually accrue. Put another way, AI is not the thesis – it is what great businesses do with it that matters.
