9  The Japanese Lost Decade (1990–2000)

9.1 What Happened

9.1.1 The Peak of All Peaks

On December 29, 1989, the Nikkei 225 stock index closed at 38,915 — its all-time high and the culmination of one of the most extraordinary asset price expansions in economic history. The Japanese property market had reached valuations so extreme that the grounds of the Imperial Palace in central Tokyo were estimated to be worth more than the entire state of California — approximately half a trillion dollars for a few dozen acres. The land under the Ginza shopping district sold at prices that implied a value greater than the entire landmass of Canada.

Both the Nikkei and the property market would fall, and keep falling, for years.

The bubble’s construction had taken a decade. The Plaza Accord of September 1985 — an agreement among the G-5 countries (the US, UK, Japan, West Germany, and France) to coordinate intervention in currency markets to depreciate the US dollar — deliberately engineered an appreciation of the yen. The yen rose from approximately 250 per dollar before the accord to 150 per dollar by 1987. Japanese export competitiveness fell. The Bank of Japan, alarmed by the economic slowdown that yen appreciation was causing, cut interest rates sharply — from five percent in 1985 to two and a half percent by 1987 — one of the lowest rates in the developed world at that time.

9.1.2 The Credit Expansion

Cheap money created an asset boom that overwhelmed the Bank of Japan’s intentions. Banks, awash in deposits and facing artificially depressed lending rates, lent aggressively against rising asset collateral. The circularity was self-reinforcing: banks lent against land; land prices rose because banks were lending against them; rising land prices increased the collateral value of existing loans, enabling more lending; more lending drove land prices higher.

Japanese corporations participated eagerly. With collateral rising, companies borrowed against their land holdings to purchase other companies, to acquire financial assets, and to fund expansion. The Nikkei tripled between 1985 and 1989. Land prices in the six largest Japanese cities rose ninety percent in four years. Stock portfolios and real estate holdings appeared to have become one-way wealth creators — and since they were collateralizing bank loans, the banks that financed them appeared to be strengthening.

The international dimension amplified the domestic bubble. Japanese companies and investors, flush with capital and encouraged by the yen’s appreciation, went on a buying spree overseas: Rockefeller Center in New York, Columbia Pictures in Hollywood, Pebble Beach golf course in California, trophy assets across Europe and America. The purchases reinforced the perception that Japanese finance had reached a level of strength that made normal valuation metrics irrelevant.

9.1.3 The Bank of Japan Tightens

The Bank of Japan raised interest rates five times between May 1989 and August 1990, taking the official rate from two and a half percent to six percent. The tightening was intended to cool the asset price inflation and address concerns about speculative excess. It worked too effectively. Asset prices that had been elevated solely by cheap money collapsed when money became expensive.

The Nikkei fell forty percent by October 1990 — losing more in ten months than it had gained in the preceding three years. The property market, more illiquid, declined more slowly but more persistently. Residential land prices in Tokyo would fall for fifteen consecutive years. Commercial property values in the major cities fell more than eighty percent from their peaks, a decline without precedent in any advanced economy in the postwar era.

9.1.4 The Emergence of the Zombie Economy

By 1992, Japan’s banking system was technically insolvent. Banks had extended loans against real estate collateral that was now worth a fraction of the outstanding loan balances. A bank that had lent one hundred million yen against commercial property now worth twenty million yen faced an eighty million yen loss on that single transaction — multiplied across tens of thousands of similar loans across the banking system.

The rational response — the response that would have maximized recovery and minimized long-term costs — was to recognize the losses, write down the loan books, recapitalize the banks through equity issuance or government intervention, and allow the banking system to resume its normal function of channeling savings to productive investment. This is what the United States would do, imperfectly but more decisively, when its own banking crisis arrived in 2008–2009.

Japan did not do this. Japanese banking culture, regulatory pressure, and political economy combined to produce a different response: the losses were not recognized. Banks continued to carry the impaired loans at their original values, extending and rolling over the credits rather than writing them off, keeping technically insolvent borrowers technically alive. These became the “zombie loans” that would define Japan’s lost decade — and zombie companies, kept alive by zombie banks, continued to occupy market share and labor that productive new entrants needed.

