AI is a structural driver of equity returns, but concentration risk remains a key consideration. US equities could be vulnerable to corrections if the Fed changes course. Investors should therefore look across a broader opportunity set, from infrastructure providers to AI adopters across sectors and regions.
Artificial intelligence is rapidly becoming a key driver of transformation across industrial sectors. From an investment perspective, the AI value chain is broadening beyond a pure technology theme, becoming more capital intensive and increasingly shaped by geopolitics.
AI opportunities across the value chain
| AI value chain | ||||||
| Upstream | Upstream / Midstream | Midstream | Downstream | |||
| Build the infrastructure | Bridge compute into systems | Deploy at scale | Deliver AI applications | |||
| Semiconductor manufacturing (chips, tools, hardware) | Advanced packaging (memory, package) | Server assembly (network, server) | AI system integration | Data centre construction (cloud, data centres, power) | AI deployment and operations | End-user applications (platforms, software, services) |
Sources: Amundi Investment Institute, internal elaboration as of 26 June 2026.
Seek breadth across the AI value chain
"As the AI story shifts from who can build the frontier to who can scale it, investing will be about seeking breadth across the full value chain and diversifying* against technological, geopolitical and physical risks. This backdrop calls for a selective, diversified*, and protective approach."
The upstream segment1, where leadership is concentrated among a limited set of players, continues to capture the strongest pricing power, with the US remaining best placed in terms of innovation, particularly focusing on chips, semiconductor capital equipment and cloud infrastructure. However, we recognize that converting that lead into scale is becoming more difficult as constraints begin to bite. As the ‘winner-takes-all’ dynamic fades, value is likely to migrate downstream2. Dispersion across sectors and regions is likely to widen, which, combined with future technological, policy and geopolitical shifts, makes diversification* essential for portfolio resilience.
AI fragmentation, different opportunities
| AI strategic regions | ||
| Country | AI value chain segment | AI edge |
| US | Full chain | Frontier AI & Cloud leader |
| Mainland China | Mid-downstream | Deployment & industrial scaling leader |
| Taiwan | Up-midstream | Foundry & packaging leader |
| Japan | Up & downstream | Industrial enabler |
| South Korea | Up-midstream | AI memory leader |
| India | Downstream | Services & adoption |
| EU | Downstream | Enablers & regulated adoption |
Looking across developed markets and beyond AI, the US investment and earnings cycle is expected to continue supporting equities in H2, with gains broadening beyond Big Tech. Nevertheless, elevated valuations and concentration risk leave the S&P 500 vulnerable to a correction if the Fed shifts course.
European equities are caught between competing forces. Strong fiscal stimulus in Germany, the EU’s greater focus on strategic domestic sectors and attractive valuations all provide support. However, the region will remain vulnerable in the near term if the energy crisis persists for longer.
Japan, by contrast, is supported by medium- to long-term tailwinds. The new political direction and economic backdrop should help profitability improve this year, assuming oil supply normalises relatively quickly. Corporate leverage appears appropriate, earnings growth is expected to be in double digits, the yen should be a less significant headwind, and valuations stay attractive relative to peers.
| US | Neutral overall, with a preference for equal-weight exposure |
| Europe | Slightly positive with focus on sectors benefiting from investments |
| Japan | Remain constructive as corporate reforms continue |
Source: Amundi Investment Institute Mid-Year Outlook 2026, “Power of endurance”, June 2026.
* Diversification does not guarantee a profit or protect against a loss.
Views and opinions are as of end June 2026 and are subject to change without prior notice
1 Upstream AI refers to the enabling layer of the AI ecosystem, including chips, compute infrastructure, data infrastructure and foundation models.
2 Downstream AI refers to the applications layer of the AI ecosystem, where AI capabilities are embedded into products, services and workflows used by businesses and consumers.
Unless otherwise stated, all information contained in this document is from Amundi Asset Management S.A.S. and is as of 20 July 2026. Diversification does not guarantee a profit or protect against a loss. The views expressed regarding market and economic trends are those of the author and not necessarily Amundi Asset Management S.A.S. and are subject to change at any time based on market and other conditions, and there can be no assurance that countries, markets or sectors will perform as expected. These views should not be relied upon as investment advice, a security recommendation, or as an indication of trading for any Amundi product. This material does not constitute an offer or solicitation to buy or sell any security, fund units or services. Investment involves risks, including market, political, liquidity and currency risks. Past performance is not a guarantee or indicative of future results.
Date of first use: 20 July 2026
Doc ID: 5728517
