10 11 12 werner

How to Prolong a Recession

Seven Lean Years

By mid-1995, Japan’s recession had already lasted far longer than most economists had predicted. Analysts and investors who had been holding out for a full-blown recovery—and there were many in the first half of the 1990s— became gloomy as they surveyed the economy. The yen had risen to around ¥80/$—unthinkable for many just half a year earlier. Exporters were under pressure, demand in the economy faltered, production growth slowed, inventories built up, and firms cut costs to stay in business. Increased competition and deregulation put further deflationary pressure on the economy. Price destruction made consumers postpone purchases as the layoffs pushed up unemployment to a new postwar high. Meanwhile, the banking system was weighed down by bad debts.
Unexpected by most observers, the economy staged a sudden recovery in 1996, growing by around 4 percent. But this was not sustained: The economy slumped again in 1997 and 1998. It seemed to take the yen with it this time. It collapsed to nearly ¥147/$ on June 15, 1998, around 80 percent weaker than its peak in April 1995. Yet the weak yen did not help the Japanese economy. To the contrary, most analysts now considered it as a sign of weakness and of capital flight from a country that seemed headed toward economic meltdown.
Attempts by the authorities to stimulate the economy had been to no avail: The downward spiral was accelerating and had turned into a vicious cycle of contracting demand, falling prices, squeezed companies, and further contracting demand. Few economists thought about recovery.
Yet most observers were once again surprised by a sharp recovery of the economy in 1999 and a more than 50 percent rise in the Tokyo stock market. But the stock market peaked in the first quarter of 2000 and both market and economy slumped once again in mid-2000 and 2001. By early 2002, most commentators had given up hope of a speedy recovery. There had been too many false starts. Each time the economy recovered, it seemed to sink back into recession soon after.
Who Is the Perpetrator?

Since 1991, the government and the Ministry of Finance have been trying to boost the economy by using interest rates. The Bank of Japan lowered the official discount rate (ODR) ten times in the 1990s, beginning with the first reduction in July 1991, before which it stood at 6 percent. Prior to September 1993, it was lowered seven times, reaching a historical low of 1.75 percent. The ODR was further lowered to 1.0 percent in April 1995 and to 0.5 percent in September 1995. In October 1995, the uncollateralized overnight call rate (officially declared the “target operational rate”) was “guided” below the ODR for the first time (at around 0.47 percent). Three years later, in October 1998, the Bank of Japan lowered the call rate further to a new record low of 0.33 percent. In February 1999, it fell to 0.1 percent—what at the time was called a “zero interest rate policy.” After a temporary hike in August 2000, the call rate was lowered again to 0.12 percent in March and 0.02 percent in April 2001. In September of that year, the ODR was lowered to 0.1 percent and the call rate to 0.003 percent. After that, it fell to a wafer-thin 0.001 percent.
Due to the apparent failure of monetary policy, the politicians had been pushing for Keynesian fiscal stimulation. During the 1990s, over a dozen large-scale government spending packages had been implemented, amounting in aggregate to over ¥145 trillion—also apparently to no avail.
If both the monetarist and the Keynesian prescriptions did not work, many economists thought, what was there left to do? They started to listen to those voices that argued that the recession was due to Japan’s economic system. The only way out was to introduce deep structural changes, such as deregulation and opening of markets. By 1998 a broad consensus had emerged in favor of a historic structural transformation. Business leaders, politicians, and, surprisingly, even members of the bureaucracy argued that fundamental change was necessary.
Such a conclusion had become very tempting, especially as the U.S. economy went from strength to strength during the 1990s. High economic growth, record low unemployment, low inflation, and rising asset prices seemed to usher in a new economic era in America. This was said to have been the result of productivity gains stimulated by free markets. Since 1996, the annual G7 summit had become a platform for the U.S. president and his treasury secretary to assert the superiority of U.S.-style capitalism. If a country wanted to be successful, they would frequently proclaim, deregulation, liberalization, and privatization were necessary. With Japan and the rest of Asia in a slump, U.S. pressure, and with it the pressure of international organizations, mounted for them to abandon their old economic systems and introduce the successful model demonstrated by the United States.
It had been conveniently forgotten that only a decade earlier the tables were turned. In 1991, the U.S. economy was in recession. U.S. banks had lent too much to real estate speculators in the 1980s, and by 1990 bad debts were threatening even the largest U.S. banks. The banks had become risk-averse, less able and less willing to lend. As a result, small firms, dependent on bank funding, received insufficient funding. As they laid off staff, demand slumped. With credit creation shrinking, the economy contracted in 1991.
At the time, pessimism about the U.S. economic structure was about as widespread as optimism a decade later. Many authors were even advocating that the United States introduce the Japanese system, which, shortly after the peak of its 1980s bubble, seemed superior. In 1991, most commentators expected Japan to overtake the U.S. economy by the turn of the millennium. The twenty-first century was going to become the Japanese century.1
What we learn from this is that the assessment of what constitutes a successful economic structure is not independent from the business cycle. During times of boom, commentators are quick to credit the economic system. A slump is seen as proof that the economic structure is at fault. In actual fact, both are merely reflections of the business cycle. And that is determined by credit creation.
Government Spending Ineffective

During much of the 1990s, however, most observers analyzing Japan argued that credit growth was slow only because there was no demand for loans in the economy. Their policy prescription: Domestic demand had to be boosted by government spending, and then loan demand would also rise. For a decade, the government followed their advice, thus boosting government debt to historic levels and ruining Japan’s fiscal virtue.
Yet we saw already in chapter 4 that the credit market is supply-determined. Money is different from apples and oranges—there is always demand for it. There are always enough entrepreneurs who would like to borrow money and invest in risky projects. Potential credit demand is so large that if banks raised interest rates to equalize demand and supply, the interest rates would rise enough to disqualify conservative and sensible investors, leaving only the high-risk entrepreneurs as bank clients. That is why banks keep interest rates below what would be the market clearing rate and instead select their borrowers: Banks ration credit. The mac-roeconomic result is the virtually permanent supply-determination of the credit market.
Meanwhile, fiscal spending could not boost demand, because it does not create money. It transfers purchasing power into the hands of, for instance, the construction industry, which receives large-scale government orders. Many economists simply add up these amounts of extra government spending and expect that GDP will be boosted by that amount.2 Again, the fallacy of composition has struck, which is due to the neglect of the government’s need to fund its fiscal expenditure. The question is how the fiscal spending is funded. In the case of pure fiscal policy, dominant during the 1990s, the Ministry of Finance would issue government bonds to raise the money. Thus the money for the fiscal stimulation of the private sector is taken from the private sector itself. Investors, such as life insurers, have to pull the money for the purchase of government bonds out of other investments. We see that fiscal policy does not create new purchasing power but merely reallocates already created purchasing power. Pure fiscal policy is largely growth-neutral.3 Indeed, over the 1990s it has been shown that for every yen the government spent in fiscal stimulation, private demand shrank by one yen.4
Put simply, credit creation determines the size of the economic pie. Fiscal policy determines how that pie is divided up between the private sector and the government. For unchanged credit creation, increased fiscal spending must therefore reduce the amount of purchasing power available in the private sector. Hence, without an increase in credit creation, the private-sector share of the national income pie must shrink (quantitative crowding out). In particular, the main customers of banks, the small firms, suffered from the credit crunch for most of the 1990s. That depressed consumption and hence GDP.
Print Money

For more net new transactions to take place, more purchasing power is necessary. This allows an increase in the economic pie. The necessary and sufficient condition for an economic recovery is the creation of new purchasing power. Purchasing power is created by the banking system and the central bank. Policies to create a recovery therefore had to aim at increased credit creation of either one or both of these. Even if policies to help banks were slow in showing results, this would not prevent an immediate recovery—if the central bank fulfils its mandate and creates new purchasing power instead. Since 1992, a recovery in Japan could have been triggered at any time. A sufficient condition would have been for the Bank of Japan to switch on the printing presses.5
Inflation would not have resulted from such money creation. If the economy were operating at full capacity, printing too much money would indeed lead to inflation. That is why under circumstances of deflation and unemployed resources, printing more money will increase demand and reduce deflation. Inflation occurs only once the economy has expanded sufficiently for all factors of input to be fully used, for unemployment to be reduced to a minimum, and for all factories to operate at full capacity; on top of that, demand is boosted beyond this full capacity. In other words, once an economy is fully reflated and growing at the maximum potential growth rate, the central bank would have to slow the printing presses. But in Japan’s predicament of the 1990s, there was no such worry.
Money Printing Increases Demand

