Polar Capital Technology Trust plc (the "Company"): The Company is an investment company with investment trust status and its shares are excluded from the Financial Conduct Authority’s (“FCA”) restrictions on the promotion of non-mainstream investment products. The Company conducts its affairs, and intends to continue to conduct its affairs, so that the exemption will apply.
The Company is an Alternative Investment Fund under the EU's Alternative Investment Fund Managers Directive 2011/61/EU as it forms part of UK law by virtue of the European Union (Withdrawal) Act 2018.
The Investment Manager: Polar Capital LLP is the investment manager of the Company (the "Investment Manager"). The Investment Manager is authorised and regulated by the FCA and is a registered investment adviser with the United States' Securities and Exchange Commission.
Key Risks
- Investors' capital is at risk and there is no guarantee the Company will achieve its objective.
- Past performance is not a reliable guide to future performance.
- The value of investments may go down as well as up.
- Investors might get back less than they originally invested.
- The value of an investment’s assets may be affected by a variety of uncertainties such as (but not limited to): (i) international political developments; (ii) market sentiment; and (iii) economic conditions.
- The shares of the Company may trade at a discount or a premium to Net Asset Value.
- The Company may use derivatives which carry the risk of reduced liquidity, substantial loss and increased volatility in adverse market conditions.
- The Company invests in assets denominated in currencies other than the Company's base currency and changes in exchange rates may have a negative impact on the value of the Company's investments.
- The Company invests in a concentrated number of companies based in one sector. This focused strategy can lead to significant losses. The Company may be less diversified than other investment companies.
- The Company may invest in emerging markets where there is a greater risk of volatility than developed economies, for example due to political and economic uncertainties and restrictions on foreign investment. Emerging markets are typically less liquid than developed economies which may result in large price movements to the Company.
Important Information
Not an offer to buy or sell: This document is not an offer to buy or sell or a solicitation of an offer to buy or sell any security, and under no circumstances is it to be construed as a prospectus or an advertisement. This document does not constitute, and may not be used for the purposes of, an offer of the securities of, or any interests in, the Company by any person in any jurisdiction in which such offer or invitation is not authorised.
Information subject to change: Any opinions expressed in this document may change.
Not Investment Advice: This document does not contain information material to the investment objectives or financial needs of the recipient. This document is not advice on legal, taxation or investment matters. Prospective investors must rely on their own examination of the consequences of an investment in the Company. Investors are advised to consult their own professional advisors concerning the investment.
No reliance: No reliance should be placed upon the contents of this document by any person for any purposes whatsoever. None of the Company, the Investment Manager or any of their respective affiliates accepts any responsibility for providing any investor with access to additional information, for revising or for correcting any inaccuracy in this document.
Performance and Holdings: All data is as at the document date unless indicated otherwise. Company holdings and performance are likely to have changed since the report date. Company information is provided by the Investment Manager.
Benchmark: The Company is actively managed and uses the Dow Jones Global Technology Index (total return, Sterling adjusted) as a performance target. The benchmark is considered to be representative of the investment universe in which the Company invests. The performance of the Company is likely to differ from the performance of the benchmark as the holdings, weightings and asset allocation will be different. Investors should carefully consider these differences when making comparisons. Further information about the benchmark can be found at: https://www.spglobal.com/spdji/en/indices/equity/dow-jones-us-technology-index/#overview.
Third-party Data: Some information contained in this document has been obtained from third party sources and has not been independently verified. Neither the Company nor any other party involved in compiling, computing or creating the data makes any warranties or representations with respect to such data, and all such parties expressly disclaim all warranties of originality, accuracy, completeness, merchantability or fitness for a particular purpose with respect to any data contained within this document.
