Carmignac Investissement: Three routes to structural growth

The semiconductor rally has broadened well beyond Nvidia. The logic is simple: demand for computing power is exploding, supply remains constrained, and the bottlenecks are no longer limited to the chip itself. This is where Nitto Boseki, or Nittobo, becomes interesting. At first glance, it does not look like an AI company. It was originally a Japanese textile business. Today, however, it has become a critical supplier to the advanced semiconductor supply chain.

Its key product, known as T-glass, is a highly specialised glass cloth used in semiconductor packaging substrates. These materials reinforce and stabilise the platforms on which advanced chips are assembled. They help control thermal expansion, reduce signal loss and improve reliability at very high levels of computing intensity. In plain terms: as chips become faster, hotter and more complex, the surrounding materials become increasingly critical to performance and reliability as the silicon itself. This is not a commodity material. Producing it requires decades of know-how in glass formulation, fiber spinning, chemical composition and manufacturing precision. The result is a niche market where Nittobo holds a very strong position and where new competitors cannot simply arrive by spending more capital.

The company’s recent earnings reinforced this picture. Management has provided guidance that implies more than 70% growth in operating profit over two years¹. More importantly, T-glass capacity is expected to increase threefold between March 2024 and March 2028¹. Yet even these investments may not be sufficient to fully relieve supply tensions. Demand for advanced semiconductor materials remains so acute that companies such as Apple and Nvidia reportedly travelled to Japan to secure additional production capacity directly.

This partly explains the strong share price performance already seen this year, but we believe the company can continue to benefit over the medium term from a clear shortage of advanced materials that are becoming essential for next-generation semiconductors, alongside growing demand for its specialty glass as electronics become smaller, more powerful and increasingly complex to manufacture reliably.

Nittobo is therefore a good example of an “off-the-beaten-path” AI beneficiary: relatively small compared with the giants of the semiconductor ecosystem, yet strategically embedded in one of its most constrained and valuable layers.

Safran offers a very different route to structural growth: not through AI infrastructure, but through the long-duration economics of the aerospace installed base.

The company recently reported exceptional first-quarter results, with organic revenue growth of 23%, well ahead of expectations. Management also indicated that it is now targeting the upper end of its operating profit guidance. Despite this, the share price has remained under pressure, as investors have focused on geopolitical tensions in Iran and the potential impact on airline capacity and travel demand.

The key question is whether the conflict could translate into a lasting deterioration in aircraft utilisation and aftermarket demand. The immediate impact appears limited. Groundings have increased in the Middle East, but the effect remains regional. Globally, fleet retirement rates are still historically low, suggesting that, so far, there is no clear evidence of a broad-based deterioration in global aircraft utilisation.

The more relevant risk would come from a prolonged period of higher oil prices. Persistently elevated fuel costs could prompt airlines to reduce aircraft utilisation and accelerate the retirement of older narrowbody aircraft. This would be negative for the CFM56 aftermarket — the older-generation commercial aircraft engine family produced by CFM International, the joint venture between Safran Aircraft Engines and GE Aerospace. Even under a more adverse scenario, however, we believe the earnings impact would remain manageable, particularly as the market may still be underestimating the ramp-up of LEAP, the newer-generation successor to the CFM56.

Beyond this short-term risk, the long-term investment case remains powerful. Safran is one of Europe’s highest-quality industrial companies, exposed to structural growth in global air traffic, generally estimated at 4% to 6% per year over the coming decades². The entire aerospace value chain is benefiting from the post-Covid recovery, but engine manufacturers are capturing an increasing share of value. In 2024, they generated around 42% of global commercial aerospace profits, roughly twice the level of a decade ago³.

In this context, Safran’s position is exceptional because of CFM International, its joint venture with GE Aerospace. CFM dominates the narrowbody aircraft engine market through the CFM56 and, increasingly, the LEAP engine, which powers the Airbus A320neo and Boeing 737 MAX. The persistent reliability issues affecting Pratt & Whitney’s GTF engine further strengthen LEAP’s competitive position.

