The Perils of Vietnam’s Land-led Growth

INTRODUCTION

In April 2026, Ho Chi Minh City (HCMC), Vietnam’s commercial capital, broke ground on four infrastructure projects worth a combined VND142 trillion (USD5.5 billion). Rather than pay cash, the city set aside 33 prime plots for the investors. The event matters less for the projects than for the mechanism behind them. It marks HCMC’s return to the Build-Transfer (BT) scheme, under which a private developer builds public works and is paid in land, as a primary form of contract. This took place five years after Vietnam’s own 2020 Public-Private Partnership (PPP) Law excluded BT on the grounds of corruption risk and weak oversight of land transfers. The reversal signals Vietnam’s larger strategy: to finance an ambitious growth and infrastructure programme — the 2026–30 public investment plan alone runs to VND8.22 quadrillion (USD316.2 billion) — by monetising the one asset the state controls in abundance, land.

The model runs through four self-reinforcing mechanisms. Land is fiscalised, becoming a core source of government revenue. Credit is collateralised and expands, as rising land values loosen borrowing constraints. Credit is diverted, flowing into property and away from more productive uses. And land rents are captured, as discretionary control over conversion and zoning hands value to connected insiders. Each mechanism is manageable on its own; together they are hard to escape, because a downturn strikes public budgets, bank balance sheets, developer solvency and household wealth at once. East Asia offers ample precedent for both outcomes: Singapore and South Korea show that land-led development can support growth when rules-based institutions capture land-value gains and recycle them into infrastructure, industrial land and mass housing; China, Japan and Thailand, by contrast, show how binding public finance to the property cycle can destabilise the wider economy.

This article argues that Vietnam is assembling all four mechanisms faster than the safeguards needed to contain them can be put in place, with the revived BT scheme sitting at their intersection. It examines each mechanism in turn, and the structural risks they compound, before setting out how the land-for-infrastructure model can be governed rather than merely deployed.

THE FISCALISATION OF LAND

Vietnam’s budget has quietly become a land budget. Revenues from housing and land rose from VND123 trillion (USD4.68 billion) in 2016 to VND575.5 trillion (USD22 billion) in 2025, almost double the 2024 final-account figure and around five times the 2016 level (Figure 1). Measured against the economy, land- and housing-related revenue rose from 2.75 per cent of GDP in 2016 to an estimated 4.48 per cent in 2025. Measured against the budget revenue, it jumped from a range of roughly 11–14 per cent of total state revenue over 2016–24 to an estimated 21.7 per cent in 2025.

The pattern matters as much as the level. Land revenue did not simply rise year after year: it held broadly steady from 2016 to 2023, then surged over the past two years after Vietnam under General Secretary To Lam adopted an ambitious growth target of averaging 10 per cent annual GDP growth over 2026–2030. That jump shows how central land finance has become to the growth strategy (Figure 1).

Figure 1. Vietnam’s land-related state revenue, 2016–2025

The national figure understates an even more intense dependence concentrated at the provincial level. Because revenue from land-use rights is earmarked for capital spending, land has become the workhorse of provincial investment budgets as infrastructure needs grow. AMRO’s 2025 Country Fiscal Review, prepared with Vietnam’s Ministry of Finance, finds that land-related revenue accounts for around 80 per cent of provinces’ own-source revenue, and that local governments’ share of total state revenue rose from about 30 per cent in 2013 to nearly 50 per cent in 2022. Provinces typically finance infrastructure by converting agricultural or peri-urban land, auctioning or allocating the rights, and recycling the proceeds into roads, bridges, and new urban districts. Hanoi alone aims to triple its land-related revenue to around VND800 trillion (USD31 billion) over the next four years, up from just VND20 trillion (USD760 million) in 2022.

This is understandable. Vietnam’s infrastructure needs are vast, and ordinary taxation cannot meet them. Land is the one asset the state can mobilise quickly: it can be auctioned, leased, converted, swapped for infrastructure or used to draw in private capital. Properly governed, land-value capture is sound public finance, allowing the state to recover part of the value created by planning and infrastructure. The trouble begins when land stops being a development input and becomes the fiscal engine itself.

