Vietnam’s banks emerged from the 2007–2011 credit boom with bad loans that a central-bank inspection put at roughly 17% of the system, more than three times the figure the banks themselves reported. Rather than let lenders fail, the state built a machine for hiding, warehousing and slowly digesting the damage: the Vietnam Asset Management Company (VAMC), which swapped bad loans for special bonds; the zero-dong nationalisations of three broken banks in 2015; a 2017 law that let banks seize collateral; and, a decade later, the compulsory transfer of those banks to stronger rivals. The machine worked, in the sense that no depositor lost money. The bill was paid in time, in credit that did not flow, and in a habit of rescue that the SCB collapse tested at a scale nobody had planned for.
Vietnam has never let a bank fail, and the way it has avoided doing so explains more about its financial system than any balance sheet. Between 2012 and 2025 the country processed a banking crisis without a formal bail-out, a deposit-insurance payout or a single bank liquidation. It did so with a set of instruments that look strange to outsiders — an asset manager that buys loans with paper it prints, a central bank that buys banks for nothing, a resolution that lets creditors take the keys to collateral without a judge — and that, taken together, describe how the state thinks about risk. This article explains how the debt piled up, what each instrument did, what happened when the largest fraud in the country’s history broke the model’s assumptions, and what founders, lenders and investors should take from it. It is part of the Vietnam Company Stories hub.
What is VAMC?
A state asset-management company set up in 2013 that buys bad loans from banks and pays with special bonds. The bank provisions against the bond over five to ten years, which spreads the loss and keeps the bad loan off its balance sheet in the meantime.
What is a zero-dong bank?
One of three lenders — Construction Bank, OceanBank and GPBank — whose shareholders were wiped out in 2015 when the State Bank compulsorily bought all their shares for zero dong after fraud left them with negative equity.
Did the approach work?
Depositors were protected and the system stabilised, but resolution took over a decade, the weak banks needed subsidised funding throughout, and the SCB case of 2022 showed the cost of a rescue-first culture when the hole is measured in tens of billions of dollars.
How did Vietnam’s banks end up with a mountain of bad debt by 2012?
Through four years of credit growth that averaged well above 30% a year, a wave of new bank licences handed to industrial groups that lent to themselves, and a property and state-enterprise boom that collapsed when inflation forced the central bank to slam on the brakes in 2011.
Credit expanded by about 54% in 2007 alone, and by more than 30% in each of 2009 and 2010 as the government pushed stimulus lending after the global crisis. Credit to the private sector rose from roughly 60% of GDP in the mid-2000s to around 120% by 2010, an extraordinary pace for a lower-middle-income country. Much of the money went into land, into shipbuilder Vinashin and other state groups, and into a circular pattern in which banks lent to companies owned by their own shareholders. Rural credit cooperatives that had been converted into urban commercial banks in the mid-2000s were particularly exposed: they had grown their balance sheets tenfold with capital that, in several cases, had itself been borrowed.
When inflation reached 23% in 2008 and about 18% in 2011, the State Bank tightened hard, introduced the annual credit quotas described in our piece on Vietnam’s credit-growth quota system, and the property market seized up. Vinashin defaulted on a USD 600 million syndicated loan in December 2010 with total debts on the order of USD 4.5 billion; its shipping cousin Vinalines followed. Banks reported non-performing loans of about 4.9% in 2012. An inspection by the State Bank’s supervisory agency, made public in September 2012, put the true figure at 17.2%, and even that excluded loans that had been quietly rescheduled.
What is VAMC and how does the special-bond mechanism work?
VAMC is a state-owned company created by Decree 53 in 2013 with VND 500 billion of capital, an amount so small relative to the debts it was meant to absorb that it made clear from the start the company was a warehouse, not a buyer. It purchases bad loans at book value net of provisions and pays with “special bonds” that carry no interest and are not tradable.
The trick is in what the bank does next. It records the special bond as an asset in place of the bad loan, which drops out of its reported NPL ratio. Each year it must provision 20% of the bond’s face value (later relaxed to 10% for banks under restructuring plans, stretching the term to ten years), so the loss is recognised in instalments rather than at once. The bank can pledge the bond to the State Bank for refinancing at a discount, which provides liquidity, and it remains responsible for chasing the borrower; if the loan is recovered, the bond is cancelled and the bank keeps the proceeds. VAMC itself does almost nothing with most of the loans it holds.
By 2017 VAMC had bought loans with a face value on the order of VND 300 trillion from more than forty institutions, the equivalent of roughly USD 13 billion at the time, and had resolved a small fraction of them. Its usefulness was in time: banks that had been technically insolvent on an honest mark got years to earn their way back to health. The strongest used the time well. Vietcombank fully provisioned and repurchased its VAMC bonds in 2016, the first to do so; Techcombank, ACB and MB followed by 2018. The weakest simply rolled their bonds and hoped.
What happened when the State Bank bought three banks for zero dong?
