Executive Summary

The Japanese asset bubble of 1986–1991 remains the single most important case study in the mechanics of credit-fuelled asset inflation and its aftermath. At its peak on 29 December 1989, the Nikkei 225 closed at 38,915.87. By August 1992, it had lost more than 60 per cent of its value, and the index would not surpass its 1989 high until February 2024 — a recovery period spanning thirty-four years.

At the height of the mania, the theoretical land value of the Imperial Palace grounds in central Tokyo exceeded the market value of all real estate in the state of California. Japanese commercial land prices rose 302% between 1985 and 1991 before collapsing and declining for fourteen consecutive years. The total wealth destruction in equities and real estate exceeded 1,500 trillion yen, roughly three times Japan’s GDP at the time.

The critical lesson for today’s investors is deceptively simple: when monetary policy is persistently too loose for the real economy, and when the resulting asset inflation becomes self-reinforcing through collateral-based lending, the bust is not merely a correction — it is a structural fracture that can take a generation to heal.

The Conditions That Made It Possible

The Plaza Accord and Its Consequences

The story begins not in Tokyo but in New York, on 22 September 1985, inside the Gold Room of the Plaza Hotel. Finance ministers and central bank governors from the United States, Japan, West Germany, France, and the United Kingdom — the Group of Five — signed what became known as the Plaza Accord.

The agreement’s explicit purpose was to depreciate the US dollar against the yen and the Deutsche Mark through coordinated intervention. The Americans, led by Treasury Secretary James Baker III, needed a weaker dollar to address a yawning trade deficit that had become politically toxic. Japan’s export juggernaut, powered by Toyota, Sony, and the zaibatsu-descended keiretsu conglomerates, had made it the primary target.

The intervention worked with devastating efficiency. The yen surged from roughly 240 to the dollar in September 1985 to 150 by the end of 1986 and below 130 by 1988. This 45 per cent appreciation in under three years created an economic crisis for Japan’s export sector, which the Japanese government and the Bank of Japan chose to address in the worst possible way: by flooding the domestic economy with cheap money.

The Bank of Japan’s Fatal Pivot

Bank of Japan Governor Satoshi Sumita slashed the official discount rate five times between January 1986 and February 1987, taking it from 5.0 per cent to 2.5 per cent — the lowest level in Japan’s post-war history. The cuts were made in response to endaka fukyo, the strong-yen recession of 1986, when GDP growth briefly faltered and industrial production fell.

But the medicine was far too aggressive for the disease. By 1987, the Japanese economy was growing at better than 4% annually, yet the discount rate remained at its emergency low until May 1989 — a full two-and-a-half years after the economy had recovered. This was the monetary policy error that inflated the bubble.

The transmission mechanism was straightforward. Japanese banks, operating under a system where the Ministry of Finance guided lending through ‘window guidance’ (madoguchi shido), dramatically expanded their loan books. Total bank lending grew at double-digit rates from 1986 through 1989. Critically, a rising share of that lending was collateralised against real estate. Japanese banks were permitted to count unrealised gains on their equity portfolios as Tier 2 capital under the new Basel I framework — a regulatory gift that turned rising stock prices into lending capacity. The more stocks rose, the more banks could lend; the more they lent, the more borrowers bid up land and stock prices. It was the most elegant doom loop in financial history.

The Credit Explosion

Between 1985 and 1990, Japanese bank lending to the real estate sector roughly doubled. The aggregate loan-to-GDP ratio surged past 100%. The seven largest city banks grew their balance sheets by roughly 60% in five years. But the most dangerous lending came from institutions that operated largely outside the regulatory spotlight: the jusen (housing loan companies), agricultural cooperatives, and non-bank financial intermediaries.

The jusen, originally established to provide housing mortgages, had pivoted aggressively into speculative commercial real estate lending. By 1990, the jusen’s combined loan book totalled approximately 13 trillion yen, the overwhelming majority backed by land whose valuations bore no relationship to its income-generating potential.

Corporate leverage was equally extraordinary. Japanese corporations used a mechanism called zaitech — financial engineering — to raise cheap capital through equity-linked bond issues (typically warrant bonds and convertible bonds issued in the Euromarket) and reinvest the proceeds in tokkin funds (specified money trusts) that speculated in equities and bonds. At its peak, zaitech profits accounted for a larger share of total corporate earnings at some major companies than their core industrial operations. The tail was wagging the dog.

