Carmignac Portfolio Tech Solutions: Winning beyond the obvious

From semiconductors to software, Carmignac Portfolio Tech Solutions has delivered strong returns since launch by investing across the full technology value chain. This highlights the importance of selectivity in an increasingly polarised market.

Launched in June 2024, at the peak of the so-called “Magnificent 7” era—when the seven mega-cap technology stocks accounted for nearly 80% of the S&P 500’s returns— Carmignac Portfolio Tech Solutions was designed to capture the accelerating adoption of artificial intelligence (AI) and the broad investment cycle supporting it. Rather than focusing on a narrow set of mega-cap AI winners, the strategy was designed to capture innovation across the entire ecosystem, from upstream materials to downstream applications, reflecting a conviction that technological disruption extends far beyond traditional sector boundaries.

Strong early outperformance has reflected this deliberate positioning away from index concentration and towards underappreciated segments of the technology value chain combined with a disciplined approach to valuation.

Since launch, the fund has delivered robust absolute and relative performance (+92.9% vs +74.1%¹), outperforming its reference indicator². In 2025 alone, the fund returned +29.4% compared with +14.8% for its benchmark, placing it in the first quartile of its peer group since inception.

The Fund’s performance has been rooted in the strategy’s ability to identify opportunities across the full breadth of the AI ecosystem. Over the past two years, the AI investment theme has evolved from a GPU-led story into a full-stack infrastructure build-out spanning compute, networking, storage, power, cooling and software.

Core AI infrastructure
The mission-critical bottlenecks, these companies without which AI training and inference simply cannot scale, have been core holdings of the portfolio since launch and have been among the strategy’s strongest performers.

TSMC has ranked among the top three holdings since day one, reflecting its pivotal role in the AI ecosystem. Building a leading-edge logic foundry from scratch requires decades of process expertise, materials science know-how, and deep relationships across the semiconductor supply chain—capabilities that cannot be replicated quickly. TSMC’s competitive moat strengthens with each successive node generation.

Beyond TSMC, companies such as Nvidia, SK Hynix and Broadcom, among the primary beneficiaries of surging demand for advanced semiconductors—have also been major contributors to performance.

AI Infrastructure buildout, networking and interconnects
The Fund also benefited from its positioning in industrial technology and digital infrastructure, particularly as the AI investment theme broadened. Each new generation of AI systems requires more networking, power infrastructure, cooling capacity, advanced packaging, and specialised materials. As a result, an increasing share of value is accruing to the critical bottlenecks that determine the speed and scale of deployment.

Holdings such as Amphenol, Celestica, and Comfort Systems contributed meaningfully, an area whose importance is often overlooked by traditional sector classifications, despite being essential to the industry's expansion.

The Hidden Enablers Driving Value Creation
Beyond the obvious winners, the strategy also captured value in less visible parts of the supply chain. Companies involved in electronic components and materials, such as those producing copper-clad laminates, delivered outsized returns as demand for AI infrastructure expanded rapidly. Exposure to companies such as Elite Materials and Nitto Boseki illustrates the strategy’s ability to identify off-benchmark beneficiaries of the AI cycle (e.g. laminates, substrates, connectivity components).

This reflects a deliberate effort to monetise second-order effects of AI capex rather than focusing solely on front-end beneficiaries.

Far from being a peripheral player, Asia is the essential enabler of the AI revolution, a reality that investors cannot afford to overlook. While some critical nodes, such as TSMC, are well-known and strategically critical, other companies are far less visible to investors, controlled by a handful of mid-sized firms producing key components. These businesses often combine underappreciated profitability with robust balance sheets, enabling them to capture a compelling growth trajectory and sustained earnings expansion. Emerging market exposure accounts for 43% of the performance of the strategy (vs 19% for the index)³.

A key differentiator has been the integration of valuation discipline within a growth-oriented universe. This approach has been particularly relevant in an environment where parts of the AI ecosystem have experienced rapid multiple expansion disconnected from fundamentals.

Carmignac Portfolio Tech Solutions has sought to monetise this dispersion through a disciplined valuation framework, applied across the investment universe.

The portfolio has been actively managed, with several key actions contributing to performance:

• Capitalizing on the opportunities created by Liberation Day: We actively added to positions during the market weakness in April 2025 and subsequently took profits as markets recovered later in 2025. This disciplined approach was a significant contributor to the portfolio's strong outperformance relative to its benchmark.
• Adjusting exposure to hyperscalers: Over recent months, investor focus has shifted from the scale of AI-related investments to the sustainability of the returns these investments can generate, leading to increased volatility across the sector. In response, we selectively reduced our exposure.
• Actively managing semiconductor portfolio beta: Following the sharp re-rating of semiconductor valuations in 2026, we rebalanced the portfolio by rotating from higher-beta semiconductor names into higher-quality companies with more attractive risk-adjusted return profiles.

Not all parts of the portfolio contributed positively. Software exposure detracted from performance over 2026, as parts of the market grew concerned that AI would disrupt incumbent business models. These positions had been mostly built at the end of 2025/ beginning of 2026, on the contrarian view that the disruption wave would remain limited for some of the well-established players. We have reduced some of these positions lately, as skepticism toward the segment may persist until markets gain better visibility on the true scale and timing of any disruption.

Markets remain largely momentum-driven, with narrow leadership and signs of excess in parts of the market, particularly among smaller names that have doubled or tripled in recent months despite unclear fundamentals. That said, the broader picture remains reassuring: the MSCI IT index trades at around 22x P/E, about 10% below its five-year average⁴ and well below dot-com bubble levels, while recent share-price gains have been driven more by earnings than multiple expansion. The market is also already discounting a moderation in growth next year.

In this environment, selectivity is critical: across all tech segments, we have shifted towards high-quality stocks that offer a valuation safety net, like Nvidia and Broadcom.

Beyond this near-term caution, we believe the medium-term outlook remains robust. The cycle appears to be extending rather than peaking, with companies increasingly planning capacity around 2027–28 demand and signing three-to-five-year supply agreements. AI demand is becoming more additive than cyclical, as edge AI, agentic AI, and physical AI create new compute endpoints while infrastructure deployment remains constrained. This means demand is not simply shifting from one chip generation to the next: each generation is likely to serve a growing base of use cases and customers. As an example, demand for Nvidia’s current Blackwell AI chip platform could remain strong for longer than usual, even as customers begin to prepare for its next-generation Rubin platform.

At its core, the fund is built on a broad and unconstrained investment universe. The strategy invests across sectors and geographies, with a meaningful allocation to emerging markets (c.30%) and small- and mid-cap companies (c.20%), areas which are often underrepresented in traditional technology benchmarks.

This is reflected in the portfolio’s structure. Compared to its benchmark, the Fund exhibits greater exposure to semiconductors, materials and components, alongside a more diversified geographical footprint, including a strong presence in Taiwan and South Korea.

The early success of Carmignac Portfolio Tech Solutions highlights a key reality of today’s technology market: the opportunity set is broader and more complex than ever.

As AI reshapes industries, the next phase of performance is likely to be driven not by being invested in technology, but by being invested in the right parts of it.

Carmignac Portfolio Tech Solutions has demonstrated that navigating this complexity requires more than exposure. It demands selectivity, flexibility, and a willingness to look beyond the obvious.