In 2015 the State Bank of Vietnam compulsorily acquired all the shares of Vietnam Construction Bank (February), OceanBank (April) and Global Petro Bank, GPBank (July) for a price of zero dong each, after inspections found their capital had been wiped out and shareholders had failed to recapitalise them. It was the first time the state had expropriated bank shareholders, and it did so without a formal bankruptcy.
Each case was a fraud as much as a failure. At Construction Bank, chairman PhαΊ‘m CΓ΄ng Danh had used the bank to fund his own property group through fake loans and deposits; the courts later found losses in the region of VND 9 trillion and sentenced him to 30 years. At OceanBank, chairman HΓ VΔn ThαΊ―m and the bank’s former chief executive Nguyα» n XuΓ’n SΖ‘n, who had also run state oil group Petrovietnam, were convicted over illegal interest payments and the loss of an VND 800 billion Petrovietnam investment; SΖ‘n received a death sentence, later reduced, and ThαΊ―m life imprisonment. GPBank had lent itself into insolvency on real estate. All three had negative equity running into the trillions of dong, and in each the shareholders, including state enterprises that had been forced to invest in them, lost everything.
The central bank then assigned the strong to mind the weak: Vietcombank sent managers to Construction Bank, VietinBank to OceanBank and GPBank. DongA Bank was placed under special control in August 2015 after its own chief executive, Trần PhưƑng Bình, was found to have concealed losses. Depositors were told their money was safe, and it was. What the state had not decided was what to do with four banks that had no capital, no viable franchise and a mounting need for cheap central-bank funding just to stay open. That question took nine years to answer.
Why did Resolution 42 matter and what did it change?
Resolution 42/2017/QH14, passed by the National Assembly in June 2017 as a five-year pilot, gave banks and VAMC the right to seize and sell collateral on bad loans without going through the courts, to sell loans below book value, and to prioritise the secured creditor over other claims. It turned a legal system that had made foreclosure almost impossible into one where a determined lender could recover something.
Before 2017 a bank trying to repossess a mortgaged house or factory could spend years in litigation, and a borrower could stall indefinitely by refusing to hand over the property or by opening a dispute. Resolution 42 let a lender take possession if the loan contract allowed it and the borrower had defaulted, with police and local authorities obliged to assist. The State Bank later reported that the resolution had helped resolve on the order of VND 440 trillion of bad debt between 2017 and 2023, much of it through borrowers paying up rather than face seizure, which is precisely how such rules are supposed to work.
The pilot was extended to the end of 2023 and then, in the 2024 Law on Credit Institutions, partially made permanent — but the seizure right itself was dropped in the final drafting after objections about due process. Banks lobbied hard, bad-debt ratios rose in 2024, and in mid-2025 the National Assembly amended the law again to restore collateral-seizure rights on a permanent basis. The episode shows the tension at the heart of Vietnamese bank resolution: the state wants banks to recover loans but is reluctant to hand private lenders a power that resembles enforcement.
What happened with SCB, and how big was the hole?
In October 2022 depositors ran on Saigon Commercial Bank after the arrest of property tycoon TrΖ°Ζ‘ng Mα»Ή Lan, and the State Bank placed it under special control. The 2024 trial found that Lan had controlled more than 90% of SCB through nominees and drawn out loans worth roughly VND 677 trillion — on the order of USD 27 billion — over a decade, making it the largest financial fraud in Vietnamese history and one of the largest anywhere.
SCB was itself a product of the earlier rescue machine, formed in a 2012 three-way merger of weak lenders that the State Bank had encouraged. Its balance sheet had grown to become the fifth largest in the country, largely on deposits attracted by high rates and on bonds sold through its branches, while its assets were loans to shell companies in the VαΊ‘n Thα»nh PhΓ‘t orbit secured on overvalued property. The details of how it was run and who was convicted are in our companion article on VαΊ‘n Thα»nh PhΓ‘t and SCB.
What matters here is the response. The State Bank lent SCB special loans to meet withdrawals, on a scale reported in the hundreds of trillions of dong, and has kept it open under administration since. Lan was sentenced to death for embezzlement in April 2024 and ordered to repay; asset recovery is expected to take years. A restructuring plan involving a private investor has been discussed publicly but not, as of mid-2026, completed. The zero-dong banks of 2015 had negative equity in the low tens of trillions; SCB’s hole is an order of magnitude larger, and the tools built for the smaller problem have been stretched to fit it.
How did the forced transfers of 2024–2025 finally close the zero-dong chapter?
The 2024 Law on Credit Institutions created a compulsory-transfer mechanism under which the State Bank can hand a failed bank to a healthy one as a wholly owned subsidiary, with the acquirer receiving zero-interest refinancing, extra credit-growth room, exemption from consolidating the subsidiary’s losses into its capital ratio and, eventually, freedom to merge or sell it. Within three months all four legacy cases were placed.
On 17 October 2024 Construction Bank went to Vietcombank and became VCBNeo; OceanBank went to MB and became MBV. In January 2025 GPBank was transferred to VPBank and DongA Bank to HDBank, which renamed it Vikki Bank. The acquirers were chosen because they had capital, technology and, in most cases, a decade of experience seconding managers to the banks they were now taking on. Each has framed the subsidiary as a digital-banking venture, which is a reasonable way to describe a licence with no legacy franchise worth preserving.