Market Psychology and the Myth of Japanese Exceptionalism

What made the bubble psychologically durable was a deeply held conviction — shared by Japanese policymakers, investors, and much of the Western financial establishment — that Japan’s economic model was structurally superior and that Japanese asset prices reflected genuine fundamental value. The narrative had surface plausibility. Japan had achieved miraculous post-war growth.

Its manufacturing quality was world-leading. Its savings rate was extraordinarily high, providing a seemingly bottomless pool of domestic capital. Its corporate governance model, built on cross-shareholdings within keiretsu groups, was praised as fostering long-term thinking.

“Japan has demonstrated that a capitalist economy can achieve sustained growth rates previously thought impossible outside of centrally planned economies.”

— Ezra Vogel, Japan as Number One, 1979

By the late 1980s, this narrative had hardened into dogma. The Nikkei 225 traded at a price-to-earnings ratio that peaked above 60x in 1989 — more than triple the historical average of the S&P 500.

Japanese bulls argued that traditional valuation metrics were irrelevant because of cross-shareholdings, hidden asset values in corporate real estate, and the lower ‘required rate of return’ in Japan’s low-interest-rate environment. The argument was circular: prices were justified because Japan was different, and Japan was different because prices kept rising.

Regulatory Gaps

The Ministry of Finance, under successive ministers from the Liberal Democratic Party, operated as both regulator and industrial champion. There was no independent financial regulator — the MoF supervised banks, securities firms, and insurance companies simultaneously, with endemic conflicts of interest.

Accounting standards allowed banks to carry assets at historical cost, masking the true extent of their exposure to inflated property values. Capital adequacy ratios were flattered by the inclusion of unrealised equity gains. Securities firms routinely engaged in tobashi — the practice of hiding client losses by transferring them between accounts at fiscal year-end — with the implicit tolerance of regulators. The system was optimised for growth, not for stability.

The Key Players

The Architects

Satoshi Sumita — Governor, Bank of Japan (1984–1989)

Satoshi Sumita presided over the most consequential monetary policy error in Japan’s post-war history. A career MoF bureaucrat rather than a monetary economist, Sumita was widely regarded as a placeholder governor who deferred to political pressure. His decision to hold the discount rate at 2.5 per cent from February 1987 until his term expired in December 1989 — long after the economy had overheated and asset prices had entered manic territory — provided the fuel for the bubble’s most dangerous final phase. Sumita’s defenders argue he was constrained by the Louvre Accord of February 1987, which committed Japan to supporting the dollar and preventing further yen appreciation, effectively tying the BOJ’s hands.

The defence has merit but is not exculpatory: a more independent and courageous central banker would have prioritised domestic financial stability over exchange rate diplomacy.

Noboru Takeshita — Prime Minister (1987–1989) and the Political Class

Prime Minister Noboru Takeshita and the broader LDP establishment treated rising asset prices as evidence of Japan’s ascendance and as a political asset. Takeshita’s government launched fiscal expansion alongside the BOJ’s easy money, including a 6 trillion yen spending package in 1987. The LDP’s political funding model was deeply entangled with real estate developers and construction interests — the so-called ‘construction tribe’ (kensetsu zoku) within the party ensured that any policy that threatened land prices faced intense lobbying resistance. Takeshita himself would later resign over a bribery scandal involving the Recruit company, whose share dealings were emblematic of the era’s speculative excess.

The Keiretsu and Their Bankers

The great industrial groups — Mitsubishi, Mitsui, Sumitomo, and their associated banks — were both architects and victims. Their cross-shareholding structure, where companies within a keiretsu held each other’s shares as a demonstration of mutual loyalty, meant that rising equity prices automatically strengthened the entire group’s balance sheet. The main banks at the centre of each keiretsu expanded lending aggressively, confident that land values could only rise. Ichiro Isoda, chairman of Sumitomo Bank through much of the 1980s, presided over an institution that became one of the most aggressive lenders in the commercial property market, including controversial loans linked to the Itoman affair — a sprawling scandal involving art speculation, extortion by sokaiya (corporate racketeers), and billions in bad loans.

