Global foreign direct investment rose 6% to $1.6 trillion in 2025, ending two years of decline. The headline suggests a broad return of confidence. UN Trade and Development’s World Investment Report 2026 shows something more selective: capital is concentrating in a small group of countries, strategic sectors and megaprojects.
Developed-economy inflows increased 11%, while developing economies recorded growth of only 2%, reaching $901 billion. The top 20 host economies attracted more than 80% of global FDI. Least developed countries received $43 billion—up 21%, but still only 2.7% of the total, with flows concentrated largely in resource-rich economies.
The sector shift is even sharper. AI infrastructure, advanced technologies, critical minerals, energy-transition technologies and semiconductors accounted for 44% of global greenfield project value in 2025, up from 16% in 2020. Announced value in these strategic sectors rose from $109 billion to $576 billion over that period.
This is not simply a cyclical investment recovery. It is a reordering of what companies build abroad and which locations can compete for it.
Megaprojects can distort the headline
Much of the rise in FDI reflected a limited number of very large projects, particularly data centres and related digital infrastructure. Oil and gas and semiconductor projects also contributed. Meanwhile, many other sectors—including manufacturing, infrastructure and renewable energy—registered declines.
A single project can transform a country’s annual inflow statistics without producing a similarly broad economic transformation. Capital-intensive facilities may create substantial construction demand and tax revenue but relatively few permanent jobs. Their imported equipment and specialised inputs can limit local supplier gains unless host economies prepare for them.
Governments and corporate strategists should therefore look beneath aggregate flows. The relevant questions are how many projects were started, which assets were created, where inputs come from, what skills are developed and whether the investment connects to domestic firms. A rising dollar total does not automatically mean broader productive capacity.
Location decisions now combine economics and security
Strategic projects depend on reliable electricity, water, digital networks, transport, specialised labour and predictable regulation. They also sit inside national-security and industrial-policy debates. Semiconductor plants, data centres and mineral processing facilities are increasingly evaluated for resilience and geopolitical alignment as well as cost.
Governments adopted a record 229 investment-policy measures in 2025. Most were favourable to investors, but incentives and screening are becoming more targeted toward domestic priorities and economic security. This changes the location contest. A tax incentive alone cannot compensate for an unreliable grid, slow permitting, scarce engineering talent or uncertainty about data and trade rules.
For companies, public support can improve project economics but may come with conditions on employment, sourcing, technology, capacity or future expansion. Investment teams should test whether incentives survive political change and whether compliance obligations are compatible across jurisdictions.
The development gap is a capability gap
Low-income and lower-middle-income economies attracted only about 10% of strategic-sector investment between 2020 and 2025, compared with more than 20% of investment in other sectors. The imbalance matters because today’s projects influence where future industrial clusters, technical skills and supplier networks form.
Countries cannot replicate every subsidy offered by large economies, nor should they attempt to host every strategic industry. More realistic competition begins with an entry point: component manufacturing, business services, renewable power, mineral processing, testing, maintenance or regional logistics. Reliable implementation can be more valuable than an ambitious strategy without infrastructure or institutional coordination.
Regional markets also matter. A small domestic customer base may not justify a major facility, while harmonised standards and cross-border infrastructure can create investable scale. Supplier-development programmes and vocational training can connect foreign projects to local businesses instead of leaving them as isolated enclaves.
Boards need a wider definition of investment risk
Traditional site selection compares wages, taxes, transport and market access. Strategic investment adds energy security, water stress, export controls, cyber risk, investment screening, community acceptance and the durability of industrial policy.
Concentration creates portfolio risk as well. If the same small group of locations attracts data centres, chips and advanced manufacturing, constraints in electricity, construction labour or permitting can spread across projects. Companies may need to sequence commitments, develop alternative sites or invest jointly in enabling infrastructure.
The objective is not geographic diversification at any price. Some activities benefit strongly from clusters of suppliers, talent and research. The board-level task is to distinguish productive clustering from dependency on a single policy regime, network or geopolitical route.
UNCTAD expects the 2026 outlook to remain clouded by trade uncertainty, conflict, high financing costs and fragmentation. Yet competition for strategic projects is likely to intensify. The investment rebound therefore offers no return to the old assumption that mobile capital will spread automatically.
Capital is moving toward places that combine physical capacity, institutional credibility and strategic relevance. For host economies, attracting a megaproject is only the first step; the larger prize is turning it into skills, suppliers and lasting productive capability. For companies, the winning location will be the one that works after incentives, headlines and geopolitical conditions change.
Featured photograph: NASA Glenn Research Center via Wikimedia Commons, a United States federal government work in the public domain.
