How much investment there was, and what it was made of
Global FDI inflows rose from $1.5 trillion in 2024 to $1.6 trillion in 2025. UNCTAD says that conduit flows, meaning money routed through financial centres such as Switzerland and Ireland, subtracted about $40 billion from the 2025 total (against $70 billion in 2024), and that Switzerland and Ireland accounted for most of the change (Ch. I, p. 5). Even so, global inflows excluding conduits were about 10% above their 2010-2024 average (p. 5).
Other measures were weaker. Cross-border mergers and acquisitions fell 7% (p. 9). Announced new-build ("greenfield") projects stayed close to 2024's high, but mainly because of megaprojects in developed economies. Rates of return on inward FDI fell from 10.2% in 2024 to 7.0% in 2025, even though the profits of the largest multinationals rose (p. 15, Table I.2 and Figure I.6). UNCTAD reads this as a growing concentration of profits in a small group of dominant firms in strategic sectors.
Where it went
The United States stayed the largest host at $277 billion. Singapore, Hong Kong (China) and China followed. Developing economies were half of the top 20 hosts (p. 9). Four regional points stand out:
- Europe rose 39% to $285 billion, driven largely by financial-centre movements and a few large deals. Germany's inflows rose from $21 billion to $74 billion (p. 18).
- Developing Asia rose from $623 billion to $644 billion (pp. 25-26). China fell from about $116 billion to $105 billion, and UNCTAD describes a shift from scale to quality. West Asia rose from $92 billion to almost $111 billion (p. 29).
- Latin America and the Caribbean rose to $188 billion, helped by Brazil. Mexico's inflows rose from about $38 billion to $41 billion, but its announced new-build projects fell from about $44 billion to $24 billion, which UNCTAD takes as a sign that nearshoring was delayed, not abandoned (pp. 17, 33, 36).
- Africa fell from an exceptional $94 billion to $70 billion, because 2024 was lifted by the Ras El-Hekma megaproject in Egypt. It was still the third-highest level since 1990 (p. 21).
On the investor side, the top five home economies were unchanged, and the United States, Japan and China were the top three (p. 11). Gulf investors and sovereign wealth funds are becoming more prominent, including in Africa, where the United Arab Emirates has accelerated since 2021 (pp. 13, 25).
Within Africa the picture was uneven. Guinea's inflows rose more than fivefold to about $8 billion on bauxite and iron ore, Mozambique's rose to about $6 billion on gas, and Egypt remained the largest recipient at about $15 billion. South Africa recorded negative inflows of about $2.3 billion because of intracompany financial flows, profit repatriation and acquisitions (pp. 22-23). The top 10 announced projects made up roughly 40% of African new-build value (p. 23).
Which sectors
Announced projects and project-finance deals are growing in a narrow set: data centres, oil and gas, and semiconductors. Megaprojects of $1 billion or more were 44% of announced new-build value in 2025, against 22% in 2017 (p. 41). In infrastructure, digital assets attract capital while the rest struggles. Renewable energy saw a fourth annual fall in international investment, though it is still the largest infrastructure segment at over $600 billion (p. 43).
Manufacturing tied to global value chains weakened: announced project value fell 13% and project numbers 15%. Semiconductors were the exception, rising 11%, and automotive fell by more than a quarter (pp. 45-46). Critical minerals bucked the weak trend in extractive industries, with announced projects rising from $22 billion to $26 billion (p. 44).
Investment in sectors tied to the Sustainable Development Goals recovered, with announced projects up 11% to $233 billion and project-finance deals up 26% to $386 billion. Gains were uneven: telecommunications and food and agriculture rose strongly, while announced renewable energy projects in developing economies halved from $109 billion to $55 billion (pp. 46-48).
Smaller and poorer economies
Inflows to least developed countries (LDCs) rose 21% to $43 billion, driven by a small number of African LDCs that received about $33 billion, mostly in resources, energy and large infrastructure. Landlocked developing countries were roughly flat at about $25 billion, and small island developing States rose 4% to about $10 billion (p. 38). Announced new-build projects in LDCs rose from $3.5 billion to $14.3 billion and project-finance deals by 164% to $21 billion, but UNCTAD notes the gains rest on a narrow set of projects and sectors, while transport and water investment in LDCs stayed weak (pp. 46, 50).
Investors
The top multinationals increased foreign assets by 8% in 2025, led by technology, pharmaceuticals and electric vehicles (p. 51). More than a quarter of the top 100 are State-owned, and private equity accounted for about 20% of global mergers and acquisitions in the last two years (pp. 3, 51). The average foreign share of assets of United States multinationals fell from 43% to 39% between 2017-2019 and 2022-2024, which points to a tilt towards the home market (p. 54).