The Cassandras

Yasushi Mieno — Governor, Bank of Japan (1989–1994)

Yasushi Mieno took office on 17 December 1989, twelve days before the Nikkei’s all-time high, and immediately declared war on the bubble. A career central banker who had risen through the BOJ’s ranks rather than through the MoF’s revolving door, Mieno believed that asset inflation posed an existential threat to Japan’s economic health and that the BOJ had been negligently accommodative for too long.

He raised the discount rate five times in fifteen months, from 2.5 per cent to 6.0 per cent by August 1990. He simultaneously imposed direct quantitative restrictions on bank lending to real estate. Mieno was the right man arriving too late: by the time he acted, the bubble was so large that its deflation would cause a banking crisis regardless. His aggressive tightening has been criticised for accelerating the collapse, but the counterfactual — allowing the bubble to inflate further — would almost certainly have produced an even worse outcome.

The Foreign Shorts

Several foreign hedge funds and proprietary trading desks identified the bubble in its late stages and positioned short. Most famously, George Soros through his Quantum Fund had significant short positions in Japanese equities and was openly sceptical of Japanese valuations by 1989. The American financial press, particularly Barron’s, ran increasingly critical coverage of Japanese market valuations in 1989. But the foreign bears were early and often wrong during the bubble’s final melt-up: the Nikkei gained 29 per cent in 1989 alone, punishing premature shorts severely.

The Regulators

The Ministry of Finance

The MoF, under a succession of ministers including Kiichi Miyazawa (who later became Prime Minister during the crisis) and senior bureaucrats like Makoto Utsumi (Vice Minister for International Affairs), oscillated between complacency during the bubble and panicked intervention during its deflation. The MoF’s most damaging legacy was its role in delaying recognition of bad debts at Japanese banks throughout the 1990s. The practice of ‘forbearance’ — allowing banks to avoid writing down non-performing loans by rolling them over, often to zombie companies that were already insolvent — transformed what should have been a painful but finite crisis into a decades-long structural malaise. The MoF was finally stripped of its financial supervisory role in 1998 with the creation of the Financial Supervisory Agency.

The Casualties

The casualties were legion. Ordinary Japanese families who purchased homes at peak prices in 1989–1990 found themselves trapped in negative equity for years, often decades. A salaryman who bought a modest flat in Tokyo’s suburbs for 60 million yen in 1990 might have watched its value decline to 25 million by the mid-1990s while his mortgage remained at the original amount. The jusen collapsed entirely, requiring a taxpayer-funded bailout of 685 billion yen in 1996 that provoked public fury.

Smaller regional banks and credit cooperatives failed by the dozen. Most dramatically, several of Japan’s largest financial institutions eventually succumbed: Yamaichi Securities, one of the ‘Big Four’ brokerages, collapsed in November 1997, its president weeping on national television as he announced the firm’s closure. The Long-Term Credit Bank of Japan and the Nippon Credit Bank were both nationalised in 1998.

The Trigger and the Cascade

The Final Melt-Up: 1989

The last year of the bubble was its most frenzied. The Nikkei 225 began 1989 at approximately 30,159 and rose nearly without interruption through the year. Trading volumes on the Tokyo Stock Exchange were enormous — on peak days, the TSE’s turnover exceeded that of the New York Stock Exchange. New equity issuance, particularly through warrant bonds, reached record levels as corporations rushed to capitalise on stratospheric share prices. The real estate market was even more divorced from reality: Grade A office rents in central Tokyo’s Marunouchi district had reached levels where the rental yield was barely 1%, a fraction of the cost of capital even at Japan’s low interest rates. Investors were buying land purely for capital appreciation, not income.

On 25 December 1989, the final trading day before the New Year holiday, the Nikkei 225 closed at 38,915.87. The mood in Tokyo’s financial district was euphoric. Japan had just completed the decade in which its companies had acquired Rockefeller Center, Columbia Pictures, and the Pebble Beach Golf Links. Japanese banks held seven of the ten top positions in global banking by assets. The prevalent view was that minor corrections might occur but that the structural story was intact. Almost nobody expected what came next.

