FDI saturation point: when does the manufacturing tailwind exhaust?

Vietnam's manufacturing FDI story, on paper, is in better shape than at any point in the country's modern history. The General Statistics Office reports that disbursed FDI in the first quarter of 2026 reached USD 5.41 billion, up 9.1 percent on the year and the highest first-quarter level in five years. Total registered FDI rose 42.9 percent year-on-year to USD 15.2 billion. Manufacturing and processing absorbed 82.8 percent of disbursement and 69 percent of newly registered capital. Singapore led source countries with USD 5.32 billion, followed by South Korea with USD 4.4 billion. Cumulative disbursed capital across the country's modern FDI history has now passed USD 355.7 billion, equivalent to roughly 65.6 percent of total registered capital. Q1 GDP printed 9.73 percent manufacturing growth, contributing 32.52 percent to the gross value added of the economy. The headline picture is one of a manufacturing engine running hot.

The composition underneath is more nuanced. Newly registered capital in January-February 2026 was concentrated in smaller projects: 620 newly registered projects in two months, with average ticket size considerably below 2024 norms, and capital adjustments to existing projects falling 52.3 percent year-on-year. Q1 capital adjustments were down 55.1 percent. The interpretation is not that interest is waning. The interpretation is that the composition of incoming projects is shifting toward smaller, more selective operations while existing investors are pausing on capacity expansion until the global tariff and tax environment clarifies. That is a different story than the headline number tells.

The desk's framework here is not a forecast of FDI numbers. It is a structural test: at what point do the friction costs that have been quietly building in Vietnam (wages, power, water, real estate, the OECD Pillar 2 minimum tax, and increasingly the political economy of US tariff scope) catch up with and partially offset the country's genuine and durable advantages? The short answer is that Vietnam is past the easy phase, where any large multinational with a Chinese supply chain made the +1 trip almost reflexively, and is now in the harder phase, where the marginal investor has to be more selective and the host country has to absorb the higher-spec end of the global value chain to keep the engine running. Both sides face execution risks that did not really exist five years ago.

Fact one: registered FDI growth has decoupled from disbursement growth, and the meaning of that gap has changed. In 2025, total registered foreign investment rose 0.5 percent year-on-year, while disbursed FDI grew 9 percent. In Q1 2026, registered FDI surged 42.9 percent (driven by a small number of large manufacturing, energy and high-tech projects) while disbursement grew a more measured 9.1 percent. Excluding the largest projects in Q1, underlying registered FDI growth was modest. The interpretation is that Vietnam continues to attract pipeline commitments, but the realisation rate (disbursement against registered) is becoming more dependent on a handful of mega-projects rather than the broad cohort of mid-sized investors that dominated the 2018-2023 phase.

Fact two: foreign-invested enterprises now drive an outsized share of exports, and the dependency is deepening. The OECD's 2025 Vietnam Economic Survey reported that foreign firms accounted for nearly three-quarters of the country's exports in 2023, with shares as high as 90 percent in headline product categories such as phones, computers, machinery, apparel and textiles. Samsung alone accounts for around one-fifth of national exports. In Q1 2026, the foreign-invested sector contributed USD 98.46 billion to Vietnam's USD 122.93 billion in exports, an 80.1 percent share. The domestic export sector contracted 16.6 percent year-on-year in the same quarter. This concentration is the most underweighted feature of the Vietnam story; any single anchor dislocation produces national-level numbers.

Fact three: wages have been rising faster than productivity in light manufacturing. Average monthly wages in industry and construction reached around VND 9.1 million (USD 348) in 2025, up roughly 8 to 10 percent on the year. Foreign-invested enterprises pay a 10 to 20 percent premium over local employers. In Ho Chi Minh City, Binh Duong and Dong Nai, FIE wages now run VND 9 to 10 million per month (USD 344-383). Average factory hourly wages of around USD 3 are still well below China's USD 6.50, but the gap with low-cost ASEAN peers (Cambodia, Bangladesh, parts of Indonesia) has narrowed materially. For the parts of the value chain where labour is the decisive cost, Vietnam is no longer the obvious winner.

Fact four: power is the binding constraint that nobody talks about until it isn't. Electricity demand is forecast to expand 10.5 to 13 percent in 2025 versus 2024, and 12 to 15 percent annually through 2030. The June 2023 blackouts that hit industrial parks in Bac Ninh and Bac Giang, two key northern manufacturing hubs, were not a one-off. They were a stress test that the system passed only just. Power Development Plan 8 (revised April 2025) targets total generation capacity of around 236 GW by 2030, up from just over 80 GW at end-2023. Whether that is achievable in five years, given offshore wind delays, transmission constraints, and the financial pressures on Electricity of Vietnam from low regulated tariffs, is an open question. Direct Power Purchase Agreements introduced under Decree 57/2025/ND-CP help large industrial consumers source renewables directly, but this is a partial fix.

