Apr 22, 2026

Vietnam Real Estate: How Big Has the Bubble Gotten?

Macro ยท Vietnam Real Estate & The Financial System
Core Thesis
Not just real estate - the whole economy is sitting on a bomb

In 2007, Ray Dalio saw the crash coming before the rest of the world. In 2026, every one of those warning signs - mismatched cash flows, leverage flashing red, disguised junk assets, regulators trapped between conflicting goals - is repeating in Vietnam. Real estate is only the symptom. The real disease sits in the financial system, in credit, and in the structure of the economy itself.

Disclaimer: This piece is not neutral - it carries a deliberate bearish bias, meant as a counterweight to a Vietnamese real estate information space saturated with bullish signals. Read critically, cross-check other sources, and form your own judgment.

On the data: Macro data in developing economies is often lagged and difficult to verify independently. The figures here are compiled from the World Bank, IMF, OECD, CEIC, and analyst estimates. This is not investment advice.

Numbers That Don't Lie

34
Years of salary to buy a home (Saigon)
61%
Household debt / GDP
~150%
Private credit / GDP
~20%
Overnight interbank rate (peak)
โˆ’8%
Net loss/year holding real estate
The Hook
Even if you never touch real estate, your bank deposit is still exposed.

70% of all bank collateral system-wide is real estate. If a bank fails, inflation rises and purchasing power erodes. The SBV is carrying two mutually canceling goals at once - stabilizing the exchange rate and pushing credit growth.

A Saigon resident needs 34 years of income to buy a home. China has already cut prices 20โ€“30%, tier 3โ€“4 sales are down to 53% of their 2021 peak, and 1.4 billion people still aren't buying (the 33-month chronicle of falling prices โ†’). Vietnam, with 100 million people and income 3x lower - what cushion does it have?

Vietnamese households were once estimated by HSBC to owe 61% of GDP as of 2020 - far higher than the Philippines. Bank lending has surpassed ~150% of GDP (credit +19.1% in 2025) - double the average for developing countries. When banks sell collateral to recover debt, many real estate assets get discounted 30โ€“50% and still go unsold through multiple auction rounds.

Credit โ†’ Real Estate
25.5%
Real estate business lending alone is ~$78.5B; including home loans, real-estate-related credit is ~25.5% of outstanding loans; ~70% of collateral is real estate. VNBA Q4/2025
Loan / Deposit
~112%
Credit +19.1% vs deposits +14.1% (2025); LDR at a multi-year high - a textbook maturity mismatch. The Investor
Corporate Bonds Tied to Real Estate
62%
Real estate accounts for 62% of cumulative bond defaults; real estate bond yields spiked to 13.5%. VIS Rating 2024
Stalled Projects
~3,000
VND 2.4 quadrillion (~$100B) buried alive. VietnamNet
Reading it differently

The economy is "all in" on real estate at every level - households, banks, corporate bonds, all the way to local government budgets. Everyone is betting on a single asset class whose true value nobody knows, whose liquidity has effectively frozen, and whose cash flow comes from domestic credit - not real economic strength.

How is that โˆ’8%/year figure calculated? - Breaking down the cash flow of holding real estate

Take a common real-world scenario in HCMC/Hanoi: a mid-tier apartment worth VND 5 billion, 30% equity (1.5 billion), 70% bank loan (3.5 billion). We roll up all associated costs and income to arrive at the "total real return" on the asset's value.

Cash flow on a VND 5 billion apartment: rent doesn't cover the debt
Unit: million VND/year. Rent +170, costs โˆ’587, shortfall โˆ’417 before accounting for asset price swings.
0 200 400 600 MONEY IN Rent +170 Rental yield ~3.4% COSTS/YEAR Loan interest โˆ’420 Opportunity cost โˆ’90 โˆ’587 Fees + vacancy โˆ’60 Rental tax โˆ’17 Loan principal + equity locked up not enough to cover SHORTFALL โˆ’417 million/year 587 costs โˆ’ 170 rent โ‰ˆ โˆ’8.3% of apartment price To break even, the apartment price must rise at least ~8%/year just to offset the negative cash flow
Reference sources: VnExpress / Savills on yield; Vietnam Briefing on tax; Trading Economics on deposit rates.

Opportunity cost is the profit forgone by not putting the money into another channel. That 1.5B in equity, if put into a 12-month savings deposit currently paying ~5โ€“6%/year - burying it in real estate instead means turning down 90 million in safe interest. Economics calls this a real cost, not an "imaginary" one as many mistakenly assume.

The real number could be worse

This math doesn't yet include: (1) furniture depreciation of 1โ€“2%/year, (2) 6โ€“12 months of vacancy risk between tenants, (3) one month's rent in brokerage fees per lease, and (4) asset price volatility. If prices go sideways or fall - as with China's 20โ€“30% drop from its 2021 peak or Japan's 17-year wait to recover its peak - the real loss can easily climb to โˆ’15 to โˆ’20%/year.

So why do people still buy? They're betting on a single variable: that the asset price will rise faster than the โˆ’8% net annual loss. When P/I already sits at 31โ€“40x (HCMC 31x, Hanoi 39.6x per Numbeo) and income has been flat for three straight years - the price has to rise โ‰ฅ8%/year just to break even on cash flow, before even talking about profit. This is exactly Hyman Minsky's definition of Ponzi finance: the asset can't service its own debt, and survives only on future price appreciation.

2007 - The Man Who Saw the Crash Coming

While the rest of the world partied on, Ray Dalio and the team at Bridgewater Associates sounded the alarm - not on a hunch, but with quantitative models.

1
Mismatched cash flows

Starting early 2007, Bridgewater sent memos to clients and the US Treasury: debt-service costs were rising faster than income. Subprime defaults were beginning, and the economy's debt structure was already overstretched.

2
Leverage gauge flashing red

Bridgewater's Leverage Gauge in 2007 was "flashing red" - the financial system was carrying record-high leverage. It would only take a small jolt to housing prices to knock over the entire row of dominoes.

3
Meeting regulators directly

Dalio met with the Treasury and the Fed: banks were holding "junk assets" disguised inside complex financial instruments, with insufficient capital buffers. Most policymakers ignored him - dismissing him as overly pessimistic.

4
Predicting a liquidity crunch

When leverage runs too high, a liquidity crisis becomes inevitable - everyone needs cash at the same time, and no one has enough. This happened exactly as predicted in 2008, after the collapse of Lehman Brothers.

Not just the US - China had its own warning voices too

"A hard landing is virtually unavoidable. I'm not predicting it - I'm observing it happen. China's credit debt is a mirror of the US before 2008."
Chinese state media attacked him immediately.
"When everyone simultaneously realizes that the assets they bought - real estate, financial products - aren't worth that much, everyone will rush for the exit. This is China's Minsky Moment."
The speech was scrubbed from the Chinese internet within 24 hours.
"China needs to guard against a Minsky Moment - the point where asset values collapse suddenly after unsustainable debt accumulation. Total debt/GDP has reached 247%."
A warning from the head of the central bank himself. Evergrande defaulted four years later.

What is a "Minsky Moment"?

According to economist Hyman Minsky, an economy naturally moves through three phases of borrowing. When the credit line decouples from real GDP, a bubble forms - and once no one is willing to borrow any more, it bursts.

The Phases of Borrowing
Credit vs. real GDP
THE PHASES OF BORROWING REAL GDP TIME HEALTHY BORROWING (Hedge) SPECULATIVE BORROWING (Speculative) PONZI BORROWING (Borrowing to repay old debt) MINSKY MOMENT CREDIT GDP
Credit (steepens, then reverses) Real GDP (steady growth) Minsky Moment (turning point)
When credit decouples from real GDP, the system enters the Ponzi phase - surviving only if asset prices keep rising.
01
Hedge
Healthy borrowing

Income covers both principal and interest. Banks lend selectively. Example: buying a home with income 3x the monthly payment.

02
Speculative
Speculative borrowing

Income covers only interest, not principal - requiring constant debt rollover. Borrowers count on rising asset prices to keep the cycle going.

03
Ponzi
Ponzi borrowing

Income can't cover interest or principal - new debt is taken on just to pay off old debt. Prices stop rising โ†’ sell-off โ†’ Minsky Moment.

Where does Vietnam stand?

Rental yield of 3.4% < loan rate of 12โ€“15%. Income doesn't cover interest, let alone principal. The real estate market survives entirely on the expectation that "prices will keep rising" - that's the textbook definition of Ponzi Financing.

Concretely, in the HCMC market: from 2011 to 2026, mid-to-high-end apartment prices in central districts have risen ~5.5x (from ~22 to ~120 million VND/mยฒ, Savills / Batdongsan). Over the same period, GDP per capita rose only ~2.5x (World Bank), average HCMC income rose ~2.1x (GSO Living Standards Survey), apartment rents rose ~2.3x nominal (real rent is roughly flat after inflation, Savills/GlobalPropertyGuide), while the M2 money supply - the fuel of the bubble - was pumped up ~8.7x (from ~2,773 to ~24,150 trillion VND, Trading Economics/IMF IFS / SBV). Even more concerning, foreign exchange reserves - the buffer against capital flight - grew only ~6.3x (from ~$13.5B to ~$85B, World Bank / IMF IFS), and have fallen 24% from their 2021 peak of $109B as the SBV was forced to sell USD to defend a depreciating VND. Meanwhile, commercial electricity output - the measure of real economic activity that doesn't lie - rose only ~3.3x (from 94.65 to ~315 TWh, EVN / MOIT), less than half the pace of M2. Assets rose 5.5x, money printed rose 8.7x, the safety buffer weakened, production rose only 3.3x, yet cash flow rose only 2.1-2.3x. This is no longer growth - it's decoupling.