9.2 The Failures

9.2.1 The Paradox of Thrift at Institutional Scale

The core economic problem Japan faced after the bubble burst was what economists call the paradox of thrift — a concept first identified by Keynes but now operating at a scale that exceeded anything Keynes had analyzed. The paradox is simple: what is rational for an individual saver is destructive for the aggregate economy. When every household, every corporation, and every bank simultaneously attempts to reduce debt and increase savings, the collective result is that income falls faster than debt can be repaid.

In Japan after 1990, this dynamic operated with brutal consistency. Households that had borrowed to buy equities or real estate were selling assets to repay loans. Corporations that had pledged their land as collateral for acquisitions were selling assets to reduce leverage. Banks that recognized, even informally, the deteriorating quality of their loan books were pulling back on new lending. Every individual balance sheet decision was rational; the aggregate effect was to drain demand from the economy continuously.

This was not simply a negative wealth effect from falling asset prices — though that effect was real and substantial, reducing household wealth in ways that suppressed consumption for years. It was a structural shift in the behavior of the private sector: from net borrower to net saver, from credit-financed expansion to credit-financed contraction. Richard Koo, an economist at the Nomura Research Institute, would later name this phenomenon the “balance sheet recession” — a recession driven not by inadequate monetary stimulus but by the private sector’s overwhelming desire to reduce debt regardless of the price of borrowing.

9.2.2 The Broken Monetary Transmission Mechanism

The Bank of Japan’s standard response — cutting interest rates to stimulate borrowing and investment — failed to produce recovery because the transmission mechanism through which monetary policy normally works was broken. In a normal economy, lower interest rates reduce the cost of borrowing, incentivizing businesses to invest in new projects and households to purchase homes and durable goods. The credit thus extended flows through the banking system into productive activity.

In Japan, this mechanism failed at multiple points. Businesses did not want to borrow at any interest rate — they were using every available cash flow to reduce existing debt. Banks did not want to lend — they were quietly absorbing losses on existing portfolios and had limited appetite for new exposure. And the balance sheets of both the banking sector and the corporate sector were so impaired that lower interest rates did not change the fundamental calculation: the first priority was not new investment but debt reduction.

By 1995, the Bank of Japan had lowered its policy rate to effectively zero — a condition unprecedented in the postwar era for any major central bank. Japan had reached what economists called the “zero lower bound” — the point at which nominal interest rates cannot fall further without becoming negative. The liquidity trap that Keynes had theorized in the abstract in the 1930s was operating in concrete form in a major modern economy. Cheap money was available; no one would borrow it.

9.2.3 The Political Economy of Denial

The political economy of banking reform was dysfunctional in ways that extended the crisis by nearly a decade beyond its economically necessary duration. The Ministry of Finance, the ruling Liberal Democratic Party (LDP), and Japan’s major banks were deeply intertwined through a system of relationships — the “iron triangle” — that had governed Japanese economic policy since the postwar reconstruction. The same institutional relationships that had coordinated Japan’s postwar economic miracle now prevented the honest accounting of failure.

Forcing banks to recognize their losses would have required recapitalizing them through government injection of public funds — acknowledging both the scale of the failure and the need for taxpayer support. In Japan’s political culture, this was extraordinarily difficult. Admitting that the banks were insolvent would trigger political crisis for the ministry officials who had supervised them. The LDP governments that relied on banking sector support could not easily legislate against the interests of their principal financiers. And the banks themselves had every incentive to extend and pretend — to continue carrying impaired loans at par, rolling over non-performing credits, and hoping that rising asset prices would eventually restore solvency.

The result was a decade of zombie banking: institutions that appeared to function but had stopped performing their core economic purpose of channeling savings to productive investment. Credit growth remained persistently below the levels needed for economic recovery. New businesses could not get loans; existing zombies absorbed the limited credit available. Japan had an expensive banking system that was providing the services of a free one.

9.2.4 The Deflationary Spiral

Deflation compounded every other problem. Japan’s GDP deflator — the broadest measure of economy-wide prices — began falling in the late 1990s and would continue falling for fifteen years. Consumer prices fell persistently through the first decade of the twenty-first century.

Deflation is particularly dangerous because it creates a rational incentive to postpone spending. A household contemplating a major purchase — a car, home appliance, electronic equipment — knows that the same purchase will be cheaper next year. A business considering capital investment knows that its costs and revenues will both be lower in the future. This rational postponement reduces current demand, which reduces prices further, which reinforces the incentive to postpone. The dynamic is self-reinforcing and has no natural floor.

