1997 asian financial crisis · reconstruction & analysis

When the "East Asian Miracle" Collapsed in 6 Months - And Where Each Country Stands 29 Years Later

In July 1997, Thailand floated the baht after burning through 23 billion USD in foreign reserves defending its fixed exchange rate. Over the following six months, the wave of collapse spread to Indonesia, Malaysia, the Philippines, then South Korea - sweeping away governments, banks, and faith in the "East Asian model" along with it. This article reconstructs the core causes, the country-by-country unfolding, and where each economy has recovered to 29 years later - including two special cases: Vietnam (indirectly hit) and Singapore (lightly wounded).

1 · Background - The "Asian Miracle" Before the Storm

In the early 1990s, Southeast Asia was dubbed by the West as the "Asian Tigers" and the "Asian Miracle". Thailand grew at an average of 9% a year through 1985-1996. Indonesia at 7%. Malaysia at 8.7%. South Korea had become an OECD economy in 1996. In 1993 the World Bank published a report titled "The East Asian Miracle", celebrating the formula of "high growth + low inequality" as a new model.

Beneath the dazzling numbers lay a fragile structure that few paid attention to:

What is the current account - and why did it run so deeply negative before the crisis?

The current account (CA) is one of the two main components of a country's balance of payments, measuring all "real" transactions (as opposed to financial ones) between a country and the rest of the world over a given period. A simple formula:

CA = Exports of goods & services − Imports of goods & services + Net foreign income + Net transfers

Where: net income = profit, interest, and dividends the country's residents receive from assets abroad − the equivalent that foreigners receive from domestic assets. Net transfers = remittances, aid. For most countries, merchandise trade is the largest component.

An intuitive way to read it: if the CA runs a surplus (positive), the country "earns net USD" from the world - able to self-finance its own consumption and investment. If the CA runs a deficit (negative), the country spends more than it earns - the shortfall must be covered by foreign capital (via the financial account). A country with a CA of −8% of GDP has to pull in foreign capital equal to ~8% of GDP every year just to balance.

Here is where the danger lies: foreign capital comes in two types. The "good" kind (long-term FDI, reinvested profit) is stable. The "bad" kind (short-term bank credit, portfolio investment, hot money) can flee within weeks. If a deep CA deficit is financed mainly by hot money - as in Thailand in 1995-96 - then once confidence collapses and capital reverses, the country no longer has the USD to pay for essential imports: oil, food, medicine, components. That is the moment a balance-of-payments crisis explodes.

Why did Southeast Asia's CA run so deeply negative in 1995-1996?

Four compounding reasons:

1. Currencies became overvalued (due to the USD peg). The USD strengthened 50% against the Japanese yen between 1995-1997. Because the baht/rupiah/ringgit were pegged to the USD, they automatically strengthened against the yen too - making Southeast Asian goods more expensive in Japan (then the region's #1 market). Imports became relatively cheap → consumers and businesses ramped up imports while exports stalled.
2. China devalued the yuan in 1994 (from 5.8 to 8.7/USD - a ~33% drop). Chinese goods suddenly became significantly cheaper, seizing market share in textiles, footwear, toys, and low-end electronics - exactly the export basket of Thailand, Indonesia, and Malaysia. Thai exports posted negative growth for the first time in 10 years (−1.3% in 1996).
3. An investment boom & property bubble sucked in imports. As Bangkok, Jakarta, and Kuala Lumpur built high-rises, airports, and industrial parks at a furious pace, imports of equipment, cement, steel, and machinery surged. An "overheating" economy → domestic demand outstripping production capacity → import demand growing faster than export supply.
4. Carry trade created an "illusion of wealth". Cheap short-term USD capital pouring in created a feeling of "easy money", encouraging consumption and investment beyond real capacity. This was a self-amplifying mechanism: capital in → exchange rate up → cheap imports → worse CA → need for even more capital in.

The warning rule: macroeconomists have an informal convention - a CA deficit exceeding 5% of GDP and lasting >2 years is a red zone (Edwards 2002, Calvo-Reinhart). Thailand stood at 8.1% in 1996, already 5 consecutive years above 5%. Indonesia, Malaysia, and South Korea were all in the 3-5% zone but deteriorating fast. This was the clearest early warning sign - ignored by markets during the "Asian Miracle" era.

The growth model rested on three pillars:

1. A soft fixed exchange rate against the USD - the Thai baht pegged at ~25/USD, the Indonesian rupiah drifting gently within a narrow band, the Malaysian ringgit at ~2.5/USD. A stable exchange rate meant businesses were confident borrowing in USD, and foreign investors were confident pouring money in.
2. Borrowing cheap short-term USD, lending at high domestic rates - a 5-8 percentage point spread between LIBOR (~5%) and VND/baht/rupiah lending rates (~12-15%) created a massive carry trade. Thai, Indonesian, and Korean banks raced to borrow short-term USD to fund long-term domestic loans.
3. Free foreign capital flows - Thailand opened the Bangkok International Banking Facility (BIBF) in 1993, Indonesia liberalized its capital account from 1989. Hot money poured in - and could flow out at any moment.

