AI boom sends Polar Capital Technology Trust’s NAV up 66% in six months
Polar Capital Technology Trust’s half-year numbers are a thumping endorsement of its AI-first stance. For the six months to 31 October 2025, net asset value (NAV) per share jumped 66.0% to 539.97p, comfortably ahead of the benchmark’s 48.4% gain. The share price rose 65.2% to 476.50p, leaving the shares on an 11.8% discount to NAV.
NAV means the value of all assets minus liabilities, divided by shares in issue. In short, the portfolio did the heavy lifting as AI infrastructure spending surged across the globe.
Key half-year numbers investors need to know
| Metric | 31 Oct 2025 | 30 Apr 2025 | Change |
|---|---|---|---|
| Total net assets | £6,107,951,000 | £3,804,889,000 | +60.5% |
| NAV per share | 539.97p | 325.20p | +66.0% |
| Share price | 476.50p | 288.50p | +65.2% |
| Discount to NAV | (11.8%) | (11.3%) | Wider |
| Benchmark return | +48.4% (Dow Jones Global Technology, sterling adjusted) | ||
| Cash and cash equivalents | £419,822,000 | £188,911,000 | n/a |
| Debt | ¥15bn fixed loan (£74,109,000) | £78,730,000 | 1.2% of NAV; 2.106% fixed to Sep 2027 |
| Shares bought back | 38,847,019 into treasury in the period; a further 10,698,878 after period end | ||
| Dividend | No interim; no intention to pay a dividend for year ending 30 April 2026 | ||
Where the outperformance came from: an ‘AI maximalist’ tilt
The Trust outperformed thanks to a deliberate tilt to the nuts and bolts of AI. Management calls this being an “AI maximalist” and, importantly, they see AI as a hardware-heavy cycle. That informed overweight positions in semiconductors, memory, networking and power-cooling suppliers that feed the data centre buildout.
- Semiconductors led. Overweights in Advanced Micro Devices (+103 basis points to relative return) and Taiwan Semiconductor (+20bps) helped, alongside Broadcom (+9bps). A basis point is 0.01%.
- Memory was spectacular. SK Hynix and Micron Technology contributed +113bps and +56bps respectively, helped by high bandwidth memory shortages and firmer DRAM and NAND pricing.
- Networking names were standouts: Celestica (+160bps), Credo Technology (+154bps), Fujikura (+121bps) and Ciena (+109bps) all benefited from the demand to wire up ever-denser AI clusters.
- Power and cooling exposure also paid off, with positive contributions from Delta Electronics (+57bps), Vertiv (+48bps), Asia Vital Components (+37bps) and GE Vernova (+31bps).
Being underweight Apple was helpful too, adding +224bps as tariff worries and questions over its AI strategy weighed on the shares earlier in the period.
What didn’t help
- Underweights in NVIDIA and Alphabet hurt, costing 116bps and 53bps respectively. Management did use equity call options on Google that added back 127bps.
- Cash averaged about 4% and was the biggest drag at -228bps in a roaring market. Put options on the NASDAQ, used for downside protection, cost 82bps during the rally but had helped during the prior selloff. A put option is a contract that gives the right to sell an index or share at a set price.
Portfolio now geared to AI hardware, not software
Sector weights tell the story. Semiconductors and equipment rose to 35.9% of investments (from 28.8%), while software fell to 9.7% (from 19.2%). Management is explicit: the AI cycle is a hardware cycle, and they remain wary of traditional software where seat-based models and margins face pressure.
Top holdings are anchored in AI infrastructure: NVIDIA (11.2% of NAV), Microsoft (5.9%), Broadcom (5.5%), Taiwan Semiconductor (4.5%), Alphabet (4.0%) and AMD (3.8%). The Trust has also reintroduced areas like hard disk drives and components as training and inference data volumes explode, with new or larger positions in Lam Research, Western Digital and Lumentum.
Geography and market-cap mix
- US and Canada remain dominant at 69.1% of assets, with Asia Pacific ex-Japan at 13.3% and Japan rising to 6.3%.
- Large caps rule: 97.0% of invested assets are in companies over $10bn market cap.
Discount, buy-backs and balance sheet discipline
The shares trade at an 11.8% discount to NAV, slightly wider than April’s 11.3%, despite a 65.2% share price gain. The Board continued to buy back stock, repurchasing 38.85 million shares into treasury during the period and a further 10.70 million after the period end. There is no formal discount policy, but the aim is to reduce share price volatility and add a small uplift to NAV per share.
Gearing is low and cheap. The sole borrowing is a ¥15bn fixed-rate term loan at 2.106% due in September 2027, equivalent to roughly 1.2% of NAV. Cash and cash equivalents stood at £419.8 million at period end, giving flexibility if volatility resurfaces.
Income seekers should note there is no interim dividend and no intention to pay a dividend for the current financial year to 30 April 2026.
Concentration and risk: less ‘Big Tech’, more plumbing
The managers flag market concentration risk. In the S&P 500, the 10 largest companies account for 42% of market cap and trade on 31x forward P/E, compared with 19x for the rest. The Trust has moved further underweight the largest technology companies, with the so‑called Mag7 at about 30% of the portfolio versus 54% in the benchmark. P/E is a price-to-earnings ratio, a common valuation yardstick.
Key macro risks they monitor include a weakening US labour market, sticky core inflation, higher term premiums in bonds, tariff noise and Taiwan-related geopolitics. Taiwan is critical to the AI supply chain. At period end, Taiwanese and Chinese equities represented 11.2% of NAV.
Outlook: still constructive, but expect bumps
The managers remain bullish on AI’s momentum, adoption and monetisation, while acknowledging bouts of volatility. They continue to hold cash and use out-of-the-money NASDAQ puts to cushion sharp drawdowns. The base case is positive into 2026, supported by monetary and fiscal settings, fund flows and, above all, accelerating AI investment. Beta remains above one, so the Trust will move more than the market both ways. Beta gauges sensitivity to market moves.
Importantly, the team argues today’s setup differs from the dot-com era: returns have been driven by earnings upgrades, AI capex is a smaller share of GDP than prior tech buildouts, and spending is being funded largely by strong corporate cashflows.
What this means for investors
- Clear thesis and strong execution. A 66.0% NAV rise versus a 48.4% benchmark shows the AI hardware bet is working.
- Discount opportunity. An 11.8% discount offers a potential kicker if sentiment continues to improve and buy-backs stay active.
- Know the trade-off. Low income, higher volatility, and a concentrated tilt towards AI infrastructure. Great if you believe in the multi‑year AI capex cycle, uncomfortable if you expect a pause.
- Risk controls are in place. Modest gearing, healthy cash, and index put options provide some ballast.
- Watch the moving parts. Software could remain sluggish; memory and networking are hot. Geopolitics and labour market data are the near-term swing factors the managers highlight.
Net-net, this is a high-conviction way to back the ongoing AI buildout. The managers are picking the shovel-makers of this gold rush, and for the half-year just gone, that positioning delivered in style.
