Understanding Tomorrow from Japan’s Great Recession

Up till recently, Japan had been in a 20 year recession (1990-present) that has lasted as long as it did because the asset bubble that burst in 1990 was arguably the biggest loss of wealth that a developed nation has suffered in the last 100 years.

There have been many theories as to the causes and remedies to the prolonged Japanese recession, but few have been as convincing to me the argument laid out by Richard C. Koo in his book The Holy Grail of Macro Economics: Lessons’ From Japan’s Great Recession: Japan’s recession lasted as long because the burst asset bubble severely damaged the balance sheets of its corporations, forcing them to switch from profit maximization mode of borrowing and spending (investing) to debt minimization mode of paying down debts. When corporate demand for credit diminished by 23%, the country’s GDP and money supply would have severely declined, resulting in devastating economic consequences, if not for government spending programs that effectively replaced the private sector’s decline in borrowing and spending.

Koos’s main thesis is that the economy can be in two states, called yin (post-bubble) and yang (‘normal’). Most economists have been trained to see how the economy works in the yang state, the economy that behaves just like predicted in the neo-classical models where monetary policy works, and fiscal policy is not so much effective, because of its tendency to crowd out private investment. But Koo has discovered the other side of the coin, the yin phase of a post-bubble economy that has firms trying to decrease debt, and the way they do that is by redirecting the cash flow from investment to debt repayment. Therefore, investment slumps, pulling the economy downwards and fiscal policy can and must come to the rescue and stabilize demand, and monetary policy does not work as effectively when there is less demand.

Japan’s Asset Price Bubble of the Late 1980s

In order to understand the post bubble era from 1990-2005, we should revisit the size and magnitude of the asset bubble that peaked in 1990.

This asset bubble could be seen in three areas:

  1. The five fold rise of the Nikkei 225 Stock Index
  2. The four fold rise in land value
  3. The five fold rise in commercial real estate and two fold rise in residential real estate

The Nikkei 225 Stock Index began to rise in the early 1980s, and the continued to rise more than five times the 1980 level (from 7000 to 40,000). Then, from 1990 it started a long period of decline with medium-term fluctuations, falling back to early 1980 levels.

Meanwhile, Japanese urban land prices rose throughout the 1980s, rising more 4 times their 1982 level. Prices peaked in 1991, and since then has continued to decline even to date.

Here is a combined picture of stock market and land prices overlayed together next to nominal GDP:

Bear in mind that urban land prices in Japan had already risen 238% from 1970 to 1980, and during the 1980s they rose another 400%.

Unsurprisingly, with the stock market and land prices inflating to record highs, the urban residential and commercial real estate markets inflated in lockstep. Commercial real estate values rose close to 550% their 1980 level, while residential real estate values rose 200% their 1980 level.

In 1990, the bubble’s collapse destroyed ¥1500 trillion in wealth:

This loss of wealth represented the equivalent of 3 years of Japan’s GDP.

Put in perspective, as horrible as was the US Great Depression (with GDP falling to nearly half its 1929 peak in four years, 25-50% unemployment, shares dropping to 1/8th their value), the national wealth lost was equivalent to only a year’s worth of 1929 GDP.

The bursting of the Japanese Bubble was thus the biggest loss of wealth seen in the last century, three times greater than the loss of wealth suffered during the US Great Depression.

The fact that Japan did not fall into a Great Depression with massive unemployment and falling GDP despite it losing 3 years worth of GDP, and instead the GDP continued to grow from 1990 onwards, points up to the fact that the Japanese government did something right rather than wrong in dealing with their crisis.

However, outside observers have been quick to point out the causes and solutions to end the Japanese recession.