Country Specific Disclaimers
United States: The information contained within this document does not constitute or form a part of any offer to sell or issue, or the solicitation of any offer to purchase, subscribe for or otherwise acquire, any securities in the United States or in any jurisdiction in which such an offer or solicitation would be unlawful. The Company has not been and will not be registered under the United States Investment Company Act of 1940, as amended (the “Investment Company Act”) and, as such, the holders of its shares will not be entitled to the benefits of the Investment Company Act. In addition, the offer and sale of the Securities have not been, and will not be, registered under the U.S. Securities Act of 1933, as amended (the “Securities Act”). No Securities may be offered or sold or otherwise transacted within the United States or to, or for the account or benefit of U.S. Persons (as defined in Regulation S of the Securities Act). In connection with the transaction referred to in this document the shares of the Company will be offered and sold only outside the United States to, and for the account or benefit of non-U.S. Persons in “offshore- transactions” within the meaning of, and in reliance on the exemption from registration provided by Regulation S under the Securities Act. No money, securities or other consideration is being solicited and, if sent in response to the information contained in this document, will not be accepted. Any failure to comply with the above restrictions may constitute a violation of such securities laws.
Further Information about the Company: Investment in the Company is an investment in the shares of the Company and not in the underlying investments of the Company. Further information about the Company and any risks can be found in the Company’s Key Information Document, the Annual Report and Financial Statements and the Investor Disclosure Document which are available on the Company's website, found at: https://www.polarcapitaltechnologytrust.co.uk.
Fund Manager Commentary As at 30 October 2026
Key events
Market review
Global equities fell further in October, with the MSCI All Country World Net Total Return Index returning -2.4%. In the US, the S&P 500 returned -1.8%, while in Europe the STOXX Europe 600 Index returned -0.6% (all returns in sterling terms).
US economic data presented a mixed picture. US job numbers rose by just 29,000 in September, well short of the 90,000 expected, while revisions to the previous two months’ figures cut the total by a combined 60,000. Annual consumer price inflation accelerated to 3.7% in September from 3.4% in August, modestly above expectations but still below the 4.2% peak reached in May.
The price of Brent crude oil rose +15.7% as the Iran war continued to disrupt supply. Despite these inflationary pressures, the 10-year US Treasury yield fell -0.4% during the month, having reached post-2002 highs earlier in the month, reflecting softer employment data and increased demand for safe havens amid share price volatility and stress over French government debt.
Geopolitical tensions remained elevated, with continued disruption to oil shipments through the Strait of Hormuz. President Trump ruled out further US strikes on Iran ahead of the midterm elections, but the possibility of renewed military action after the vote kept energy prices and the inflation outlook uncertain. Investors became more cautious ahead of the midterms, even though US shares have historically tended to perform well in the 12 months that follow.
The Federal Reserve (Fed) held interest rates at 3.75-4.00% in October following September's hike of 25 basis points1 (bps). The vote was 10-2, with two policymakers favouring a further increase. Fed Chair Kevin Warsh pointed to robust consumer spending, broadening business investment and a resilient labour market, while acknowledging that inflation remained persistently high. Nonetheless, he argued that tighter financial conditions and improving productivity justified a pause, while leaving the door open to further rate rises should inflationary pressures persist.
Technology review
Large technology stocks underperformed their small and mid-sized peers: the Russell 1000 Technology Index (large-cap) and Russell 2000 Technology Index (small-cap) returned -9.6% and -6.8% respectively.
September's rebound reversed sharply as investors rotated out of AI infrastructure. The Philadelphia Semiconductor Index (SOX) fell -16.1%, its worst month since July, while the NASDAQ Internet Index (QNET) and iShares Expanded Tech-Software Sector ETF (IGV) returned -3.5% and -3.0%, respectively. The NASDAQ 100 fell -7.5%, its weakest month since March 2025 (all returns in sterling terms).
The SOX fell for eight consecutive trading days, its longest losing streak since October 2008. A higher-than-expected September consumer price inflation figure revived expectations of interest rate rises. Reports that China's CXMT had made progress on a higher-density design for memory chips pressured memory stocks. Rising concerns about whether AI neoclouds, which rent out computing capacity, could repay their borrowing added to the pressure. A rebound into month end was then undone on 30 October when disappointing results from Alphabet, Amazon and Intel saw the SOX fall -7.3% in a single trading day.
That said, AI fundamentals remained strong. Anthropic filed its paperwork on 26 October for what could be the largest initial public offering (IPO) ever, reportedly seeking to raise around $100bn. OpenAI responded by accelerating a funding round, later trimmed to $20bn at a $1.25trn valuation amid the market selloff, while telling investors that annualised revenue had reached $50bn following the GPT-6 Astra launch and should hit $70bn by year end.