Of course, “printing money” does not merely mean an increase in paper money. We have seen that nowadays the majority of money takes the form of “book money” or, more correctly, “computer money.” The central bank can increase that at any time, without limit, by simply buying assets from the private sector and paying with newly created credit. Economically speaking, it does not matter what the central bank buys. It could buy neckties, toothpaste, or real estate.
The Bank of Japan could, for instance, go out and purchase the house of Mr. Harada. It could entice him to sell by offering a price above the market rate. That would not be a problem for the Bank of Japan, because it could print the money, or, more precisely, create new purchasing power that previously did not exist. It does not matter to Mr. Harada whether he gets the money in the form of paper currency or a BoJ transfer to his bank (which simply means that his bank will get a credit in its books with the Bank of Japan and he in turn will get a credit in his books with his bank). Mr. Harada now has more purchasing power available, and he most likely will use at least a small part of it to buy something else— another house, for example. He transfers the newly printed cash to the seller. That person then goes out and buys something from someone else, and so on. Suddenly, more economic transactions take place, and the reverberations are felt throughout the economy. Increased demand has been created by the BoJ out of nothing.
Central Bank Credit Creation

In reality the Bank of Japan does not buy much real estate (although it has acquired many real estate properties, such as houses, clubs, and recreation facilities for the use of its staff). In order to inject large amounts of money in a short time, the central bank tends to buy government bonds, bills, and commercial paper issued by corporations. When the Bank of Japan buys such paper in the markets, it helps the economy just as much as if it purchased a piece of land.
This can easily be visualized: With the banks paralyzed by bad debt, many medium-sized and small firms are suffering from the credit crunch. One way out is for them to issue debt certificates, such as commercial paper or corporate bonds. This paper can then be bought by the Bank of Japan, which in exchange hands over new yen notes to the firms. Banks may act as intermediaries by first discounting the bills, which the central bank rediscounts. But this does not change the analysis. As a result, the firms are able to receive money that did not exist before. Smaller firms can also receive the funds indirectly, in the form of trade credit from larger firms that issue such debt paper. The result would be the same. When the banks are not doing their job of lending and creating new money, the Bank of Japan can step in and act as a banker to the nation.
There is another way to illustrate how simple “money printing” helps the economy. We have found that pure fiscal spending funded by bonds that are bought by investors cannot stimulate new economic growth. No new purchasing power is created; old purchasing power is merely diverted. But fiscal policy can be made effective if it is backed by credit creation. If the government bonds are not sold to private investors but are bought or underwritten by the central bank, then credit creation increases, and the fiscal stimulation serves to inject this new money. What makes the difference in that case is not the fiscal spending but the action of the central bank to create money. Alternatively, the government can switch funding of the public sector borrowing requirement from bonds to simple loan contracts from banks.
Print Money and Create Parks

London boasts 26.9 square meters of park space per capita, New York 29.3 square meters, and Paris 11.8 square meters. Tokyo, however, comes last in a long list of the world’s major cities, with 5.3 square meters per head.6 Moreover, Tokyo has the least park space of all the big Japanese cities. A good way to boost demand, stimulate the economy, invigorate the real estate market, and at the same time increase the quality of life in Tokyo would be for the Bank of Japan to print money and buy up land all over Tokyo to turn into parks and facilities that can be used by the public.7 Printing money to boost park space per head to the relatively low Parisian level could, depending on area and price, inject almost ¥70 trillion into the economy—not dissimilar to one estimate of the size of the bad debts. Of course, other, even more productive uses could be made of newly printed money. Facilities could be established that address public needs, such as an improved medical system or welfare infrastructure for the elderly. In a sense, the recession of the 1990s represented an opportunity to print enormous amounts of money and use them in a beneficial way without what would normally be the price to pay, namely, inflation. Even direct handouts by the central bank to each taxpayer—for instance of ¥2 million each—would be feasible, without any costs. They could simply be considered refunds from the central bank (for failing to deliver the goods).
All these examples serve to demonstrate just how easy it would have been to create an economic recovery as early as 1992 or 1993 to the benefit of Japan and beyond. Millions of unemployed would have found jobs. It was entirely feasible to create a recovery throughout the lost decade of the 1990s if the right policies had been taken.
History Proves That It Works

Printing money to boost demand is not just a nice theoretical idea. It has been tried and tested. We have already seen how the BoJ under Ichimada and the government’s Economic Stabilization Board successfully reflated Japan’s economy right after 1945, when the banks were in far worse shape than in the 1990s, and when the economy had been devastated by carpet bombing. There are other examples, for instance, the 1930s, when the world was gripped by the Great Depression, which triggered the structural transformation of Japan. Just as in the 1990s, the problem was that banking systems shut down, first in the United States, then Germany, Japan, and other countries.8 As we saw in chapter 4, banking systems are fundamentally fragile because they are based on what many would consider fraud: Banks do not actually have the money that they guarantee is being deposited with them. This becomes clear particularly when the money is lent out simultaneously over ninety times for unproductive, speculative purposes and hence in aggregate there is little hope of its being paid back. U.S. banks during the 1920s, for instance, had lent too much to speculators, driving up stock and land prices.9
However, Germany and Japan were the first countries to pull out of the Great Depression. While the U.S. central bank failed to reflate and allowed many banks to go bankrupt, the German and Japanese central banks started to print money sooner. Although it is often said that it was fiscal policy that stimulated the Japanese and German recoveries, it was in fact the creation of new credit that made fiscal policy effective. There is no known example of a country where aggressive central bank money printing did not stimulate demand. Whenever a credit bust follows an excessive credit boom, a recovery happens only after the banks or the central bank expands credit creation again.
Solving the Banking Problem

While the central bank has to kick-start the economy, simultaneously the problem in the banking system needs to be solved. After a decade of failed attempts, it may be appealing to think of this bad debt problem as being complex beyond imagination, but in actual fact it is an issue that could be solved immediately, at zero cost to anyone. And it should have been solved long ago. While banks are burdened with significant amounts of bad debt, they will not fulfill their role of lending and creating money. The only solution is for banks to write off their bad debts and delete them from their books. Since accounts are made up of assets and liabilities (loans are assets for banks, and deposits are liabilities; equity is on the liability side) and the two must always balance, simply deleting the bad assets will not do. Liabilities would exceed assets—which is one definition of insolvency. So in order to be able to write off the bad debts, the banks need to put something else on the asset side of their balance sheet. These are called reserves, and they are put in place of the hole that the write-offs would create in the balance sheet. Put simply, the banks need money.
So what we need to do is to give money to the banks. We are relieved to find that the problem is not more complicated than that. For money, as we know, can be created, either by banks themselves or by the central bank. The simplest solution is therefore for the Bank of Japan to print money and give it to the banks.10 Of course, the Bank of Japan would like to obtain something in return, in order to list it on the asset side of its own balance sheet. These are details. The banks could issue a debt paper, which states that they borrowed the money from the Bank of Japan (for instance, at zero interest). Or they could issue new shares, such as preferred shares, which the Bank of Japan would then buy. Alternatively, they could transfer ownership of the land they own to the Bank of Japan.
Bad Debt Problem Can Be Solved in a Day