The outlook and the Middle East
UNCTAD expects real investment activity to stay subdued in 2026, held back by geopolitical tension, trade policy uncertainty and fragmentation (p. 5). A box on the 2026 Middle East conflict calls it a major adverse shock: capital-intensive projects in energy, infrastructure and industry are the most exposed, shipping routes and costs are disrupted, and higher oil and gas prices feed inflation. It also notes a dual response to high oil prices, with more fossil-fuel investment and also more renewable investment (p. 14). Financial market analysts quoted in the report expect mergers and acquisitions to recover gradually in 2026, but selectively (p. 16).
Policy: promotion and control together
Governments in 104 countries adopted 229 measures, a record. Incentives made up half of favourable measures, and developed countries used them more (60% of their favourable measures, against under 25% in 2016-2020) (pp. 71, 77). Incentives are also becoming more targeted and tied to jobs, technology and local value, partly because of the global minimum tax (p. 77).
Examples of targeting include Brazil's special tax regime for data centres, Thailand's double deduction for large electric-vehicle investment, and a Saudi programme covering up to 35% of initial project cost in selected industries (pp. 77, 79). Every liberalising measure in 2025 came from a developing country, mostly in Asia, and mostly in services such as finance, telecoms and tourism (p. 82).
Restrictive measures rose to 62. National security screening made up almost 40% of all restrictive measures in 2025, and every one of those measures came from a developed country. In developing countries, restrictions more often tightened incentives or added local-content rules (pp. 83, 86). UNCTAD stresses that screening mostly adds scrutiny rather than refusals: of 1,259 projects given an in-depth review, 4.8% were rejected and 17.6% were approved with conditions. It also warns that the real effect may be larger, because some investors may never apply (p. 84).
Treaties and disputes
Countries signed 44 investment agreements in 2025, 24 bilateral treaties and 20 broader economic agreements with investment chapters, bringing the total to 3,369 (p. 88). Newer agreements cover facilitation (77%) and cooperation (77%) more than before, and protection provisions, though still common, fell to 62% (p. 89). The Investment Facilitation for Development Agreement now covers 129 WTO members (p. 94).
Investors brought 56 known treaty-based arbitration cases in 2025, taking the total to 1,463. About 80% were against developing countries, and Latin America and the Caribbean had about 45%. Around a third concerned extractive activity, including critical minerals. Most cases rely on old treaties signed before 2010 (pp. 98-100).
Chapter III: the strategic-sector contest
AI infrastructure is the largest strategic segment, rising from under $50 billion in 2020 to almost $350 billion in 2025, about three fifths of the strategic total. Semiconductors grew fastest, at 54% a year, while critical minerals and energy transition technologies each grew about 17% a year (pp. 117-118).
Concentration is high. In 2025 the top three investor economies accounted for an average 72% of strategic-sector project value, and the top three recipients 56%, against 27% and 34% in other sectors (p. 118). Critical minerals is the one strategic area where low-income countries get a noticeable share, because of where the resources are (p. 119).
UNCTAD traces industrial policy through several eras, from state-led import substitution after the 1950s to today's focus on resilience and economic security. Of the stated motives it can observe, national security and geopolitical concern rose by 21 percentage points between 2015 and 2024, and security of supply by 13, while strategic competitiveness fell by 19 and climate change mitigation by 15. It cautions that motives are recorded for fewer than 35% of measures (pp. 132-134).
Subsidies are nearly half of industrial policy actions recorded over the past decade. They are 91% of China's interventions and 72% of the European Union's, but only 18% in developing countries, many of which cannot sustain large subsidies (pp. 107, 136). Outside strategic sectors, announced manufacturing investment fell 17% between 2015-2019 and 2021-2025. The decline was sharper for investors from the United States (27%) and Japan (51%), while investors from China and the Republic of Korea increased theirs (pp. 153-154).
Tariffs also create exposure. UNCTAD gives country-sector examples in which tariff increases weigh on export platforms in Cambodia, Thailand and Viet Nam, and on narrowly based exporters in Central America (p. 157).
UNCTAD sees selective adjustment, not a broad move of factories to new low-cost countries. In its survey of investment promotion agencies, 60% reported cancelled or downsized projects and 50% reported projects relocating in, with LDC agencies much more affected: 85% reported cancellations or downsizing against 47% elsewhere in the developing world (pp. 110, 153). Political alignment, more than distance, has so far driven where manufacturing moves (p. 156).
What UNCTAD recommends
For developing countries it advises picking a few realistic entry points rather than competing everywhere, spending scarce funds on bottlenecks such as power, infrastructure and skills, and replacing broad tax holidays with targeted, capped and performance-based incentives (pp. 173-178). On security, it suggests clear and proportionate rules and, for low-capacity countries, lighter tools than a full screening system, such as listed controlled activities (pp. 181-182). It also urges regional corridors and supplier networks (pp. 184-186).
For the international system it proposes a search for "the commons": a global mechanism to monitor investment-relevant measures, guidelines for security-related investment rules, and shared project-preparation and risk-sharing platforms among development banks and others (pp. 188-189).