The pace of new AI model releases continued to accelerate. Meta Platforms (Meta) launched Muse Wonder, its first frontier (leading) model, on 12 October, scoring 56 on the Artificial Analysis Intelligence Index, behind only Anthropic's Claude Opus 5.5 at 58. Google followed Gemini 4 Argon with Gemini 4 Carbon barely two weeks later, OpenAI released GPT-6.1 Terra on 27 October and Anthropic's Claude Haiku 5.5 launched at a 90% lower price than its predecessor.
October was the busiest month of the Q3 reporting season and, in aggregate, results and forecasts from businesses benefiting from AI capital expenditure (capex) remained robust. Four of the Magnificent Seven2 (Mag7) technology giants reported, all noting that computing power is constrained, order backlogs are at record levels and 2027 capex will rise again, with analysts now forecasting 2027 capex at hyperscalers (the largest cloud computing companies) of around $1.2trn.
Microsoft delivered a strong quarter with revenue of $92.3bn, up +19% year-on-year (y/y). Azure growth accelerated to +48% in constant currency3 terms, versus the 44-45% forecast, and the company forecast 49-50% growth. Capex of $45.2bn came in below the $50bn forecast due to some third-party capacity delays. The company pointed to faster spending on central processing units (CPUs) and graphics processing units (GPUs), so capex expectations for its 2027 financial year (FY27) remain largely unchanged.
Alphabet posted strong Google Cloud revenue of $30.7bn y/y driven by enterprise AI infrastructure and a notable contribution from sales of tensor processing unit (TPU) systems. However, Search and YouTube missed expectations. Search was 2-3 percentage points below expectations and management pointed to tough comparisons in Q4 and Q1. The market remains concerned about the longer-term impact on the Search business during the move to an agentic world. Capex more than doubled to $51.4bn, free cash flow – the cash it has after covering its capex – turned negative at -$9.1bn, and management signalled a “meaningful increase” in 2027 capex on top of the $195-205bn for 2026. Google began using third-party compute at the end of the quarter given its own supply constraints versus AI demand.
Amazon reported in-line revenue and a record 13.9% operating margin but Amazon Web Services (AWS) growth of +39% y/y (its fastest in almost five years, to a $184bn annual run rate) fell short of the 42% investors expected. Management reiterated $220bn 2026 capex with “meaningfully more” in 2027. CEO Andy Jassy also indicated expectations for AWS growth above 40% in both 2027 and 2028, and a $1trn revenue business within 5-10 years.
Meta Platforms (Meta) grew revenue +26% y/y and forecast Q4 above analyst expectations, while raising the low end of its 2026 capex forecast to $135-145bn. Shares were weighed down by concerns about 2027 spend which management declined to quantify. Meta aims to “maximise capacity” as it values acquiring computing power now rather than later and wants to keep flexibility for 2028 and beyond. Large language models (LLMs) are making a tangible impact on the business, improving metrics such as time spent on Instagram (up double digits in the first half) as the use of Muse Spark in Meta AI resulted in 60% more users.
In semiconductors, TSMC delivered results and an outlook above expectations (a ‘beat and raise’), with revenue up +41% y/y above the top end of its forecast range. Full-year growth was raised for a third time, to 44% in dollar terms. Management described demand as extremely robust, with no signs of a slowdown. TSMC is now sending its own teams to verify US data centre buildouts and power availability, and its N2 chipmaking process jumped from 3% to 8% of wafer revenue in a single quarter with capacity more than doubling in 2027.
Intel reported a poor set of Q3 results. Growth of +78% y/y in data centre and AI revenue was offset by client computing, its PC business, which fell -9% quarter-on-quarter (q/q) and missed expectations by $800m as PC makers began cutting production from late August. Management cut its 2026 PC market forecast to -12% (from -8%) with the second half of 2026 down -20% and 2027 down a further -10%, stating it does not expect a sustained recovery “until memory availability improves and pricing begins to normalise”. Server CPUs remain supply-constrained through the first half of 2027 as demand continues to strengthen.