If it so wished, the Bank of Japan could have solved the bad debt problem in its entirety within one morning. What it needed to do was to purchase all bad debts from all banks at face value, and pay for them through the creation of new money. The banks would have welcomed this idea, for they would have received cash in excess of market value for loans that had gone bad. What about the Bank of Japan? Would it not suffer huge losses? Actually, no. By purchasing the bad debts at the nominal face value of the 1980s, the central bank would appear to make a loss (since their market value is now much lower). However, the central bank, having a license to print money, always makes a gain: It has zero fund-raising costs and can obtain something that has some value (even if only 10 cents on the dollar) for free. The true cost for the Bank of Japan is zero. It creates the money out of nothing and therefore always gets a good deal. In practice, transferring the cash to the banks does not even involve printing presses, as most of the money is created online in the Bank of Japan’s computers. Since the banks all have accounts with the BoJ, it could use its electronic transfer system to rid the banks of all the bad debts within seconds—instead of taking over a decade.
There are of course many variations on this theme. For example, if the central bank is reluctant to show any such assets on its balance sheet, a government institution could be used that buys the bad debts from the banks and itself is funded by issuing bonds or bills to the central bank. The possibilities are there, if there is a will to solve the bad-debt problem.
Not only would there be no costs to the Japanese central bank, more importantly, there would also be no costs to the economy or society at large. If, instead, government money (i.e., tax money) is used to bail out banks, then the taxpayers will have to refund the money in the future. If the BoJ simply prints the money, taxpayers do not incur a liability. Since the economy has been in the grip of deflation, this would also not produce what is normally the cost of excessive credit creation, namely, inflation. In this situation, at best we would get less deflation— which would be a good thing.
How to Stimulate Bank Credit

Alternatively, money could be transferred to the banks by helping them make sizable profits. There are several ways in which this can be achieved. One way is for the central bank to corner a market to help the banks—in effect creating a mini bubble in a certain market in which banks invest heavily, providing large profits for them. This turns out to be a relatively common technique by central banks to help their banking systems. Another, more transparent way would be to use the banks’ ability to create credit to fund fiscal spending. As we saw, the main reason fiscal spending has been ineffective is that it was not linked to greater credit creation. By borrowing from the private sector through bond issuance, quantity crowding out occurred. This would not be the case, however, if the government changed its method of funding the public-sector borrowing requirement. Instead of issuing bonds, it could enter simple loan contracts with the banks. The banks would be eager to lend, as the government is a zero-risk borrower. Unlike the bond market, bank credit creates new purchasing power. The money spent by the government would not be withdrawn from the economy but would be newly created—thus addressing the cause of the recession, namely, a lack of credit creation. No crowding out would occur.11 Net demand would increase. Such a method is of course particularly useful at times when the central bank refuses to monetize fiscal spending by buying government bonds. Finally, if the Bank of Japan really wanted to create a recovery, it could also have used its window guidance mechanism to simply “guide” bank lending higher.
Moral Hazard Principle

Should the central bank bail out the banks and create a recovery? Economists are concerned with an incentive problem, referred to as “moral hazard.” To avoid it, those who create problems should know that they face some kind of penalty. If banks expect to be bailed out, there would be no incentive for them to avoid reckless lending. This principle already tells us that taxpayers should not be made to fund any bank bailout, because the bad debts were not their respsonsibility.12 Thus it is often argued the Japanese banks should not be bailed out at all. But there are also problems with this argument. First, it is now almost twenty years too late. We should have worried about it in the early 1980s, when banks engaged in excessive lending. However, since the mid-1990s, the main problem has been too little lending. Thus one could bail out the banks on this occasion, and after the bailout make suitable institutional changes to avoid future reckless lending of the type that occurred in the 1980s. To do that properly, however, one would have to examine closely just why the banks were lending so aggressively during the 1980s.
Second, the above argument assumes that the banks were the main perpetrators of the lending-driven bubble of the 1980s. Indeed, it has been shown that bank lending explains the bubble.13 However, we have not yet established just why the banks were lending so much.
How Much Money Has the BoJ Created?

Many of the above ways to boost bank credit would have taken some time to implement. Since Japan’s economic downturn became painful for small firms and hence the majority of the Japanese people from 1992 onward, already at that time the fastest method to kick-start the economy should have been adopted. In Figure 9.2 in chapter 9 we measured bank credit creation in the real circulation (i.e., without the bubble sectors) and saw that it had fallen sharply from 1990 onward. A year later, nominal GDP growth also fell. Credit creation remained minimal and even turned negative in late 1994—resulting in negative nominal GDP growth in early 1995—for the first time in postwar history; indeed, the first time since 1931.14 In such circumstances, where bank credit creation is collapsing, it is the duty of the central bank to counteract this and do what banks are not doing—step in and create more credit.

Figure 10.1  Bank of Japan Credit Creation (as measured by Profit Research Center’s Leading Liquidity Index)

Source: Bank of Japan; Profit Research Center Ltd.
No doubt, the BoJ has been holding in its hands the key for an economic recovery. Let us therefore check just how much money the Bank of Japan has been creating during the 1990s. First, we need to measure its credit creation correctly. Since the central bank has to buy something from the private sector when it creates money, and since it has to sell something to neutralize purchasing power, a more accurate measure of central bank credit creation is found by simply adding up all its transactions in all markets.15 Many economists, when analyzing the central bank, confine themselves to adding up what the Bank of Japan calls “short-term money market operations,” because they are conveniently announced on a daily basis. However, these operations do not represent the total net credit creation of the BoJ. Instead, the net credit creation of the BoJ is best measured by adding up all its transactions. Figure 10.1 shows one such measure—based on figures released by the central bank. Assuming that the Bank of Japan has supplied accurate data, this should provide a reasonably useful measure of the Bank of Japan’s credit creation.16
Aggressive Money Printing in 1998 and 2001

We see from this chart that in the 1980s, from around 1986 onward, the Bank of Japan stepped up its credit creation significantly. This was to be followed by a sharp reduction in credit creation: In 1992, our index fell into negative territory. This implies that the Bank of Japan was withdrawing purchasing power from the economy. It engaged in the opposite of credit creation. Credit creation remained minimal still in 1993; in 1994 it rose somewhat, only to fall sharply again and turn negative in March 1995. From around May 1995 until early 1997, credit creation rose, but fell once more later in 1997. While the index has been on an up trend, this was marred by frequent periods of significant credit reduction. In other words, for most of the 1990s, the BoJ did not print money aggressively, or sufficiently for a lasting recovery.
In March 1998, the Bank of Japan suddenly boosted credit creation sharply. Our index reached the highest level since January 1974, when the BoJ was supplying the funds for the 1970s real estate bubble. This was good news for Japan and, indeed, was followed by a sharp economic recovery in 1999, and a stock market rise exceeding 50 percent. Unfortunately, the central bank turned off the taps with almost equal vigor in 1999, and went beyond this by actively withdrawing money from the economy for much of that year. This could not fail to end the nascent recovery. Indeed, by 2001 the economy was once again in the midst of another deflationary round, with demand falling and prices declining faster again. In June 2001, the central bank changed its monetary policy once more and sharply increased the quantity of its credit creation.17 As expected, this contributed positively to the economy in 2002, despite the severe slump in 2001 and early 2002. But will any recovery last? The Bank of Japan’s policy so far has not been one that is aimed at creating a sustained economic recovery. Instead, we had temporary minor recoveries within one long recession.
Why Did the BoJ Not Fully Reflate?

It is often said that the job of central banks is to counteract business cycles and create a stable economy. That is also what the proponents of the U.S. Federal Reserve argued in the early twentieth century when they wanted to persuade Congress that a central bank was necessary. However, upon analyzing the monetary policy of the Bank of Japan over the past decade, it becomes clear that Japan’s central bank did not engage in countercyclical monetary policy. To the contrary, it created more purchasing power at times when there was already too much—the late 1980s—and it created far too little purchasing power, even decreased purchasing power, at times when there was already too little and a credit crunch squeezed the entire economy—the 1990s. Why was the BoJ following such a policy course? It is time to take a closer look at just what the BoJ has been up to.