Hanoi alone aims to triple its land-related revenue to around VND800 trillion (USD31 billion) over the next four years, up from just VND20 trillion (USD760 million) in 2022.

That risk is now visible. A province funding its capital programme through land conversion is betting not just on development but on the property market’s continued liquidity. In an upswing, land finance bolsters the budget: revenue grows, infrastructure gets built, developers expand and growth targets appear attainable. In a downturn, the cycle reverses: auctions fail, land payments slow, developers struggle and capital spending is squeezed. Fiscal strength in the boom becomes fiscal fragility in the bust.

The 2024 Land Law sharpens the dilemma. By scrapping the old five-year price frame and moving towards annual, more market-based provincial land-price tables, helps narrow the gap between official and market valuations. This gap that has long enabled underpriced land allocation. But more realistic prices also raise the fiscal yield of each conversion. That makes the reform double-edged, it may improve valuation while encouraging local governments to monetise more land. At its Third Plenum in July 2026, the Party’s Central Committee approved directions for amending the Land Law and instructed the government to carry the amendments forward. The amended Land Law would retain state-determined prices while consolidating the valuation system around provincial price tables and adjustment coefficients and reducing reliance on project-specific valuations and the inconsistencies they can produce. But the plenum still cast land as a strategic resource for the double-digit growth target. Thus, while the valuation mechanism may be reworked, the underlying reliance on land-led growth remains.

China offers the clearest warning of such reliance. After the tax-sharing reform of 1994 left local governments with heavy spending duties but limited revenue, they turned to land-use-right sales and land-backed borrowing. By 2018 land-leasing revenue had reached almost 40 per cent of local-government revenue, and many cities financed more than half their infrastructure investment from land sales. The model built cities at an extraordinary speed. But when prices turned after 2021, land-sale revenue fell by roughly half from its peak, leaving local finances and their off-budget vehicles deeply stressed. The fiscalisation of land is self-reinforcing on the way up, and on the way down.

THE LAND-CREDIT LOOP

Land’s second mechanism works through bank balance sheets. As land prices rise, so do collateral values: borrowers gain access to more credit; and that credit flows back into land. The loop looks benign in an upswing, because both sides of the balance sheet strengthen at once: borrowers look wealthier, banks better secured, developers more bankable. It turns dangerous in reverse, because the collateral judged safest at the peak is often the first to sour when prices fall.

Vietnam is unusually exposed to this loop because its financial system is already stretched for its income level. The credit-to-GDP ratio climbed above 145 per cent in 2025, among the highest in ASEAN and far above similarly rated peers. That need not mean an imminent banking crisis, but it leaves less room to use credit as a shock absorber, especially when so much of that credit ultimately rests, directly or indirectly, on land.

Properly governed, land-value capture is sound public finance, allowing the state to recover part of the value created by planning and infrastructure. The trouble begins when land stops being a development input and becomes the fiscal engine itself.

The warning lies in a mismatch. Real estate is a modest part of production: narrowly defined real-estate business activity is only around 3-4 per cent of GDP, and even with construction added, the direct contribution is roughly a tenth of output. Yet property looms far larger in the banking system. By end-2025, State Bank of Vietnam (SBV) data put property-related credit at about VND4.74 quadrillion (USD182.3 billion), up 36.24 per cent year on year and equal to 25.53 per cent of total outstanding credit. Since 2018, property-related credit has risen from VND1.30 quadrillion (USD50.1 billion), or 18.1 per cent of system credit, to more than a quarter of the banking book (Figure 2).

Figure 2. Property-related credit growth in Vietnam, 2018–2025.

The mix is also shifting toward the riskier end. Lending for real-estate business alone reached VND2.16 quadrillion (USD83.1 billion) in 2025 — 45.6 per cent of all property-related credit, and roughly 11.6 per cent of system credit — after growing by nearly half in a single year. Earlier SBV data, for 2015-23, put property credit mostly in the 18–21 per cent range, with consumer and self-use loans as the larger share. Mortgage lending can be risky, but it is at least tied to household demand. Developer lending is a leveraged bet on approvals, absorption, bond refinancing and future land prices. When it outpaces the rest of the property book, banks are no longer merely financing homes; they are underwriting the land cycle itself.