Warning signs that were ignored

From 1996, the warning signs were already clear:

• Thai exports posted negative growth for the first time in 10 years (−1.3% in 1996), partly because a weak yen made Thai goods less competitive, partly because China had begun taking market share.
• Bangkok's property bubble burst - office prices fell 30% in 1996. Finance One, Thailand's largest finance company, ran into liquidity trouble from March 1997.
• South Korea: Hanbo Steel went bankrupt in January 1997 ($6 billion in debt), followed by Sammi Steel and Jinro. Chaebol began to fall.
• The BIS published a warning report about East Asia's "credit boom" in late 1996, but capital inflows kept coming.

Why didn't markets react? Because faith in the "Asian Miracle" was too strong. Moody's and S&P still rated Thailand at A2/A. Goldman Sachs and Merrill Lynch continued to recommend buying. Global fund managers viewed East Asia as "must-allocate" in their portfolios. When a crowd believes the same story together, warnings get dismissed as noise.

2 · Four Core Causes

The academic debate over the crisis's causes has lasted to this day. The two largest schools of thought - Krugman (1998), who argues this was a structural crisis (crony capitalism, moral hazard), and Radelet & Sachs (1998), who argue this was a liquidity crisis (financial panic, self-fulfilling crisis). In reality both are right - the following four factors all existed and compounded each other:

2.1 · An unsustainable fixed exchange rate

The USD strengthened 50% against the Japanese yen between 1995-1997. Because Southeast Asian currencies were pegged to the USD, they automatically strengthened against the yen too - making exports less competitive in their largest market, Japan. At the same time, China had devalued the yuan in 1994 (from 5.8 to 8.7/USD), raising the competitiveness of everything from textiles to electronics.

Southeast Asia's current account swung from surplus to severe deficit. To defend the exchange rate, central banks had to sell USD to buy domestic currency - but foreign reserves are finite. Once speculators sensed the breaking point, they opened large short positions against local currencies.

2.2 · Debt mismatch - the "original sin"

Southeast Asian corporates and banks borrowed short-term USD (3-12 months) to lend long-term domestic currency (5-20 years, especially for property projects). Two mismatches stacked on top of each other:

• Currency mismatch: debt in USD, revenue in local currency. When the local currency drops 50%, debt obligations automatically double.
• Maturity mismatch: short-term debt, long-term assets. When creditors refuse to roll over, there is no way to repay quickly.

2.3 · Crony capitalism & weak banks

Many loans were approved not on creditworthiness but on relationships - between banks and South Korean chaebol, between Indonesian banks and the Suharto family, between Thai finance companies and politicians. The consequences:

• Overinvestment in "vanity" projects - airports, high-rises, industrial parks - with no real demand behind them.
• Actual non-performing loan (NPL) ratios were far higher than the published figures. When the crisis broke, Thailand's NPL hit 45%, Indonesia's 50%+, South Korea's 25-30%.
• Bank capital adequacy was illusory - "dressed up" through cross-shareholding between banks and corporations.

What are NPL, CAR, and Basel - and what do the numbers in this article mean?

These three concepts appear throughout the article (Thailand's NPL at 45%, Indonesia's 50%+, South Korea's 25-30%, Malaysia's 18%, Vietnam's 4.1%; Singapore's capital at 20% "double Basel"; Indonesia's 2026 CAR at ~25% "triple Basel"). This is a quick reference glossary for making sense of the figures.

1. NPL (Non-Performing Loan) - Bad debt

What NPL is: a loan is classified as "non-performing" when the borrower has failed to pay interest or principal for ≥90 days, or when the bank has sufficient evidence it will not be recovered. NPL ratio = total non-performing loans ÷ total outstanding loans.

How to read the numbers:

• <2%: a healthy system (Singapore, HK).
• 2-5%: a normal level for a developing market, worth monitoring.
• 5-10%: a sign of stress - partial restructuring needed.
• >10%: a banking crisis - systemic intervention needed.
• >20%: the system has already collapsed - a large-scale bailout needed (IMF/government).

