29 Dec 2025
The development and adoption of AI will continue to drive economic transformation in 2026 and beyond. A selective approach and valuation discipline are central to capturing long-term opportunities amid volatility, say portfolio managers Alison Porter, Graeme Clark and Richard Clode.
For the last two years in our outlook, we have talked about the importance of artificial intelligence for the economy and the markets. At the end of 2023 we pointed out that while the US Federal Reserve’s (Fed) Chair might be the economy’s pilot, it was AI – the emerging “co-pilot” – that was really going to drive the economy. At the end of 2024 we expected technology to cement its reputation as the vampire sector – sucking growth from other sectors of the economy. And in 2025 this all became evident, with the Fed taking a backseat to AI in markets. While AI spending is still a small share of GDP, it was responsible for the vast majority of US economic growth in the first six months of 2025.1
As we enter 2026, we see a constructive backdrop for equities with a new Fed Chair in May likely providing support. Should headwinds of government spending cuts and tariffs in 2025 reverse, they will become tailwinds to the economy, driving incremental demand across broader sectors, beyond AI. With the Trump administration and change afoot, a smooth path is never certain but industrials, autos, and housing-related demand for technology have shown signs of bottoming. The delineation of AI demand as separate to the rest of the economy will fade as the focus shifts to its potential for improving overall productivity. We expect that this will drive further technology share gains in the economy.
AI is a long-term wave – not just a theme. A technology wave (AI is the fourth wave after the mainframe, PC internet, and mobile cloud) is defined by the fact it touches every aspect of the economy. It requires investment in every layer of the technology stack from the silicon (semiconductors), to platforms, devices, and models with every company becoming an AI user in some way. These waves take multiple years to evolve and for AI, the pace of build out of capacity is being gated by deglobalisation, permitting, power availability, construction limitations as well as availability within the compute supply chain.
There is a circular problem in that the gating factor on demand for compute power has been the capacity available to train and develop new models. As we shift from generative AI to agentic AI, more reasoning capability and memory are needed to provide greater context. This requires significantly more compute power for increased token generation (units of data processed by AI models). We see areas such as physical AI developing quickly with a broadening of testing of autonomous driving and robotics worldwide. In short, as we look towards 2026 and 2027, we believe demand for compute power will continue to outstrip supply.
Revenue opportunities emerging
Source: Janus Henderson Investors.
The debate on the spending magnitude on AI has come much earlier than we would have expected given the impressive revenue ramps seen at start-ups, such as OpenAI and Anthropic, with revenue growth at a pace that we have not seen in our team’s combined 100+ years of investing in the technology sector.
Recent signs of investing circularity, combined with the outperformance of perennially unprofitable tech names this year, have brought healthy skepticism to the sector. We are monitoring this closely and recognise that there are pockets of hype appearing, for example in quantum computing. This aligns with our core belief that valuation discipline and assessing true unappreciated growth are essential for a rewarding long-term investment in tech stocks.
Tech sector valuations remain within the range of the last five years, significantly below that of the internet bubble. We see opportunity as we expect positive earnings revisions and growth above that of the broader equity market in 2026.
Source: Bernstein, as at 30 September 2025. Forward P/E = Price to forward earnings. (Orange line) MSCI ACWI Information Technology Sector, price-to-forward earnings relative to MSCI ACWI Index from December 1999 to November 2018 pre GICS sector changes in MSCI Global indices, (Red line) represents the change to (Grey line) MSCI ACWI Information Technology + ACWI Communication Services relative to MSCI ACWI Index to 3 September 2025 post GICS sector changes in MSCI Global indices. Past performance does not predict future returns.
We still see underappreciated growth and disparity in valuations within the Magnificent 7 stocks; the Mag 7 are certainly not a monolith. When innovation and disruption are accelerating at such a pace, market leadership can change. This heightens the importance of identifying the leaders of tomorrow versus being passively tied to the winners of a prior era.
This is why active management is crucial in tech investing – to ensure portfolio diversification as well as balance exposure to the best of the mega caps, while also broadening into names where we see emerging leadership and underappreciated earnings growth. Be it tied to next generation AI infrastructure layers or in areas such as fintech, automation, agentic ecommerce (internet 3.0), or enabling the electrification of the economy.
We believe that the magnitude and duration of AI is still underappreciated. Our experience of investing in past technology waves has taught us that the build of capacity and the emergence of new applications will not be linear – investors must be prepared for the ensuing volatility. We are no longer at the beginning of the AI transformation but there is still a long way to go. Stock selectivity, active management, and valuation discipline will be key to navigating the hype from the reality of this transformative technology wave in 2026 and beyond.
1 Fortune.com; “Without data centers, GDP growth was 0.1% in the first half of 2025, Harvard economist says”; 7 October 2025.