January–March 1990: The First Crack

The Nikkei opened the new decade on 4 January 1990 and immediately began to fall. The initial trigger was the Bank of Japan’s rate hike on 25 December 1989 — Governor Mieno’s first move, raising the discount rate from 3.75% to 4.25%. The hike had been signalled, but its timing — right at the market peak — shattered the complacency that had supported the final rally.

By the end of January, the Nikkei had fallen to roughly 37,200, a decline of about 4.5%. In February, the fall accelerated. The BOJ hiked again on 20 March 1990, to 5.25%. By the end of March, the Nikkei had plunged to approximately 29,980 — a 23% decline from its peak in just three months.

The Mechanism of Collapse

The cascade operated through three interlocking channels. First, the BOJ’s rate hikes directly increased the cost of leveraged speculation. Margin borrowers and zaitech operators who had funded equity positions with cheap short-term debt faced rapidly rising carrying costs. Second, the quantitative restrictions on real estate lending, implemented through administrative guidance beginning in April 1990, cut off new credit to the property sector, freezing transaction volumes and beginning the deflation of land prices. Third, the doom loop operated in reverse: falling equity prices reduced banks’ unrealised gains, shrinking their capital base and forcing them to curtail lending. Reduced lending put further downward pressure on both equity and property prices.

1990–1992: The Slow-Motion Crash

Unlike the 1929 Wall Street Crash, which concentrated its devastation into a few terrifying days, the Japanese market’s decline was a grinding, relentless affair that lasted years. The Nikkei rallied briefly in the summer of 1990, recovering to about 33,000 by June, giving false hope that the correction was over. Then Iraq’s invasion of Kuwait on 2 August 1990 triggered an oil price spike that hit Japan — almost entirely dependent on imported energy — with disproportionate force. By October 1990, the Nikkei had crashed through 20,000, a level it had last seen in 1987. The index closed 1990 at approximately 23,849, down 38.7 per cent for the year — one of the worst annual performances of any major stock market in history.

The year 1991 brought the first visible cracks in the banking system. Toho Mutual Life Insurance became the first major insurer to fail. The jusen’s bad debts were becoming impossible to conceal. Land prices in Japan’s six major cities began their long descent, falling 1.7 per cent in 1991 — the first decline in the national index since the data began. The Nikkei oscillated between 22,000 and 27,000 through 1991, never recovering meaningfully.

By August 1992, the Nikkei hit its initial trough at approximately 14,309 — a 63 per cent decline from its peak. The bear market had lasted thirty-two months. But this was not the end. The index would make even lower lows in the years ahead, bottoming at 7,607 in April 2003 — an 80 per cent peak-to-trough decline that took over thirteen years to fully register.

The Aftermath and Resolution

The Banking Crisis and the Lost Decade(s)

The aftermath of the Japanese bubble was defined by a single catastrophic policy failure: the refusal to recognise and address bad debts in the banking system promptly. Japan’s regulators, political class, and banking executives collectively chose to engage in an elaborate pretence that the loans backing peak-valued real estate were still performing. Banks rolled over non-performing loans rather than foreclosing, keeping insolvent borrowers alive as ‘zombie companies’ that consumed capital, depressed productivity, and prevented the creative destruction necessary for economic recovery.

The scale of the bad debt problem was staggering. Official estimates of non-performing loans peaked at roughly 35 trillion yen, but independent analysts and the eventual audits by the Financial Supervisory Agency suggested the true figure was closer to 100 trillion yen — approximately 20% of GDP. The drip-feed of institutional failures continued throughout the 1990s: Cosmo Credit Cooperative (1996), Hokkaido Takushoku Bank (1997), Yamaichi Securities (1997), Long-Term Credit Bank (1998), and Nippon Credit Bank (1998). Each failure eroded confidence further and reinforced the deflationary psychology that gripped the nation.

Peak-to-Trough Losses

The Nikkei 225 fell from 38,916 to its ultimate low of 7,607 in April 2003, an 80.5% decline. Japanese commercial land prices fell for fourteen consecutive years from 1991, with six-city commercial land prices declining by approximately 87% from peak to trough. Residential property in the Tokyo metropolitan area fell by roughly 65%.