The Battle of the Yen

The Powers of MoF

Until 1998, Japan’s Ministry of Finance controlled—legally speaking—most aspects of economic life in Japan. It controlled taxes through the National Tax Agency and the tax bureau, the government budget through the budget bureau, government bond issuance via the finance bureau, foreign exchange intervention and international capital flows through the international finance bureau, imports and exports through the customs bureau, securities transactions via the securities bureau, and the banking sector via the banking bureau. For good measure, the Ministry of Finance prints government documents and even the paper money issued by the Bank of Japan at its printing bureau (the central bank pays a small fee for this, but the ministry merely acts as a printer and has no say over the quantity printed).
MoF’s power was exerted through formal rules as much as through informal guidance; some were published as tsutatsu (administrative ordinances), while some were unwritten, extralegal “traditions” handed down over generations of bureaucrats. MoF influence seemed to permeate economic and political life, because former finance bureaucrats obtained influential posts as heads of government financial institutions, public corporations, private banks, securities houses, large firms, or as politicians. A substantial number of Diet members are former finance ministry bureaucrats.
The Law Said: MoF Is in Control

Although monetary policy was implemented by the Bank of Japan, the laws gave the oversight to the Ministry of Finance. Legally, the highest decision-making organ at the Bank of Japan was the Policy Board, which, as we saw, was instituted in 1949 by the U.S. occupation to “democratize” the central bank and dilute the wartime character of its setup. Apart from the BoJ governor and deputy governor, the policy board also included outsiders. Policies were discussed and then voted on. The well-known truth, however, was that the board merely rubber-stamped decisions previously made by the executive board of the Bank of Japan, which was composed of BoJ staff and headed by the governor.
It is generally accepted that decisions about lowering or raising the official discount rate (ODR) were made only in close consultations with the Ministry of Finance. That was not simply due to the “informal” power of MoF, but also perfectly legal: The Bank of Japan Law, which established the responsibilities and competence of the Japanese central bank, stated that the Bank of Japan can be told what to do in most areas of its business by the Ministry of Finance. Article 42 put it bluntly: “The Bank of Japan shall be under the supervision of the competent Minister.” Article 43: “The competent Minister may, if deemed particularly necessary for the attainment of the object of the Bank of Japan, order the Bank to undertake any necessary business, or order alterations in the By-Laws as well as other necessary actions.” And article 44: “The competent Minister may … issue orders or take such actions as are necessary in the exercise of his supervision.” And the list continues. In the first forty-seven articles of the Bank of Japan Law that dealt with its rights and duties, the phrases “with the permission of the competent Minister” or “in accordance with the prescriptions of the competent Minister” appeared twenty-nine times. Although the “competent Minister” is the finance minister, in practice this has meant the highest-ranking MoF bureaucrat, namely, the vice minister of finance. Monetary policy was therefore made “in accordance with” MoF. This is not surprising, given the wartime origin of the law.
MoF Thought It Was in Charge

In the postwar era, MoF frequently exerted direct influence over ODR policies.1 The jurisdiction over foreign exchange intervention was even more clear-cut, as it is governed by the separate Foreign Exchange Law, which grants sole authority for foreign exchange intervention to the Ministry of Finance. The Bank of Japan merely acts as the executive branch of MoF that implements whatever size foreign exchange intervention is decided by the international finance bureau at MoF. This is also why foreign exchange markets watched the actions of the flamboyant vice minister of international finance, Eisuke Sakakibara, so closely in the mid-1990s.
Given these legal facts, there seemed little doubt to most observers: MoF was in charge. The bubble was created by the Ministry of Finance through excessively low interest rates. From 1984 to 1989, an ex-MoF vice minister, Satoshi Sumita, was even governor of the Bank of Japan. The BoJ, with its professional knowledge, knew better, but given its weak legal position, it could not help implementing the wrong policies forced on it by the Ministry of Finance. Even the average salariman now understands that in the 1980s, banks have been the key accomplices of speculators who squandered sums in excess of a quarter of annual national income. The banks are supervised by MoF, and MoF is their patron. During the 1990s, the Japanese weekly papers ran frequent stories about how MoF’s banking supervisors dined and were entertained in hostess restaurants in exchange for leaks about when the next bank audit would occur. When MoF responded to the banking problems (which initially erupted in the jusen, or housing loan institutions) by suggesting using more tax monies, its reputation was thoroughly tarnished. It seemed that the prosecution’s case could be rested: Japan’s economic policies were disastrous and the culprit was MoF. Thus MoF had to be punished by being stripped of its powers and broken up.
Hero Mieno

This is similar to the picture painted of Yasushi Mieno, governor of the Bank of Japan from December 1989 to 1994. When he came to power in 1989, he noticed that the monetary policies of his predecessor, the former MoF vice minister Sumita, were too loose. Mieno hated the outcome. Japan, he lamented, was becoming a nation of “haves” and “have-nots,” as landowners became extremely rich at the expense of ordinary employees. He decided to end the bubble and raised the official discount rate only a fortnight after becoming governor, delivering the infamous “Christmas present” of 1989. Asset prices, led by stock prices, began to tumble in 1990. They were still hitting new lows more than a decade later, in 2002. Land prices had come down by around 80 percent and were still falling in early 2002. Seven fat years of waste and distortions were followed by more than seven lean years that sobered the economy.
A public dispute between Makoto Utsumi, the MoF vice minister of international finance, and BoJ governor Mieno erupted in 1990. Mieno, criticizing the injustice of the excesses of the bubble, gained the moral upper hand. Proclaiming that he had never owned stock in his life, this clean pair of hands appeared to be creating fair policies that were good for Japan. The media portrayed him as “onihei of the Heisei era,” a modern-day Robin Hood fighting for the rights of the poor.
Soon after his retirement from his position as governor in December 1994, Mieno embarked on another campaign. Giving speeches to various associations and interest groups across the country, he lobbied for a change in the BoJ Law. His line of argument was to subtly suggest that MoF pushed the BoJ into the wrong policies. To avoid such problems in the future, the BoJ needed to be given full legal independence. According to Mieno, making central banks independent “reflects the human wisdom that has been nurtured by history.”2 Eventually, his case found a sympathetic reception from the coalition government under Prime Minister Hashimoto. Like many observers, its project team for administrative reform blamed the Finance Ministry for Japan’s troubles of the 1990s and hence proposed to revoke many of the ministry’s powers. This included the proposal to make the Bank of Japan independent and to take away the power of MoF to appoint and dismiss Bank of Japan officers. In 1997, the coalition submitted a bill to revise the Ministry of Finance Establishment Law and the Bank of Japan Law to the Diet. The new BoJ Law became effective in April 1998, stripping MoF of the power to set monetary policy and making the BoJ legally totally independent.
MoF Tried to Create a Recovery

There can be little doubt that the ministry wanted to create a recovery. In charge of the government budget, it was deeply averse to fiscal deficits and to increasing government debt. Deterioration in these indicators was seen as a blemish on MoF’s image. A recession would reduce corporate, income, and sales taxes. It would boost unemployment benefits and social welfare expenditures. On top of that, politicians would demand government spending packages.3 Expenditures would balloon and revenues shrink, pushing the budget into deficit. Deficits are funded by bond issuance, and that is also not popular with MoF. Most of all, MoF did not like recessions because it was legally in charge of the economy. It would thus also get the blame for a recession. Along recession could cost it all its long-standing power.
MoF’s worst-case scenario occurred: The economy failed to recover. As a result, far-reaching administrative reform was implemented. The year 1998 went down in Japanese history as the year in which the Ministry of Finance lost its main power—its monopoly on budgeting. For the first time in the postwar era, it was politicians who drew up the stimulus packages. MoF lost control over banking supervision to an independent financial supervisory authority. It lost its licensing power thanks to the Big Bang deregulation program. Finally, it lost control over monetary policy to an independent Bank of Japan. To add insult to injury, in January 2001 MoF even lost its grand old name. Today, the Ōkurashō does not even exist.
While MoF clearly had a good incentive to create a recovery, the case is less clear-cut for the Bank of Japan. That is why it is time to take another look at the evidence.
Reexamine the Evidence, Dr. Watson