The barriers to entry are almost impossible to overcome. Developing a new aircraft engine requires 5 to 15 years, tens of billions of investment, rare engineering expertise and lengthy certification processes. This explains why the industry has so few credible competitors. Even China is unlikely to offer a truly competitive alternative before the next decade.

Safran also benefits from outstanding visibility. Airbus and Boeing both have record order backlogs, representing more than ten years of deliveries at current production rates. For Safran, each new aircraft delivery is not just a one-off engine sale. It expands the installed base that will generate maintenance, repair and spare parts revenues for decades. This is the core of the model. Engine manufacturers typically make limited profits when engines are first sold. The real economics come later, through the aftermarket, over 20 to 30 years, with high margins and recurring cash flows. Every engine delivered today is therefore a long-duration revenue stream.

Current aerospace constraints may even reinforce this dynamic. Boeing production delays, supply chain pressure and labour shortages are forcing airlines to keep aircraft in service for longer. This “older for longer” fleet dynamic increases the need for heavy maintenance and supports an aftermarket super-cycle that is highly favourable for Safran.

There is also optionality in defense, which represents around 12% to 13% of sales. Safran supplies the Rafale engine, military helicopter engines for Airbus platforms, and propulsion, guidance and navigation systems for European missile programmes. In the current geopolitical environment, this provides an additional source of resilient demand.

In short, Safran should therefore not be viewed simply as a play on air traffic recovery. It is a strategic industrial asset built on a massive installed base, deep technological barriers and a business model that becomes more profitable with time.

The third example is Tradeweb, a US-listed electronic market infrastructure company. Founded in 1998 as an electronic marketplace for US Treasuries, it has since expanded across rates, credit, ETFs, equities and money markets. Rates remain at the heart of the business and continue to be the main driver of its earnings. The investment case is straightforward: fixed income is still significantly under-electronified. Voice trading is steadily losing ground to electronic platforms, especially in rates, swaps, liquid credit and money markets. Around 60% of rates trading and 45% of credit trading is electronic today, leaving a large runway for further migration, particularly in less standardised products². The company earns commissions on executed volumes, but it also monetises data, subscriptions, analytics and workflow tools. This means it is more than a cyclical broker. It is a platform with transactional revenue at its core, complemented by a recurring, sticky, software-like layer.

Its moat is built on network effects. Tradeweb serves a broad range of clients, including asset managers, hedge funds, banks, wholesalers, retail channels and, following the acquisition of Institutional Cash Distributors, corporate treasurers. The more it connects different pools of flow (institutional, wholesale, retail and corporate treasury) the stronger its liquidity and execution quality become. This is particularly valuable in fixed income, where liquidity remains highly fragmented.

Once clients are embedded in the workflow, switching away becomes difficult. Tradeweb’s multi-asset offering, spanning Treasuries, swaps, credit, ETFs and money markets, also creates significant cross-selling opportunities, strengthens client retention and deepens the platform’s role in daily market activity.

The company has a strong culture of innovation, with products ranging from automated execution and multi-asset trading to Treasury dealer algorithms, emerging-market swaps and new data partnerships. Its innovation is focused on areas where Tradeweb can offer differentiated, proprietary solutions to clients, rather than simply replicating products already available elsewhere.

The business also has defensive characteristics. Exchanges and electronic marketplaces often benefit from periods of volatility, as these tend to increase hedging, repositioning and trading activity.

The stock has corrected in recent months for two main reasons. First, the market has rotated aggressively towards the most direct AI beneficiaries, leaving even high-quality financial infrastructure companies behind. Second, the valuation had become demanding: Tradeweb was previously priced as a near-perfect compounder. The multiple has now fallen significantly, shifting the focus back to fundamentals.

Tradeweb is therefore a market infrastructure compounder: less spectacular than the AI leaders, but exposed to a durable shift in the way capital markets operate.