Artificial intelligence, emerging markets and SMID caps: Carmignac Investissement’s recipe

The big structural shifts reshaping global equity markets, from artificial intelligence to emerging markets, don't hand investors easy portfolio decisions. They demand a method for telling companies that will truly benefit apart from those that only look exposed. We put that question to Kristofer Barrett, manager of the Carmignac Investissement fund.

In the interview, Barrett walks through the fund's highest-conviction views, from the AI cycle to the structural case for emerging markets, explains why he favors EV/FCF over traditional multiples when picking stocks, discusses how the fund balances concentration and diversification across around 80 holdings, and tackles the SMID-cap question along with the challenges agentic AI poses for the software sector.

Looking at the global equity universe as a whole, what are your strongest convictions today in shaping the portfolio's positioning?

Our strongest conviction remains the AI investment cycle. Demand for computing power continues to expand rapidly, with strong earnings growth across the semiconductor ecosystem. However, we are increasingly focused on where the economic value of this investment ultimately accrues. Hyperscalers are committing unprecedented amounts of capital to AI infrastructure, putting pressure on free cash flow. The key question is whether they are investing because they want to or because competitive pressure means they have to. What is clearer is where this capital is flowing: into hardware, semiconductors and the broader AI infrastructure value chain. Beyond AI, we also maintain exposure to more defensive areas such as healthcare distribution. These businesses have tended to perform well during periods when AI-related stocks come under pressure.

Emerging Markets remain a structural component of the portfolio. What is the rationale behind this exposure today, and where within EM do you find your strongest convictions?

Emerging markets represented around 26% of the portfolio at the end of August, roughly twice their weight in the MSCI ACWI. This reflects bottom-up stock selection rather than a top-down allocation. We invest in the region because we believe it offers a broad range of high-quality companies with attractive long-term growth prospects, despite often being overlooked by some investors due to perceived volatility. Emerging markets are also central to the AI value chain. Taiwan, representing around 15% of the portfolio, is critical to advanced semiconductor manufacturing, while South Korea, at around 6%, is a leader in high-bandwidth memory. Beyond AI, we see compelling opportunities in consumer and financial businesses, including MercadoLibre, Sea, Allegro and Itaú. These companies benefit from structural trends such as digital adoption, financial formalisation, favourable demographics and the continued development of underpenetrated domestic markets across Latin America, Asia and Central Europe.

The fund uses EV/FCF (Enterprise Value/Free Cash Flow) as its central metric rather than traditional earnings multiples. How does it translate into balancing higher-growth names against others with more moderate growth but attractive valuations?

Using EV/FCF allows us to compare companies with very different growth profiles on a common basis: the cash they can ultimately generate relative to their enterprise value. Unlike traditional earnings multiples, it also captures balance-sheet structure and capital intensity, making it particularly useful across different business models. In practice, this supports a balanced approach. On the one hand, we own companies such as McKesson and Cencora2, which offer moderate but highly visible growth, strong free-cash-flow generation and attractive valuations. Their defensive healthcare businesses provide recurring demand and resilient cash flows. On the other hand, we are prepared to accept higher valuations where we believe structural growth can drive a much stronger expansion in future free cash flow. Through sustained investment in AI infrastructure, Nvidia and Broadcom2 are prime examples of this .

The fund holds around 80 stocks, a number designed to balance conviction and diversification. How do you decide how much risk to concentrate in a high-conviction idea versus keeping a broader portfolio?

Position sizing is not simply a function of conviction. It also reflects valuation, downside risk, liquidity and each stock’s contribution to overall portfolio risk. We want our highest-conviction ideas to have a meaningful impact on performance, while avoiding excessive dependence on a single company, theme or market factor. Diversification therefore comes from combining businesses with different earnings drivers and risk profiles, rather than simply owning a large number of stocks. This is particularly important when market leadership can shift rapidly, even within strong structural themes such as AI. We also preserve flexibility to invest outside our core convictions when valuations become compelling. After a sharp sell-off, we may initiate a small position when we believe the market is overly discounting a company’s fundamentals, and build it progressively as our conviction increases. Zoetis and Uber are good examples of this2. Conversely, when a stock reaches what we consider fair value, we are willing to crystallise gains and reduce or exit the position.