Applying it to the cases in this article:

Country / period	NPL	Meaning
Thailand 1998-99	45%	Nearly half of all outstanding loans could not be recovered. The banking system had to be "torn in half" - 56 of 91 finance companies were closed. Restructuring cost = 43.8% of GDP.
Indonesia 1998-99	50%+	The worst of the group. 67 banks were taken over by IBRA. Restructuring cost = 56.8% of GDP - the highest in the world at the time.
South Korea 1998	25-30%	Chaebol collapses dragged down their creditor banks. Daewoo's bankruptcy left $80 billion in debt. The government created KAMCO to absorb bad debt.
Malaysia 1999	~18%	Much lower because the system had less USD exposure and imposed capital controls early. Danaharta cleaned up in 5 years.
Vietnam 2024	4.1%	A "needs monitoring" level. But if COVID-restructured debt plus property bonds extended under Decree 08 that haven't been reclassified are counted, the "broad" figure under international standards could be significantly higher.
Vietnam property 7/2024	3.7%	Up from 2.8% at end-2023 - a worsening trend. This is specifically the real estate segment.

An important detail: "reported" NPL and "actual" NPL can differ enormously. Thailand reported NPL of 7-8% before the crisis, but after the collapse the true figure was revealed to be 45%. Indonesia reported 9% before 1997, later revealed at 50%+. This gap comes from: (1) loose classification criteria (restructured debt not counted as NPL), (2) evergreening - banks issuing new loans to pay off interest on old ones, so the loan still appears "current", (3) hidden debt - SOEs guaranteeing subsidiary companies.

2. CAR (Capital Adequacy Ratio) & Basel - Capital adequacy ratio

What CAR is: the ratio of a bank's own capital ÷ risk-weighted assets (RWA). This is the "life raft" - when a loan turns into bad debt, the bank's own capital absorbs the loss first to protect depositors. The higher the CAR, the bigger the "cushion" a bank has to absorb losses.

Simple formula:

CAR = Own capital (Tier 1 + Tier 2) ÷ RWA × 100%

Where: Tier 1 = "core" capital (paid-in capital, retained earnings) - absorbs losses while the bank remains operating. Tier 2 = "supplementary" capital (subordinated bonds, general provisions) - absorbs losses when the bank is wound down. RWA = assets multiplied by a risk weighting factor (government bonds weighted 0%, corporate loans typically 100%, property loans higher).

3. Basel - the international capital standards

The Basel Accords are a set of standards issued by the Basel Committee (BIS - Bank for International Settlements), which have become the global standard for bank capital. There have been 3 generations:

• Basel I (1988): minimum CAR of 8%, distinguishing only 4 risk categories. Simple - which is also why it was easy to "game".
• Basel II (2004): CAR still 8% but with more complex RWA calculations, built on three "pillars" (capital requirements, supervision, market transparency).
• Basel III (2010, finalized 2017): a response to the 2008 crisis. Raised Tier 1 capital from 4% to 6%, added a 2.5% conservation buffer, a 0-2.5% countercyclical buffer, a liquidity coverage ratio (LCR), and a 3% leverage ratio. Total effective CAR requirement rises to 10.5-13% depending on the cycle.

Applying it to the cases in this article:

Singapore DBS/UOB/OCBC (1997)

~20%

"Double Basel" - this refers to Basel I (8%). Double that is 16%; 20% is genuinely very high. That's why Singapore didn't fall into crisis: a large capital cushion.

Indonesia 2026

~25%

"Triple Basel III" (~8%). Indonesian banks learned an expensive lesson after the 1997 crisis and are now among the most conservative in Southeast Asia.

Thailand / Indonesia / Korea (pre-1997)

~8% reported

Met Basel I on paper, but actual capital was "illusory" (cross-shareholding, RWA underestimating property risk). When NPL was revealed at 45-50%, capital was "eaten alive" within months.

Why illusory capital happens:

• Cross-shareholding: Bank A holds 10% of Bank B's shares and vice versa. When calculating CAR, both count that stake as "own capital" - but in fact it's the same money counted twice.
• Inflated collateral: property is valued at peak-bubble market prices. When prices fall 30-40%, RWA values aren't adjusted in time, so CAR on paper still looks fine.
• Off-balance-sheet exposure: lending through subsidiaries or guarantees (finance companies, SPVs). It doesn't show on the balance sheet but the risk is still there - when it blows up, the parent bank has to absorb it.

4. Why do these three metrics matter so much to a crisis?

A typical banking crisis cycle unfolds as follows:

1. Overheated credit growth → lending to property projects and connected businesses.
2. Asset bubble → property and stock prices rise sharply → reported NPL stays low because everyone can still pay interest.
3. A shock (exchange rate, interest rate, trade) → businesses struggle → NPL starts rising.
4. Banks start evergreening (issuing new loans to pay old interest) → reported NPL still pretends to be low, but actual NPL rises fast.
5. The truth surfaces (external audit, crisis trigger) → actual NPL jumps → CAR falls below 8% → banks lose the ability to operate.
6. Bank run / government intervention → restructuring cost = % of GDP (Thailand 43.8%, Indonesia 56.8%, Malaysia 5%, Vietnam ?%).

