Polar Capital Technology Trust (PCT) has been one of the standout performers in the technology investment trust space over the past year.
For the trust’s financial year to date, its Net Asset Value (NAV) total return of 72% and share price return of 77% has outperformed its benchmark, the Dow Jones Global Technology Index, which has returned 40% over the same period (as at 27 February 2026, all figures in Sterling terms. The Company’s financial year ends on the final day of April each year).
If there were ever an investment trust to demonstrate the benefits of active management over passive investment, Polar Capital Technology Trust could well be it.
The managers pursue a high-conviction, high-turnover approach to investing in the AI revolution, and its lead manager, Ben Rogoff, describes it as “an active, active manager.”
The thesis: discontinuous progress
At the heart of PCT’s strategy is the view that AI is not following the normal arc of technological adoption. Rather than advancing incrementally, AI is exhibiting what Rogoff calls “discontinuous progress,” where decades of development compress into years.
The trust draws parallels with the printing press, railway construction, and mass production. These are moments where technology didn’t just improve people’s lives, but fundamentally rewired how economies function.
This framing matters because it shapes how PCT allocates capital. Where a conventional technology fund manager might wait for clear evidence of mainstream adoption before committing, PCT’s team is incredibly forward-thinking and prepared to spend time exploring areas that others may overlook. We look at what this means in practise later in this article.
Exploding enterprise revenues, enormous capital expenditure commitments, and rapid AI model improvement are all cited by the PCT team as confirmation that the inflection point has arrived.
In their view, 2026 is the year AI transitions from a topic of debate to a key driver of economic performance.
Gauging growth: revenue, capex, and adoption
PCT tracks several key indicators to validate its thesis. The most striking is the revenue trajectory of the leading AI model providers. OpenAI’s annualised recurring revenue grew from roughly $10 billion to over $20 billion in less than twelve months through to the end of 2025, and had reportedly reached $25 billion by early March 2026. Anthropic’s growth has been even more dramatic on a relative basis, growing from $1 billion in annualised revenue in late 2024 to $9 billion by the end of 2025, before accelerating to $14 billion by mid-February 2026 and $19 billion by early March.
Capital expenditure from the hyperscalers Alphabet, Amazon, Meta, and Microsoft, provides a second critical signal. Consensus estimates now project combined hyperscaler capex of $667 billion in 2026, around 62% higher than the prior year and $127 billion above where expectations stood at the start of Q4 earnings season. Google alone guided to $180 billion at the mid-point, considerably above the $130 billion the market had expected. Amazon raised its outlook to $200 billion, spanning AI, chips, robotics, and low earth orbit satellites.
The adoption of AI tools worldwide reinforces the thesis. OpenAI confirmed 900 million weekly active users and 50 million paying consumer subscribers. Alphabet processes well over a quadrillion tokens monthly across its AI services, a figure that has grown several times over in under a year. PCT’s manager argues these metrics point to a structural shift in how businesses and consumers interact with technology. Importantly, the demand curve is steepening, not flattening.
Portfolio Construction: Hardware Over Software
PCT’s portfolio construction reflects a core belief that the AI cycle is fundamentally a hardware story, not a software one. Their view was validated in early 2026, when software stocks were ravaged by fears of AI disruption, sector by sector. Some of the world’s leading SaaS names have lost well over 50% of their value in recent years.
But PCT appears to have been ahead of the game, and the trust has dramatically reduced its software exposure, increasing its allocation to semiconductors, power infrastructure, optical networking, and memory. These are the physical building blocks of AI, sometimes known as the ‘AI enablers’.
Semiconductors and semiconductor equipment represent the largest sector allocation at roughly 37% of the portfolio, followed by interactive media and services, and technology hardware. Software, by contrast, has been cut to less than 5%, a negligible weighting, driven by the team’s conviction that agentic AI, combined with outcome-oriented pricing, will disrupt traditional headcount-based software-as-a-service models.
Geographically, the trust has been shifting away from the US at the margin, with Asia Pacific exposure growing. The US and Canada still account for around 64% of the portfolio, but the trust holds meaningful positions in Asian names like TSMC and Samsung Electronics, reflecting the concentration of advanced semiconductor manufacturing in the region.
Perhaps the most telling feature of the portfolio is its active share, which sits near recent highs at around 49%. The trust holds approximately 94–98 positions but makes large individual-stock bets and is willing to have zero weightings in major index constituents when the growth case doesn’t stack up. The manager is significantly underweight the “Magnificent Seven” group of mega-cap US technology companies and is content to use call options to manage the upside risk of being underweight names like Apple, Microsoft, and Alphabet.
Current holdings: Following the bottlenecks
PCT’s top ten holdings read as a map of the AI supply chain. NVIDIA remains the largest position at around 9–10% of NAV, though the trust is notably underweight relative to the benchmark. Alphabet has risen sharply in the portfolio following the success of its Gemini 3 model and the commercialisation of its custom TPU chips. TSMC, Broadcom, and Samsung Electronics round out the semiconductor-heavy upper echelons of the portfolio.
The more interesting story, however, lies in the relatively under-the-radar overweight positions. These may be the positions that distinguish PCT from a passive technology vehicles. LAM Research, a memory chip equipment manufacturer benefiting from acute DRAM and NAND supply bottlenecks, was the trust’s largest overweight at the end of 2025.
Ciena and Lumentum Holdings, both optical networking specialists, represent high-conviction positions driven by the manager’s belief that power-intensive electrical networking is increasingly threatened by fibre-optic alternatives. Ciena’s shares have risen over 300% in the past year. Lumentum’s have risen over 850%. These companies’ returns would have had a negligible impact on the majority of passive alternatives.
Even more unconventionally, PCT took a position in Caterpillar on the basis that its industrial generators provide a near-term solution to the power squeeze constraining data centre expansion, and that its mining equipment exposure gives it leverage to rising copper demand.
Ben Rogoff has explained in a recent video interview that PCT is always looking for the most important companies in the AI story. This may not always be the ‘Mag 7’ companies responsible for the capex driving the industry forward.
The risks and the conviction
PCT’s manager is clear-eyed about the risks. If AI model progress stalls, investment momentum could reverse quickly. But this risk is considered small enough not to materially detract from the investment case.
Geopolitical uncertainty, including tensions in the Middle East and shifts in US trade policy, adds further complexity. The trust’s concentrated sector focus means that a broad downturn in technology sentiment would hit hard.
Yet the trust continues to invest with a high level of conviction and doesn’t buy into the idea of an ‘AI bubble’. Most of the performance in the technology sector in recent years has been driven by earnings growth rather than valuation expansion, a crucial distinction from the late 1990s bubble.
AI spending currently represents around 1% of global GDP. Historical precedent suggests this could rise to 2–5% at peak intensity over a cycle lasting five to ten years. Should this come to pass, the companies in the PCT portfolio will be direct beneficiaries.
For those that share PCT’s view on the continued expansion of AI, the recent bout of volatility in equity markets has presented an interesting opportunity in a widening of the discount to NAV around 9.2%.
Past performance is not indicative or a guarantee of future returns.