Roughly a quarter of the fund's holdings sit outside the MSCI AC World. What kind of research process do you use to identify these less-followed opportunities?

The fact that roughly a quarter of the portfolio sits outside the MSCI AC World is not a target in itself, but the result of our bottom-up research process. When we identify a structural trend, we analyse the entire value chain rather than focusing only on the largest or most obvious beneficiaries. For each theme, we look for bottlenecks, pricing power, barriers to entry and products that are difficult for customers to replace. This often leads us towards smaller or less-followed companies that play a critical role in an ecosystem but receive limited analyst coverage. Our research is highly fundamental and collaborative. Together with Carmignac’s wider investment team, assumptions are tested through discussions with company’s management teams, suppliers, customers, competitors and local industry specialists. The goal is to understand not only a company’s fundamentals, but also how its position within the value chain is evolving and whether its competitive advantage is sustainable.

What do you look for in a SMID that analyst coverage hasn't yet priced correctly? Within the semiconductor ecosystem, the fund has found value in suppliers of critical materials along the supply chain

One distinctive feature of the portfolio is its exposure to SMID caps. Like emerging-market companies, these businesses can be underappreciated because they receive less analyst coverage and have lower visibility than developed-market large caps. We look for niche leaders with a durable competitive advantage, strong balance sheets, attractive cash generation and products or services that are difficult to replace. At the end of August, there were over 20 SMID caps representing around 10% of the portfolio. The Japanese company, Nitto Boseki. manufactures specialised glass fibre and electronic materials, including T-glass used in advanced semiconductor packaging and is a prime example of a SMID. As processors and high-bandwidth memory are packed increasingly close together, these materials help improve reliability and control thermal expansion. Its manufacturing expertise is difficult to replicate, making it an underappreciated beneficiary of AI infrastructure investment. This SMID exposure also helps us build a more agile and diversified portfolio, particularly in an environment where global equity indices remain highly concentrated.

Where do you see the biggest "off-the-beaten-path" opportunities in the AI value chain today, beyond chipmakers and hyperscalers?

Beyond chipmakers and hyperscalers, we see compelling opportunities among less visible companies providing the tools, materials and infrastructure needed to turn AI demand into physical computing capacity. As compute becomes more powerful and complex, bottlenecks are emerging across semiconductor manufacturing, advanced packaging, connectivity and cooling. The scale of investment is significant. TSMC spent $15.7 billion on capex in the second quarter of 2026, while Intel continues to invest heavily in manufacturing capacity. This supports demand across the broader semiconductor supply chain. Within equipment, DISCO provides precision tools used to cut, grind and thin wafers, while Grand Process Technology supplies cleaning, etching and plating equipment for advanced packaging. We also see opportunities in enabling materials and connectivity: JX Advanced Metals produces high-purity semiconductor materials, Sumitomo Electric develops optical components for faster data transmission, and LOTES supplies high-speed connectors and liquid-cooling components for AI servers2.

Software stocks have suffered a broad de-rating despite solid competitive positions. As agentic AI advances, what characteristics distinguish software companies that will defend their value from those at risk of being disintermediated?

Agentic AI will create clear winners and losers within software. We favour companies embedded in mission-critical workflows, with proprietary data, high switching costs and direct access to large installed customer bases. These advantages should help them integrate AI into existing products and deliver measurable productivity gains. By contrast, generic point solutions and thin wrappers around third-party models appear more vulnerable to commoditisation. Atlassian illustrates the characteristics we look for. Jira, Confluence and Jira Service Management2 sit at the centre of critical workflows, giving the company access to the organisational context AI agents need to automate processes, while its installed base provides a natural distribution channel for new AI features. However, competition from AI-native players remains intense and monetisation is still uncertain. Valuations are also not particularly cheap: the S&P 500 Software Index trades at around 25x one-year forward earnings, versus roughly 19.5x for the S&P 500. We therefore maintain a limited exposure to software stocks.