We are back in 1991. MoF, in the belief that the key monetary policy tool is the rate of interest, had been very unhappy with the BoJ’s high interest rate policy and had been exerting influence to obtain a lower official discount rate. In 1991, the BoJ gave in and lowered the official discount rate for the first time in the 1990s. When the economy did not improve and, to the contrary, slowed down further, MoF used its legal status to lean on the Bank of Japan several more times to lower the discount rate. When the economy deteriorated further in 1992 and 1993, and strains in the financial markets had been reflected in sharp drops in the stock market, MoF’s frequent fiscal stimulation packages seemed to betray signs of panic. Yet however badly the ministry wanted a recovery, it was elusive.
Throughout its postwar history, the Bank of Japan had relied on the quantity of credit to monitor and control the economy. From its experience of the 1950s and 1960s, when it kept interest rates especially low to ration credit and allocate funds to selected industries, the BoJ knew that with credit rationing, lower interest rates do not lead to an increase in the quantity of credit. In the 1990s, the bad debt problem paralyzed the banks and created severe credit rationing. Hence, the policy prescription must have been obvious to the Bank of Japan. Since credit can be created only by banks and the central bank, a decline in bank credit growth due to quantity rationing must be countered by an expansion in Bank of Japan credit creation. However, Bank of Japan credit creation did not increase significantly. To the contrary, as late as March 1995, the Bank of Japan withdrew money from the economy. The central bank failed to support government policy as the Bank of Japan Law mandated.
Fiscal Stimulation Neutralized

The same fate awaited the government’s fiscal policy. Between 1992 and 1994, four massive fiscal stimulus packages amounting to ¥45 trillion were added on to regular government spending. More than ¥100 trillion has been added in a string of further spending packages since.4 The enormous size of the stimulation programs reflects just how desperately MoF and the politicians wanted a recovery. Once again, the goal eluded them. The problem was that fiscal spending was funded by bond issuance and the money largely taken from the private sector. Without simultaneous central bank (or bank) credit expansion, fiscal spending had to reduce private demand. Yet the Bank of Japan kept credit tight. Thus, the fiscal stimulus packages of the 1990s were wasted and merely increased government debt.
Why would MoF let the Bank of Japan get away with failing to support its policies? The answer seems to be that MoF officials believed that they were in control of monetary policy, since they could manipulate interest rates, fiscal policy, and exchange rate policy. The old generation of MoF control bureaucrats, who had briefly competed with the Bank of Japan for increasing credit creation in the early postwar era, had long gone. The bureaucrats of the Economic Stabilization Board had been dissipated and their institutional knowledge lost. Neoclassical economics had become the mainstream approach also at Japanese universities. Credit creation was not taught anymore. The knowledge existed inside the Bank of Japan, but the central bank failed to share it with MoF or the general public. To the contrary, Bank of Japan staff has spent considerable resources on publications and PR, telling the public that its policies are determined by manipulating interest rates, and such measures as credit are not interesting.5 As a result, by the early 1980s MoF was no longer concerned with the quantity of credit as a monetary policy tool.
The Battle of the Yen

In late 1994, when both interest and fiscal policy had proven to be ineffective in stimulating the economy, MoF attempted to play its last trump card: exchange rate policy. The hope was that if the yen could be weakened, at least external demand could support the economy. The policy tool employed was so-called foreign exchange intervention, official purchases or sales of foreign exchange. According to the laws, the Bank of Japan merely acts as the executing organ, while MoF makes the decisions.
So the ministry ordered the Bank of Japan to sell large amounts of yen and buy U.S. dollar assets (mainly U.S. Treasuries). In the second half of 1994, this amounted to U.S. $7 billion. In February and March 1995 it reached U.S. $19 billion, one of the largest quarterly foreign exchange interventions on record. The cumulative dollar purchases had boosted the official foreign exchange reserves to U.S. $131 billion at the end of March 1995, the largest in the world.6 The theory was that the demand for dollar assets would strengthen the dollar, hence weaken the yen. But again MoF was disappointed. Despite such unprecedented intervention, the yen did not weaken. To the contrary, it strengthened. During the months from January to April 1995, the yen rose by 20 percent, hitting a historic high of ¥79.75/$ on April 19, 1995. Such previously inconceivable yen strength dealt yet another blow to the battered economy.
Exchange Rate Theory Built on Interest Rates

Just what determines the yen-dollar exchange rate? The leading theoretical models of international capital flows and exchange rates are based on interest rates as the main determinant variable. If interest rates are higher or rising in the United States compared with Japan, money will flow from Japan to the United States and this will weaken the yen and strengthen the dollar. Analysts and the financial press refer to this theory almost daily.7
Although the theory is so widespread that it has become common knowledge, frequently quoted by the financial press, it does not stand up to an empirical test. Figure 11.1 presents the yen-dollar exchange rate and the differential between ten-year government bond yields in the United States and Japan. One is tempted to conclude that things are working out; as can be seen, there seems to be a weak link between the two. Unfortunately, reality dares to ignore theory. Whenever the U.S.Japan interest differential widens (i.e., U.S. interest rates are increasingly higher than Japanese rates), the dollar does not strengthen, but often even weakens. In the first half of the 1990s, for instance, the U.S.-Japan interest differential widened, as U.S. rates rose and Japanese rates dropped. But instead of a stronger dollar, this was accompanied by a stronger yen—peaking in April 1995. Over the past two decades, the correlation between exchange rates and interest rates has not been supportive of the interest theory of exchange rate determination. The same result holds for short-term rates.8
Poring over the empirical literature on exchange rates, the largely unanimous conclusion of dozens of research papers is that exchange rates cannot be explained, let alone predicted. The “random walk” so far beats anyone’s currency forecasts, we are told. In other words, there is no specific model or set of explanatory variables that predicts the exchange rate any better than the last few actual exchange rate observations. “Economists do not yet understand the determinants of short- to medium-run movements in exchange rates,” concluded one experienced researcher.9 What exchange rates will do in the future is anybody’s guess.

Figure 11.1  Yen/U.S.$ Rate and U.S.-Japan Interest Differential

Source: Board of Governors of the Federal Reserve System
Total More than the Sum of Its Parts

Mainstream interest theory falls afoul of the fallacy of composition: When looking at only one investor, interest rates are a given variable that influences the behavior of this investor. Economists simply add up all investors. The fallacy lies in the aggregation. What is true for one individual investor cannot hold for all. If, for instance, all investors shift their portfolios in favor of one specific asset, interest rates will be affected. For all investors in the United States and Japan taken together, the interest rate differential is not a given, external variable. To the contrary, it is the result of everybody’s investment activity.10 Put simply, all the information that is in the bond markets (and hence interest rates) is also at any moment in time in the foreign exchange markets. Both are driven by a third factor and one cannot explain the other. A model based largely on interest rate differentials therefore can’t have much predictive power.
We have already found in chapter 4 that due to unrealistic assumptions (perfect information) and the fallacy of composition, modern economic theory has misguidedly placed the emphasis on the price of money (the interest rate). In reality, the most important determinant is the quantity of new purchasing power. Without perfect information, markets cannot be expected to be in equilibrium. Instead, they are likely to be quantity-rationed. Hence quantities determine the outcome, not prices. We should expect the same to hold for exchange rates. Instead of the interest differential, they should be determined by the differential in credit creation. Put simply, if the BoJ prints yen faster than the Fed is printing dollars, then the yen should be expected to fall and the dollar to rise.11 While foreign exchange dealers focused on every word uttered by Eisuke “Yen” Sakakibara of the Ministry of Finance, the yen-dollar rate has been mainly controlled by the Bank of Japan and the Fed.
BoJ and Fed in Control of Yen-Dollar Rate