Proposed “causes” and “solutions” of the Japan’s Great Recession:

Over the years there have been a number of journalists, financial analysts and academics that have weighed in on the probable (though problematic) causes and solutions of Japan’s Great Recession, and below is a table that summarizes the four leading ideas:

Causes of 15-year Recession Proponents Proposed Solutions  Problem with Diagnosis
1) Structural Problems Journalists, Alan Greenspan, former Prime Minister Junichero Koizumni & finance ministor Heizo Takenaka Push for supply-side reforms (e.g., privatization of public corporations, solving labor issues, weeding out “zombie companies”) because the usual demand-side monetary and fiscal stimulus had seemed to fail to turn economy around. Such economies typically have
large trade deficits, frequent strikes, and weak currency, which forces higher interest rates. Japan was the exact opposite: its interest rate fell to near zero, no labor strikes, world’s largest trade surplus, and strengthening currency. Japan’s products were in high demand everywhere but in their domestic market. The unleashing of the “Takenaka shock” (late 2002) to reform the banking sector did more harm than good, causing a nationwide drop in asset and stock prices, serving only to prolong the cleanup of the debt-ridden corporate balance sheets.
2) Banking Sector Problems Financial Analysts, Problems in the banking sector and resultant credit crunch created a bottleneck that choked off the flow of money into the economy, turning willing borrowers away. The problems at the banks are a result of the recession, not a cause. Bankers were willing lenders but there were no borrowers. More than 3800 Japanese corporations could have issued debt or equity securities on the capital markets if they were unable to borrow from banks but they did not. Instead, from 1990 to 2007, with zero interest rates, corporate bond debt fell sharply. Moreover, if there were few willing lenders but many willing borrowers, borrowers would have competed for the limited supply of loans by offering to pay higher interest rates, but bank interest also fell sharply during this period.
3) Improper Monetary Policy Academics,
Krugman, Burnanke,
Eggertsson, the Koizumi administration (Heizo Takenaka), IMF, and OECD
Krugman (1998). Deflation was the root cause of difficulty, and to counter it, there needs to be quantitative easing and inflation targets.
Burnanke (2003). Monetization of government of debt.
Eggertson (2003). Various combinations of price-level targeting and currency depreciation.
Acedemics generally thought that the prolonged economic downturn and liquidity trap of the Great Depression could have been avoided if the U.S. central bank had injected reserves more aggressively.
In a balance sheet recession, monetary policy becomes useless: since private sector is busy paying down debt to repair their balance sheets, there are no borrowers, even though interest rates remained at near zero from 1995 to 2005. BOJ did pump ¥25 trillion into the system, between 2001-2006 (tripling the amount of money available to the system), but it ended up being a non-event: the money supply available to the private sector rose only by 50%, and this happened only because of government borrowing (if Money supply were based only on private sector demand, it would have fallen by 37%, which is more than the 33% that happened during the Great Depression).
4) Poor Debt Management Journalists and media Many casual observers latched onto the view that the government must have spent the the ¥140 trillion inappropriately–after all, GDP remained stuck at ¥500 trillion, and the economy was unable to stage a healthy recovery despite massive economic stimulus in the form of investment in public works.  In reality, it was only because the government increased fiscal expenditures to the extent that it did that the nation’s standard of living did not plummet. Indeed, it is nothing less than a miracle that Japan’s GDP remained at above peak bubble-era levels despite the loss of ¥1500 trillion in national wealth (3 years of GDP) and corporate demand equal to 20 percent of GDP.


Best Idea on Cause of Fifteen Year Recession: Richard C Koo’s Research on Japan’s Balance Sheet Recession and Fallacy of Composition

In his book The Holy Grail of Macroecnomics: Lesson’s from Japan’s Great Recession, Richard C Koo argues persuasively that the situation in Japan was a balance sheet recession. Because of the massive fall in the stock market, urban land values and urban real estate, corporations were looking at tattered balance sheets with liabilities exceeding assets by many times:

When the value of the properties collapsed, but the loans used to buy them – or the loans obtained by using properties as collateral – remained, companies all over Japan suddenly found that they not only lost a lot of wealth, but that their balance sheets were underwater. A business that had acquired land valued at ¥10 Billion, for example, might have found itself with land worth ¥1billion and a residual loan balance of ¥7billion. In other words, this asset-liability pair suddenly had a net worth of ¥6billion, opening a large hole in the firm’s balance sheet.