Japan’s deflation interacted with the debt overhang in the same way that deflation had interacted with debt in the Great Depression: it increased the real burden of existing debt even as nominal interest rates fell. A company carrying nominal debt at two percent interest faced an effective real interest rate of four or five percent when prices were falling at two or three percent annually. The cost of carrying the debt, in real terms, exceeded the returns available on any safe investment. Debt repayment remained the dominant priority regardless of the nominal interest rate.

9.3 The Response

9.3.1 Seventeen Stimulus Packages

The government’s response to Japan’s stagnation was fiscal — enormous, incremental, and largely ineffective for a decade. Between 1992 and 2001, the Japanese government launched seventeen separate fiscal stimulus packages, spending a cumulative amount estimated at more than one hundred and thirty trillion yen — roughly forty percent of one year’s GDP — on public works, infrastructure, and other government spending.

The construction was often surreal in its economic rationale. Highways to regions with minimal traffic. Bridges connecting small islands whose populations could easily use ferries. Sea walls along coastlines with no significant storm risk. Public buildings in villages too small to require them. The construction was driven not by identified economic needs but by the political imperative to spend money and by the construction industry’s political leverage within the LDP coalition. Japan became home to some of the world’s most elaborate infrastructure and some of the most economically unjustifiable.

The infrastructure spending kept unemployment from reaching depression-era levels — unemployment peaked at approximately five and a half percent in 2002, elevated for Japan but modest by international standards — and maintained a floor under economic activity. It did not restart private sector growth. Each stimulus package provided a temporary boost followed by renewed stagnation. The underlying problem — an overleveraged private sector unwilling to borrow and an insolvent banking sector unable to lend — was unaffected by roads to nowhere.

9.3.2 Zero Rates and Quantitative Easing

The Bank of Japan lowered its policy rate to zero in 1995 — the first major central bank to do so since the Great Depression. The policy rate remained effectively zero for much of the following decade, periodically raised and then cut back when the economy weakened. Japan demonstrated empirically that near-zero interest rates, maintained for extended periods, could coexist with persistent economic stagnation. The zero lower bound was not a theoretical curiosity but an operational constraint that the world’s second-largest economy was experiencing in real time.

When zero rates proved insufficient to stimulate borrowing, the Bank of Japan introduced a more radical innovation: quantitative easing, announced in March 2001. Rather than targeting the policy interest rate, the Bank would target the level of reserves held by commercial banks at the central bank, committing to purchase Japanese government bonds in whatever quantity was needed to reach the target. By expanding the quantity of central bank money in the system, the Bank hoped to stimulate lending and spending through channels that interest rate policy could no longer reach.

Quantitative easing was a new tool with no track record. It was developed, tested, and refined in Japan before being adopted — after 2008 — by the Federal Reserve, the Bank of England, and the European Central Bank. Japan invented the instrument. Its experience also demonstrated the instrument’s limits: QE could prevent deflation from worsening and could support asset prices, but it could not by itself restart private sector borrowing in the face of a balance sheet recession.

9.3.3 Banking Reform: Finally

The genuine resolution of the banking crisis required what had been politically unavoidable for eight years: forcing banks to recognize their losses and recapitalizing them with public funds. The Financial Reconstruction Law of 1998 — passed after a wave of major bank failures that finally made the denial politically untenable — authorized the government to inject capital into troubled banks and to nationalize those too damaged to rescue through recapitalization alone.

Long-Term Credit Bank and Nippon Credit Bank, two of Japan’s major banks, were nationalized in 1998 — the first nationalizations of major Japanese financial institutions since the postwar period. Other banks accepted government capital injections in exchange for commitments to clean up their loan books. The recognition of losses that had been delayed for eight years was forced, concentrated, and painful: the banks that received capital injections did so at the cost of equity dilution, management replacement, and public acknowledgment of failure.

The lesson about bank resolution was stark and would be applied — with varying degrees of speed and completeness — in subsequent banking crises elsewhere: zombie banks do not recover on their own. The longer losses are hidden, the larger they grow. The pain of honest accounting, concentrated in the short term, is less than the pain of extended pretense, distributed over a decade. Japan’s 1998 banking reform came eight years after the bubble burst. The United States, when its own banking crisis arrived in 2008, would compress a similar process into two years — still painful, but dramatically shorter.

9.4 The Legacy

Full analysis of the long-run consequences of the Japanese Lost Decade will be published as part of the complete series.