This is why the article focuses so heavily on the gap between "reported vs. actual": that gap is itself a measure of future crisis risk. Thailand in 1996 had a 6-7 point gap (reported 7-8%, actual 45%); Vietnam in 2024 has a gap that isn't yet clear because many loans haven't been reclassified - an unknown worth watching.

2.4 · Hot money & herd behavior

From 1990-1996, net private capital inflows into the five East Asian countries (Thailand, Indonesia, Malaysia, the Philippines, Korea) reached $220 billion USD. Most of it was short-term capital - international bank credit, portfolio investment.

In 1997, the net capital flow reversed to −$12 billion. In 1998: −$28 billion. Total reversal over 18 months: nearly $130 billion - equal to ~10% of the combined GDP of the five countries. No economy could absorb a capital-flow shock of that size without a crisis.

3 · The Unfolding - An 18-Month Timeline

4 · The Scale of Devastation, By Currency

Comparing the peak-to-trough decline of each currency against the USD, from June 1997 to June 1998:

GDP - the real impact on the real economy

Exchange rates are a financial-market matter. GDP is a matter of jobs, poverty, and shuttered businesses. Real GDP growth, 1996-2002:

5 · Thailand - Ground Zero

6 · Indonesia - The Worst-Hit

7 · Malaysia - The Distinctive Fork in the Road

8 · Singapore - Lightly Wounded, Fast Recovery

9 · Vietnam - Indirect Impact

10 · How Long to Recover - The Answer Depends on What You Measure

"Recovery" is a slippery concept - the answer depends heavily on which metric you use. The table below compares recovery time across three measures:

Why did recovery times differ?

Four factors determine the speed of recovery:

1. The scale of the initial destruction of financial assets. Indonesia recovered the slowest because its currency fell the deepest - turning the private sector's USD debt burden into a permanent monster.
2. The quality of the policy response. Malaysia recovered faster than Thailand despite a similar crisis - largely because it escaped the IMF's cycle of "high rates + fiscal tightening".
3. The capacity for institutional reform. Indonesia took 7 years because it also had to get through a political crisis; Thailand took 5 years because its banking system needed major surgery; Singapore took 2 years because it needed almost no reform at all - its system was already sound.
4. Fiscal buffers & reserves. Singapore, Taiwan, and China had ample FX reserves and budget surpluses - absorbing the shock easily. Indonesia, South Korea, and Thailand did not - and had to borrow from the IMF under strict conditions.

11 · Post-1997 Policy Reforms - And Their Impact Today

Each country "escaped" the crisis in its own way, but all were forced to redesign economic policy from the ground up. The reforms of 1998-2003 still shape development trajectories through 2026. Here is a comparative picture:

🇰🇷 South Korea - Root-and-branch restructuring, birthing global Samsung-SK Hynix

South Korea was the most radical reform case. The crisis forced Seoul to do what 30 years of "miracle" growth had not:

• Breaking up first-generation chaebol: Daewoo (the 2nd-largest) went bankrupt in 1999, with $80 billion in debt. Hyundai was split into 5 companies. The total number of top-30 chaebol fell from 30 to ~17. Cross-shareholding was restricted. The principle of "focus on core business" was imposed - resulting in Samsung abandoning cars and returning to electronics.
• Opening up to foreign investment: the foreign-ownership cap rose from 26% to 100% in many industries. Cross-border M&A was legalized. For the first time in history, foreign investors could hold large stakes in chaebol.
• An independent Bank of Korea (1998): inflation targeting, no longer forced by the Ministry of Finance into policy lending.
• A formal deposit insurance scheme (KDIC), ending the vague "implicit guarantee".
• Investment in R&D & semiconductors: the Kim Dae-jung government shifted priorities from "growth in quantity" to "growth in quality" - laying the foundation for Samsung Electronics and SK Hynix to rise to the global top in DRAM/NAND/HBM. By 2026, the two firms hold ~75% of the global HBM market - the memory inside every AI GPU.

The price: sharply rising inequality (the Gini coefficient rose from 0.27 to 0.34), the breakdown of stable employment (lifetime employment disappeared), a household-debt explosion (105% of GDP - the highest in the OECD), and a birth rate that fell to 0.72 (2024) - the lowest in the world. Free-market reforms succeeded macroeconomically, but produced a young "geukgi sedae" ("hell generation") - and a demographic time bomb that will go off within the next 20 years.

