Carmignac Portfolio Tech Solutions: Letter from the Fund Manager - Q3 2026

Over the third quarter, the Fund gained +6.3%, compared with +2.6% for its reference indicator, bringing year-to-date performance to +49.1%, versus +47.9% for the reference indicator.

Global equity and sovereign bond markets moved in opposite directions during the third quarter of 2026. While equities proved resilient and globally delivered positive returns, government bond markets came under significant pressure as persistent inflation, elevated public debt levels and stronger-than-expected economic activity pushed yields sharply higher. Equity markets, by contrast, continued to benefit from resilient corporate earnings, with profit expectations revised higher in both the US and emerging markets, particularly within the technology sector. In several markets, earnings growth actually outpaced share-price appreciation, allowing valuations to compress despite positive equity returns.

Artificial intelligence (AI) remained the dominant theme, although leadership changed significantly during the quarter. July saw a correction in semiconductors and Asian AI hardware following their exceptional performance in the first half of the year. As the quarter progressed, market leadership broadened towards hyperscalers and software companies.

More broadly, style leadership remained unstable, rotating repeatedly between high-growth technology stocks and more defensive or value-oriented segments. This created a less directional market environment than in the first half of the year.

Geopolitical tensions also returned to the forefront through energy markets. Renewed instability in the Middle East pushed Brent crude back above $100 per barrel, adding to inflation concerns. The broader macroeconomic backdrop nevertheless remained resilient. US growth continued to be supported by strong business investment, notably AI-related capital expenditure, while September business surveys pointed to improving activity in both the US and the euro area.

Over the third quarter, the Fund gained +6.3%, compared with +2.6% for its reference indicator, supported by strong stock selection across both software and AI-related infrastructure.

The largest contributor was Atlassian. After having been heavily penalised earlier in the year amid concerns that agentic AI could disrupt traditional software business models, the share price rebounded strongly following better-than-expected results. Revenue growth accelerated to 28%, cloud revenue grew by 31% and remaining performance obligations increased by 44%. These results reinforced our view that software companies deeply embedded in mission-critical workflows, with proprietary data and context, can benefit from AI rather than simply be disrupted by it. GitLab also contributed positively.

Microsoft was another important contributor. Strong quarterly results helped reassure investors that the company is beginning to translate its significant AI investments into tangible growth, with Azure continuing to benefit from strong demand for cloud and AI services.

On the hardware side, Nvidia and Asia Vital Components (data-centre cooling solutions) also contributed positively, reflecting continued strength across the AI infrastructure value chain, from accelerated computing to thermal management and cooling.

These positive contributions were only partly offset by weakness in Grand Process Technology (semiconductor wet-process equipment) and Nitto Boseki (advanced semiconductor packaging materials). Both companies had previously performed strongly, leaving market expectations particularly demanding. While quarterly developments fell somewhat short of these expectations, we continue to see both as important beneficiaries of the AI infrastructure build-out.

We remain constructive on the long-term outlook for technology, and AI in particular, but the opportunity set is becoming increasingly selective. The debate is no longer whether AI will matter, but where the economics of this investment cycle will ultimately accrue. In our view, the most attractive opportunities still lie in the companies supplying the infrastructure required to build and deploy AI rather than in those financing the build-out.

Hyperscalers continue to commit unprecedented levels of capital to AI infrastructure. While this spending supports the broader technology ecosystem, it is also becoming increasingly capital intensive, putting pressure on free cash flow and raising questions around future returns on investment. Competitive dynamics make a sharp reduction in spending unlikely in the near term – underinvesting risks falling behind – but the balance between investment and monetisation is becoming increasingly important. We have therefore continued to considerably reduce our exposure to hyperscalers and favour the beneficiaries of their capex instead.

Hardware remains our strongest conviction. Demand visibility remains robust, while bottlenecks are broadening across the value chain, from advanced semiconductors and memory to packaging, connectivity, power and cooling. The rise of agentic AI, AI-enabled cybersecurity applications and increasingly compute-intensive workloads should continue to support strong demand for both processing power and memory. As inference requirements scale, the hardware intensity of AI infrastructure should continue to rise. Importantly, many of these companies are monetising the AI build-out today: orders translate into revenues, earnings and, for the businesses we favour, free cash flow.

At the same time, valuation discipline remains central to our portfolio construction. Nvidia and TSMC remain among our highest-conviction holdings given their technological leadership, competitive advantages and strong cash generation. However, after significant share-price appreciation and the expansion of their market capitalisations, we believe the prospective upside has become less asymmetric than it was previously. We have therefore taken some profits and reallocated capital towards areas where expectations and valuations offer greater upside.

Korean memory is a good example. Following recent weakness, we increased exposure to selected names where we believe market expectations have become too cautious relative to the earnings outlook. High-bandwidth memory remains a critical bottleneck, while disciplined industry supply should support pricing and cash-flow generation. More broadly, we continue to seek less obvious beneficiaries across the technology value chain where valuations still offer attractive upside.