Thus, corporations all over Japan moved collecively to repair balance sheets by paying down debt:

When a company’s liabilities exceed its assets, it is technically bankrupt, but this is not ordinary. Typically, business finds its products no longer selling as well as they used to to. But Japan did not follow. For most of this period Japan boasted the world’s largest trade surplus – implying that consumers all around the world still wanted to buy Japanese products, and that companies still had good technology and the ability to attract products. Cash flow was robust and companies were generating annual profits. Yet many had negative net worth because of the huge hole left in their balance sheets by the plunge in domestic asset prices. Thousands fell into this category. First priority is no longer profit maximization but debt minimization. Use cash flow to pay down debt as quickly as possible. A firm’s debt overhang will eventually disappear as long as it continues to reduce the liability. At that point the business will return to the profit-maximizing mode assumed by economic textbooks.

Which ultimately leads to the Fallacy of Composition: Behavior that would be right for one person (or company) leads to an undersirable outcome when engaged in by all people. In Keynesian macroeconomics, the “paradox of thrift” illustrates the fallacy of composition: increasing saving (or “thrift”) is obviously good for an individual, since it provides for retirement or a “rainy day,” but if everyone saves more, Keynesian economists argue that it may cause a recession by reducing consumer demand. In a post-bubble economy, the fallacy of composition shows up with corporations working to reduce their debts, which is a good thing when conducted individually, but becomes a bad thing when all corporations are doing it, because then it leads to a recession by seriously reduces aggregate demand.

Koo nicely illustrates the nuts and bolts of this fallacy in reference to household savings and borrowers:

In a national economy, banks and securities houses act as intermediaries to channel household savings to corporate borrowers. A household with ¥1000 of income spends ¥900 and saves the remaining ¥100. The ¥9000 that is spent becomes income for someone else, and continues to circulate in the economy. The ¥100 of savings is deposited in a bank or another financial institution, and eventually is lent to a business, which spends (invests) it. Thus the original ¥1000 is passed on to others. The economy remains in motion because every ¥1000 in income generates ¥1000 (¥900+¥100) in expenditures.

But in Japan there was no willing borrowers, even with interest rates at zero. Companies paid down debt at the rate of several tens of trillions of yen a year despite interest rates that were close to zero. In these conditions, the ¥100 in savings that our hypothetical household deposits with the bank will be neither borrowed or spent. Instead, it will pile up in the form of bank deposits, for which — in spite of the bank’s best efforts — there are no borrowers. As a result, only ¥900 of the original ¥1000 is spent to become income for someone else.

Now assume that the next household also spends 90 percent of its income, which amounts to ¥810, and saves the remaining 10 percent, or ¥90. Once again, the Y80 becomes income for others, while the remaining ¥90 simply accumulates in the banking system because there is no one to borrow it. As this process is repeated, the initial ¥1000 of income is reduced to ¥900, ¥800, ¥729, and so on, sending the economy into a deflationary spiral. The downturn in the economy depresses asset prices furhter, redoubling the urgency of business’s efforts to pay down debt. Although repaying loans is the correct (and responsible) course of action for individual firms, when pursued by all firms at once it leads to a disasterous fallacy of composition.

Japan suffered this fallacy over from 1990 to 2005.

The net result was that corporate demand fell by 22% of GDP:

In 1990, at the height of the bubble, the corporate sector was borrowing and spending 9% of GDP, or about Y41 trillion. But by 1998, firms had become net savers, and in 2003, their debt repayments amounted to 9% of GDP, or Y44trillion. Therefore, the total swing during this period was Y85trillion, or 18% of GDP. A loss of aggregate demand equal to 18 percent of GDP will send any economy into a recession, if not outright depression.

The government then rushed into replace the gap in corporate spending with fiscal spending, which dramatically minimized the crisis.