🇹🇭 Thailand - A weak recovery, stuck in the middle

Thailand reformed more slowly and half-heartedly than South Korea:

• An independent Bank of Thailand, under the 2008 BoT Act (it took 11 years after the crisis to get a formal law).
• Banking-sector consolidation: from 15 domestic banks + 91 finance companies (1996) → 11 banks (2010). Bangkok Bank, Kasikorn, and Siam Commercial became the "big 3".
• The FIDF (Financial Institutions Development Fund) absorbed all of the system's NPL - a cost of ~1.4 trillion baht (~12% of GDP), becoming a public-debt burden lasting into the 2030s.
• Aggressive FX accumulation: from $30 billion (1996) to ~$240 billion (2025). The BoT never wants to repeat the 1997 situation.
• Tobin-style controls on short-term capital flows: a 30% unremunerated reserve requirement was imposed in late 2006 (later scrapped after it shocked markets).

The big problem: economic reform was not matched by political reform. Internal conflict over who pays the price of the crisis (the Bangkok elite vs. the rural north) erupted into the Thaksin conflict (2001-2006) → the 2006 coup → red-shirt/yellow-shirt protests → the 2014 coup → persistent instability through 2026. In the 20 years after the crisis, Thailand has had 4 coups and 6 constitutions. This is why long-term growth has reached only 3.5% a year - far below the ~9% of pre-1997. Thailand's middle-income trap traces directly back to how the crisis was "resolved".

🇮🇩 Indonesia - Reformasi, democracy, banks reborn

Indonesia reformed the most of the group because the crisis led to the collapse of the Suharto regime:

• An independent Bank Indonesia (1999 Law): shifting from a tool of the Presidential Palace to a central bank meeting international standards. Inflation targeting from 2005.
• IBRA (the Indonesian Bank Restructuring Agency): took over 67 banks, processing ~$70 billion in bad assets. Fiscal cost: ~50% of GDP - the highest in the world at the time. Bank Mandiri was created by merging 4 failed state-owned banks.
• Deposit insurance (LPS) in 2004, ending the blanket guarantee.
• Decentralization (the Big Bang of 2001): fiscal power devolved to 500+ districts - reducing Jakarta's concentration of power. Both a reform and a legacy that has fueled local corruption.
• Democratic transition: direct presidential elections in 2004 (SBY wins), 2014 (Jokowi), 2024 (Prabowo) - political stability that surprised those who predicted otherwise in 1998.

Impact on the present: Indonesia in 2026 is Southeast Asia's largest economy, with GDP of ~$1.5T, a sturdy democracy (29 consecutive years), and sustainable growth of 5% a year. Having learned an expensive lesson, its banking system today is very conservative - a capital ratio of ~25% (triple the Basel requirement), the highest CAR in the region. A notable legacy: Indonesia is one of the countries that has weathered COVID and the 2022-23 Fed rate-hike shock best - partly thanks to the "expensive but thorough" reforms after 1997.

🇲🇾 Malaysia - Bank consolidation, upgraded into an Islamic finance hub

Malaysia reformed in a "state-directed" style:

• Forced bank consolidation: 54 banks → 10 "anchor banks" (2000-2002). Maybank, CIMB, and Public Bank became regional groups. This process was directly led by BNM - there was no choice involved.
• Danaharta & Danamodal: two agencies for restructuring bad debt and injecting bank capital. They completed their mission in 5 years - the most effective in the group.
• Khazanah Nasional (the sovereign wealth fund) was expanded to hold strategic stakes in regional businesses, and the EPF (pension fund) grew to ~25% of GDP.
• Gradually easing capital controls: rules were loosened in 1999, fully shifted to a managed float in 2005, opened further in 2009. But BNM still keeps its macroprudential "toolkit".
• An Islamic Finance Hub strategy: leveraging its Muslim population advantage + strong domestic banks. By 2026, Kuala Lumpur is the largest Islamic finance center in the world outside the Gulf - a sukuk market worth >$300 billion.
• Benefiting from China+1: Penang became a semiconductor back-end manufacturing cluster for Intel, AMD, and Infineon. Semiconductor FDI in 2023-2025 reached an all-time high.

The price: Malaysia has never escaped the "intertwined government-business" structure - leading to the 1MDB scandal (2009-2018, $4.5 billion embezzled, Najib Razak jailed). Long-term growth of 4-5% - better than Thailand but never a breakout.

The biggest policy lesson

There is a hidden pattern in the story of these four countries: whichever country reformed most painfully after 1997 ended up 29 years later with the strongest foundation. South Korea accepted the destruction of first-generation chaebol → and got Samsung-SK Hynix leading the world in HBM. Indonesia accepted the fall of Suharto + 50% of GDP spent on banks → and got a stable democracy and Southeast Asia's soundest banking system. Thailand reformed half-heartedly → and got stuck in the middle-income trap. Malaysia "reformed while holding on" → stable but never breaking out.