Total wealth destruction in equities and real estate is estimated at 1,500 trillion yen (approximately $13 trillion at the time), or roughly three times Japan’s annual GDP.

Policy Response

The BOJ eventually cut the discount rate back to 0.5 per cent by September 1995 and effectively to zero by 1999. Japan became the first major economy to adopt quantitative easing in 2001 under Governor Masaru Hayami. Fiscal stimulus was deployed repeatedly — Japan launched ten major fiscal packages between 1992 and 2000, totalling over 100 trillion yen in nominal terms.

The result was a surge in government debt from approximately 60 per cent of GDP in 1990 to over 100 per cent by 1998 and eventually beyond 250 per cent, making Japan the most indebted developed nation on earth.

The regulatory response was slow but eventually comprehensive. The Financial Supervisory Agency was established in 1998, stripping the MoF of its regulatory role. The Resolution and Collection Corporation, modelled partly on the US Resolution Trust Corporation, was established to manage and dispose of bad assets. Mark-to-market accounting was phased in. Banks were eventually forced to raise capital and write off bad debts, a process that was not substantially complete until the mid-2000s.

Investor Lessons and Modern Parallels

Five Lessons for the Active Investor

Lesson 1: Collateral-Based Lending Spirals Are the Most Dangerous Force in Finance

When rising asset prices automatically generate the collateral and capital that fund further purchases of those same assets, the system has created a perpetual motion machine that will run until it doesn’t. Japanese banks counted unrealised equity gains as capital. Rising land served as collateral for loans used to buy more land.

The mechanism is identical in every major bubble: mortgage-backed securities in 2008, crypto lending protocols in 2022. The investor’s task is to identify when credit growth is being driven primarily by asset appreciation rather than by income growth. When the ratio of credit growth to nominal GDP growth persistently exceeds 1.5x, extreme caution is warranted.

Lesson 2: Valuations Always Matter, Eventually

A price-to-earnings ratio of 60x on a major national index has never been sustained. Every narrative constructed to explain why traditional valuation metrics no longer apply — Japanese cross-shareholdings in 1989, ‘new economy’ eyeball metrics in 2000, ‘software is eating the world’ in 2021 — has eventually been refuted by the mean-reverting force of real returns.

The Japanese experience teaches that the market can remain irrational for years, but that the magnitude of the eventual reversion is directly proportional to the magnitude of the overvaluation. There is no such thing as a permanently high plateau.

Lesson 3: Central Banks Can Inflate Bubbles Far More Easily Than They Can Manage Their Deflation

The BOJ held rates at 2.5% for twenty-seven months too long, then hiked aggressively into a bubble of its own creation. The asymmetry is structural: easy money creates diffuse, positive-feedback dynamics that are politically popular, while tightening creates concentrated, identifiable losers who lobby ferociously. The modern investor should pay close attention to periods where real interest rates are persistently negative relative to asset price growth — this is the signature condition for bubble formation.

Lesson 4: The Aftermath Is Always Worse When Bad Debts Are Hidden

Japan’s decision to allow zombie banks to roll over zombie loans to zombie companies transformed a financial crisis into a generational economic malaise. The contrast with Sweden, which experienced a similar banking crisis in the early 1990s but forced rapid recognition of losses, nationalised failing banks, and recovered within five years, is instructive.

For investors, the lesson is that the speed and decisiveness of the policy response to a banking crisis is the single best predictor of recovery duration. When regulators favour forbearance over resolution, extend your timeline for recovery by a factor of three to five.

Lesson 5: Demographic Headwinds Compound Financial Crises

Japan’s working-age population peaked in 1995 and has been declining ever since. The bubble burst precisely as Japan’s demographic dividend was exhausting itself. Falling asset prices, rising dependency ratios, and declining aggregate demand reinforced each other in a vicious cycle that monetary and fiscal policy proved powerless to break.

Modern Parallels

The parallels to China’s current situation are striking and uncomfortable. China’s property sector, which at its peak accounted for roughly 29 per cent of GDP (including related industries), has undergone a pronounced correction since 2021. Chinese banks hold vast quantities of property-related loans, local government financing vehicles have accumulated debts that may exceed 50 trillion yuan, and the demographic outlook — China’s working-age population peaked around 2015 — mirrors Japan’s with a lag of approximately twenty years.