Available Solution to Balance Sheet Recession: Government Becoming Borrower and Spender of Last Resort

In US Great Depression, the economy suffered three great shocks as a result of a plunge in aggregate demand:

  1. A massive fall in the money supply: U.S. money supply shrank by 33 percent as businesses and households drew down their bank deposits to pay back loans
  2. A massive fall in the GDP: in 1933, GDP shrank by 46% from 1929;
  3. A massive rise in unemployment: 50% in some cities, 25% nationwide

Given that Japan lost three times more wealth relative to GDP in the bursting of its bubble relative to that of the US Great Depression, it would follow that Japan should have experienced a likewise massive fall in money supply and GDP, with a consequent rise in unemployment. But not so.

In fact, the GDP continued to rise after the bubble burst:

The fact that the GDP continued to rise after losing ¥1500 trillion in wealth, or 3 years of GDP, or three times more damage than the US Great Depression, was nothing less than a miracle.

The thanks must be heaped on the shoulders of the government and its willingness to borrow and spend to the same amount that the private sector had decreased its own borrowing and spending.

As can be seen in Financial Surplus /Deficit Sector chart above that  in 1990, the government was in surplus of 3% while the corporate sector was in deficit of -12%. Then in 2003, the corporate sector was in surplus by 10%, while the government sector was in deficit by -10%.

You can see this role switching take place in another tabular comparison of Japanese bank balance sheets:

You can see that in the space of 15 years, banks had dropped off the credit extended to corporations by ¥-99.4 trillion, because corporations were busy paying off loans, and instead credit was increased to the public sector by ¥133.5 trillion when the government became the borrower of last resort. The net result was that Money supply (M2+CDs) increased ¥116.8 trillion over those years, thanks to the increased borrowing of the government. If the government had not increased its borrowing, the money supply would have severely contracted.

The table below shows that different money aggregates over the same period:

Richard C. Koo explains the above chart:

Private sector credit is outstanding credit and loans extended to financial instutitions to the private sector. as noted, bank deposits cannot increase without a corresponding rise in bank lending. Under ordinary circumstances, therefore, private-sector credit should be the key determinant of the money supply. But by June 2006, private sector credit had fallen to 95 from 100 in 1990. This means that if the money supply was determined solely by private-sector demand for funds, Japan’s money supply would be 95 instead of 150, or about 37 percent less than the current money supply. For the past fifteen years, in effect, Japan’s economy has been experiencing the same difficulties faced by the U.S. during the Great Depression, when the money supply shrank by 33 percent. Japan has avoided falling into depression-like conditions only because the government has continued to borrow and spend.

Thus, the private and public sectors had switched roles: the government was moving into debt at the same pace that the private sector was paying off its debts. This move alone kept the GDP and money supply rising. Japan’s GDP and money supply grew at a steady pace, instead of falling off a cliff, and unemployment was kept in check as well.

The financial debt of the goverment sector mounted sharply, leaving in its wake the national debt we face today. But it was precisely because of these expenditures that Japan was able to sustain GDP at above peak-bubble levels despite the drastic shift in corporate behavior and a loss of national wealth equivalent to three years of GDP. Government spending played a critical role in supporting the eocnomy, and only through these annual stimulus packages was the government able to prevent a deflationary gap from emerging (in economics, a deflationary gap is the difference between potential and actual GDP).

If the government had simply stood by and watched, the economy will fall into a kind of catestrophic deflationary spiral seen in the U.S. between 1929 and 1933. To stop this vicious circle, the government has only one option: it must do precisely the opposite of what the private sector is doing. In other words, it must borrow (and spend) the savings that the private sector can no longer use. This is what Japan ulimately chose to do, and it is why the money supply did not contract and GDP remained steady about ¥500 trillion despite the loss of ¥1500 trillion in national wealth and a decline in corporate demand totalling more than 20 percent of GDP.

