Technology volatility, the AI cycle, and what's next for investors

The long-term potential of AI is undiminished, but evolving market conditions mean we should keep testing our assumptions.

Key takeaways 

  • AI’s long-term growth story is intact, despite political scrutiny and the prospect of greater regulation slowing the pace of US data centre development.
  • Recent technology weakness creates opportunities, with indiscriminate selling and attractive valuations offering entry points for long-term investors.
  • Diversification and active management are critical as investors navigate AI disruption, geopolitical uncertainty, and persistent market volatility.

A backlash against data centre development was one of the most important takeaways from the US midterm elections in early November. Until now, hyperscalers in the US have been able to proceed with AI infrastructure buildouts with few constraints. That could be about to change. More stringent licensing requirements and the potential removal of lucrative tax breaks in states like Texas, Ohio, and Georgia - which account for a large proportion of AI data-centre development - could increase costs for hyperscalers and erode returns on investment.

Investors are concerned. The resulting sell-off in technology stocks has been severe in some cases, but existing AI capabilities offer untapped potential, even in a tougher political environment, says Fidelity Chief Investment Officer, Equities, Niamh Brodie-Machura.

“Adoption will continue and demand strength will continue,” she told a webinar discussion, taking stock of the action in markets a week after the US polls. “We now need to figure out what a change in environment and atmosphere looks like for different companies and different subsectors going forward”.

Matt Quaife, Global Head of Multi Asset, echoed this point. The sell-off was driven substantially by technical sellers from the passive and systematic strategies that now dominate global equity market flows. Exceptional share price growth in semiconductor and technology hardware names in the first three quarters of 2026 had also been driven by retail investors, some of whom had used leverage to magnify their exposure. Many of these investors were forced to liquidate their positions as the market turned against them, amplifying the weakness in the tech complex.

“We've been adding a little bit of risk in the last few days, but quite cautiously,” he said.

Throwing the baby out with the bathwater

While share price weakness among household IT names has dominated headlines, Austin Kelly, one of Fidelity’s equity research analysts covering the sector, suggested the long-term appeal of selected stocks in the sector remains unchanged: “several companies have a great structural, long-term outlook from widening AI adoption, can generate very high returns on capital from that growth, and are trading on very reasonable valuation multiples given their long term outlook”.

Valuations are very reasonable now because of the short-term fear around the impact of the US midterms. The elections have created uncertainty around how much data centre capacity will come online and, by extension, what the impact to earnings in the technology sector might be.

The market hates uncertainty and fundamentals that it struggles to quantify, he said. “I suspect the fear could be worse than what the reality will eventually end up being.”

Ultimately, both Democrats and Republicans are incentivised to keep the wheels of the AI machine turning. Fidelity’s Global Head of Macro and Strategic Asset Allocation, Salman Ahmed, explained that AI investment has been among the major drivers of GDP growth in the US in recent quarters and that the economy would be close to recession without this tailwind.

Separately, he argued, there is a desire for the US to be the global leader in AI development and utilisation. Allowing other countries - particularly China - an advantage in the race to develop the technology is unpalatable on both sides of the political divide and suggests further AI-related developments will be permitted to proceed, albeit potentially with some additional regulation or political oversight.

Undemanding valuations

If the recent market jitters turn out to be a bump in the road rather than a more serious roadblock, current valuations for selected companies look appealing for long-term investors. The forward price-to-earnings ratios of the Nasdaq and the broader MSCI AC World Index are below their historic averages, largely reflecting strong earnings growth forecasts for 2027 and beyond.

Notwithstanding current uncertainty over the AI infrastructure rollout, Brodie-Machura pointed to earnings in the IT sector in 2027 that are forecast to be around 150 per cent higher than their 2022-2026 average. At more than 50 per cent, consensus earnings growth forecasts demand attention for the broader market too (as measured by the MSCI AC World).

Whether these expectations are realistic is therefore the key question right now for tech stocks as well as those in other industry sectors.

How resilient will earnings be against a potentially more challenging political backdrop? What signposts might indicate that exceptionally strong growth rates in some areas of the market are being eroded? These are the questions our research teams and portfolio managers are wrestling with, and which will influence portfolio positioning from here.

Broader implications

As well as considering the implications for semiconductor and technology hardware stocks directly, the panel discussed how the evolving outlook for AI and economic momentum is affecting thinking and portfolio positioning more broadly.

Quaife and his team remain cautious. In fixed income, they prefer short-duration credit over sovereign debt. They do not currently anticipate a meaningful deterioration in GDP growth, and ongoing geopolitical and inflation risks mean it is too difficult to forecast projected returns from sovereign bonds with any degree of confidence.

Within the equity space, the implications of potential delays to the AI build-out for infrastructure firms and utilities, for example, are important variables. So too is how AI adoption might benefit users over time. The financial sector is a good example: AI tools are increasingly replacing legacy systems for banks and insurers, with associated implications for productivity and margins.

“We know there are many industries and sub-sectors that will benefit from volume growth,” Brodie-Machura said. “In many cases, we are finding old-world, old-economy business models with strong investment cases underpinned by an existing business, but where technological developments provide additional optionality.”

Ultimately, one of the most important jobs for discretionary asset managers is to provide sufficiently diversified portfolios and mitigate the risk of individual themes dominating investment outcomes. With high correlations persisting regionally and across industry sectors, this task is more important than ever and requires careful management of position sizing as well as stock selection.

Market developments like this require investors to re-test their assumptions and adjust when ideas prove wrong. At Fidelity, that means constantly challenging one another with lively, constructive debate and stress-testing existing investment theses. Responding decisively to new information, being prepared to change positioning as market conditions evolve, and remaining true to tried-and-tested processes helps navigate periods of market volatility all support the pursuit of long-term returns.