The policy response thus far has featured the same instinct toward forbearance and gradual adjustment that characterised Japan’s approach. If Chinese policymakers fail to force rapid recognition of bad debts and instead allow a zombie lending dynamic to take hold, the Japanese template suggests the downturn could last far longer than consensus expects.

The broader lesson applies to any market where prolonged low interest rates have inflated asset values beyond what income fundamentals support. The post-2020 surge in US commercial real estate valuations, funded by floating-rate debt issued at near-zero rates, bears a structural resemblance to Tokyo’s office market in 1989: rental yields compressed to levels that only make sense if capital appreciation continues indefinitely. When that assumption fails, the correction can be severe and protracted.

What Would a Contrarian Have Done?

The optimal strategy with perfect hindsight was straightforward: go long Japan through the mid-1980s, exit all Japanese equity and property exposure by late 1988 at the latest, and establish short positions in early 1990 after Mieno’s first rate hike confirmed the policy shift. In reality, this was extraordinarily difficult. The Nikkei gained 29% in 1989 alone, meaning an investor who sold at the ‘obviously expensive’ levels of early 1989 left massive gains on the table. Short-selling Japanese equities was logistically difficult for foreign investors due to securities lending constraints, and the cost of carry on short positions during a melt-up was punishing.

A more realistic contrarian strategy would have been to reduce Japan exposure progressively from 1988, moving into undervalued markets (the US equity market, which was relatively cheap in the late 1980s after the 1987 crash), and to hedge remaining Japanese exposure through put options on the Nikkei, which were relatively cheap due to suppressed implied volatility during the bubble’s complacent final phase.

The key signal was the BOJ’s rate hike cycle beginning in May 1989: once the central bank shifted to tightening after two years of emergency-level stimulus, the bubble’s days were numbered regardless of how long the market took to notice.

Key Data Table

Key Lessons and Investment Principles

This final section distils the Japanese experience into a set of concrete, reusable principles. These are not abstract observations. They are operational rules, drawn directly from what happened between 1985 and 1995, that can be applied to portfolio construction, risk management, and macro analysis in any market environment. Revisit it whenever you find yourself reaching for a narrative to justify valuations that arithmetic cannot support.

I. The Mechanics of Bubble Formation

Principle 1: The Credit–Collateral Reflexivity Trap

When the value of collateral determines the volume of lending, and the volume of lending determines the value of collateral, you have a system that is inherently unstable.

In Japan, unrealised equity gains counted as bank capital; bank capital permitted more lending; lending inflated the assets that generated the gains. This is not a Japanese phenomenon — it is a universal feature of leveraged financial systems. The investor’s operational rule is this: track the ratio of asset-secured lending to income-secured lending in any economy. When asset-secured lending grows faster than household and corporate income for more than three consecutive years, the system is building a reflexive trap. The exit will be disorderly.

Principle 2: The Structural Superiority Narrative Is the Most Expensive Story in Markets

Every great bubble is accompanied by a compelling narrative explaining why this time is genuinely different. Japan’s version — superior management culture, keiretsu coordination, high savings rates, MITI-guided industrial policy — was more intellectually respectable than most.

It convinced not just retail punters but Nobel-calibre economists, the world’s largest pension funds, and the leadership of every major investment bank. The narrative was not entirely wrong: Japan did possess real competitive advantages. But the narrative was used to justify prices that had completely detached from any plausible stream of future cash flows.

The principle: when a market’s valuation premium is justified primarily by qualitative structural arguments rather than quantitative cash flow analysis, the premium is fragile. Narratives reverse overnight. Compound interest does not.

Principle 3: Watch the Discount Rate Relative to Asset Price Growth, Not in Isolation

The BOJ’s discount rate of 2.5 per cent in 1987–1989 was not low by historical standards in absolute terms. What made it explosive was the context: nominal GDP growth above 5 per cent, equity returns above 25 per cent annually, and land price appreciation exceeding 20 per cent per annum. The real policy rate, measured against asset price inflation rather than consumer price inflation, was deeply, catastrophically negative.

This is the measure that matters for identifying bubble conditions. Consumer price indices are useful for monetary economists; asset price adjusted real rates are useful for investors. If the policy rate minus the rate of asset price appreciation is negative by more than 5 percentage points for an extended period, you are in bubble territory regardless of what the CPI says.