Placed side by side on the same 2011 = 100 baseline: HCMC on the left, Jakarta on the right - two ASEAN countries of similar GDP per capita, similar reliance on FDI inflows, and a similar credit-easing period. Click the legend to hide/show both charts at once. (Sources: SBV ยท EVN ยท GSO ยท World Bank WDI ยท Savills/GlobalPropertyGuide ยท Bank Indonesia SEKI ยท PLN Statistik ยท Colliers Jakarta ยท BPS DKI Jakarta.)

Shared legend - click to hide/show both charts at once
HCMC (Vietnam)
Mid-high end ยท 2011 = 100
Jakarta (Indonesia)
Citywide strata-title ยท 2011 = 100
Endpoint 2026: HCMC decouples from Jakarta at precisely the most dangerous point
2011 = 100. The left bar is HCMC, the right bar is Jakarta; the red zone is the housing-price-minus-income gap.
HCMC Jakarta Quick-read gap Apartment price 545 224 +321 points M2 money supply 871 362 +509 points Avg. income 214 226 Jakarta slightly ahead GDP/capita 255 144 +111 points Commercial electricity 333 208 +125 points Rent 228 140 +88 points Real estate price โˆ’ income gap: HCMC +331 points; Jakarta โˆ’2 points one decoupled, one โ‰ˆ 0
The year-by-year data table remains in the expandable section below so readers can check sources and assumptions.
Quick read on both charts: On the left (HCMC), the lines fan apart from each other - purple M2 shoots highest, red real estate price trails it, amber income and blue GDP lag behind. On the right (Jakarta), the lines stay tightly bunched - red real estate price and amber income nearly overlap, purple M2 is only slightly ahead. Same seven indicators, same 15 years - one side shows asset-price inflation, the other shows growth tracking real cash flow. HCMC's real estate price rose 5.5x while income rose only 2.1x (a 2.6x gap); Jakarta's real estate price rose 2.2x, income 2.3x (a 0.95x gap). Note on Jakarta data: "Apartment price" is the citywide strata-title average (Colliers); the CBD alone was ~53 million IDR/mยฒ in 2025 but doesn't reflect the mass market. "Average income" is the Sakernas formal-worker net wage for DKI Jakarta (BPS Table 453); 2011โ€“2017 is estimated using the DKI/national ratio of ~1.65x (the BPS DKI provincial table is only continuous from 2018). The 2024โ€“2025 dip reflects actual Sakernas volatility (the Feb-vs-Aug release carries sampling error).
Year-by-year data - detailed table for HCMC & Jakarta (2011โ€“2026)
Year HCMC price (mn/mยฒ) M2 (trillion VND) FX Reserves ($B) City electricity (TWh) GDP/capita (USD) HCMC income (mn/mo) Yield gross Months' salary / mยฒ
2011~222,77313.5~94.71,951~3.65~6.7%~6.0
2015316,02028.6143.32,578~4.97~5.4%~6.2
2019~4510,57678.3209.43,4416.77~4.5%~6.6
2021~5513,701109.4 โ†‘peak225.33,7046.01~3.5%~8.3
2022~6314,22786.5 โ†“โˆ’21%242.74,1486.39~3.6%~9.9
20249117,91483.1276.4 (+9.2%)4,717~7.0~3.3%~13.0
2025111~21,00083.6287.9 (+4.9% โ†“)4,745~7.4~2.9%~15.0
2026 (proj.)~120~24,150~85~315~4,965~7.8~2.8%~15.4
Year Jakarta price (mn IDR/mยฒ) M2 (trillion IDR) FX Reserves ($B) PLN electricity (TWh) GDP/capita (USD) Jakarta income (mn IDR/mo) Yield gross Months' salary / mยฒ
2011~17.02,877110.1158.73,688~2.30~7.0%~7.4
201526.04,548105.9202.83,3683.45~7.0%~7.5
201932.06,137129.2245.54,1934.49~6.0%~7.1
202132.07,867144.9257.04,3514.75~5.6%~6.7
202233.08,528137.2273.84,7845.20~5.3%~6.3
202436.49,120155.7 โ†‘peak306.24,9585.24~4.6%~6.9
202537.09,783156.5~3185,0705.00~4.5%~7.4
2026 (proj.)~38.0~10,400~158~330~5,300~5.20~4.4%~7.3
4 Minsky indicators, all flashing red

(1) Affordability has collapsed: 2011 - 1mยฒ = ~6 months' salary; 2026 - ~15.4 months, 2.5x higher and far past the "healthy" 5โ€“7 month threshold for a developing market. A 70mยฒ apartment now demands ~90 years of average HCMC income if bought without debt. (2) Yield has compressed: 6.7% (2011) โ†’ 2.8% (2026, below even VND deposit rates) - the gap with the cost of capital keeps widening. The asset's own cash flow no longer covers its cost of capital - textbook Ponzi. (3) Money supply up 8.7x but output only 3.3x, cash flow only 2.1x: M2 multiplied 8.7x (from 2,773 to ~24,150 trillion VND), prices multiplied 5.5x, commercial electricity - the real measure of production - only 3.3x (94.7โ†’315 TWh, EVN), income only 2.1x. In 2025 alone, credit grew +19.1% while electricity grew only +4.9% (the electricity/GDP elasticity coefficient fell to 0.61 - the first time below 1, GSO). Money printed isn't flowing into factories, it's flowing into assets. The purple line (M2) sits above the red line (real estate price), while the yellow line (electricity) lags behind - the signature of post-2008 asset-price inflation. (4) The safety buffer has reversed: FX reserves fell from a $109B peak (2021) to $83B (2024โ€“2025) - down 24% in three years as the SBV was forced to sell USD to defend the VND against a tightening Fed. Meanwhile M2 kept growing ~30% over the same period. The M2/reserves ratio has clearly deteriorated - the blue line reversed far earlier than the other five lines. With M2 +15% in 2026 while reserves stay flat, the SBV's room to defend the exchange rate keeps narrowing. That's the last fragility point before the Minsky Moment.

Evidence 1 - Credit has "decoupled" from real GDP (Credit/GDP 146%, up +19.1%)

The credit/GDP ratio measures how dependent an economy is on borrowed money - exactly what Minsky called "leverage tension." Vietnam has hit ~146% of GDP by end-2025 (Fitch Ratings called this "an unusual level of leverage for a developing economy") - the highest among low-middle-income economies by the central bank's own assessment. Full-year 2025 credit grew +19.1% (full-year final figure, outstanding balance ~18,580 trillion VND / ~$731B) - more than 2x nominal GDP growth of ~9โ€“10% and the highest level in years. That's the definition of the red credit line decoupling from the blue GDP line on the Minsky chart above.

Credit/GDP: Vietnam sits in the red zone - and is still climbing fast
Horizontal axis is credit/GDP; the circle size is 2025 credit growth. Above 100% means the economy is already heavily debt-dependent.
developing-economy range high warning 80% 100% Indonesia 35% credit +7.9% Philippines 48% credit +10.3% India 59% credit +12.2% Thailand 115% credit โˆ’0.6% Vietnam 146% Vietnam has both the highest ratio and the fastest growth in the group
Source: Fitch / SBV, BIS/FRED, World Bank, RBI, BSP, Bank Indonesia.
Credit / GDP - How credit-saturated the economy is
% GDP
BIS warning zone (>100%) High (80โ€“100%) Developing (<80%) BIS threshold (80%)
BIS Working Paper 352 (Drehmann et al.) treats a credit/GDP ratio that exceeds its long-run trend (the Credit-to-GDP Gap) as an early-warning indicator of banking crises - accurate in 2/3 of cases over 40 years of history.
Evidence 2 - Credit exploded, but "electricity stood still" (Credit +19.1% vs Electricity +4.9%)

The key question: if credit grew +19.1%, where did the money go? If it flowed into factories, construction sites, production - electricity output should have risen correspondingly. Electricity is the measure of real economic activity that doesn't lie: no factory runs without power. The 2025 data reveals a gap that's hard to explain away.

2025: credit up, GDP up, but electricity broke rhythm
When money flows into production, electricity usually tracks GDP. In 2025 electricity grew only 4.9% while credit grew 19.1%.
0% 5% 10% 15% GDP 7.09 8.02 Credit 15.1 19.1 EVN electricity 9.24 4.9 0.61 electricity/GDP elasticity first time below 1.0 2024 2025 Source: VietnamNet, VietnamPlus, The Investor, EVN.

In 2000โ€“2010, every 1% of GDP growth came with 1.7โ€“2.0% electricity growth. 2011โ€“2019: 1.2โ€“1.4. 2021โ€“2024: 1.0โ€“1.1. In 2025 it suddenly dropped to 0.61 - the first time electricity growth has lagged GDP. With GDP +8.02% and Vietnam's own 2011โ€“2024 elasticity of 1.0โ€“1.4, electricity should have grown ~8โ€“11%. It actually grew only +4.9% - a shortfall of ~3โ€“6 percentage points versus its own history, while credit kept surging +19.1%. That extra credit didn't produce any kWh.

