Key Takeaways
- AI is a structural shift, not a normal tech cycle: Generative AI represents a rare discontinuous change that could reshape multiple industries.
- Valuations look broadly justified, though pockets of speculative excess are emerging, especially in private markets.
- Alpha is broadening beyond mega-caps, creating greater opportunities for active, stock-specific returns.
After a powerful rally driven by artificial intelligence and mega-cap technology stocks, global tech equities have entered a more complex phase. This is not a normal tech cycle: Generative AI represents a rare discontinuous change, with adoption now reaching an inflection point that could reshape multiple industries.
Leadership is broadening, valuations are under renewed scrutiny, and investors are increasingly focused on which parts of the technology value chain will deliver sustainable returns as AI investment moves from hype to hard economics.
Nick Evans is lead manager of the Polar Capital Global Technology Fund, which has outperformed its benchmarks and peers since 2023, and has consistently ranked in the top quartile of its category in the past three years.
Here, he talks about where the AI cycle stands today, how regional and sector dynamics are shifting, and where active investors can still find alpha in global technology. This conversation has been edited and condensed.
A New Phase for Tech Markets
Valerio Baselli: After a period of strong performance and sharp rotations, how would you describe the current state of global technology equities?
Nick Evans: Three years into the AI cycle that began with the launch of ChatGPT in late 2022, something structural has changed. In 2024, the Magnificent Seven dominated performance, but in 2025 they lagged broader technology indices and global equities. That trend has continued into early 2026, with mega-cap and software stocks underperforming the sector and the wider market.
We believe we are entering the next phase of the AI cycle. Early on, AI complemented existing technologies; now, improving models are becoming credible substitutes, challenging long-term terminal values. This is not a typical tech cycle. Generative AI is a general-purpose, discontinuous innovation that compresses decades of progress into years. We believe we are at an inflection point for rapid adoption.
Polar Capital Funds PLC - Polar Capital Global Technology Fund
- Morningstar Medalist Rating: Silver
- Morningstar Rating: ★★★★
- Fund Size: €10.1 billion
- Ongoing Charge: 1.11%
- Morningstar Category: Equity Technology
Valuations and the AI Reality Check
Baselli: Many technology stocks are again trading at elevated valuations. Are prices still attractive for long-term investors?
Evans: Valuations have expanded, but we think they are broadly rational given technology’s superior growth outlook. The S&P IT sector trades close toa 1.2x premium to the S&P 500, which we see as justified. Performance has largely been driven by earnings growth rather than multiple expansion, similar to 1996–97, and there is no sign yet of the excesses seen in 1999–2000.
That said, there are pockets of excess, particularly in private AI markets where both valuations and capital deployment look aggressive. Volatility is a feature of early-cycle growth: between 1995 and 1998, the NASDAQ rose 354% but experienced seven corrections of more than 15%.
Baselli: How do you separate genuine AI value creation from hype?
Evans: Investor skepticism around AI returns is actually healthy. In 2025, fewer than 22% of global active equity funds and 17% of technology funds outperformed, suggesting many investors have not fully embraced AI yet. This echoes 1997–98, when fears of excess kept many managers on the sidelines.
Our experience managing a dedicated AI strategy for eight years helps us distinguish substance from hype. We focus on infrastructure enablers such as compute, networking, memory and data-center power. AI capabilities are advancing faster than expected, and we are already using AI agents internally. That said, there will be casualties: parts of software have struggled to monetize AI investments so far.
Baselli: What explains the divergence between US, Asian and European technology stocks?
Evans: The divergence reflects structural and policy differences. US rate cuts have supported long-duration sectors, and a new Fed Chair may be under pressure to cut again, although we think most of that tailwind is now behind us. The US economy is entering 2026 in relatively good shape, supported by fiscal stimulus and measures such as accelerated depreciation and tax rebates.
While the US dominates AI infrastructure, we remain deliberately globally diversified. Asian technology outperformed last year, with the Dow Jones World ex-US Tech Index gaining 39% versus 21% for the NASDAQ 100. Global diversification allows us to capture these relative opportunities and reduce exposure to any single policy environment.
Portfolio Construction in an AI-Driven Market
Baselli: Your fund has outperformed peers over the past three years. Which factors contributed most, and which detracted?
Evans: Our outperformance reflects our focus on AI and the upside surprise to AI capex driven by rapid user adoption. We believe this trend is still in its early stages. In 2025, mega-cap and software stocks lagged, and we reduced exposure to these areas as we felt they were no longer the best conduits for AI returns.
Generative AI usage has expanded rapidly, with weekly active users exceeding one billion. Scaling laws remain intact, and newer models from Google and Anthropic show meaningful performance gains, which should support broader enterprise adoption in 2026.
Baselli: A handful of mega-caps now dominate index returns. Is that concentration justified by fundamentals, or does it increase downside risk?
Evans: The Magnificent Seven represent around 30% of our fund versus 54% of the benchmark. We are seeing a healthy broadening of opportunity beyond the largest names. While there are still opportunities in mega-caps, this requires a highly dynamic approach given the pace of innovation.
In 2025, the majority of our holdings outperformed both NVIDIA and the global technology index. Only Alphabet and NVIDIA outperformed the tech sector among the mega-caps. Going forward, stock selection within mega-cap will be critical, with markets likely to reward companies that can demonstrate returns on AI spending and penalize those facing the “incumbent’s dilemma”.
Baselli: If growth expectations fall, which parts of tech are most vulnerable, and which are more resilient?
Evans: The most vulnerable would be speculative AI names with high valuations and unclear monetization, particularly in application software exposed to generative AI disruption. Small and mid-cap technology could also lag further.
More resilient areas should be AI infrastructure providers with contracted demand, including semiconductors, memory and data-center equipment. These businesses benefit from supply constraints, growing backlogs and hyperscale demand that is outpacing capacity additions.
Baselli: How do you decide when to trim winning positions?
Evans: We use our AI framework to assess when technological progress changes a company’s risk-reward profile. When valuations move beyond fundamentals or disruption threatens competitive positioning, we reduce exposure. Portfolio turnover increased in 2025 as AI-driven change accelerated.
Looking Ahead: What to Expect from Tech Stocks
Baselli: For investors allocating to global technology now, what return assumptions should they temper?
Evans: We expect 2026 to be another strong year for AI-driven capex, particularly in infrastructure. AI spending remains around 0.5–1% of GDP, versus 2–5% in prior infrastructure buildouts that lasted 5–10 years. We are only three years into this cycle, and usage trends suggest upside risk to spending.
Most AI capex [capital expenditure] comes from well-capitalized companies with strong insight into future AI progress. If new models deliver meaningful performance gains, sentiment and adoption could improve further in 2026.
Baselli: What belief about investing in technology have you revised over the past three years, and how has that changed your strategy today?
Evans: We underestimated the speed of AI capability development. What we expected to see in 2027–28 is happening now, with AI agents already performing complex coding tasks autonomously. This has increased our confidence in sustained AI demand, returns on investment and productivity gains.
We see this as the “end of the beginning” of the AI cycle rather than the end of the opportunity. As adoption accelerates, we believe the impact of AI across sectors will become increasingly difficult for investors to ignore in 2026.