12 · 29 Years Later - Where Each Country Stands (2026)

By 2026, each Southeast Asian economy has traveled far from its 1997 starting line - but in very different directions. Here's a snapshot of today:

Who "won" after 29 years?

Ranking by the increase in living standards (GDP per capita, PPP) from 1996 to 2025:

A fascinating historical paradox: the country least hurt by 1997 (Vietnam) grew the most. The country hit hardest (Indonesia) achieved a genuinely sustainable recovery. The country that "did everything right" economically (Thailand) got stuck the longest. Sometimes a crisis destroys models that were truly weak but had been masked by good times.

13 · Lessons Still Alive - And What's Been Forgotten

What has been learned

• Enormous FX reserves. East Asia has stockpiled ~$5 trillion USD by 2026 - partly "self-insurance" from the 1997 lesson. China alone holds ~$3.2T, Japan ~$1.2T, Taiwan ~$580B, South Korea ~$420B.
• Floating (managed) exchange rates. Almost no country still runs a hard peg after 1997. Hong Kong is the sole exception.
• Macroprudential regulation. Bank capital ratio rules (Basel III), limits on currency mismatches, limits on property leverage - all partly stem from the 1997 lesson.
• The Chiang Mai Initiative (CMI). An ASEAN+3 currency-swap network established in 2000, expanded in 2010 ($240B) and 2024 ($600B). Its purpose: never depend on the IMF again.
• Vigilance against "Asian Miracle 2.0". Every overheated growth cycle since gets flagged by analysts with 1997 as a "warning".

What has been forgotten or repeated

• Original sin hasn't disappeared. 2018 (Turkey, Argentina), 2022 (Sri Lanka, Pakistan, Ghana) - the same mechanism: a weak currency + USD debt = default. History repeats.
• Hot money is still hot money. The 2010s USD-EM carry trade, the 2013 taper tantrum, the 2020 COVID flight - capital still flows in and out at breakneck speed.
• China's 2020-2024 property bubble. Evergrande, Country Garden - the same script: rapid borrowing, overbuilding, hidden leverage. The difference is scale: China cannot be "attacked" because of its strict capital controls.
• Belief in "this time is different". Reinhart-Rogoff wrote an entire book about this. Investors keep repeating the 1997 mistake in new forms.

14 · Lessons For Vietnam - Looking Into the 2026 Mirror

Vietnam escaped 1997 because its financial system hadn't opened up in time. But 29 years is a long time - the scale and complexity of today's financial system is worlds apart from 1997. Large private businesses now carry significant international debt, the stock market has a capitalization of several hundred billion USD, a corporate bond market has been born and nearly collapsed (2022-23), and foreign investors hold stakes in banks and bluechips. The question is: if a wave similar to 1997 arrived today, would the "closed shield" still work?

The best way to answer is to hold up this article's own checklist against Vietnam's 2026 indicators - not to pass judgment, but to spot early warning signs.

14.1 · Eight warning signs - drawn from the 1997 experience

These are 8 indicators that every 1997 victim had "flashing red" before it collapsed. The more red flags a country accumulates, and the longer it holds them, the higher the probability of crisis:

14.2 · Applying the checklist to Vietnam in 2026

Based on the most recent public reports from the IMF Article IV 2025 (published 9/2025), Fitch Ratings, S&P Global, the AMRO Annual Consultation Report 2024, World Bank Vietnam Macro Monitoring, UOB Global Economics, MUFG Research, and data from the SBV/General Statistics Office/Ministry of Finance. Each figure is sourced. Assessed on a 3-tier scale: Safe · Needs watching · Red zone.

The updated tally: 2 safe boxes, 2 needs-watching boxes, 4 red-zone boxes. This paints a picture less optimistic than what much market commentary is currently describing. Vietnam is not in Thailand's position from June 1997, because (1) its capital account is still controlled, and (2) its external debt is low. But the boxes that have already gone red - credit/GDP at 146%, house-price/income at 24-34x, FX reserves under 3 months, an NPL gap between reported and actual - are precisely the same motifs as Bangkok in 1996. On top of that, the sudden Q1/2026 CA reversal is a signal that even many optimistic macro analyses have not yet fully digested. The current position is a fork in the road: the old "closed" shield is still holding, but endogenous risk has grown right underneath it.

14.3 · Four key observations worth pondering

One - a credit/GDP ratio of 146% is a figure worth thinking hard about. Fitch Ratings (Vietnam News 12/2025) notes that Vietnam has risen to have "the highest credit/GDP ratio among lower middle-income economies" (most of which remain below 80%), and warns that "heavy reliance on bank credit could pose systemic risks". Thailand in 1996 had a credit/GDP ratio of ~150% before collapsing. Not every country that crosses 100% has a crisis - Japan, South Korea, and China are all at higher levels - but the difference is that developed economies have deep capital markets (bonds, equities, investment funds) that share the load with banks. Vietnam in 2026: the banking system carries ~85% of the economy's credit flow. When a single channel bears everything, a shock to that channel has nowhere to hide. A rhetorical question: can a country with per-capita income under $5,000 USD sustainably run a credit/GDP system of 146%?