The fingerprint of an asset bubble

When credit grows ~3.9x faster than electricity output, most of that money isn't going into factories - it's flowing into things that inflate asset prices without creating production: mortgaged real estate, real-estate corporate bonds, stocks, debt rollovers, land bought purely to wait for appreciation. This is exactly Minsky's "Ponzi financing": the economy borrowing more to hold up asset prices, not to borrow for production.

"But the government calls this a positive sign?" - a rebuttal

VietnamNet itself - the online outlet of the Ministry of Information and Communications - ran the headline "Vietnam's power consumption decouples from GDP growth". State-aligned experts explain it as: (1) the economy shifting to "higher quality," less energy-intensive growth; (2) rooftop solar self-consumption of ~10 TWh that doesn't show up in EVN's statistics.

"Electricity growth coming in lower than GDP isn't abnormal. It reflects improved growth quality."

The problem with this explanation:

  • Rooftop solar self-consumption of ~10 TWh: even counting all of it as new this year (in reality only part of it is), it explains at most ~4 percentage points - and doesn't explain why elasticity fell from 1.0โ€“1.3 to 0.61 in just 12 months.
  • "Higher quality" doesn't arrive in a single year. South Korea and Taiwan took 10โ€“20 years to reduce their economies' electricity intensity. Vietnam had no technological breakthrough in 2024โ€“2025 large enough to cut elasticity from 1.30 to 0.61 in just 12 months.
  • Industry & construction still account for 53% of electricity demand and grew only +7.2% - while system-wide credit grew +19.1%. Where did the rest - ~2.7x production growth - go? The answer: real estate credit, land mortgages, real-estate corporate bonds - loans that don't turn any turbines.
  • VietnamNet itself admits that GDP was driven by "large infrastructure projects - expressways, airports - that consume relatively modest electricity." That's exactly the fixed-asset speculation channel, not factory output.

In other words: an asset bubble repackaged as "high-quality growth." A pretty label, but the substance is exactly what Minsky described half a century ago.

US 2007. China 2019. Vietnam 2026?

Cash flow doesn't add up? Yield 3.4% vs. loan rate 12โ€“15% โ†’ net loss of 8%/year.
โœ“ Check
Leverage flashing red? Private credit ~150% of GDP, household debt estimated by HSBC in 2020 at 61% of GDP.
โœ“ Check
Junk assets in disguise? Real estate = 62% of defaulted corporate bond volume. Banks discount 30โ€“50% across multiple auction rounds and still can't sell.
โœ“ Check
Regulators looking away? The SBV is stuck between defending the exchange rate and pushing credit growth.
โœ“ Check

The Global "Overpriced" Ranking

The Price-to-Income Ratio (P/I Ratio) - how many years of working without eating to buy a home. The comparison shows Vietnam is "overpriced" on par with economies many times larger.

P/I Ratio - 16 Global Cities
Years of income
China's peak (already burst) Bubble / overstretched Stretched but cushioned Rich country, can bear it Affordable
Source: Numbeo Property Price-to-Income Ratio, current as of 05/2026; the two China bars are historical peak benchmarks for cycle comparison.

City-by-city breakdown

Why does the same high P/I ratio lead to a bust in some places and not others? Click each city to see the structure underneath.

๐Ÿ‡ฐ๐Ÿ‡ท Seoul - Overpriced but hasn't blown up? 24.9ร—

Don't use a rich country's numbers to console your own wallet.

  • The underlying economy is different from froth: South Korea's GDP/capita is >$33,000, Vietnam's is ~$5,000. Korea has massive manufacturing, technology, and export capacity acting as a capital cushion.
  • The chaebols carry the team: Samsung, Hyundai, SK pull huge amounts of foreign currency back into the country.
  • Conclusion: A rich country can bear high home prices on a century of accumulated "real strength." A low-middle-income country bearing home prices on par with the first world = strangling its own domestic production.
๐Ÿ‡ต๐Ÿ‡ญ Manila - Even more overpriced than Vietnam? 35.2ร—

A P/I of ~35x is among the highest in ASEAN, higher than HCMC but lower than Hanoi per Numbeo current as of 05/2026. The structure underneath is completely different from Vietnam's:

Compared with Vietnam: Vietnam's remittances are ~$16B (3.4% of GDP), with only ~$3.5โ€“4B flowing into real estate. The Philippines has real USD remittance inflows into housing plus household debt of just 12.8% of GDP โ†’ a partial justification. Vietnam's case is a credit bubble plus speculation, with no foreign-currency cushion.

๐Ÿ‡น๐Ÿ‡ญ Bangkok - Overpriced but not frothy? 32.8ร—

On par with HCMC by P/I, but the structure is completely different:

  • Real demand from foreign investors: Bangkok condos are propped up by real money flows from China, Hong Kong, Singapore - not domestic credit recirculating.
  • No zombies: Thailand already went through the 1997 crisis and reformed deeply. Thailand's central bank is more independent, with tighter real estate lending standards.
  • A disciplined BoT: Prioritizes inflation stability, isn't forced to floor the credit accelerator.

Vietnam's prices are high because of domestic speculation + leverage + zombies. Bangkok's because of real demand from international capital flows. Same P/I, one side is froth, the other is fundamentals.

๐Ÿ‡ฎ๐Ÿ‡ฉ Jakarta - Got hurt, then got healthy 23.5ร—

Lower than Vietnam despite a comparable GDP/capita (~$5,074 in 2025). The 1997 crisis taught Indonesia a brutal lesson:

Lesson for Vietnam: A painful crisis can be the necessary "medicine" - forcing central bank reform, tightening credit, building a healthier market. Vietnam has never gone through a comparable shock.

๐Ÿ‡ฒ๐Ÿ‡พ Kuala Lumpur - A lesson in deflating slowly 8โ€“9ร—

A P/I of only 8.7x sounds "healthy," but in reality KL is deflating painfully:

  • Overbuilt: Malaysia built high-end condos en masse through the 2010s, oversupplying demand. 4,234 unsold units in KL alone in Q1/2025, concentrated in the >RM500K segment.
  • Developers are bleeding: Forced into deep discounts, free furniture, 5-year free maintenance just to move inventory.
  • Low P/I โ‰  a healthy market: Prices fell because of oversupply, not rising income. Liquidity is weak, speculative money has fully exited.

Lesson for Vietnam: KL is a preview of what happens when a bubble deflates slowly - a P/I that looks "nice" on paper, while developers go bankrupt en masse and the market takes a decade to recover.

Where It Breaks - The Liquidity Black Hole

High home prices are only the symptom - the root cause sits in the financial system behind them. Real estate lives on bank credit; banks live on liquidity. When liquidity dries up, credit contracts, and real estate - the most leverage-dependent asset class - takes the first hit.

1
The Trigger
The free-market USD rate blows through the ceiling

In 10/2025, the gap between the free-market USD rate and the official rate spiked to 6โ€“8% - red-alert territory. The SBV sold reserve USD to close the gap while simultaneously mopping up VND to ease devaluation pressure. The gap narrowed, but the domestic banking system got drained of VND liquidity.

FX reserves fell from a peak of ~$110B (2022) to $83.6B (12/2025) - ~2.3 months of import cover, below the IMF's 3-month safety threshold.

2
The Domino Chain
Interbank rates explode

As the SBV mopped up VND, banks went "bloodless." The interbank rate spiked to a record 16.4%/year (VNIBOR overnight, 2/2026) - 3โ€“4x normal, and by 6/2026 briefly approached 20% (see the Live Updates section at the bottom). Banks fought each other for household deposits - pushing deposit rates up to 8โ€“9%/year.

What is the interbank rate?

Every day, commercial banks lend to each other overnight or short-term (1โ€“2 weeks) to balance liquidity - banks with surplus cash lend to banks short on cash. The rate on these loans is called the interbank rate (VNIBOR).

It's the system's "thermometer": a low rate (2โ€“5%) means banks trust each other and liquidity is plentiful. A spiking rate means banks are afraid to lend to each other - fearing a counterparty is "sitting on a bad-debt bomb," or that they themselves are badly short of cash.

3
The Kill Shot
Floating loan rates hit 12โ€“15%

As funding costs rose (8โ€“9%), banks pushed floating loan rates up to 12โ€“15%/year. On a 3 billion VND loan, the gap between 9% and 14% adds ~12.5 million/month in interest - nearly a full month's salary. This is the Forced Selling mechanism: the market doesn't need to collapse - individual cash flow buckles first.

4
Special Support
When the ventilator runs too long

Zombie bank: a bank that's still operating on paper but has effectively lost the ability to survive on its own. It stays alive on a continuous SBV "transfusion" via special lending under the 2024 Law on Credit Institutions - since 15/10/2025 allowing the SBV to lend at 0%, uncollateralized. The paradox: on one hand the SBV mops up VND to defend the exchange rate; on the other it prints 0% VND to pump into zombies - the two moves cancel each other out.

What is a zombie bank? + Vietnam case study

A "zombie bank" is a bank that's still operating on paper - still taking deposits, still has offices and staff - but has effectively lost the ability to survive on its own. Bad debt has piled up far beyond its equity capital, and its collateral (mostly real estate) has lost value while the books still carry the old price.

Illustrative example: A bank lends 10,000 billion, of which 4,000 billion is unrecoverable real-estate bad debt, leaving equity at only 2,000 billion - effectively negative capital. Yet it's still allowed to operate and gets special injections to keep liquidity going.