Fact five: the corporate income tax framework changed in 2025, and the OECD Pillar 2 minimum tax broke part of the incentive model. The amended CIT Law of June 2025 left headline rates unchanged but restructured how corporate incentives are granted, particularly for manufacturers in industrial parks. More consequentially, Vietnam's adoption of Pillar 2 has rendered previous tax incentives largely ineffective for global enterprises with revenues above EUR 750 million. No new compensatory policy has been introduced. Several major corporations have postponed new investment decisions until the framework clarifies. This is the main mechanical reason behind the deceleration in capital adjustments by existing investors, and it is the most underweighted policy variable in the desk's coverage of the country.

Fact six: the US tariff scope is now a binding rather than a contingent risk. Recently announced higher US tariffs on Vietnamese exports affect roughly one-third of the country's exports to the US market, by OECD estimate. Some categories have seen six consecutive months of declining exports since the tariff warning was issued in April 2025. Traditional export categories such as textiles, footwear, and wooden products grew only 5-10 percent year-on-year in late 2025. The risk is not the tariff level itself, which is manageable; the risk is scope creep, where US Treasury or USTR action redefines transhipment to include intermediate inputs commonly sourced via Vietnam. That move would compress FIE margins across multiple categories simultaneously.

The China-plus-one curve, redrawn

The shorthand "China-plus-one" has done useful work in the supply-chain consultancy market and almost no useful work for capital allocators. It treats Vietnam as a single decision: invest or do not invest. The reality of how multinationals are now sequencing manufacturing capacity across Vietnam, Thailand, India, Mexico, and increasingly Indonesia is more granular. Vietnam now wins reliably on the "northern electronics with Korean and Japanese suppliers" decision, less reliably on the "garments at the lowest hourly cost" decision (where Bangladesh and parts of Africa are now competitive), and is increasingly being asked to compete on a "higher-tech component manufacturing with semiconductor adjacency" decision against Malaysia, Thailand, and India. The +1 has become a +n.

This recomposition is reflected in the Q1 2026 source-country mix. Singapore at 52 percent of total Q1 registered FDI is partly a routing artefact for global capital, but South Korea at nearly 36 percent is a real industrial signal: Korean investors are deepening exposure to electronics, semiconductors, AI infrastructure, energy, and smart urban infrastructure. Chinese investment, which became more visible after the 2018 trade-war diversion, is now politically sensitive in 2026 given US scrutiny over circumvention of tariff designations; its share has dropped to around 4 percent of Q1 registrations. Japan's contribution remains steady but unspectacular. The interesting movers are Indonesia (USD 1.7 billion in Q1) and Sweden (a notable European entrant in 2025 figures), suggesting that the next leg of incoming investment is broader-based geographically but more selective sectorally.

Vietnam's industrial geography has decentralised faster than most observers appreciate. Ho Chi Minh City and Hanoi, the two long-standing FDI magnets, have been joined by an arc of second-tier provinces with serious manufacturing density. The Q1 2026 distribution shows just how far the centre of gravity has moved: Thai Nguyen led all provinces with USD 5.7 billion (37.6 percent of total Q1 FDI), driven by Samsung-anchored capacity. Ho Chi Minh City was second with USD 2.9 billion (19.2 percent). Nghe An, in the central north, came third at USD 2.3 billion (14.8 percent), a province that would not have appeared in this league table five years ago. Bac Ninh, Tay Ninh, and Hanoi rounded out the top destinations.

The northern cluster, anchored by Samsung in Bac Ninh and Hai Phong and now reinforced by Foxconn, Pegatron, and LG, has become the country's electronics core. Hai Phong's Deep C industrial zones and the Yen Phong I-II and Que Vo industrial parks in Bac Ninh form the densest electronics supplier base outside China itself. The southern cluster remains broader-based, with consumer goods, food processing, and the LEGO Binh Duong manufacturing complex now operational. Central Vietnam, anchored by Da Nang and the central high-tech corridor, is the emerging tier; provinces like Nghe An have begun absorbing capacity that previously would have gone north. The administrative restructuring of Vietnam's industrial parks in 2025, shifting from a three-tier to a two-tier structure, accelerated investor processing time and reduced friction. Korean investors, the largest single nationality cohort outside Singaporean intermediaries, have been the most visible beneficiaries.

The geographic question that matters for 2026-2027 is whether the second-tier provinces can absorb the next round of capacity without running into the same power and water bottlenecks that pressured Bac Ninh in 2023. The desk's view is mixed: the north has more industrial-park headroom than is widely understood, but central-region and southern provincial grids will be the binding constraint by 2027 if PDP8 transmission build does not keep pace. The 600 MW Monsoon Wind Power Project (operational August 2025, electricity exported from Laos's Sekong and Attapeu provinces directly to Vietnam) is a credible example of cross-border ASEAN-cooperation problem-solving, but a single 600 MW project does not move the dial against the 2,200-2,500 MW of additional capacity required annually.