Two - a house-price/income ratio of 24-34x far exceeds international warning levels. The World Bank / UN-Habitat standard considers a house-price-to-annual-household-income ratio of >5x "severely unaffordable". Hanoi currently sits at ~24-28x, Ho Chi Minh City at ~32-34x (The Diplomat 11/2025), putting Vietnam among the top 5 least affordable of the 103 countries surveyed (Numbeo 2025). For comparison: Bangkok before the crisis in 1996 was at ~13x; Tokyo at the peak of its 1989 bubble was ~15x. A 25-35 year-old in Hanoi would need to save more than 20 years of income - spending on nothing else (AIA Research) - to buy an average apartment. This figure alone is not a crisis - people in Hanoi and Ho Chi Minh City can live with it for many more years. But structurally, when apartments priced above 80 million VND/m² are pledged as collateral for credit, a 20-30% price correction would drag the banking system down with it - exactly the mechanism of Bangkok in 1996. AMRO's Annual Financial Stability Report 2024 devoted an entire chapter to vulnerabilities from the property sector, including Vietnam.

Three - the Q1/2026 trade reversal: a Thailand 1995-96 motif. This is the most recent and also the most worrying data point. In Q1/2026, Vietnam's trade balance suddenly swung from a $3.57 billion surplus (Q1/2025) to a $3.64 billion deficit - a swing of $7.2 billion within a single year (SGGP / General Statistics Office). Exports still grew 19.1% - healthy - but imports grew 27.0%, 8 percentage points faster than exports (VnEconomy). January 2026 alone posted a $1.78 billion deficit - versus a $3.1 billion surplus in January 2024.

Why does this resemble Thailand in 1995-96? Re-read section 2.1 of this article: "an overheating economy → domestic demand outstripping production capacity → import demand growing faster than export supply". Vietnam in Q1/2026: GDP grew 7.83% - the highest in Southeast Asia, with a full-year target of 8.3-8.5%. Credit is growing nearly 20% a year. Public investment disbursement is large. The inevitable result: imports of machinery, raw materials, petroleum, and construction equipment have surged. In good times, this reads as "investing in the future". In bad times (say, a US tariff shock, a Chinese devaluation, or the Fed failing to cut rates), it becomes "shooting yourself in the foot" - the currency comes under downward pressure, FX reserves come under selling pressure, and a currency spiral begins.

What figure is most worth pondering? The 2024 CA surplus was $27 billion (6.6% of GDP). If imports keep outpacing exports at the Q1/2026 pace, the full-year 2026 trade balance could slide to a deficit of $10-15 billion. Add in services (usually a $5-8 billion deficit) and investment income paid abroad, and the $16 billion in remittances may not be enough to offset it. The full-year 2026 CA could go from a 6.6%-of-GDP surplus to near balance or a mild deficit - a structural change within a single year. This is exactly the kind of "drift" the BIS warned about in Thailand in late 1996 - but which was ignored because of faith in the "Asian Miracle". The question: is the market of 2026 alert enough to notice this signal, more than the market of 1996 was?

Four - a record-thin FX reserve buffer - thinner than in 1996. Vietnam's FX reserves at end-2025 stood at ~$84 billion (SBV / Trading Economics), equivalent to just 2.0-2.2 months of imports (CEIC) - below the IMF's recommended 3-month minimum. Regional comparison (2025): Thailand $240 billion (~9 months), Indonesia $155 billion (~6 months), Malaysia $120 billion (~5 months), Singapore $420 billion. Thailand collapsed in 1997 when its "usable" reserves turned out to be only ~$2.8 billion instead of the reported $30 billion. Vietnam doesn't have a similar forward-swap transparency problem - but in absolute terms, Thailand in 1996 had reserves equivalent to 5-6 months of imports; Vietnam in 2025 sits at 2 months. Combined with the VND's 3.55% decline over the first 9 months of 2025 and "almost no room left" within its 3-5%/year band (MUFG Research 12/2025), the SBV is walking a tightrope between exchange-rate stability on one side and protecting reserves on the other. The question this raises: if the Fed doesn't cut rates as quickly as forecast, or if the USD strengthens again in 2026, how long can the buffer hold?