In Figure 11.2, we have calculated the difference between the credit creation of the Fed and the BoJ and plotted it against the change in the exchange rate. As can be seen, the exchange rate does not follow a random walk, but is correlated with the quantity of credit creation by the central banks. Their relative credit creation seems to explain exchange rates over the last twenty years reasonably well. From 1990 onward, the yen strengthened and the dollar weakened. At the time, the U.S. Fed created dollars rapidly to reflate the U.S. economy; the Bank of Japan did the opposite, to prick the bubble. With fewer yen and more dollars, the dollar weakened and the yen strengthened. The same happened again in early 1995. It also held for the period of extraordinary yen weakness in 1998 and the subsequent strengthening in 1999. Each time, the relative credit creation of the central banks to a large extent explained and predicted the exchange rate.
It is now time to go back to the mysterious episode of March and April 1995, when the yen soared to a record high. Was this primarily due to the policies taken by the Fed, or was it more the result of the policies taken by the Bank of Japan? We know that the foreign exchange intervention of the Bank of Japan, as ordered by the Ministry of Finance, had reached double-digit billion-dollar figures. So surely the Bank of Japan was creating a lot of yen?
To find the perpetrator, we have broken the index into its two components—Fed credit creation and BoJ credit creation (Figure 11.3). We find that the Fed relatively consistently reduced its dollar creation between 1993 and 1996. That alone would tend to weaken the yen. So the yen strength of 1995 was not due to the Fed’s actions. A look at the BoJ’s credit creation uncovers the smoking gun: Net credit creation by the Japanese central bank dropped sharply from August 1994 onward. The tightening became worse and worse, culminating in the events of March 1995. Despite the painful recession and the desperate attempts by MoF to weaken the yen, the Bank of Japan actually reduced credit in this month. It withdrew money from the economy. As our exchange rate model demonstrates, this sudden reduction in yen creation meant that the yen had to strengthen drastically relative to the dollar. Had the Bank of Japan wished to support government policy and weaken the yen, it would only have had to increase its credit creation. Yet it did the opposite. Since the net credit creation of the Bank of Japan is a policy variable that the Bank of Japan can change at will, we have established that the Bank of Japan was directly responsible for the sudden surge of the yen to ¥79.75/$ on April 19, 1995.

Figure 11.2  Relative Credit Creation by Federal Reserve and Bank of Japan (Fed — BoJ LLI) and the Yen/U.S.$ Rate

Source: Board of Governors of the Federal Reserve System; Bank of Japan

Figure 11.3  Credit Creation by Bank of Japan and Federal Reserve (as measured by Profit Research Center’s Leading Liquidity Indices)

Source: Board of Governors of the Federal Reserve System; Bank of Japan; Profit Research Center Ltd.
MoF Fooled by Forex Interventions

How could the Bank of Japan get away with such a policy? Again, lack of knowledge of the facts by the Ministry of Finance seems to have been the reason. The bureaucrats of the International Finance Bureau in the Ministry of Finance may have believed that they were firmly in control of exchange rate policy, because they decide the official foreign exchange intervention. Their statements in public, including those by Sakakibara, certainly seem to suggest that. But what actually moves the exchange rate is the net credit creation of the central banks. That is determined by the total sum of all transactions of each central bank. Foreign exchange intervention represents merely one part of this.
When the BoJ was told by MoF to buy U.S. dollars, it went ahead and did so. MoF perhaps assumed that the BoJ would print the yen needed to buy U.S. dollars or U.S. Treasuries and that thus the yen would weaken. But this is not necessarily the case. In simple terms, the BoJ sold government bonds or other paper to investors in the domestic economy and used the proceeds to implement the foreign exchange intervention. Instead of printing the money, it took it from the economy. Economists call such an operation “sterilization.” Since the BoJ’s credit creation does not rise under such a policy, the yen will not weaken.12
In February and March 1995, when MoF ordered the BoJ to buy U.S. $20 billion worth of U.S. paper, the BoJ “oversterilized” by withdrawing more money from the economy than was needed for the foreign exchange intervention ordered by MoF. Net credit shrank in March 1995, and the yen shot up to its record high.
A replay of the battle between MoF and the BoJ over the yen exchange rate occurred in 1999. Having fallen dramatically in 1998, the yen was widely expected to remain weak in the following year. However, the Bank of Japan withdrew credit from the economy at a record pace. This credit shrinkage strengthened the yen to close to ¥100/$ by the end of 1999, despite new record foreign exchange interventions ordered by MoF. The Bank of Japan again oversterilized them.
Misery Created by the BoJ

Perhaps the Bank of Japan deemed the economy strong enough. In the first quarter of 1995, unemployment had reached a postwar high, creating human misery. The actual number of jobless probably topped five million people in early 1996.13 That is the size of the whole population of Denmark. Recession-related suicides jumped to a postwar high as well. Meanwhile, the banking system was still in deep trouble. Public money was not forthcoming, as even the bailout of the tiny jusen housing loan companies met fierce public resistance. Fiscal stimulation was wasted, as it was not backed by money creation. To pay back the record high debt resulting from it, future generations will be burdened with rising taxes. Meanwhile, the low-interest-rate policy, which could be avoided by increasing the quantity of credit creation, transferred wealth from savers to the banks. The total sum of transferred purchasing power likely exceeded ¥100 trillion—more than the annual GDP of the United Kingdom.
Was the BoJ Afraid of Inflation?

Why did the BoJ for so many years fail to implement the most obvious, simplest, and most suitable policy, namely, to print money? The first explanation might be that the Bank of Japan did not realize it was the right policy. Governor Mieno’s declared policy aim in 1990 was to fight asset price inflation. The BoJ might still have been fighting yesterday’s war, though the economy was already in the grip of deflation. But this explanation is not plausible. By 1992, it had already become painfully clear that asset prices were falling sharply. Moreover, there was never a danger of consumer price or wholesale price inflation. Inflationary pressures can come about only when credit creation is pushing the economy beyond its full capacity and when unemployment is at a very low level. But to the contrary, capacity utilization had been falling, inventories were rising, and many factories had been idle for much of the recession. With unemployment at postwar highs, it defies logic to explain the BoJ’s tight money policy by a fear of inflation.
Is the BoJ Just Incompetent?

Could the BoJ’s policies simply be due to incompetence? If one uses some of the publications by the Bank of Japan as an indication, then one might indeed be tempted to come to this conclusion. Many of its studies use a highly stylized and abstract form of neoclassical economics, which at the same time is based on oversimplifying, counterfactual assumptions about reality. Moreover, BoJ officials have repeatedly published research that claims that the Bank of Japan cannot control the money supply. However, we saw in earlier chapters that while the central bank’s injections of funds into the interbank market may be outside its control, this is not true for all its transactions. While these models may well reflect the views of hired economists trained in neoclassical economics, the likelihood is that they are far removed from any decision making or the type of work experience that would familiarize them with credit creation, such as at the Banking Department. The history of the BoJ suggests that the decision makers know very well from five decades of institutional experience that the quantity of credit creation determines the state of the economy and the exchange rate. I, for my part, have met staff members who were not on the research side, and they told me that their economists did not know about the true policy implementation, which is based on credit creation.14
BoJ staff would only have to reflect on the successful reflation of 1975, when Japan’s first boom-bust recession was ended by aggressive credit creation by the Bank of Japan. The man in charge of increasing the quantity of credit at the time, the head of the Banking Department, was an up-and-coming young central banker named Yasushi Mieno. Central bankers could also have recalled their actions during the even more dramatic banking problem and reflation after 1945.
BoJ Can’t Plead Ignorance

There is another reason why since 1991 the Bank of Japan must have been aware of the solution to Japan’s looming troubles. In 1990, the United States faced a problem similar to Japan’s, in which a boom funded by excessive credit creation had turned into a bust. Huge bad debts paralyzed the banking system. The ensuing credit reduction brought the wheels of the nonfinancial economy to an abrupt halt. In 1991, real GDP contracted by approximately 1 percent. How long this recession was going to last was in the hands of one institution—the Federal Reserve. It needed to print money. That is what it did, from 1990 onward. Credit creation rose and the United States moved out of the recession faster than expected.
Then the Fed embarked on the second phase of its reflation program—its clandestine bank bailout. It transferred money to banks by helping them create substantial profits in the bond market. Thanks partly to these profits, banks could afford to write off bad debts and begin to create credit again. This way, the Fed quickly reflated the economy. While most observers remained pessimistic until early 1994, the U.S. economy had already fully recovered by 1992. By acting immediately, the Fed had ended the U.S. banking crisis and recession within a year. While the Fed reflated, the Bank of Japan did not. Yet at the time the BoJ studied the Fed’s actions in detail.15 Finally, there are testimonies by leading Bank of Japan staff indicating that they knew early on what the problems were.16 It is therefore not plausible to argue that the Bank of Japan was unaware that printing money and creating a bond market boom was the right policy to end the recession.
Since its actions were consistent for many years (from 1990 to 1997, with another significant tightening in 1999), there also can be no argument that BoJ policy was just a temporary slip-up. There is no evidence to support the hypothesis of irrationality on behalf of the central bank.