Vietnam case study - SCB: The SBV has injected $24B (CNBC/Reuters 04/2024), totaling $25.6B (652.7 trillion VND) as of 18/02/2025 - equivalent to ~6% of Vietnam's GDP. SCB's deposits fell from 669 trillion (10/2022) to just 19.2 trillion by end-2024 (a 97% drop). Its capital adequacy ratio (CAR) fell to โˆ’176% - meaning negative capital equal to 1.76x its risk-weighted assets. The restructuring plan runs 15 years, with the SBV only starting to be repaid from year 14.

Vietnam case study - the 4 "zero-dong" mandatory-transfer banks:

5
The Black Box
Nobody knows how deep the pit goes

Debt rescheduling without reclassifying bad debt โ†’ official NPLs sit at 2โ€“3%, the real figure is 3โ€“5x higher. Valuations carry conflicts of interest, collateral is layered and cross-pledged, and there's no real enforcement. When the market freezes and there's almost no trading, what reference price do you even mark-to-market against?

What the mechanism is called

In the banking literature, this is extend and pretend / loan evergreening: extending a loan, keeping its debt classification unchanged, or issuing a new loan so the old one never shows up as an NPL. It doesn't erase the risk - it just pushes today's loss into the future, and usually deepens the hole, since interest keeps accruing while the collateral has no real liquidity left.

The Full Domino Chain
๐Ÿ”ด Free-market USD gap 6โ€“8% ยท 10/2025
โ†“ SBV sells USD, mops up VND
๐Ÿ”ด Interbank rate spikes 16.4% โ†’ ~20% ยท 2/2026โ€“6/2026
โ†“ Banks fight for deposits
๐Ÿ”ด Deposit rates rise to 8โ€“9%
โ†“ Funding cost gets passed to borrowers
๐Ÿ”ด Floating loan rates hit 12โ€“15%
โ†“ Homebuyers run out of breath
๐Ÿ’€ Forced Selling โ†’ Real estate sell-off

Checkmate - No single way out

โš–๏ธ An impossible problem
The central bank has to choose - defend the exchange rate or rescue real estate?
Every single-path exit locks the other one shut
Option A ยท Defend the exchange rate
Raise rates to anchor the VND
โ†“
  • Real estate borrowing costs spike to 12โ€“15%
  • Mass defaults among real estate firms
  • Banks absorb a wave of bad debt
  • System-wide liquidity dries up
Option B ยท Rescue real estate
Keep rates low, pump credit
โ†“
  • The VND devalues ever faster
  • Foreign capital flees toward USD
  • FX reserves (~$84B) get drained
  • Imported inflation breaks out
โ†“ Both paths converge on the same endpoint โ†“
๐Ÿ’€
FINANCIAL CRISIS
A pattern that repeats in every economy that meets the conditions - Thailand 1997, South Korea 1997, Indonesia 1998, Argentina 2001.
Money-printing indicators - M2/FX ยท Import Cover (2 charts + comparison data)

Central banks and the IMF typically use three indicators to gauge a country's monetary "health": (1) money-system leverage against FX reserves, (2) the depth of the import-cover buffer, and (3) how credit-saturated the economy is. Indicator (3) was already covered in the Minsky Evidence section above - Vietnam blows past the BIS threshold with credit/GDP at 146%. The two remaining indicators below are also flashing red.

1. M2 / FX Reserves

Vietnam's M2 at end-2024 was ~17,900 trillion VND, or ~$700โ€“760B (~138% of GDP), while FX reserves stood at ~$83.6B. The M2/FX ratio โ‰ˆ ~850% - and with 2025 M2 already at ~21,000 trillion (~$800B), the ratio is heading toward ~950%. For every $1 in the SBV's vault, it's "carrying" $9+ of circulating VND. If just 10โ€“15% of the population panics and converts to USD, reserves would evaporate entirely.

How much VND does one USD in reserves have to carry?
Red block area is M2/FX; blue circle is import cover. Vietnam has both the highest money leverage and the thinnest import buffer.
M2/FX Import cover India 124% 11 Thailand 310% 7.1 Philippines 304% 7.0 Indonesia 408% 6.0 Vietnam ~850% 2.3 IMF 3-month threshold
Source: Trading Economics, BSP, IMF Article IV 2025. The charts below are kept for reading each indicator individually.
M2/FX leverage - how much domestic money is each $1 of reserves "carrying"?
% (M2 / FX Reserves)
Extreme leverage (>600%) High (300โ€“600%) Moderate (200โ€“300%) Safe (<150%)
The higher this indicator, the easier for the central bank to lose control if the public converts VND to USD. Vietnam is ~7x India, ~2.8x Thailand/Philippines, over 2x Indonesia.

2. Import Cover

Import Cover - how many months of imports can reserves pay for?
Months (IMF standard: โ‰ฅ3 months)
Below IMF threshold (<3 months) Thin buffer (3โ€“5 months) Safe (>6 months) IMF threshold (3 months)
Source: IMF Article IV 2025. Vietnam sits below the warning threshold - the thinnest buffer in the region.
A red trinity - all three indicators flashing warning

M2/FX โ‰ˆ 850โ€“950% (highest leverage in the region) + Import Cover 2.3 months (below the IMF threshold) + Credit/GDP 146% (already covered in the Minsky Evidence section above - highest among lower-middle-income economies, per the SBV itself). This is the classic trio of indicators that an IMF Working Paper (2018) lists across 70% of the 1997 Asian and 1980sโ€“90s Latin American currency crises. No country has ever had all three flashing red at once and still managed a "soft landing."

The central bank: referee or player?

In modern monetary theory, an ideal central bank needs to be independent of the executive branch - its #1 job is controlling inflation and stabilizing the exchange rate, even when that means decisions that run against the executive's short-term wishes. This is the standard followed by the Fed (US), the ECB (EU), and the BoJ (Japan).

In most developing economies - Vietnam included - the central bank doesn't have that same degree of independence. When a GDP growth target is fixed in stone (say, 8%/year), monetary policy usually has to yield to the aggregate-demand target. The IMF's Article IV consultations (2025) have flagged this tension between credit-driven growth and macro stability for several years running.

A multi-objective problem

Controlling inflation and the exchange rate while simultaneously pumping credit for growth - this is what Mundell-Fleming theory calls the "impossible trinity": you cannot simultaneously have a stable exchange rate, free capital flows, and independent monetary policy. At least one has to go.

Regional history shows that upgrading central-bank independence usually only happens after a crisis - once external pressure is large enough to overcome internal inertia:

Thailand ยท 11 years later
The Bank of Thailand only formally gained independence under the 2008 BoT Act - 11 years after the 1997 crisis. That "slow and half-hearted" reform is often cited as one of the reasons Thailand remains stuck in the middle-income trap today. See details โ†’
Indonesia ยท 2 years later
Bank Indonesia was granted independence under the 1999 Act, adopting inflation targeting from 2005. Indonesia's banking system today has a CAR of ~25% - considered the most solid in Southeast Asia. See details โ†’

From the regulator's seat - the political logic of growth

It's worth understanding: from the position of any developing-country government, each of the choices below has its own internal logic - this is a general pattern in the political economy of growth, not unique to Vietnam.

Push public investment?
When the private sector shrinks and real estate freezes, government spending is the one remaining lever for aggregate demand. Public investment = jobs = social stability. This is classic Keynesian logic - the same path the US took in 2008, the EU in 2009, and China in 2008, all at massive scale.
Hand land to large conglomerates?
The state needs private capital to do what the budget can't afford. In exchange, conglomerates get land, permits, credit - and in return build infrastructure and attract FDI. This loop is common across every East Asian model: Korean chaebols, Japanese keiretsu, Indonesian conglomerates. The common consequence: the more power they're handed, the more these groups become "too big to fail."
Build bridges and mega-infrastructure?
Infrastructure pulls the economy out from its two locomotive cities, balances development across regions, and spreads the benefits of growth. Large-scale projects also serve as growth symbols - the same role Japan and South Korea leaned on during their own industrialization phases.

This pattern is universal across developing economies - and it's exactly why so many bubbles get "nursed" far longer than they would if left to correct on their own. Cutting early = short-term pain = instability risk = nobody wants to be the one who pushes the button. The lesson of the 1997 crisis: when the adjustment finally becomes unavoidable, the damage is far greater than if intervention had come early - and the one who ultimately pays, as in every financial crisis, is always the ordinary citizen.

The final scenario - knocking on the IMF's door

If every internal measure fails, the IMF is the lender of last resort. South Korea 1997, Thailand 1997, Indonesia 1998, and Argentina 2001 all went through this scenario. The IMF doesn't rescue for free - the bailout package comes with: fiscal tightening, sharp rate hikes, opening up financial markets, restructuring/closing weak banks, and privatizing state-owned enterprises.

Koreans call 1997 "National Humiliation Day" - selling off national assets on the cheap in exchange for a $57B bailout package. The price of delaying until you've lost the right to decide for yourself.

โš–๏ธ The Great Deleveraging - Beautiful or Ugly?

Ray Dalio lays out a framework: every deleveraging cycle uses a mix of 4 tools. It's "beautiful" when balanced; "ugly" when heavily skewed toward one side. A textbook "ugly" scenario played out in East Asia in 1997 (tightening + bank closures + devaluation with no way out) and in Japan in the 1990s (dragging zombies along โ†’ a lost two decades). Vietnam is heavily skewed toward the same Japanese pattern - stretch it out, hide it, pump it, but don't dare cut:

1 In progress
Cutting spending

Streamlining the apparatus: ministries 18โ†’14, ministry-level agencies 4โ†’3, provinces 63โ†’34, district level abolished. But severance payouts run ~130 trillion VND (~$5.1B), against projected savings of ~190 trillion (~$7.3B) over 2026โ€“2030 โ†’ it takes years just to break even.