II. Warning Signs and Timing

Principle 4: The Four Reliable Early Warning Indicators

The Japanese bubble, like every major bubble before and since, broadcast four signals before its collapse.

First: the velocity of new financial instrument creation accelerated — warrant bonds, convertible bonds, tokkin funds, and novel real estate securitisation structures proliferated in 1988–1989 as the conventional instruments were insufficient to absorb and deploy the tidal wave of liquidity.

Second: the marginal buyer shifted from institutions with long-duration liabilities (pension funds, insurers) to leveraged short-duration speculators (zaitech operators, jusen, individual margin accounts).

Third: fundamental analysts were systematically marginalised — Nomura’s and Daiwa’s research departments were subordinated to their sales and trading desks, and any analyst who published bearish research was sidelined.

Fourth: cross-border capital flows accelerated as domestic investors, drunk on gains, began overpaying for foreign trophy assets (Rockefeller Center, Pebble Beach), a behaviour that signals the final parabolic phase of domestic asset inflation. When you observe three or more of these signals simultaneously, reduce risk exposure immediately and accept the opportunity cost of being early.

Principle 5: Central Bank Regime Change Is the Single Most Reliable Timing Signal

Governor Mieno’s appointment on 17 December 1989 and his immediate pivot to tightening was the clearest actionable signal of the entire episode. A new central bank governor who explicitly articulates a different policy framework from his predecessor — particularly one who identifies asset prices as a policy concern — is telling you that the rules of the game have changed.

The market took three months to absorb this message. The investor who acted on day one was positioned correctly.

The principle generalises: when a central bank shifts from accommodation to restriction, believe the shift. Do not wait for confirmation from asset prices. Asset prices are the lagging indicator; the policy rate path is the leading one.

Principle 6: The Bubble Does Not End When Prices Peak — It Ends When Credit Conditions Change

The Nikkei peaked on 29 December 1989, but the real estate market did not peak until 1991, and the banking crisis did not fully manifest until 1997–1998.

This is because the deflation of a credit bubble operates in sequential phases: financial asset prices fall first (equities are marked to market daily), followed by real asset prices (property transactions slow and price discovery becomes unreliable), followed by credit quality deterioration (non-performing loans emerge with a lag as borrowers exhaust cash reserves), followed by institutional failures (the weakest lenders fail when they can no longer roll over funding).

The investor who observes Phase 1 and assumes the crisis is ‘priced in’ is making a catastrophic error. The losses in Phase 3 and Phase 4 are almost always larger in aggregate than those in Phase 1.

III. Portfolio and Risk Management

Principle 7: Diversification Across Geographies Is the Only Reliable Hedge Against a National Bubble

Within the Japanese market during 1990–1992, there was almost nowhere to hide. Equities, property, corporate bonds, and bank deposits (for those in failed institutions) all suffered. Defensive sectors fell less than cyclicals, but they still fell enormously. The only investors who preserved capital were those with significant allocations outside Japan.

An investor with a 50/50 split between Japanese and US equities in January 1990 would have experienced a blended decline of roughly 20–25 per cent by end-1990, painful but survivable. An investor with 100 per cent Japan exposure lost 39 per cent. Home bias is the most expensive behavioural error in finance, and it is most expensive precisely when it matters most — during a domestic crisis.

Principle 8: In a Post-Bubble Environment, Avoid the Value Trap

Japanese equities looked cheap throughout the 1990s by conventional metrics — price-to-book ratios fell below 1.0, dividend yields exceeded those of government bonds, and absolute price levels were a fraction of their 1989 peak. Foreign investors who bought ‘cheap’ Japanese stocks in 1993 or 1995 or 1998 watched them fall further for years.

The reason is that in a balance sheet recession, conventional valuation metrics are misleading: book values are overstated because assets are impaired, earnings are depressed because aggregate demand is collapsing, and dividends are at risk because corporate cash flows are being redirected to debt repayment. The principle: after a credit bubble bursts, do not buy assets merely because they are cheap relative to the bubble peak. Buy them only when the credit cycle has definitively turned — which means when bank lending is growing again, non-performing loan ratios are falling, and the private sector has shifted from net debt repayment back to net borrowing.