Memory companies continued to beat and raise. Micron Technology (Micron) grew +31% q/q with a gross margin of 87%. The introduction of long-term agreements (LTAs) has enabled Micron to forecast an 80% gross margin for several quarters. South Korean peer SK Hynix posted quarterly revenue above KRW100trn (South Korean won) for the first time. LTAs now cover more than half of 2027-28 output and High Bandwidth Memory (HBM) is “essentially sold out” through 2029. SK Hynix also announced a KRW100trn shareholder return for 2026 (KRW30trn+ of dividends and KRW60trn+ of share buybacks, including a special dividend from Kioxia Holdings gains). Samsung Electronics became the first global technology company to exceed KRW100trn of quarterly operating profit, as HBM4 sales tripled q/q. 2027 HBM volume is sold out, HBM4E enters mass production in the first half of 2027 and 2026 capex of KRW84trn is expected to rise again next year as there are few signs of supply and demand balancing before 2028.
SanDisk, posted another beat and raise, albeit smaller than in prior quarters. Data centre revenue rose +61% q/q to 44% of sales (from 12% a year ago). Its new business model (NBM) contracts now cover 50% of fiscal 2027 memory volume and about two-thirds of fiscal 2028 at 80% minimum gross margins. Management noted that demand exceeds supply through 2030. Consumer revenue fell -29% q/q, corroborating Intel's PC warning.
Seagate Technology beat expectations on revenue and profit, but a +1% q/q increase in exabyte shipments and a flat gross margin forecast disappointed. However, exabyte growth is expected to accelerate markedly towards year end as new products ramp. Given the strength in pricing, much of the inflated input costs can be passed through to customers.
Makers of chip-making equipment reported good numbers albeit against high expectations. ASML Holding reported record Q3 sales of €11.7bn and raised its 2026 sales forecast to €44-45bn (from €43-45bn). The company noted that customers are already seeking to place orders for 2029 and 2030 as AI demand exceeds planned capacity. LAM Research beat and raised, although supply constraints and clean-room capacity limited the December quarter forecast. The company sees mid-$150bn wafer fab equipment (WFE) spending for calendar year 2026 and provided an initial 2027 expectation of 35% growth, implying more than $200bn. KLA posted record revenue of $4.14bn (+29% y/y) and now expects WFE spending to exceed $200bn in calendar 2027, similar to LAM Research.
AI hardware and component suppliers continued to beat and raise. Amphenol posted an exceptionally strong quarter with revenue up +65% y/y as IT datacom sales grew +30% q/q versus a mid-teens forecast. Management's message was simple: next-generation architectures “need more interconnect, period”, whether copper, optical or power. Corning beat expectations on revenue and profit with the key optical communications segment growing +44% y/y to $2.4bn with generative AI (GenAI) products more than doubling.
Despite political headwinds, power companies delivered strong results. GE Vernova posted revenue growth of +27% y/y with orders of $29bn (+96% organic). Its backlog was up $20bn on the previous quarter, while gas turbine backlog and reserved production slots rose from 116 gigawatts (GW) to 122GW (expecting at least 130GW by year end) and 2026 gas equipment pricing per kilowatt (kW) was running 50% above 2025. Fuel cell maker Bloom Energy reported record revenue, up +138% y/y, with a gross margin of 33.8%. Like GE Vernova, it raised its 2026 revenue forecast to growth of 122% at the mid-point, and sees substantial growth ahead as its solutions help customers get power quickly.
Space Exploration Technologies (SpaceX) reported a strong quarter, with revenue +161% y/y as Super Intelligence revenue (Cursor; Grok; neocloud; X) rose to $7.6bn (+829% y/y) on newly activated cloud agreements, the largest of which has been revealed in its 10-Q regulatory filing as being with Anthropic, at 35% of group revenue. Nameplate compute capacity reached 1.7GW, with 2.5-3GW expected by December and 13GW (potentially 20GW) by the end of 2027, versus prior commentary of 5-10GW. Elon Musk stated that short-term compute contracts are renewing at higher dollars per watt and that SpaceX “could rent out our compute several times over”.
In software and internet, ServiceNow fell sharply as its current remaining performance obligations faced a further headwind in Q4. A strong public sector drove more customers to on-premise software and some business brought forward from Q4 flattered the numbers. As a result, investors now worry that the core growth algorithm may be challenged by competition from agentic AI.
Outlook
“We did have a governor that won last night who did make data centres a big part of their platform say, ‘As soon as we get past the inauguration, let’s meet because we don’t want to lose this investment in our town. We don’t want these jobs to go elsewhere.’”