Razor Note:
Government spending only works during a Balance-Sheet Recession when the private sector is paying down debt, and does not work in a regular Business-Cycle Recession when private sector is maximizing profit. From 1940 through the 1970s, most of the leading nations adopted an active, Keynesian approach to fiscal policy, believing that any recession could be averted by the proper administration of fiscal policy. In the end, these policies resulted only in inflation, higher interest rates, crowding out of private sector investment, and a misallocation of resources. Crowding out of private sector investment by government is the biggest drawback of fiscal policy, but it cannot occur during a Balance-Sheet Recession because the private sector is busy paying down debt.

Monetary Policy was a Non-Event

Bank of Japan (BOJ) initiated two monetary policies to deal recession, both of which had limited effect:

  1. ZIRP (Zero Interest Rate Policy): From 1990 to 2006, interest rates were near zero. These rates were lowered to such levels in order to lure the private sector to continue to borrow and spend. The problem was that because it was a balance sheet recession, the private sector was more interested in paying down debt than it was in borrowing more debt. Corporate demand for loans fell by 22% over fifteen years, despite near zero interest rates.
  2. Quantitative Easing: More than ¥25 trillion in excess reserves was pumped into the banks under QE. Banks may lend money against these reserves, but only about ¥5 trillion in reserves is actually required under the law to sustain the current money supply and loans outstanding. Consequently, the additional ¥25 trillion in reserves could potentially support a money supply six times as large as the existing one. A 500 percent increase in the money supply translates to a potential 500 percent increase in price levels. A central bank charged with keeping inflation in check cannot countenance such a scenario. So it ended QE as soon as a rebound in private sector demand was confirmed.

In a balance sheet recession, monetary policy becomes useless: since private sector is busy paying down debt to repair their balance sheets, there are no borrowers, even though interest rates remained at near zero from 1995 to 2005. When there are no borrowers, the money supply cannot grow, because liquidity injected by the central bank cannot leave the banking system. Instead, there is an antagonist force at work; when the entire private sector is paying down debt, bank deposits lessen and the money supply contracts. During the great depression, U.S. money supply shrank by 33 percent as bussinesses and households drew down their bank deposits to pay back loans.

BOJ did pump ¥25 trillion into the system, between 2001-2006 (tripling the amount of money available to the system), but it ended up being a non-event: the money supply available to the private sector rose only by 50%, and this happened only because of government borrowing (if Money supply were based only on private sector demand, it would have fallen by 37%, which is more than the 33% that happened during the Great Depression).  Note: as long as there are no borrowers, no amount of QE will harm the economy. But if the policy is continued after borrowers return to the market, it can lead to dangerously high money supply growth and inflation.

Under QE, the BOJ supplied liquidity to the market by purchasing government bonds held by commercial banks, and crediting money to their current accounts. This process was repeated until the aggregate value of bank’s current accounts had risen to more than ¥30 trillion. To terminate the policy, this process had to be reversed. In theory, this would involve the bank selling government bonds to commercial banks to absorb the excess funds in their current accounts.

In terminating QE, the ¥25 trillion in surplus funds that the Bank of Japan sought to mop up was already sitting in commercial bank’s current accounts with the central bank, which pays no interest.  Facing an absence of private sector borrowers, the commercial banks could do nothing with these funds. Because the banks did not need to raise funds elsewhere, the operation had none of the negative impact of normal tightening operation, and interest rates did not rise significantly.

2005 marks the beginning of new Japanese recovery

For the most part, Japan’s unfolding recovery from the burst asset bubble of 1990 started to happen in 2005 when firms finally stopped paying down debt and began to borrow again for the first time in 15 years.

Net repayments began to decline in 2004, and by the end of 2005, they had fallen to zero for the corporate sector as a whole.  Now companies have started to borrow again in what represents a historic turning point after 15 years of recession. Now debt has fall back to 1956 levels and leverage has come down to the Western norm.

The 1990s were by no means a “lost decade” because Japenese companies made huge strides to acquire a clean bill of financial health, signalling an end of the bubble aftermath and high leverage.

As of 2012, the Japan recession is still continuing because Japan has been hit indirectly by the newer financial recessions in US and Europe.