1 Source: Carmignac, 31/08/2026. Reference indicator: MSCI AC World NR index. The portfolios of Carmignac funds may change without previous notice.2 Reference to certain securities and financial instruments is for illustrative purposes to highlight stocks that are or have been included in the portfolios of funds in the Carmignac range. This is not intended to promote direct investment in those instruments, nor does it constitute investment advice. The Management Company is not subject to prohibition on trading in these instruments prior to issuing any communication. The portfolios of Carmignac funds may change without previous notice. 3 Source: Yahoo finance, 31/08/2026. This is not intended to promote direct investment in those instruments, nor does it constitute investment advice.

Marketing communication. Please refer to the KID/KIID, prospectus of the fund before making any final investment decisions. This document is intended for professional clients.

Carmignac Investissement A EUR Acc : Recommended minimum investment horizon 5 years ; Risk indicator 4/7 (Risk Scale from the KID (Key Information Document). Risk 1 does not mean a risk-free investment. This indicator may change over time) . Main risks of the Fund: equity, currency and discretionary management. The Fund presents a risk of loss of capital. Please refer to the Fund’s prospectus to view the exhaustive list of risks.

This material may not be reproduced, in whole or in part, without prior authorisation from the Management Company. This material does not constitute a subscription offer, nor does it constitute investment advice. This material is not intended to provide, and should not be relied on for, accounting, legal or tax advice. This material has been provided to you for informational purposes only and may not be relied upon by you in evaluating the merits of investing in any securities or interests referred to herein or for any other purposes. The information contained in this material may be partial information and may be modified without prior notice. They are expressed as of the date of writing and are derived from proprietary and non-proprietary sources deemed by Carmignac to be reliable, are not necessarily all-inclusive and are not guaranteed as to accuracy. As such, no warranty of accuracy or reliability is given and no responsibility arising in any other way for errors and omissions (including responsibility to any person by reason of negligence) is accepted by Carmignac, its officers, employees or agents. Reference to certain securities and financial instruments is for illustrative purposes to highlight stocks that are or have been included in the portfolios of funds in the Carmignac range. This is not intended to promote direct investment in those instruments, nor does it constitute investment advice. The Management Company is not subject to prohibition on trading in these instruments prior to issuing any communication. The portfolios of Carmignac funds may change without previous notice. The reference to a ranking or prize, is no guarantee of the future results of the UCIS or the manager. Access to the Funds may be subject to restrictions regarding certain persons or countries. This material is not directed to any person in any jurisdiction where (by reason of that person’s nationality, residence or otherwise) the material or availability of this material is prohibited. Persons in respect of whom such prohibitions apply must not access this material. Taxation depends on the situation of the individual. The Funds are not registered for retail distribution in Asia, in Japan, in North America, nor are they registered in South America. Carmignac Funds are registered in Singapore as restricted foreign scheme (for professional clients only). The Funds have not been registered under the US Securities Act of 1933. The Funds may not be offered or sold, directly or indirectly, for the benefit or on behalf of a «U.S. person», according to the definition of the US Regulation S and FATCA.The risks, fees and ongoing charges are described in the KID (Key Information Document). The KID must be made available to the subscriber prior to subscription. The subscriber must read the KID. Investors may lose some or all their capital, as the capital in the funds are not guaranteed. The Funds present a risk of loss of capital. The Funds’ prospectus, KIDs, NAVs and annual reports are available at www.carmignac.com/en, or upon request to the Management. Carmignac Investissement is a French investment fund (fonds communs de placement or FCP), a common fund in contractual form conforming to the UCITS or AIFM Directive under French law. The Management Company can cease promotion in your country anytime. Investors have access to a summary of their rights at section 5 entitled "summary of investor rights" on the following links: UK ; Switzerland ; France ; Luxembourg ; Sweden. Belgium (French) ; Belgium (Dutch).