Dina Powell McCormick, President of Meta, speaking at The Wall Street Journal’s technology conference, 4 November 2026.
The US election result has created more noise and near-term uncertainty for the AI investment cycle which is forcing prices down as investors reduce risk. We think it is unlikely to dissipate quickly and investors should be prepared for continued volatility in the weeks ahead, especially for power infrastructure-related stocks. However, once the dust settles, we believe this will ultimately prove to be a clearing event, setting the scene for a long and strong AI cycle driven by rapid model progress and business adoption, which is only just getting underway.
The near-term problem is that AI became a contentious issue in many states in the run-up to the midterm elections, largely due to rising energy prices being blamed on AI (although in truth there are many factors contributing to this; perception matters more). New governors in Texas, Georgia and Ohio campaigned on pausing or attaching conditions to data centre approvals, and Democrats now fully control Pennsylvania. Together, these states hold around 45% of planned US data centre capacity.
While there are likely to be near-term delays in permitting (buildings and environmental) and grid connection approvals, these mostly affect 2028 expansion rather than 2027. We expect this to stretch the timeline for new projects, particularly those scheduled for 2028, rather than stop them. Importantly, capacity for 2027 appears largely locked in and a central condition being sought is that data centres pay the full cost of their power. Amazon, Google, Meta, Microsoft, OpenAI, Oracle and xAI signed the White House’s Ratepayer Protection Pledge, committing broadly to that.
With AI demand still exceeding supply, we believe the industry can afford a higher clearing price for new data centre capacity. Given the importance of these projects, we expect state-level deals to begin to appear, tied to both energy prices and investment in local infrastructure. History suggests backlashes of this kind will change who pays for the network, not whether it is built. The 1870s farmer revolt against railroad freight rates produced state regulation and the Interstate Commerce Act of 1887, yet the 1880s was the biggest decade of US railroad construction.
Some jurisdictions will take advantage of this by offering incentives to migrate data centre capacity there. The combination of investment in infrastructure, including schools and essential services, tax revenue and construction jobs should still prove appealing to districts and states lacking them. Existing capacity, such as at hyperscalers, may also rise in price in the near term.
We will be more concerned about the upside to 2028 demand if there is limited sign of local progress over the next three months, but we expect sentiment and positioning could begin to swing back sharply in favour of AI as and when we see the first state-level deals. Near term, that does not prevent further volatility or an anti-AI rotation as investors reposition away from AI towards a post-Trump, lower oil/tariff world. We would be remiss not to mention the possibility – not one we believe in – that Democratic policies lead to significant constraints in 2028 AI data centre growth. We believe state-level deals will lead to a resolution of this overhang in the coming months.
Even before the midterms, AI shares suffered a sharp fall from their highs, reflecting a combination of pressures as the industry races to build the infrastructure to train and serve more powerful models. The most important of these, in our view, has been the shock from rising interest rates. The Fed raised rates for the first time since 2023 and the US 10-year Treasury yield rose above 5% in September, its highest level since 2007. Higher yields stressed some neocloud business models which depend on external financing. While this is unhelpful and poses challenges for highly indebted business models, it may not significantly change the overall AI demand or build outlook as long as AI economics remain compelling, although it may change the mix between hyperscalers with strong balance sheets and debt-funded neoclouds.
Despite the near-term volatility, we remain positive on the outlook for AI. The most important drivers are moving in the right direction and demand for AI is extremely strong. This is reflected in hyperscaler commentary from the latest results. Microsoft CFO Amy Hood said: “Customer demand continues to exceed available supply.” Azure grew +48% and is forecast to grow 49-50%, in constant currency terms, next quarter, helped by higher-than-expected token pricing (tokens are the individual units AI models process). Alphabet CEO Sundar Pichai described “huge demand” for Google Cloud, where revenue rose +103%, helped by TPU system sales, and its backlog reached $561bn. Amazon CEO Andy Jassy said that “almost as fast as we install capacity, customers are gobbling it up”. AWS growth accelerated for a sixth straight quarter to 39%, with backlog of $578bn, up more than 100% y/y.