2 Being hidden
Debt restructuring

Rescheduling, keeping debt classifications unchanged, special liquidity support - exactly the extend and pretend mechanism. More zombies pile up, the black hole gets deeper. When real restructuring is eventually forced, the damage will be far greater.

Legal trail & the numbers (6 pieces of evidence)
Result: reported NPLs look clean on paper, but many banks' actual equity capital may already be negative under proper mark-to-market. The longer the rescheduling drags on, the deeper the "black hole" gets - and when recognition is finally forced, the losses will surface all at once instead of being spread out.
3 Not yet dared
Property tax

Debated for a decade but never enacted. The tax burden instead falls on the working class through VAT and income tax - the poor carrying the load for those sitting on a dozen vacant homes.

4 Being overused
Printing money

0% support for weak banks plus monetary easing to counter exchange-rate pressure - both build inflationary pressure without fixing the root cause. The VND is caught in a pincer.

Market footprints (5 pieces of evidence)
  • The VND was Asia's 2nd-weakest currency in 2025 - behind only the Indian rupee. HSBC ยท VnExpress (11/2025) ยท VnEconomy - the VND lost ~4.5% YTD against the USD, far worse than the IDR/THB/PHP, which all appreciated back during the year.
  • USD/VND hit an all-time high - breaking the 26,000 VND/USD mark for the first time in history. VnExpress - USD/VND tops 26,000 ยท TradingEconomics USD/VND.
  • The SBV sold down reserves to defend the rate - net sales of โˆ’$9.4B in 2024. Reserves fell from ~$110B (2022) to ~$83.6B (2025) - losing nearly a quarter of the war chest in three years.
  • M2 keeps swelling while FX shrinks - M2 rose from ~14,200 trillion (2022) to ~21,000 trillion VND (2025) (+48%), while FX reserves remain nearly a quarter below their 2021โ€“2022 peak. The M2/FX ratio has clearly deteriorated - domestic money leverage is rising fast while the foreign-currency buffer hasn't kept pace.
  • Credit grew +19.1% in 2025 - VietnamPlus/SBV, nearly double nominal GDP growth. A textbook sign of treating pain with more of the same drug.
Easing monetary policy and defending the exchange rate at the same time is impossible - the Mundell-Fleming impossible trinity. The SBV is paying for it out of the reserve war chest.
Verdict: an Ugly Deleveraging is underway

Heavily skewed: cutting spending and printing money (tools 1 & 4) are being pushed hard, while real debt restructuring and a property tax (2 & 3) haven't been deployed with anywhere near the same force. The cost of adjustment falls disproportionately on people with fixed incomes, while bad debt remains unresolved.

The Multiverse of Cycles - When Does It Turn?

HCMC P/I Ratio - 3 Scenarios, 2026 โ†’ 2030
Years of income
Bullish (10%) - Soft landing Base Case (60%) - Frozen Bearish (30%) - Bubble bursts Warning ceiling (32x)
Bullish 10%
Soft Landing

Legal bottlenecks get fully resolved. Social housing takes off. A second property tax gets applied in full. Prices flatline or dip slightly; incomes slowly catch up.

Requires a macro miracle - this has never happened once P/I passes 25x.
Base Case 60%
Malignant Freeze

A K-shaped economy: incomes flatline while inflation erodes purchasing power. Supply trickles out but prices stay anchored sky-high. A slow-bleeding, hibernating market.

Hits the 32x threshold by 2028โ€“2029. The Zombie Market made manifest.
Bearish 30%
Bubble Bursts

Overdue leverage pressure snaps liquidity all at once. Panic selling, with no support from genuine buyers.

Breaks through the ceiling as early as 2027. A cross-default domino.
Why 10% / 60% / 30%?

These are subjective estimates, not a quantitative model. Bullish sits low because it requires too many conditions to align at once (legal issues resolved, the SBV solving the exchange-rate-vs-credit dilemma, social housing booming, bad debt resolving itself). In the history of global real estate bubbles, a "perfect soft landing" has almost never happened once P/I passes 25x.

The Antidote to Illusion - 6 Misconceptions

Whether it's a "soft landing" or a bubble bursting, the outcome for ordinary people is bitter either way. Real estate isn't a stock you can sell with one click - it's an illiquid asset, inflated by credit leverage.

Click each item to expand/collapse.

1
Misconception
"Bought with cash, why worry about the storm?"

Didn't borrow to buy your home? Great. But the company that pays your salary might be running on borrowed capital. Real estate freezes โ†’ credit tightens โ†’ layoffs follow. Concentrating 70โ€“80% of your wealth in an asset that's extremely hard to sell quickly is an extreme risk - when cash is needed urgently, tens of billions of dong locked in a land title can't be broken into spendable pieces.

Systemic risk doesn't discriminate: when real estate freezes, banks tighten credit across the board - the company you work for may have no real estate loans, but its working-capital credit line gets cut too. A 2023 HCMC survey found 62% of SMEs struggled to access capital even after the central bank cut rates. Your asset is a piece of paper - a land title - but your job depends on the system's overall liquidity.

Case in point - Japan's "Lost Generation": at the 1991 peak, Ginza land in Tokyo hit $200,000/mยฒ. By 2005 it had fallen 87%. Nominal prices took 17 years to recover; real (inflation-adjusted) prices still hadn't returned to 1991 levels as of 2024. 8.5 million "akiya" (abandoned homes) sit empty - owned by cash buyers who have since died, whose heirs refuse to inherit because upkeep costs exceed the asset's value.

Case in point - China's Evergrande, 2021: 1.6 million off-plan units left unfinished. Many buyers paid 100% cash upfront in installments - no home, no compensation from anyone. Cash doesn't protect you from a developer default.

The concentration-risk principle: even the world's largest pension fund (Norway's Norges Bank NBIM) diversifies across thousands of companies and markets. The average Vietnamese household holds 70โ€“85% of its total wealth in real estate - an extreme concentration by any portfolio-management standard.

2
Misconception
"Hoard cash and wait to buy the bottom!"

Markets don't crash to the floor overnight - they deflate like a table leg being sawed off, slowly, over years. Most bottom-fishers run out of capital before the market turns. As the bubble deflates, inflation and recession set in together - cash mostly ends up covering rising living costs, not "bargain hunting."

Real estate bubbles don't deflate in a straight line: look at the Case-Shiller US chart for 2006โ€“2012 - prices peaked in June 2006, bottomed in February 2012 - six straight years of decline. Buyers who thought they'd caught the bottom in 2008 (after a 15% drop) rode another 25% down through 2012. NYT - investors who bought in Phoenix and Miami in 2008โ€“2009 lost another 30โ€“40%.

Case study - Japan's '90s bottom-callers: a BIS working paper shows Japanese property prices fell for 15 straight years (1991โ€“2006), losing 70% in major cities. Foreign investors (Morgan Stanley, Goldman Sachs) tried to call the bottom repeatedly in 1993, 1995, 1997, and 1999 - and kept taking heavy losses all the way to 2003. Buffett once said: "Be fearful when others are greedy, and greedy when others are fearful" - but even Buffett admitted he was "dead wrong" about the 2010 housing bottom.

Inflation eats cash at the same time: Vietnam 2011 - CPI peaked at 23%, deposit rates were 14โ€“17%. Savers still ran a real yield of negative 5โ€“9%. If real estate falls 30% over three years but cash loses 25% of its purchasing power, the "bottom-fishing" gain shrinks to just 5%. Turkey 2021โ€“2024 is the extreme case: the lira lost 80% of its value, inflation peaked at 85% - holders of cash lira lost two-thirds of their purchasing power in three years.

The "running out of ammo" trap: you saved 3 billion dong to "buy the bottom." Prices fall 15% โ†’ an 8-billion-dong unit is now 6.8 billion. You borrow 4 billion, feeling fine. A year later prices fall another 20% โ†’ your unit is now worth 5.4B, but the debt is still 4B - LTV jumps to ~74%. Vietnam has no stock-market-style margin call, but mortgage contracts let the bank periodically re-appraise the collateral and demand extra collateral or partial repayment once LTV crosses a threshold - right as the floating rate reprices against you too. Where's your reserve cash? Already sunk into the house.

China 2026 - prices still falling, no bottom in sight after four years (the full picture of a broken machine โ†’): CNBC, Dec 2025 - "China's property market is still searching for a bottom". New-home prices across 70 cities fell 3.2% YoY in Feb 2026 - the 32nd straight month of decline. S&P, Feb 2026, raised its 2026 decline forecast from 5โ€“8% to 10โ€“14% - meaning China is now in its fifth straight year of falling prices with no bottom in sight.

Official numbers look prettier than reality - a "price floor" mechanism hides the real decline: since 2022, Chinese local governments have imposed "price guidance" - banning developers from listing prices below a certain floor (usually the average price of the prior 6โ€“12 months). The result: contract prices on paper fall only slightly, but the price buyers actually pay is far lower, delivered through non-cash incentives:

The upshot: official figures say prices are "falling 3โ€“5% a year," but Global Property Guide's compilation shows the real decline can run 2โ€“3 times higher. Some tier-3/4 cities (Hegang, Qiqihar) have fallen 60โ€“70% in real terms even as official numbers show only a 30% drop. What this means for bottom-fishers: you're watching a number that doesn't reflect the real market price, and when that number eventually gets "revised" (as in 2026, when S&P doubled its forecast decline), the "bottom" you thought you saw never actually existed.