The base case treats Vietnam's FDI engine as durable but no longer accelerating. It assumes that Pillar 2 transition friction is a temporary policy gap rather than a permanent shift, and that the replacement incentive framework arrives in time to prevent material capital reallocation. The upside scenario depends primarily on whether the country becomes a meaningful node in semiconductor back-end and packaging investment, where its existing Korean and Japanese supplier density gives it a genuine advantage over Indonesia or the Philippines. The semiconductor angle is not a fantasy: Vietnam's policy framework includes a target of training 50,000 semiconductor engineers by 2030, several universities have launched dedicated chip design programmes, and Decree 57/2021/NĐ-CP offers up to fifteen years of corporate income tax reduction for high-tech projects. The stress case, conversely, depends on the conjunction of two unrelated risks: US tariff scope creep and domestic power reliability deterioration. Either alone is manageable; both at once is not.

The most underweighted variable in the public discussion is the foreign exchange channel. The Vietnamese dong is managed against a basket but with persistent intervention to maintain export competitiveness. The State Bank of Vietnam allowed the USD/VND rate to rise approximately 3.5 percent in 2025, with sell-side projections of a further 2.0 to 2.5 percent rise in 2026. As US-Vietnam interest rate differentials narrow, the trade-off between export support and inflation control becomes harder to manage. A meaningful dong appreciation would compress FIE margins precisely when wage growth and Pillar 2 effects are also pressuring them. The desk's working view is that the SBV will continue to lean against rapid moves in either direction, but the manageable range is narrowing.

The fiscal and tax dimension is the second underweighted variable. Q1 2026 state budget revenue was up 11.4 percent year-on-year, reaching 32.8 percent of the annual estimate, while expenditure ran at 23.1 percent year-on-year growth. Vietnam's public investment plan for 2026 is VND 1.1 quadrillion, up 30 percent from 2025. The combination of expansionary fiscal policy and the OECD Pillar 2 transition creates an unusual configuration: the state is pushing capacity into the economy through public works while simultaneously removing the most powerful FDI incentive lever it had. The replacement framework, when it arrives, will determine whether large multinationals re-engage at the pre-2025 pace or settle into the smaller-ticket pattern visible in the 2025 and Q1 2026 data.

First. Maintain core long exposure to Vietnam's FDI-anchored real economy through industrial real estate, port logistics, and the small set of listed names with genuine FIE supplier relationships. The disbursement print is the cleanest signal of pipeline durability and remains supportive. The five-year-high reading in Q1 2026 should not be over-interpreted, but neither should it be dismissed.

Second. Underweight Vietnam-listed equity beta as an expression of the manufacturing thesis. The disconnect between real-economy progress and listed-equity performance has been a feature of this market for two decades and shows no immediate sign of resolving. Foreign fund access constraints, settlement frictions, and the inclusion of large state-owned enterprises in major indices distort the read-through. The State Securities Commission's increased engagement with foreign investors and international financial institutions in 2025 is constructive but not yet decisive.

Third. Treat power reliability as a binary risk to specific tier-2 and tier-3 supplier names. Stress-test exposure to manufacturers without dedicated industrial-park power reliability, on-site generation backup, or DPPA arrangements with renewable producers. The exposure is not the headline FDI story; it is the blow-up risk that travels with it.

Fourth. The semiconductor packaging trade is a multi-year position, not a 2026 trade. Build position quietly; expect headline volatility around announcement cycles but a slow, real ramp. Hai Phong, HCMC High-Tech Park, and Bac Ninh are the credible nodes. Thai Nguyen's emergence in the Q1 2026 league tables suggests that capacity is now spilling beyond the established two-cluster framework.

Fifth. Hedge against the most underweighted tail risk, which is a US Treasury or USTR action that broadens the transhipment definition to include intermediate inputs. This would compress export margins for FIEs across multiple categories simultaneously. The cleanest hedge is in the currency, not the equity. The second-cleanest hedge is in selectively long ASEAN peers (Malaysia, Thailand, the Philippines) that would absorb deflected manufacturing flow.

The honest read of Vietnam in April 2026 is that the country has done something genuinely rare in development economics: it has converted twenty years of cheap-labour FDI into the kernel of a higher-spec manufacturing economy without crashing through the middle-income trap. That achievement is not yet locked in. The transition from labour-arbitrage host to value-chain participant is exactly the phase where wages, infrastructure, and regulatory complexity tend to outpace policy capacity. Vietnam is not yet past that phase, and Pillar 2 plus US tariffs have arrived at the worst possible moment in the transition.

For the institutional allocator, the consequence is straightforward. Treat FDI flow as the headline indicator of pipeline durability, treat power reliability and tariff scope as the two non-trivial tail risks, treat Pillar 2 transition as the policy variable that resolves over twelve to eighteen months, and treat the listed equity market as a separate and only loosely correlated expression of the underlying story. The thematic narrative has done its work. The granular work is now in the names, the locations, the supplier-tier reliability questions, and the watching of the OECD Pillar 2 replacement framework that has not yet been published.