Principle 9: Sovereign Policy Response Quality Determines Recovery Duration

Japan’s fiscal and monetary response was large in aggregate but catastrophically slow and poorly sequenced. The critical error was prioritising social stability (keeping zombie firms alive, protecting bank managements, avoiding visible unemployment) over financial hygiene (forcing loss recognition, recapitalising viable banks, liquidating insolvent ones).

Sweden’s near-simultaneous banking crisis provides the counterfactual: the Swedes nationalised failing banks, forced shareholders to absorb losses, created a bad bank to manage toxic assets, and recapitalised the system within three years. Sweden recovered fully within five years; Japan took three decades.

For the active investor, the quality and speed of the policy response is the single most important variable in determining when to re-enter a crisis-afflicted market.

The checklist is simple: has the government forced loss recognition? Have insolvent institutions been resolved? Has the banking system been recapitalised with genuine equity, not accounting gimmicks? If the answer to any of these questions is no, it is too early to buy.

IV. Psychological Discipline

Principle 10: Beware the Consensus of Experts

In 1989, the overwhelming consensus among Japanese economists, government officials, mainstream Western analysts, and the global financial press was that Japanese asset prices were supported by fundamentals. The phrase ‘Japan is different’ was not a fringe view — it was the institutional consensus, backed by the apparent evidence of decades of superior economic performance.

The dissenting voices — a handful of foreign short sellers, a few maverick analysts, the occasional Barron’s column — were marginalised and mocked. The lesson is not that experts are always wrong. The lesson is that when expert consensus aligns perfectly with the financial interests of the experts themselves (sell-side analysts whose firms earn fees from equity issuance, government officials whose budgets depend on tax revenues from capital gains, central bankers whose credibility depends on economic growth), the consensus should be discounted heavily. Seek out the analysts who have no economic incentive to be bullish. Their track record will be poor during the bubble — which is precisely why they are worth listening to.

Principle 11: The Opportunity Cost of Caution Is Always Visible; the Cost of Recklessness Is Not — Until It Is

An investor who sold Japanese equities in early 1988, when the Nikkei stood at roughly 24,000, looked foolish for nearly two years as the index surged another 60 per cent to its 1989 peak. The opportunity cost was enormous and agonising. But the investor who stayed fully invested through December 1989 and held through the crash lost more than 60 per cent of their capital and did not recover their 1988 selling price until 2005 — seventeen years later.

The asymmetry is the point. In a bubble, the cost of selling too early is a missed gain that you never actually possessed. The cost of selling too late is a realised destruction of capital that you did possess. These are not symmetric outcomes. Frame every bubble-era decision in terms of permanent capital loss rather than foregone gains, and you will find it far easier to act with appropriate caution.

Principle 12: Study the Losers, Not the Winners

The financial history industry is biased toward survivors and heroes — the short sellers who got it right, the contrarians who called the top. This is entertaining but misleading. The real lessons of the Japanese bubble are embedded in the stories of the losers: the salaryman who bought a Tokyo flat at sixty times his annual income because his bank told him land prices never fall; the pension fund manager who concentrated in domestic equities because diversification felt unpatriotic; the Sumitomo banker who approved speculative property loans because his career depended on meeting lending targets; the MoF bureaucrat who delayed reform because admitting the scale of bad debts would have been politically fatal.

These people were not stupid. They were operating within incentive structures that made individually rational decisions collectively catastrophic. The principle for the modern investor: when you evaluate your own positioning, do not ask ‘Am I the smart money?’ Ask instead: ‘What incentive structures am I operating within, and could they be leading me toward the same outcome as the losers in past episodes?’ That question, asked honestly, is worth more than any amount of technical analysis.

Summary of Principles

These twelve principles are pattern-recognition tools, extracted from one of the most extensively documented financial disasters in history. No two crises are identical, but the underlying mechanics of leverage, psychology, and policy error are remarkably consistent across time and geography.

The investor who internalises these principles will not avoid every loss — but they will avoid the catastrophic, career-ending, generational losses that destroyed wealth in Japan between 1990 and 2003. That is the highest aspiration of historical financial analysis.