Chinese buyers have given up on "catching the bottom": SCMP, Sep 2025 - "Chinese homebuyers snub government incentives, betting on further price falls". When the entire market expects prices to keep falling, bottom-fishing demand disappears - which is exactly why China has gone more than four years without a clear bottom.

3
Misconception
"There's always a bailout coming"

Monetary easing doesn't create wealth - it dilutes the dong. Say a developer borrows 50 billion dong to hoard land and can't repay. The State Bank prints money to rescue the bank. More dong circulating in the economy โ†’ a 50,000-dong bowl of pho becomes 65,000, a 30,000-dong Grab ride becomes 40,000. The street vendor ends up carrying the debt of the land speculator. Inflation acts as a hidden tax - it erodes people on fixed incomes while asset owners benefit from rising nominal prices.

A bailout has three possible paths, and someone pays on every one of them:

Not everyone gets rescued - China 2021 is the lesson: Evergrande owed $300B, Country Garden owed $190B. Beijing chose not to bail them out. A Hong Kong court ordered Evergrande liquidated in Jan 2024. 1.6 million households who bought off-plan lost everything. Once bad debt exceeds 5% of GDP, even the strongest state runs out of room. Vietnam: real estate business lending alone is ~$78B (~15% of GDP); including home loans, real-estate-related credit is ~25.5% of total outstanding loans - more than a third of GDP, on top of credit/GDP already at ~146%.

Vietnam's own 2011 precedent: the government loosened credit to rescue businesses โ†’ CPI hit 18.6% for the year, peaking at 23% in Aug 2011. A one-year VND deposit paying 14% still lost 9% in real terms. Who benefited? Land hoarders (prices rose to keep pace with inflation). Who paid? Savers and wage earners. There's no such thing as a free bailout - it just shifts the burden from asset holders to income earners.

Moral hazard - a Ponzi scheme with a safety net just grows bigger: Hyman Minsky pointed out that bailouts convince investors they're "too big to fail" โ†’ they lever up even more in the next cycle. That's exactly what happened in Japan (the 1997 bailout โ†’ two lost decades) and China (the Anbang/HNA bailouts in 2018 โ†’ the bubble kept inflating through 2020โ€“2021 โ†’ and burst in 2022).

4
Misconception
"Demand is still high, how can prices fall?"

This confuses "wanting to own" with actual purchasing power (Effective Demand). A 5โ€“7 billion dong apartment against a 15โ€“20 million dong monthly income is an impossible gap. "Won't everyone rush in once prices fall?" Ask China - from 2021 to 2026, real estate sales halved, tier-3/4 cities are down to 53% of their peak, tier-2/3 prices fell 20โ€“30%. The government removed every barrier and cut rates to record lows - people still aren't buying. Feb 2026: new-home prices down 3.2% YoY, the 32nd straight month of decline.

Effective Demand - the Keynesian concept: demand only counts when backed by the ability to pay. Applied to Vietnam: HCMC's median income is ~6.2 million dong/month per person (GSO 2024); a two-earner household makes ~12.5 million/month ร— 12 = 150 million/year. A 5-billion-dong apartment equals ~33 years of gross household income (before deducting the 60% that goes to living costs). VnExpress/HoREA data shows only ~5% of HCMC households can actually afford a home. What does "high demand" even mean when 95% can't participate?

Case study - Japan's "permanent demand": from 1985โ€“1991, the argument was "Tokyo is Asia's financial center, immigration keeps rising, land is finite, demand will rise forever" - at the 1991 peak, the grounds of the Tokyo Imperial Palace were worth more than all of California. What happened next? Japan's urban population kept growing until 2010, but real estate prices fell 70% and took 17 years to recover in nominal terms. Inflation-adjusted prices still hadn't returned to 1991 levels as of 2024.

Case study - the US Sunbelt, 2006โ€“2012: Phoenix, Las Vegas, Miami - the argument was "retiring boomers, migration to the south, infinite demand." The result: Phoenix fell 56%, Las Vegas 62%, Miami 51%. Population kept growing throughout - but purchasing power dried up once subprime credit collapsed.

A 2024 Vietnam survey shows how thin real demand actually is: a Batdongsan.com.vn report found only ~30% of young households intended to buy a home in 2025, and of those, only 8% already had โ‰ฅ30% of the price in hand. The rest depend entirely on leverage - meaning "demand" is really demand for credit, not demand for housing.

The takeaway: whenever price-to-income ratios exceed 30x in any city that's had a bubble (per Numbeo's global rankings), the script is always the same - "high demand" was really just speculative demand riding the up-cycle. Once the cycle turns, demand evaporates faster than supply can contract.

5
Misconception
"Infrastructure will carry land prices on its back!"

By the time infrastructure news is all over the media, the price has already been repriced years - even decades - ahead of the potential arriving. Buying once the amenity opens means picking up the last sliver of profit at the highest risk. "Buy land as an inflation hedge" is also wrong: when inflation threatens the macro picture, the ultimate tool is raising interest rates - which strangles real estate liquidity.

Case study - Long Thanh Airport: planned in 2005, broken ground in 2021, expected operational in 2026 - 21 years of planning. Land prices in Long Thanh and Nhon Trach have already run through three cycles: 2007โ€“2010 (the first speculative peak), 2014โ€“2019 (wave two), 2020โ€“2022 (a new peak). Buyers from 2007 needed more than 15 years just to break even in nominal terms. Buyers at the 2022 peak are down 30โ€“50%, with liquidity frozen. The infrastructure arrived, but the price had already eaten 15 years of expectations in advance.

Case study - the Cat Linhโ€“Ha Dong metro line: broken ground Oct 2011, opened Nov 2021 - a decade before it ran. Land prices along the route (Ha Dong, Nguyen Trai) rose 3โ€“5x before the first train moved. After it opened in 2021, prices went flat for two years, then fell. Buyers in 2018โ€“2019 bet that "once the metro runs, prices fly" - the trains ran, and prices didn't fly.

Case study - Dubai 2008: tier-1 infrastructure didn't save prices. Dubai had the Burj Khalifa about to open (2010), Palm Jumeirah, the Dubai Metro, Emirates Airlines - every infrastructure metric led the region. The result: UBS warned of a 60% price drop; the actual decline was 55โ€“65% between 2008 and 2010. Infrastructure can't stop a liquidity crisis once leverage snaps.

Case study - Shanghai 2003โ€“2010: world-record infrastructure - the longest metro network, Pudong turned from farmland into a financial hub. Land prices rose 10x from 2003 to 2010. But by 2021โ€“2026, it still followed the national cycle - tier-2/3 prices in China fell 30% anyway. Infrastructure is a necessary condition, not a sufficient one.

The "infrastructure value trap" mechanism: a McKinsey study of 200+ major infrastructure projects found land prices typically rise 150โ€“300% in the 3โ€“5 years before the infrastructure opens, then flatten or fall once it does. The reason: the market prices in all the expected value in advance. The last buyer in inherits the risk with none of the upside left.

The "buy land as an inflation hedge" paradox: when inflation is high, the central bank is forced to raise rates. Higher rates โ†’ real estate loses liquidity โ†’ prices fall. In 1979โ€“1982, Paul Volcker pushed the Fed Funds rate to 20% - US housing fell 10% nominally, 30% in real terms. Vietnam 2011: deposit rates at 14โ€“17%, HCMC real estate prices fell 25โ€“35% between 2011 and 2013. Real estate isn't "gold" - it's a leveraged asset that moves inversely to rising rates.

6
Misconception
"The 'quit the city, buy rural land' theory"

Urbanization is a structural, near-inevitable process - young labor converges on economic centers. Owning real estate in a region people are leaving behind carries enormous liquidity risk. Many outlying areas sell "cheap" plots for a few billion dong compared to the city center, but the price-to-local-income ratio can be wildly irrational - leading to vacancy and zero rental yield.

Case study - Japan's "akiya" (abandoned homes): official 2018 statistics count roughly 8.5 million vacant homes, ~13.6% of total housing stock. Bloomberg reports the government gives many of them away for free and still finds no takers, because repair costs and inheritance tax exceed the property's value. Trend: by 2033 that could rise to 30% of total stock. Rural Japan is essentially "dissolving" - land prices in many villages have gone to zero.

Case study - China's tier-3/4 cities, Hegang (Heilongjiang): Reuters, 2023 - a 50mยฒ home in Hegang sells for 50,000 yuan (~170 million dong), down 70% from its 2013 peak. Sales across all of China's tier-3/4 cities are down to 53% of the 2021 peak. Why: young people have migrated to tier-1 cities (Beijing, Shanghai, Shenzhen), and tier-3/4 populations have begun falling in absolute terms. No tenants means no real value.

Case study - Detroit 2008โ€“2014, the $1 house: Detroit demolished 10,000+ abandoned homes; many others were auctioned for $1 and still found few takers. Why? Local taxes plus repair costs plus insurance add up to negative cash flow. Owning property in a declining town means owning a liability, not an asset.