So why didn’t the Bank of Japan reflate in 1993 or in 1994? Why did it disrupt the reflation policies of the democratically elected government and its Ministry of Finance? The quantity of credit creation by the central bank is fully decided by the Bank of Japan. The actions of the BoJ do not support the hypothesis that the BoJ wanted to create a recovery. The principle of revealed preference, used by economists, including Bank of Japan economists in their models, suggests that the BoJ has revealed its preference for tight credit creation—and the long recession that it entailed. However, it is hard to believe that the central bank would purposely want to prolong the recession. To examine this issue, it is necessary to go back to the events of the 1980s. These are the root cause of the recession, and they may provide further clues.
Who Created the Bubble?

We know that the Bank of Japan pricked the bubble in 1989. We have also found, strangely, that it was the Bank of Japan whose actions prolonged the recession. What was it up to? In our search for answers as to how and why Japan was thrown into economic upheaval, it is now necessary to analyze the root cause of the banking crisis and recession of the 1990s—and that lies in the events of the 1980s. We must next determine just who was responsible for the creation of the bubble.
In chapter 9 we learned that the bubble was due to excessive credit creation by banks. So the question is why banks created so much credit. What determined their behavior? In chapters 5, 6, and 7 we learned that from around 1940 until at least the end of the 1970s, the answer to this question was Bank of Japan window guidance. We know this as fact established by many researchers.1 But what about the crucial 1980s?
No Window Guidance in the 1980s, Says the BoJ

Some research exists on the monetary policy tools employed by the Bank of Japan in the 1980s, but it ignores window guidance altogether.2 There is, however, one research paper that directly focuses on window guidance in the 1980s (Hoshi et al., 1991). The authors conclude, “After 1982 and until 1989, window guidance played an insignificant role in the conduct of monetary policy” (p. 9). So it seems we have to look for other factors that influenced bank lending.
However, a closer inspection of the paper by Hoshi and colleagues reveals that their conclusion is not based on empirical research. It is merely asserted.3 No doubt the source for their assertion is the Bank of Japan itself: In December 1981, the central bank announced that window guidance controls would be abolished beginning in January 1982. Instead of dictating loan growth quotas to the banks, the BoJ said it was going to “respect” their lending plans. In 1984, the BoJ announced the complete abolition of any form of window guidance.4 Further official pronouncements by the Bank of Japan repeated the message: In 1986, a Bank of Japan official was quoted in the Nihon Keizai Shinbun (Nikkei) as saying that the central bank “currently is not doing window guidance.”5 In 1988, the Bank of Japan claimed that there had been no window guidance in “a narrow sense” since 1982. It argued that deregulation of interest rates, financial disintermediation, and liberalization had all rendered this policy tool ineffective.6
The Bank of Japan even set out to show that credit aggregates were no longer well correlated with economic activity and hence not useful as a tool for monetary policy. The unspoken implication was that window guidance could not have been necessary or even possible, since credit did not seem to be moving the economy anymore.7
In March 1991, the BoJ felt prompted to publish another official assessment of the role of window guidance in the 1990s. The report, “The Process of Decision Making and Implementation of Monetary Policy in Japan,” was for overseas consumption, since it appeared only in English. It claimed that, “in line with the trend of financial deregulation, window guidance has not been applied in its original form of credit control since 1982, when the Bank of Japan stopped giving individual banks instructions with respect to their lending plans.”8 Soon after its publication, the BoJ spoke its final, though somewhat contradictory, word on window guidance, announcing in July 1991 that window guidance would be abolished immediately.
Would the Bank of Japan Misinform Us?

The Bank of Japan’s version of events, that window guidance was not an important policy tool in the 1980s, has never been substantiated by empirical evidence.9 Furthermore, there are a number of reasons that should make us skeptical of the Bank of Japan’s story. First, the very nature of window guidance as an informal policy tool that is based on extralegal sanctions and shrouded in secrecy implies that it is unlikely ever to be fully disclosed in official statements. Second, when Japan joined the OECD, it was obliged to reduce direct economic controls and adopt a market-oriented economic system. Direct credit controls have been criticized by the United States. Therefore the Bank of Japan has had a political incentive to downplay such controls and emphasize market mechanisms. Third, window guidance has a pre-1980s track record of being officially declared “abolished,” although in actual fact it either continued unofficially or was reinstituted again soon after.10 Fourth, the very fact that it was abolished yet again in 1991 suggests that it did exist in a meaningful way previously. Fifth, credit growth has proven to be the cause of the economic dislocation of the 1980s and ultimately also of the recession of the 1990s.11 Anybody involved in the determination of aggregate bank credit must be assumed to have an incentive to downplay involvement and degree of control.
Call in the Witnesses

As in a court case, when in doubt, one should call in the witnesses.12 Cross-examination should establish whether the stories stack up or whether there are contradictions. A starting point for our search for witnesses may thus be a thorough check of the leading Japanese-language financial press. A search of the most highly respected financial newspapers in Japan, the Nikkei and its sister paper the Nikkei Kinyū (Nikkei Financial Daily), uncovered strong evidence that during the 1980s the Bank of Japan continued to implement its window guidance credit controls, completely unchanged from the pre-1980 era. Shown below are summaries of some of the articles that were found:13
December 1981: The Nikkei reports that the Bank of Japan has introduced a “new style of window guidance,” tantamount to the abolition of tight loan growth ceilings. According to the new procedure, the BoJ would from now on “respect” the lending plans of banks. This is “the loosest regulation since 1945.” However, BoJ officials are also quoted as testifying that window guidance was “not abolished.” One stated that “there is the intention to control lending via direct daily contact with private sector financial institutions.”14
September 1982: The Nikkei finds that despite the official BoJ claim that all bank lending plans are being “respected,” voices from the top city banks complain that “the lending plans of the individual banks are not being respected enough” and banks are not allowed to compete freely. If lending plans were fully accepted, then competition among banks would increase and the ranking of the banks would change. However, the ranking of banks by assets remains unchanged.15
December 1986: Window guidance quotas tighten. The Bank of Japan is aware of the surge in land prices, the high growth of the money supply, excess liquidity (kane amari), and the strong expansion of bank lending toward the real estate sector.16
March 1988: Window guidance loan quotas for the second quarter of 1988 are being tightened. The central bank is aware that the fastest-growing segment of bank lending is to real estate and zai-tech.17
September 1990: “Whenever the actual year-on-year loan growth has not reached the regulated quota, the Bank of Japan has until now reduced the next quota by the amount that was not used up. This resulted in the problem that each bank struggled hard to use up its loan quota, even if this was unreasonable. From now on, the Bank of Japan is studying to abolish this procedure.”18
June 1991: A bank officer from a “high-ranking city bank” is quoted in the Nikkei Financial Daily as saying that a side effect of the window guidance rule of loan increases was that banks “increased lending even when there was no loan demand. And even when there was loan demand, they had to keep [lending] in line with other banks.”19 A bank officer is quoted as saying: “So far, first the total credit amount was decided by the Bank of Japan credit allocation frame, then we decided how to divide that allocation.”20
Press Sources: The BoJ Misinformed

These testimonies from the respected press clearly suggest that window guidance did exist and that the Bank of Japan, despite public statements to the contrary, continued to impose loan growth quotas on the banks. Moreover, we find that the Bank of Japan continued to engage in qualitative credit allocation, monitoring the use of credit. In particular, the central bank seems to have been aware of the increase in credit creation for speculative financial transactions, especially real-estate-related lending.
If the Nikkei and Nikkei Financial Daily are correct, then the Bank of Japan misinformed us about the role of window guidance in the 1980s. What could be to its credit is that a few press commentaries hint that it may have used window guidance to slow the bubble, as it “suppressed” loan growth. Could it be that without the Bank of Japan’s credit controls the bubble would have been even bigger? Should we therefore be grateful to the central bank for its actions?
Direct Testimonies from BoJ and Bank Officers