Asset quality deepens the concern. Real-estate non-performing loans (NPLs) rose from 2.8 per cent at the end of 2023 to 3.7 per cent in July 2024. The ratio later eased to 2.9 per cent in 2025, but the system’s loss-absorbing buffer has thinned. Loan-loss reserves fell from 123 per cent of NPLs in 2022 to 82.9 per cent in 2025; against overdue debts, coverage was just 45.7 per cent. The thinner the cushion, the less protection if collateral values drop or restructured property loans sour again.

Japan shows the peril of treating land-backed lending as safe. In the late 1980s, weakly capitalised banks shifted lending from manufacturing to real estate as land prices soared. After the bubble burst, the banks most exposed to property fared worst on bad loans.

THE DIVERSION OF CAPITAL

For Vietnam, the binding question is who gets funded, more than how much debt the economy carries. A land-credit loop pulls capital, collateral and entrepreneurial effort toward property and away from tradables, technology, logistics and smaller firms. This is especially costly in an economy that already struggles to turn investment into output. Between 2011 and 2019, Vietnam needed more than six units of investment to generate one unit of additional output, well behind China, Malaysia and South Korea at comparable stages of development. Pouring land-backed credit into property worsens the misallocation. It also helps explain a familiar paradox: even as credit growth remains high, many manufacturers still struggle to secure bank loans.

The distortion is visible in what gets built: new urban housing supply is heavily skewed toward high-end segments, catering to investors and asset-holders rather than first-time buyers, a pattern that has prompted General Secretary To Lam to insist, in an echo of Chinese President Xi Jinping, that “houses are for living in, not for business or hoarding wealth.”

Land has also shaped Vietnam’s corporate hierarchy. Rising land values have helped spawn a generation of private conglomerates whose balance sheets are built around property. Of Vietnam’s 50 largest private firms, 31 hold substantial property or resource portfolios, and even among those whose core business is manufacturing, 73 per cent keep a foothold in property (Figure 3). This inverts the South Korean sequence, in which many chaebols began in manufacturing before diversifying into finance, services and real estate. Vietnam’s large private groups have more often moved from land, trading and finance into manufacturing.

That path is not doomed. Property profits can bankroll industrial upgrading. But the shift is hard when land stays more profitable, more easily collateralised and more politically mediated than manufacturing. More importantly, it sits awkwardly with Vietnam’s new ambition to build a technology- and innovation-driven economy. The pull can also run the other way: firms rooted outside real estate, such as the carmaker Thaco and Sovico, owner of VietJet, have expanded into property rather than deepening their core sectors.

Figure 3. Property and Resource Portfolios Among Vietnam’s Top 50 Private Firms

The dominance of property-linked conglomerates also breeds rent-seeking. Vietnam has already seen the extreme version. The Van Thinh Phat–Saigon Commercial Bank affair — USD12.5 billion in embezzlement, with estimated total damages of USD27 billion — showed how related-party lending, nominee ownership and opaque collateral chains can turn a property empire into a threat to the banking system. Such links are not exceptional; most banks among Vietnam’s 50 largest private firms sit inside wider conglomerate networks. When land is the main collateral, those networks become systemically important.

So long as land remains the easiest route to scale, profit and credit, it will keep diverting capital from the productive economy Vietnam says it wants to build.

DISTRIBUTIONAL CAPTURE

The next mechanism is political. Land-led growth decides who receives the value created by public power. A road, bridge, metro line or zoning change can turn cheap land into prime urban property. When that uplift is handed out through opaque negotiation rather than open competition, land finance becomes rent distribution.

The BT model concentrates this risk. It bundles two prices that are easy to manipulate: the cost of the infrastructure, which can be inflated, and the land that pays for it, which can be undervalued. A connected investor can thus be paid twice, once through an overpriced project, and again through underpriced land.

Vietnam knows the pattern first-hand, which is why the 2020 PPP Law excluded BT to begin with. Most BT investors had been appointed rather than chosen competitively, compensated with land allocated outside auction, and valued according to official price tables that were often well below market rates. The 2025 revival does introduce some guardrails to prevent abuse, but weaknesses persist in valuation, transparency and public participation.