Vietnam's own precedent - Lam Dong / Bao Loc, 2021โ€“2022: a "quit the city for the forest" wave of illegal land subdivision peaked in 2022, with agricultural land prices rising 5โ€“10x in 18 months. The aftermath: Lam Dong banned subdividing and selling agricultural land plots in June 2022. Thousands of hillside plots lost all liquidity, with many owners cutting losses of 40โ€“60%.

Vietnam's own precedent - Phu Quoc condotels, 2019โ€“2023: developers promised guaranteed returns of 8โ€“12% a year. A wave of developers stopped paying those guarantees starting in 2020, trapping investors. Secondary-market liquidity is near zero. Even though Phu Quoc is a famous tourist destination, "high tourist demand" never translated into "demand to own a condotel."

Principle: real estate only has value when someone actually USES it:

  • A tenant โ†’ cash flow โ†’ intrinsic value.
  • An occupant โ†’ real demand โ†’ liquidity when you sell.
  • No one โ†’ just "hope of reselling higher" โ†’ the definition of a Ponzi scheme.

Vietnam's urbanization trend hasn't reversed - World Bank data shows Vietnam's urban population share is still rising (roughly 40% today). People leaving rural areas remains a structural trend. Buying land in an emigration zone means betting against a long-term structural current.

Bright Spots - Not Everything Is Dark

The macro picture isn't all shadow. Vietnam does hold a few rare structural advantages:

FDI Disbursed 2025
$27.6B
5-year high (+9% YoY). Total registered $38.4B across 4,054 new projects.
Exports to the US 2025
~$142B
Vietnamโ†’US container volume +23% YoY. 20% tariff (vs. China's 145% peak in April 2025).
The most optimistic scenario

If the IFC succeeds, FDI keeps climbing, and the trade war stabilizes โ†’ real incomes rise gradually โ†’ the P/I ratio naturally deflates to a healthy level over 5โ€“10 years. That's the only viable "soft landing" path - but it demands patience, not speculation.

But: bright spots โ‰  a real estate rescue

FDI flows into manufacturing (82.8%), not speculative real estate. The IFC needs 5โ€“10 years before it has any real impact. A 20% US tariff plus 40% anti-transshipment duty remains a major risk.

The balance of payments - why record trade surpluses still leave FX reserves shrinking

On the surface, Vietnam's FDI and trade-surplus numbers look impressive: record foreign direct investment disbursement, a goods trade surplus in the tens of billions of dollars every year. But once the full balance of payments is unpacked - the current account, the financial account, and the errors & omissions line - the picture changes: the amount of foreign currency actually staying in the economy is far smaller than the headline trade surplus suggests, and most of the inflow is cancelled out by parallel outflows.

This is developed in more detail in Reading Vietnam's quarterly trade report: from trade surplus to FX reserves.

Six components need to be considered at once:

Balance of payments: the headline trade surplus leaks through several valves
Not every dollar of trade surplus ends up in the SBV's vault. Read the diagram left to right: inflows are cancelled out by outflows.
+$50B trade surplus mostly the FDI sector Doesn't turn into reserves right away domestic firms run a $17โ€“22B trade deficit FDI firms can repatriate profits FX reserves don't build up proportionally SBV still has to sell USD โˆ’$15.7B to โˆ’$22.5B repatriated FDI profits โˆ’$5B to โˆ’$8B services deficit ~$30B errors & omissions 2024 USDT / offshore real estate informal capital flows trade mis-invoicing transfer pricing / invoice mismatch โˆ’$9.4B SBV net USD sales 2024 Result: the trade surplus stays positive, but net USD supply in the system doesn't thicken to match
Source: World Bank on repatriated FDI profits; The Investor on 2024 FX intervention; IMF Article IV 2025 on errors & omissions.

Add it all up, and the outflows (repatriated FDI profits + services deficit + E&O + informal capital flows) have exceeded the trade surplus in recent years. That's why, even with a positive headline trade surplus, the SBV still has to sell foreign currency to defend the exchange rate, and FX reserves trend down instead of accumulating.

Counterargument: "Maybe the SBV is quietly holding more reserves than reported"

That argument could be true in theory, but it needs to be checked against market behavior. If FX reserves were genuinely ample, you wouldn't expect to see, simultaneously: interbank rates spiking to ~16%, the free-market/official exchange rate spread holding at 6โ€“8%, deposit rates pushed up to 8โ€“9%, and the VND depreciating in repeated waves. A country with a comfortable reserve cushion doesn't usually let all these stress indicators show up at once. In other words, the behavior of the domestic financial system is a more consistent signal than any published reserve figure, and that signal currently points to a genuinely tight supply of foreign currency.

The hard truth - crises are necessary

Imagine a forest that's never allowed to burn naturally. Deadwood piles up year after year. Rangers keep stamping out every small fire to "protect the forest." The fuel load keeps thickening, waiting for one spark big enough to turn the whole forest into an uncontrollable mega-fire. Economies work the same way - every round of bailing out zombie banks, extending debt, and hiding bad loans is just more deadwood piled onto the fuel load.

๐Ÿ“– A history lesson - read more

South Korea, Thailand, Indonesia, and Malaysia all lived through exactly this script during the 1997 Asian Financial Crisis: credit inflated for 5โ€“10 years, zombies were hidden, central banks burned through reserves trying to defend their pegs, and then the "East Asian Miracle" model collapsed in six months - Indonesia's GDP fell 13.1%, the Rupiah lost 85%, Thailand's NPLs peaked above 50%. 29 years later, each country drew a different lesson: South Korea and Singapore reformed thoroughly, Thailand and Indonesia are still carrying the consequences. Vietnam has never been through one - and is stacking up all the ingredients to experience its own version.

The Survival Question for Young People

Criticizing the system doesn't grow your account balance. This is the general structure of the global economy: a cycle of credit expansion โ†’ asset-price inflation โ†’ widening wealth gap. The question isn't "is the system fair" - it's "what are you going to do inside this system."

Should you buy a home right now?

The home-buying decision: only one path is still relatively reasonable
Read it as a flowchart: purpose of the purchase and loan-to-income ratio decide the risk, not the question "will prices keep rising."
Should you buy? separate the purpose first, then the price Buy to live in OK, conditionally loan < 30โ€“40% of income can withstand flat / falling prices and has a long emergency fund Buy to speculate red zone yield ~3.5% loan rate 12โ€“15% net loss 8โ€“11%/year Floating-rate trap risk rises after the teaser period 8โ€“9% โ†’ 12โ€“15% on a 3B loan: ~20M โ†’ ~35M/month If the answer depends on "selling for more later," that's leveraged speculation, not housing need
Checklist before signing the contract
  • Monthly interest < 30% of household income?
  • An emergency fund covering at least 12 months of expenses?
  • Not dependent on "selling for more later"?
  • Still fine if home prices fall 30% over the next 5 years?
  • Willing to accept long-term illiquidity?
  • Can still service the debt if rates rise to 16โ€“18%?
1Accept the rules of the game

Labor-income growth rarely keeps pace with asset-price inflation. That's not your fault - it's structural to the fiat monetary system. Stop blaming yourself and start understanding the rules.

2Find a survival niche

Diversify: stocks, gold, upgrading professional skills. Avoid leverage beyond your means. Patiently preserve capital while waiting for the adjustment cycle - don't bet everything on a single scenario.

3Renting โ‰  failure

In developed economies, homeownership rates run far lower than Vietnam's. Rent at a reasonable cost, invest the difference in yourself and in flexible financial instruments - that's a cash-flow optimization strategy during a period of irrational asset prices.

The system's cruelest paradox

The most effective hedges against asset-price inflation (physical gold, foreign currency, crypto, offshore investment) are exactly what capital controls restrict most tightly. Someone with $100 million can move assets offshore with ease. Someone with 100 million VND who wants to buy SJC gold bars has to queue up and eat an absurd markup. Capital controls turn asset diversification into a privilege of the wealthy - precisely the group that needs it least.

๐Ÿ“ฐ Live Updates - Market Developments

This section logs notable developments over time - newest on top - for readers to cross-check against the analysis above.

๐Ÿ’ฐ Interest rates

June 2026 - rates keep climbing despite SBV warnings: real deposit rates at 8โ€“9%, post-promo mortgage rates back up to 11โ€“15%

The SBV keeps ordering rates down and under control, but tight liquidity keeps pushing the cost of capital up: the posted rate says one thing, the rate actually paid says another, and post-promo mortgage rates sit in a zone real buyers can barely stomach.

Repeatedly scolded, yet real rates keep rising
  • Jun 15: SBV Region 1 ordered banks to control negotiated/add-on rates, warning against "cosmetic cuts" while real rates run above the posted figure. CafeF Jun 15
  • Jun 16โ€“17: the rate actually paid still beats the posted rate (posted 6โ€“7.3%, large depositors getting 8.3โ€“8.8%); the VNBA says the real 6โ€“12-month deposit rate has climbed to 8โ€“9%/year, up 1โ€“2 points from end-2025. CafeF/Ngฦฐแปi Lao ฤแป™ng
  • Jun 18: VDSC estimates the credit-vs-deposit gap has topped VND 2.5 quadrillion, system-wide LDR around 115%; overnight interbank rates briefly touched 20%/year. CafeF Jun 18
  • Jun 24: the Big 4 all post online 12โ€“18-month rates at 6.8%/year; many banks are still pinned near the 4.75% ceiling on short tenors. VietnamNet Jun 24
Mortgage rate table - June 2026

For floating rates, many banks use the formula base/reference savings rate + margin instead of a fixed figure - read the last column as the risk zone after the promo period, not a commitment binding on every application.