14.4 · What shields remain intact - and what has changed

What remains intact (advantages inherited from 1997):

• The capital account is still partially controlled - international hot money finds it hard to flood in en masse and hard to flee en masse.
• A flexibly managed exchange rate - avoiding the "hard peg" trap that caught Thailand and Indonesia.
• A concentrated banking system with state involvement - convenient for administrative restructuring when needed (similar to Malaysia's 1998 model).
• Public debt at 34.7% of GDP, external debt at 27.9% of GDP - substantial fiscal room to absorb a shock.

What has changed (new risks compared with 1997):

• The trade balance has reversed since early 2026 - a $3.64 billion deficit in Q1/2026, versus a $3.57 billion surplus in Q1/2025 (a $7.2 billion swing). This is the newest change, and a structural one - not seasonal noise.
• The credit/GDP ratio has exceeded 146% - a level that in 1997 only developed economies carried. A ratio of "developing income, developed leverage".
• Large private businesses (especially in real estate) have taken on significant USD borrowing and issued sizable international bonds - a new contagion channel.
• A domestic corporate bond market has emerged - and has already had a near-default precedent (2022-23, leading to Decree 08/2023 allowing a 2-year extension). ~50 trillion VND in bonds have been extended to their maximum, and the time bomb is still there.
• Property prices have risen to a price/income ratio far beyond international warning standards - putting an entire young generation outside the reach of homeownership.
• The share of foreign capital in bluechip stocks and government bonds is now significant - not as large as Thailand's in 1996, but enough to create capital-flow pressure.
• High export dependence on a few major partners (the US ~30%, China ~20%) - a trade/tariff shock could directly hit the CA.

14.5 · Five things to do - drawn directly from the 1997 lesson

1. Raise FX reserves to at least 5-6 months of imports. This is the region's post-1997 standard (Thailand 9 months, Indonesia 6 months, Malaysia 5 months). The current 2-month level is below IMF standard - not a "thin buffer" but "barely any buffer at all". The Chiang Mai Initiative (CMIM) offers Vietnam a swap line of ~$10 billion, but self-reliance remains the first line of defense. VNDirect's target of $102 billion / 3.3 months by end-2026 is only step one.
2. Rein in credit growth to a level compatible with nominal GDP. Credit growth of 18-19% a year against ~10% nominal GDP growth means leverage is ballooning at nearly double the rate of real output. This is a formula for accumulating systemic risk. The 2026 credit target should be anchored to nominal growth, not to an ambitious GDP target.
3. Bring NPL transparency up to IFRS 9 / Basel III standards. The gap between reported NPL (4.1%) and "broad" NPL (including COVID-restructured debt + Decree 08 property bonds + loans "requiring special attention") is a kind of "fog" that international markets particularly dislike. When the fog is thick, confidence collapses fast - this is exactly the mechanism that dropped the Indonesian rupiah 86%.
4. Reform the corporate bond market thoroughly. Decree 08/2023 "resolved" the 2022-23 crisis by postponing it. As the extensions expire (2025-2026), a full legal framework is needed for handling defaults, independent ratings, and bondholder protection. Without it, the cycle will repeat - and the next round could be larger.
5. Reduce the share of credit going to real estate, and diversify capital toward production. Macroprudential tools (LTV, DSR, capital surcharges for property) have existed in the region for 15+ years - they need to be applied more decisively and transparently. A house-price/income ratio of 24-34x isn't just an economic problem - it's a social one, when an entire young generation is priced out of owning a home in the major cities.

Vietnam holds a rare historical advantage: learning the 1997 lesson without having to pay the tuition. The most important question isn't "can a crisis be avoided?" - no country avoids crisis forever. The question is: when the next shock arrives, will the system be designed to absorb it, or will it still be operating on the assumption that "this time is different"?

In Summary

The 1997 Asian financial crisis was not a random accident. It was the landing point of an unsustainable structure: a fixed exchange rate + free capital flows + weak banks + short-term USD debt = a time bomb. The epicenter was Thailand, but Indonesia suffered the worst damage because its economic crisis combined with a political crisis. Malaysia took the unconventional road (capital controls) and recovered faster than expected. Singapore and Vietnam are the two "lightly wounded" cases - but for completely different reasons: Singapore because of a strong system, Vietnam because of a closed one.

29 years later, none of these countries has returned to its old exchange rate against the USD. The currency is the permanent scar of the crisis. But the economy is different - Vietnam's per-capita income has multiplied 7.4 times, Singapore has joined the ranks of the world's richest, Indonesia has become a Southeast Asian giant. Thailand, ironically, is the slowest to recover - stuck in the middle-income trap that 1997 first exposed.

The core lesson for today is not "avoid crisis" - impossible - but to design a system that can withstand crisis: adequate reserves, a flexible exchange rate, well-capitalized banks, controlled leverage, caution toward hot money. Singapore has done it. China has done it. Vietnam has done it, in its own way. But the more important question still remains: are today's rising economies - having learned the lesson - genuinely different this time?