Artificial intelligence is the talk of the town – and fund managers are paying close attention as they enter the final quarter of the year. As well as earnings, geopolitical developments and central bank decisions, the fourth quarter could also see performance pressure and potential market disruption as investors rebalance their portfolios amid talk of an AI bubble. So how are investors balancing this with the desire not to miss out on a potential $4.8 trillion market ? “Valuations and policy uncertainty make it hard to be overweight risk assets, but neither do we want to fight strong momentum and continued earnings growth,” said Rathbones’ Head of Investment Strategy John Wyn-Evans. He noted that his firm is taking a more neutral stance on risk, but its main sector overweights are technology, healthcare, media, industrials and financial services. Bubble concerns Stocks — particularly those related to AI — have soared this year. The tech-heavy Nasdaq Composite rose over 11% in the third quarter and is up around 20% over the year-to-date. It has also continued to move higher into October, hitting new record highs . Market participants — and even central banks — appear divided on whether AI-fueled investment is forming a market bubble. The Bank of England and IMF have issued warnings of soaring stock valuations, but banks such as Goldman Sachs appear less gloomy , saying the market is not in bubble territory — yet. AI investment is serving as an antidote to weaker areas of the economy, said Derek Hynes, fixed income portfolio manager at Wellington. “But any loss of confidence in this area may have an outsized impact on financial conditions and so warrants careful monitoring.” Will Mcintosh-Whyte, a fund manager on Rathbone’s multi-asset team said it’s “difficult to bet against the AI-related baskets.” “Momentum remains strong, and continues to be, in the main, backed up by earnings,” he added . “We are, however, wary of positioning, and remain watchful of stretched valuations.” .IXIC YTD mountain Nasdaq Composite Stephen Yiu, fund manager at Blue Whale Growth Fund, is maintaining significant underweight positions in the Magnificent 7 stocks, having sold out of Microsoft and Meta in the second quarter. However, he remains bullish on Nvidia. He said many big companies are spending lots on AI, but it may not pay off in the long run for investors. “Do they get enough revenue income on the back of that? Do they get a margin impact as well, even though they might be making more money?” he said. “But then we think that incremental margin for, let’s say Microsoft, charging us a subscription on Copilot, would be lower than what they are charging us a subscription on Office365, which they don’t have to pay Nvidia, so those will be things that we’re looking out for,” Yiu added. ‘Nvidia remains a bellwether’ Corporate AI spend and revenue is set to take center stage during earnings season as companies continue to seek efficiencies while others build out much-needed infrastructure, such as data centers. Now, “all eyes will be on AI capex and AI adoption, with everyone trying to separate the sheep from the goats,” said Wyn-Evans. It comes as a slew of recent AI deals have raised eyebrows , with some concerned that they are too circular. AI cloud infrastructure company CoreWeave , for example, is tied to OpenAI and Nvidia as both a buyer and seller. Meanwhile, competition is also increasing as newly spawned startups rush take advantage of AI and help other companies do the same. NVDA YTD mountain Nvidia “Nvidia will remain an important bellwether for the AI trade, and we will also be watching capex trends from the hyperscalers,” added Mcintosh-Whyte. “We will also be looking to software names to demonstrate they are able to continue to capture revenues from AI opportunities, and defend themselves from disruption.” Blue Whale Growth Fund’s Yiu has his eye on advanced memory storage, as “it’s on a receiving end of AI spending.” Advanced memory storage companies, such as Seagate and Western Digital , are focused on storing and handling the vast amounts of data needed for AI, as well as the huge amount of data generated by AI itself.