Bank Posted 12M deposit Mortgage promo, Jun 2026 Post-promo / floating
Vietcombank 6.8% online 12โ€“18M 9.6% for 6M ยท 10.5% for 12M 24M reference rate + ~3.3%/year
BIDV 6.8% online 12โ€“18M 9.7% for 6M ยท 10% for 12M 18M fixed package already at 13.5%
VietinBank 6.8% online 12โ€“18M 24M fixed already above 12%/year Per period-specific schedule/contract margin
Agribank 6.8% online 12โ€“18M 8โ€“9.8% depending on promo term Repriced to market once promo ends
Techcombank 6.75% online 12M 9.5% fixed 12M 13M reference rate + 3.5%/year
MB 6.35% online 12M 9โ€“9.5% for the first fixed period Floats to market band, 11โ€“15%/year
ACB 5.7% online 12M standard 9.5โ€“10.5% for the initial period Per reference rate/contract margin; stress-test for 12%+
VIB 7.0% online 12M 9.5โ€“12% depending on fixed term Already has packages near 12% even during the promo
OCB 6.9% online 12M 10.75% for 6M ยท 11.5% for 12M 13M reference rate (10% as of May 2026) + 3.25โ€“3.5% โ†’ ~13.25โ€“13.5%
BVBank 6.9% online 12M 4.99โ€“7.9% depending on 6โ€“24M term The low figure is only a short-term promo
Sacombank 6.6% online 12M 7โ€“8.5% per surveyed package Reverts to the general floating band once promo ends

Sources: CafeF/VTC News Jun 16, VPBank, compiled Jun 2026, VietnamNet Jun 24, OCB reference rate.

Takeaway: the SBV can order the posted rate down, but it can't order the system out of a liquidity shortage. Real deposit rates at 8โ€“9% flowing into floating mortgage rates of 11โ€“15% is a mechanical consequence; real-estate liquidity no longer hinges on "will prices keep rising" but on a simpler question: who can still afford the payment once the promo ends?
๐Ÿ›๏ธ Policy

General Secretary/President: "housing is for living in" - the right direction, but China's model can't just be copy-pasted

On May 19, 2026, General Secretary and President Tรด Lรขm stressed that housing must match affordability, that rental and social housing should be prioritized, and that speculation and policy rent-seeking must be controlled (Notice 64-TB/VPTW, released May 23). The message immediately recalls China's "houses are for living, not for speculation" slogan from the 2016 Central Economic Work Conference. CCTV

Why "housing is for living in" takes an enormous amount of money

In Vietnam, real estate stands alongside gold as the final asset - the last store of value on a household's balance sheet. To get money to stop flowing into land, the other absorption channels (infrastructure, rental housing, financial assets, new industry, social security) have to grow fast enough to take its place. China could say this after 2016 because it had a market of over 1 billion people, low CPI (World Bank), strong exports, and a fiscal base big enough to push infrastructure, new industry, and domestic consumption - and even then it was extremely expensive.

The real constraint: FX reserves and the exchange rate

For an economy that manages its exchange rate, the constraint sits in FX reserves and how much the exchange rate can absorb. China still holds about $3.41T in reserves as of April 2026 (Trading Economics/PBoC); Vietnam has fallen from a peak of about $109.6B in early 2022 to only about $83.6B at end-2025 (Trading Economics/IMF). The cost already shows up in the exchange rate: over 2010โ€“2023, USDCNY rose only about 5% while USDVND rose nearly 28% (World Bank). Once the exchange rate can no longer easily absorb the strain, interest rates are forced to become the release valve.

Takeaway: the direction is right, but copy-pasting China is wrong. Pulling money out of land requires alternative absorption channels and a budget/reserve base thick enough to fund them; with a thin FX buffer and rates already stretched, a correct slogan can still turn into liquidity pressure rather than deliver a soft landing for real estate.
๐Ÿ’ฐ Interest rates

Late May 2026 - liquidity-shortage signals: short tenors pinned at the ceiling, long tenors still need 7%+ to pull in money

Data through end-May 2026 shows the system has no spare cash: deposit competition sits right at the ceiling on short tenors, and 6โ€“12-month tenors still need the 6โ€“7%+ mark to keep money from leaving.

Six-week loop: rise โ†’ forced back down โ†’ rise again
  • Mar 30: rates rose fast enough that the SBV had to issue Official Dispatch 2342/NHNN-CSTT ordering a stable rate floor. Bรกo Chรญnh phแปง
  • Apr 9: the SBV met with 46 commercial banks, ordering cuts to deposit rates from 6 months and up, and to lending rates. Bรกo Chรญnh phแปง Apr 9
  • Apr 10โ€“29: more than 30 banks announced deposit rate cuts. VietnamNet
  • May 14โ€“21: pressure returned - the SBV had to issue further Official Dispatches 3972 and 4190 to inspect and reinforce compliance after some credit institutions raised rates again. Bรกo Chรญnh phแปง May 23
Maturity mismatch: short-term funding feeding long-term assets
  • LDR stretched thin: at end-April 2026, system-wide outstanding credit topped VND 19.4 quadrillion (+18% YoY), with the loan-deposit gap around VND 2 quadrillion; SSI estimates the real LDR at around 112% against an 85% threshold. VietnamNet Global
  • The classic maturity mismatch: roughly 80% of funding is short-term, yet it's financing medium-to-long-term loans; that group accounts for about 40% of total outstanding credit, nearly 60% of GDP. Vietnam News
  • Credit is outrunning deposits: 2025 credit grew +19.01% vs deposits +14.11%; Q1/2026 was +2.15% vs +0.44% - the gap has to be plugged with a deposit race, interbank borrowing, and valuable papers.
Deposit-market signals - May 2026
  • Short tenors hit the ceiling: as of May 20, 19 banks, including the Big 4, were posting the exact 4.75%/year ceiling for 3-month terms - "right at the ceiling" is the market's way of saying short-term capital is anything but abundant. Bรกo Hร  Tฤฉnh
  • 6โ€“12-month tenors still need the 7%+ mark: Bac A Bank paid 7.05% on deposits above VND 1 billion (6M); by mid-May, Sacombank, ACB, MBV, LPBank, PGBank, and VIB all had 7%+ (12M), with Sacombank briefly pushing its online rate to 7.7โ€“8% before trimming it back days later. Vietnambiz ยท VietnamNet May 13
Takeaway: if liquidity were genuinely abundant, deposit rates would fall on their own - no need for a string of dispatches, compliance meetings, and special inspections to force them down. Real estate runs on long-dated debt while banks fund it with increasingly expensive short-term capital; a floating mortgage rate of 12โ€“14% is therefore not a one-off anomaly but a consequence of the cost of capital, the maturity mismatch, and liquidity risk.
๐Ÿ’ฐ Interest rates

April 2026 rate table - 12M deposits spread up to 4 points, 12โ€“14% floating mortgage rates choke off real demand

An update on 12-month deposit and mortgage rates across Vietnamese banks - as posted in April 2026. Data compiled from VnEconomy, 24h, Topi.

Tier 1โ€“2 Commercial Bank Rates - Apr 2026
%/year
12M deposit Mortgage - first-12M promo Mortgage - post-promo / floating
Labels above the red bars: the gap between floating mortgage rate and 12M deposit rate - black = absolute gap (percentage points), red = relative gap (%). Example: VCB 5.9% โ†’ 13.9% = +8.0pp, but equivalent to +136% (borrowing costs 2.36ร— more than the deposit rate). Left group: Big 4 (state-owned banks); right group: the 4 largest private joint-stock banks.
Key points - Apr 2026
  • 12M deposit rates spread up to 4 percentage points: the Big 4 post 5.9% over the counter (online ~5.2%) while the highest payers sit at 7โ€“7.5% (ACB 7.3%, MBV 7.2%, Bac A 7.1%, Hong Leong 7.5%).
  • Mortgage rates hit a 2-year high: promos run 8โ€“10.5% depending on package; post-promo floating rates commonly run 12โ€“14%/year, with VCB/BIDV's longer packages hitting 13.5โ€“13.9% - the Big 4 are now as expensive as, or more expensive than, the private banks, reversing the usual pattern. Approval conditions have also tightened: income verification, lower LTV.
Why this matters for the bubble thesis: a 12โ€“14% floating rate against a rental yield of only 3โ€“4% in HCMC/Hanoi means negative carry of 8โ€“11 points a year; a real buyer carries a payment of ~VND 35โ€“40 million/month on a VND 3 billion loan. This is why developers are forced to subsidize the rate themselves: at the real market rate, demand = 0 - transaction liquidity in Janโ€“Apr 2026 ran on developer subsidies, not organic demand.

Sources

See the full source list ยท 20+ articles

Vietnam - Macro

Real Estate & P/I Ratio

FDI, Trade & IFC

China & International

Read next

More from the shelf

May 25, 2026How To Survive A Debt And Liquidity CrisisMay 18, 2026Extend and Pretend - When Banks Pretend Bad Debt Doesn't ExistMay 18, 2026Financial Repression - How Governments Borrow Cheaply By Taxing SaversApr 25, 2026Soros's Reflexivity: When Belief Creates Reality, Not Just Reflects It

Pass it on

If it found you, share it kindly

XEmail

03 Discussion

Leave a note

A considered space for questions, counterpoints, and useful additions. Civil, on-topic, signed.

Reader notes

...

Loading notes...