Demand strength is also evident in token consumption. Alphabet’s first-party model application programming interfaces (APIs) now process nearly 32 billion tokens per minute, more than four times the level a year ago, while Amazon Bedrock token volume grew almost 17x y/y. OpenRouter processed about 194 trillion tokens in the week of 26 October, an annualised pace above 10 quadrillion, and weekly usage by AI agents on the platform has risen 34-fold since February. At Vercel, AI Gateway volume rose from about two trillion to roughly 20 trillion tokens per month over six months, and 60% of Vercel deployments are now driven by AI agents, up from 3% in January. Ramp data shows 55.9% of US businesses on its platform paid for AI in September, up from 50.4% in March, with AI spend reaching a new peak at the start of October.
Strong AI demand is visible in GPU pricing as more powerful models and compute supply shortages make an hour of computing power more valuable. H100 rentals are up approximately 14% over six months and 7% over the past 30 days, while B200 rentals have risen 27.5% over six months, implying robust compute demand.
We are optimistic that the pace of AI model progress is accelerating and that the demand for AI compute will remain strong. The most important point that the bear cases on the AI trade are discounting – or even ignoring – is that we believe we are on the cusp of recursive self-improvement (RSI), where models improve themselves and the pace of improvement accelerates. The models we will have access to next year may therefore make current ones look pedestrian. We are equally excited about positive developments that have been rather lost in the torrent of negatives, including a shortening timeline to artificial general intelligence (AGI) which Demis Hassabis, chairman and co-founder of Google DeepMind, recently estimated at 2-3 years versus 5-10 years in January, and successful efforts by Anthropic at preventing distillation, potentially blunting the risk posed to the profits of frontier AI developers by open-weight models. We do not want to position too defensively ahead of these potential drivers of AI progress, painful as the volatility is.
We have reduced exposure to ‘second-liner’ AI stocks, something we began to do over the past few weeks, and have taken advantage of high correlations and what we see as an indiscriminate selloff over recent days to upgrade the portfolio. While maintaining our pro-growth and pro-AI stance, we have continued to reduce exposure to power infrastructure companies, where the greatest uncertainty lies (especially if some of the data centre buildout shifts away from the US).
Looking ahead, a combination of investors reshuffling portfolios after the US midterm election, forced selling and hedge funds reducing their exposure may have further to go, leading to more volatility. There was a sense of that when the market fell sharply following the midterms, with the broader NASDAQ Index down 4-5%, evidence of lower-quality stocks that investors had bet against rising and what appeared to be excessive and indiscriminate selling – many high-quality stocks were down 10% or more.
Normal seasonality and post-midterm election trends suggest markets may be entering a more favourable period. However, recent developments (see below) mean elevated volatility may unfortunately persist. We continue to view this as a feature of new technology cycles rather than a fundamental cause for concern.
09 November 2026
NVIDIA update
Shortly before this update was published, Jensen Huang resigned as president and CEO of NVIDIA and Chief Financial Officer Colette Kress was appointed interim CEO. This unwelcome development introduces new risk to the company’s outlook, and we reduced our position as the news began to break. NVIDIA remains a large absolute position at the time of writing, although our underweight relative to the benchmark has increased. This balances the company’s dominant market position and undemanding headline valuation against a significant adverse development. The news also presents an additional challenge for the AI sector, for both sentiment and fundamentals, given NVIDIA’s industry-leading position and the financial backing it has provided to many of the neoclouds. Despite this, we believe the impact is largely confined to the supply side of AI and our conviction in AI demand and model progress, as highlighted above, remains undiminished.
1One basis point equals 0.01%
2Apple, Microsoft, Alphabet, Amazon, NVIDIA, Meta Platforms and Tesla
3A way of measuring a company/fund's performance that takes out the effect of exchange rate movements between different currencies
Ben Rogoff
Ben joined Polar Capital in May 2003. He is lead manager of Polar Capital Technology Trust plc and is a Fund Manager of the Polar Capital Global Technology Fund and Polar Capital Artificial Intelligence Fund.
Alastair Unwin
Alastair joined Polar Capital in June 2019 as a Fund Manager. Prior to joining Polar Capital, Alastair co-managed the Arbrook American Equities Fund. Between 2014 and 2018 he launched and then managed the Neptune Global Technology Fund and managed the Neptune US Opportunities Fund. Prior to Neptune, Alastair was a technology analyst at Herald Investment Management.
Historical Fact Sheets