The economics are quiet subsidies. Zero-interest funding from the central bank for a bank that would otherwise borrow at market rates is worth the interest differential on the whole balance sheet; extra credit room for a large lender is worth billions of dong in annual income; and the exemption from consolidation means the acquirer’s reported capital adequacy does not fall. The state has, in effect, paid strong banks to take the problem off its hands using instruments that do not appear in the budget. It is a cheaper and more discreet solution than recapitalising the failed banks directly, and it has taken almost a decade to arrive at.
What are the risks in Vietnam’s bad-debt system now?
Bad debt that is real but unrecognised, a property market that has only partly recovered, an SCB resolution that has not been designed, and a system in which every participant — depositor, borrower, banker — has learned that the state will absorb the loss.
Reported system NPL ratios in 2024 and 2025 were in the range of 4–5% once loans at VAMC and restructured debt were included, up from under 3% before the pandemic. Circular 02 of 2023 let banks keep restructured loans classified as performing through the end of 2024, a pandemic-style forbearance that flattered the numbers; its expiry, the aftermath of Typhoon Yagi in late 2024, and the slow return of developers such as Novaland to solvency all fed into rising provisions. Coverage ratios at most banks fell from their 2022 peaks. None of this is a crisis; all of it is the same pattern that preceded the last one.
The unresolved SCB case is the largest single risk. Its deposit book was funded by the central bank, its assets are the subject of criminal proceedings and its franchise has been destroyed. A private investor taking it on would need concessions larger than any granted so far, and a state recapitalisation would require an appropriation the government has avoided for a decade. Until it is resolved, SCB is a standing reminder that the machine has an upper limit.
What does Vietnam’s bad-debt history mean for founders, investors and operators?
That credit in Vietnam is cheap and available in good years because the state stands behind the banks, and that this same backstop makes the cycle longer and the eventual clean-up slower. Borrowers should plan for the tightening, not the boom; lenders and investors should read the rescue history as a description of the state’s priorities; and operators should understand collateral law before they sign a loan.
For founders and CFOs the practical lesson is about collateral. Since Resolution 42 and its 2025 restoration, a bank can take a mortgaged asset without a court order if the loan is in default. A company pledging its factory, its land-use rights or its shares should understand that the days when a borrower could stall enforcement for years are over, and negotiate cure periods and cross-default clauses accordingly. The upside is that lenders who can enforce will lend more readily; the downside is that they will enforce.
For investors the history sets expectations. Bank equity in Vietnam has never been zeroed by a regulator except in cases of outright fraud, and even then depositors were protected. That asymmetry — shareholders absorb fraud, the state absorbs everything else — means bank valuations embed a state guarantee that is real but unpriced and unlegislated. The banks that benefited most from the 2024–2025 transfers were those the state trusted to execute; that trust is an asset, and the story of Techcombank shows how private banks earn it. For lenders and distressed-debt investors, meanwhile, the VAMC warehouse and the coming SCB resolution represent the largest pool of unresolved secured assets in Southeast Asia, a market that has barely opened to outside capital but that the state increasingly needs.
For operators the broadest lesson is about how the state behaves under stress. Faced with a choice between an orderly loss and a slow, hidden one, Vietnam has consistently chosen slow and hidden. That preference shapes how it handled the airline industry, as our piece on Vietnam Airlines’ rescue shows, and it will shape the next downturn too. Companies that depend on the state acting quickly will be disappointed; companies that can outlast a long, quiet workout will find the state, eventually, on their side.
Frequently Asked Questions
What does VAMC actually do with the loans it buys?
Mostly it holds them. The selling bank remains responsible for collecting the debt and provisions against the special bond it received over five to ten years. VAMC has also bought a smaller volume of loans at market value with its own capital and auctioned collateral, but its main function has been to give banks time to absorb losses gradually.
Why were the failed banks bought for zero dong instead of being closed?
Because closing a bank would have required paying out depositors, testing a deposit-insurance fund that was far too small, and risking runs on other weak lenders. Buying the shares for nothing wiped out the owners who had caused the losses while keeping the banks open and their depositors whole.
Has anyone lost deposits in a Vietnamese bank failure?
Not in any of the cases since 2012. Depositors at Construction Bank, OceanBank, GPBank, DongA Bank and SCB have been able to withdraw their money, in SCB’s case with liquidity supplied by the State Bank. Holders of bonds sold through bank branches, notably in the VαΊ‘n Thα»nh PhΓ‘t case, have fared far worse.
Is the bad-debt problem solved?
The 2015 cases have been placed with strong banks, and the system-wide NPL ratio is far below the 2012 peak. But reported bad debt rose again in 2024–2025, forbearance rules have masked part of it, and SCB remains under special control with no completed restructuring, so the answer is partly.
Discover more from Kurums | Business Intelligence
Subscribe to get the latest posts sent to your email.