Thu Thiem New Urban Area remains the clearest cautionary tale. Government Inspectorate findings in 2018–19 ordered Ho Chi Minh City to return more than VND26.31 trillion (USD1.01 billion) to the budget. The developer of four Thu Thiem roads, running 11.9 kilometres, was paid about VND1 trillion (USD38.5 million) per kilometre — roughly five times the per-kilometre cost of the North-South Expressway — and received prime land at fees inspectors found far below true value. Losses from the underpricing were put at around VND3.9 trillion (USD150 million). The Phuc Son and Thuan An cases, which involved bribery, rigged tenders and unlawful valuation tweaks, show the same logic on a smaller scale: public authority creates land value, and private actors compete to capture it.

The gains that connected firms capture in BT projects come at residents’ expense. In Thu Thiem, the lack of transparency and fairness bred decades of disputes and protest. Some residents said the state had compensated them at VND18 million (USD690) per square metre, while investors later sold land at VND350 million (USD13,500) per square metre. Such gaps may have been legal under the valuation system, but they are politically corrosive: they turn “public interest” projects into visible transfers from ordinary citizens to developers.

The risk of repeating Thu Thiem is high, this time with a much larger affected population. The proposed USD30 billion Red River Landscape Boulevard Project alone would displace more than 250,000 residents, and has already stirred discontent and rare silent protests. The nationwide push for mega-projects will touch millions and almost certainly generate more grievances. More than 80 per cent of administrative complaints in 2024 and 2025 reportedly concerned land.

Two dangers follow. The first is to legitimacy. As land rents pass from ordinary citizens to a narrow circle of connected firms, the state undercuts its own claim to uphold socialist principles of equity and fairness. The odds that silent protests will erupt into the “social volcanoes” seen in parts of China remain low, but they are rising. Land grievances accumulate; they persist across generations, and can turn local disputes into wider questions of justice.

The second danger is institutional. The entanglement of business elites and officials in capturing land rents weakens the rule of law, corrodes governance and breeds corruption. Once the rents grow large enough, firms invest less in productivity than in access. Officials gain discretion over valuation, zoning, conversion, compensation and approval, while developers gain reason to cultivate political protection. Land capture becomes, in time, political capture. A recent sign is telling: 18 “mega-infrastructure projects” by Vingroup, Sun Group and Masterise were exempted from credit-room limits, letting them borrow more from commercial banks.  

CONCLUSION

Vietnam need not choose between land finance and development. It needs to change the terms on which land is mobilised, and to stop land from becoming the default answer to every fiscal, financial and political problem.

That takes discipline before the cycle turns, and applied to each mechanism in turn. Against fiscalisation, it needs to treat land revenue as a temporary windfall, which should be buffered against the property cycle and kept out of the recurrent budget. Against the credit loop, it must hold property lending to genuine housing demand and viable projects, so that it complements rather than crowds out productivity-led growth. Against diversion, it needs to focus land-financed investment on a lean portfolio of high-return projects. Against distributional capture, it must make BT the exception, allowed only when valuation, investor selection and construction costs can survive public scrutiny.

The test is institutional rather than technical. Vietnam already has the legal tools to price land more realistically, improve disclosure and tighten developer finance. The harder task is enforcement: making auctions the norm, appointments rare, valuations contestable and audits consequential. Otherwise, reform will merely raise the price of land deals without changing who pockets the uplift.

A safer model would also be more selective. Land should be used to finance projects that raise productivity, such as transport links, urban services, industrial infrastructure and affordable housing, not scattered prestige schemes or speculative urban sprawl. The scarce asset is public trust in how land becomes value, as much as land itself. The political bargain should be explicit: if the state creates land value, the public must see the return. That means fairer compensation for those displaced, more affordable housing for new urban households, and firm limits on how much public land can be transferred to a handful of developers. Vietnam can still build growth on land. But to avoid the trap, it must make land finance more transparent, more competitive and less central to the economy’s balance sheet.

---

This is an adapted version of ISEAS Perspective 2026/62 published on 27 August 2026. The paper and its references can be accessed at this link.