I agree with Koo that Japan experienced a balance sheet recession and that the government  had “rightly” stepped to become the borrower and spender of last resort. He nails the causes of balance sheet recession: excessive private sector debt following the burst asset bubble, which forces corporations to pay down debt, which in turn reduces spending and aggregate demand. The government fills this gap in demand by becoming the borrower and spender of last resort, which prevents the economy from falling into a deflationary depression with negative GDP growth and high unemployment.

However, Koo understates the frequency of asset bubbles, the cumulative consequence of repeated fiscal spending designed to mitigate their damage, and how the government is ever able to climb out of its debt burden.

Koo states that balance sheet recessions are rare and it can take decades or even centuries for them to reoccur, which allows the government sufficient time to pay down its fiscal debt. But he fails to provide sufficient data to support his claim that balance sheet recessions are rare due to debt aversion. The private sector has two parties (corporations and individual households) that might both have to move into debt aversion in order to avoid future asset bubbles. If only one party over-leverages and learns from their mistake, the other party can still giddily blows their own bubble. This is what happened in the US. Corporations were slammed by the burst internet bubble of 2001 and possibly became more debt averse, but that did not stop individual households, enticed by low interest rates and greed, from borrowing beyond means to buy houses. Moreover, given the number of boom and bust cycles that have occurred in the last 100 years, it seems that debt aversion is a more temporary phenomomnen that is forgotten in less than a generation.

Besides thinking that there is plenty of time to pay off the debt, Koo believes that large fiscal debts can be handled just like US and the world did after WWII, by growing their way out of debt. But we are no longer living in the post war boom years.There is no Europe and Japan to rebuild, no baby boom generation, no auto industry and interstate highway system to create, no suburbia to build. These are the forces that allows America and other economies to grow their way out of WWII debt. But this time there are opposite forces in play. The developed world has a fully developed economic infrastructure, an aging population demanding massive health care and pension spending, and structurally rising costs of imported resources like oil and other commodities. All these are forces making it harder for developed economies to grow their way out of their respective balance sheet recessions.

We are living in economies of slowing growth and repeated bubbles. Most modern economies are piling up higher and higher debt levels with few or rare instances of the debt diminishing for any given year. Governments simply do not have the discipline to pay down debt once a financial crisis has been averted. No one would know when that is, except in hindsight. Moreover, it is a lot easier to pump up spending than it is to reign it in, just as it is easier to gain weight by eating more than the much more difficult task to take it off.

Elected governments are happy to borrow and spend to prime the pump in the down cycle, but no democratic government has proven able to increase taxes and pay down debt in the up cycle, so government debt only rises and is never paid down. In bad times the people want the government to spend on infrustructure and jobs, and in good times, they demand higher spending on pet programs, and politicians end up caving in to these demands rather than pay down public debt. Eventually interest payments consume such a large portion of government revenues that programs must be cut. But as Koo notes, this further reduces incomes and spending and tax revenue so the government budget ends up in worse shape, not better, by going the austerity route.

As we have seen, fiscal spending can provide a temporary solution a balance sheet recession but it presents its own set of problems. The magnitude of Japan’s public debt remains an issue of concern amongst investors and economists. Whether it will hinder Japan’s future growth remains unknown, but such high debt levels do continue to put a Japanese recovery at risk. There may be a time that the portion of debt burden that is represented by short-term treasury bonds (which will have to be refinanced frequently) puts any government at increased interest costs and additional debt as interest rates increase in a recovering economy.

Lastly, why does no one seem to acknowledge the positive and necessary attributes of debt-burdened households or corporations spending less and paying down debt? Why does everyone seem to fixate on the impact to the economy? It is absolutely vital and necessary that debts are paid and the overall debt-burdened is kept in check. Further, as governments (US & abroad) continue to expand the safety net to debtors, there will be less and less “learning” from balance sheet recessions, as the prudent parties will be paying for the acts of the debtors.


Leave a Reply

Your email address will not be published. Required fields are marked *