In a court case, juries and judges prefer evidence provided directly by eyewitnesses. Window guidance was an elaborate procedure that involved many Bank of Japan officials and an even larger number of bank officers whose job was to liaise with the Bank of Japan (the so-called nichigin-tan). Thus a sufficient pool of people exists who can be interviewed regarding the crucial question of the conduct and role of window guidance in the 1980s. However, now that the bubble has created a long recession, and since banks have received public money, it is clear that this question is politically sensitive. Both banks and their regulators have been heavily criticized for their actions. Scandals have highlighted some of the informal links that existed between Ministry of Finance officials and bankers (the “MoF-tan“). Many bank staff, and even some ministry officials, have been arrested and imprisoned. While the role of the nichigin-tan has not yet been publicly discussed, it is clear that bankers can be expected to be reluctant to talk publicly about their role. Likewise, many Bank of Japan officials might find it difficult to answer truthfully, as this might force them to implicate their employer or colleagues.
It is therefore fortunate that earlier, in 1992, I conducted a series of interviews with the aim of determining the operation of window guidance in the 1980s. I interviewed twelve Bank of Japan officials who had previously worked as window guidance officers, recording the interviews on tape in some cases and taking notes in others.21 I also took notes during interviews with six private nichigin-tan bank officers from three different banks (as well as different bank types), whose job it had been to talk to the Bank of Japan about bank lending. In 1992, neither the BoJ nor bank employees that I interviewed seemed aware of just how sensitive the question of the role of window guidance would become. Therefore, they appeared very frank and open about the questions asked.
The Jury Is In: Window Guidance Did Continue

The result was unanimous. There was full agreement on virtually all aspects of window guidance among all sources.22 The central bank and bank officers confirmed to us that window guidance was conducted without interruption during the 1980s, until at least June 1991. It took exactly the same form as the window guidance in the pre-1980 period: The Bank of Japan decided on an aggregate bank loan growth rate for the entire country, and then young BoJ staff calculated on their Excel spreadsheets how this could be divided up among all the bank types (city banks, trust banks, regional banks, etc.) and by individual bank (Fuji Bank, Sanwa Bank, etc.). Those loan growth quotas were then announced to banks all over the country in the quarterly meetings. The quarterly quota was broken into monthly increments, which were also monitored.23 “When the end of a quarter approaches, it becomes clear by how much the banks have increased loans so far. Thus around that time—for example, toward the end of March—bank people come and talk to their counterparts here at the Bank of Japan and say that they may go over the loan ceiling and want a bigger maximum. Or the BoJ man says to reduce. Thus toward the end of the quarter there is lots of talk here. They come to give reports. We actually call them. We ask them every month how much they have raised loans. If it looks that they have gone over the limit, we tell them, ‘Slow it down a bit’” (BoJ officials 5, 6).
The BoJ-Run Loan Cartel

Since banks were oblivious to cost or profitability considerations and instead aimed at market-share expansion, like other war economy industries they would engage in destructive “excess competition” if left without a cartel. Window guidance continued to serve as the BoJ-run cartel, which carved up the entire credit market among the banks and neatly preserved their pecking order. “The Bank of Japan used the yokonarabi ishiki [orientation to stay in line with others] so that the banks will always do what it wants” (bank officer 4). This not only enhanced the power of the Bank of Japan to control credit aggregates, but also was used as a way to make credit controls appear attractive to banks, as only a cartel provided relief from excessive competition: “If it was not for window guidance, we would compete until harakiri. This is not good” (bank officer 5).24 Hence banks would always use up exactly their full window guidance quotas. This meant that the Bank of Japan could determine the precise amount of credit creation and hence economic activity in the country.
The cartel was carved up as follows: “The loan increase quota was always proportional to the previous actual loans. Big banks can increase loans a lot, small [banks] a little. We do it so that bank rankings won’t change. Thus no matter how much a bank does in terms of competition, bank rankings do not change at all…. There was no free competition” (BoJ officers 5, 6). The quotas were determined according to the ranking of the banks: “For this the BoJ used the following formula: For the main four city banks, for which the volume is similar (Sumitomo, Fuji, Mitsubishi, Sanwa), it was decided first. If they are 100, then from here it was certain how much the others would get: Dai-Ichi Kangyo Bank would be 120, Mitsui and Tokai Banks would get 80. Thus the order is preserved” (bank officer 4). Then the long-term banks and other bank types are decided proportionately: “maybe 100 for Industrial Bank of Japan, 50 for Long-Term Credit Bank, and 30 for NCB” (bank officer 4). However, the Bank of Japan had far-reaching discretionary power to vary the quotas at will and favor some banks over others. So in practice, the precise loan increase quota usually differed for each bank and for each type of bank. At times, loans for some banks did not increase, even when others did.
The informal control of window guidance was also used to punish banks for other forms of “misbehavior,” such as when a Nagoya-based bank sacked a Bank of Japan amakudari (“descent from heaven”—an ex-bureaucrat who parachuted into a plush private-sector job). The punishment could also take the form of reduced loan quotas (BoJ official 1). “This has actually happened fairly often” (BoJ official 5). The decision regarding such penalties is made by the Bank of Japan official who is in charge of a specific private-sector bank and who therefore wields considerable arbitrary power (BoJ official 6).25
BoJ Also Conducted Credit Allocation

In addition to the quarterly meetings, monthly hearings also took place, during which officers from the banks came to the Bank of Japan and explained their business plans in terms of changes in deposits, loans, and investments in stocks and bonds.26 “Window guidance was very detailed” (BoJ official 5). The BoJ was aware not only of which sectors received funds, but also of the names of the major firms that did. As a bank officer put it, the BoJ was interested to see “where the money goes for what purpose” (bank officer 3).
“The Bank of Japan asked how many loans for which industrial sector, how much for short-term or long-term lending, how big is the scale of the borrower, how many loans are demanded by the real estate sector, and so on. Sometimes they ask the names of big customers, for example, Matsushita” (bank officer 4). The loan information was broken down into loans to each industrial sector of the economy, within which it was further broken down to the names and amounts of big borrowers (more than ¥100 million). All of these were subject to detailed scrutiny by the BoJ officials. “In the monthly meetings, important questions are how many nonperforming assets, how many loans to organized crime (‘yakuza company’), and so on” (BoJ official 5).
Window Guidance Was Binding

The credit controls were highly effective during the 1980s. According to the BoJ officials I interviewed, banks virtually never exceeded the loan quotas set for them, as punishment would follow immediately. “If they went over the maximum, then they would get a lower quota next time. But I have never heard of this. It virtually never happened. Window guidance has been very strictly observed” (BoJ official 5).27 More surprising and more momentous was the finding that banks also virtually never left their quotas unused in the 1980s (bank officer 2). “Banks always went to the maximum of the lending ceiling. The quota was supposed to be digested completely by the banks. If we go below it, our allotment will be decreased compared to our competitors. So we have to eat it all; it’s an o-bento [Japanese lunch box] to be eaten” (bank officer 2).
The Bank of Japan used the same penalty for exceeding window guidance as for underutilizing it: If banks did not use up their loan growth allocation for more than one quarter, then the bank’s future loan quotas would also be reduced. “I have reduced their waku [quota] because of this. Maybe not if it happens just once, but if it happens for two quarters, if they don’t use it, well, we reduced their allocation” (BoJ official 5). Banks that wanted a bigger quota in the future had to avoid a penalty and demonstrate their “strength” by always using 100 percent of the window guidance quota. “If banks do not go to the limit and do not use up their quota, their loan growth ceiling will be reduced next time. This has been a problem in the bubble [period]…. The maximum [loan growth quota] had been very high. Thus banks wanted to reach it; they tried all they could to fulfill it. Banks always strove to reach the maximum” (BoJ official 5).

Window guidance was extremely effective. We have obtained data from a private-sector research institution that apparently realized the importance of the window guidance quotas and collected all such quotas, even broken down by bank type.28 Figure 12.1 shows the total window guidance loan quotas for an aggregate of four bank types (city banks, trust bank, regional banks, and long-term credit banks). As can be seen, the window guidance quota is almost identical with the actual aggregate lending of these hundreds of financial institutions. Actual lending closely follows the ups and downs dictated by window guidance. Looking at this astonishing chart, we are prone to forget the fact that a time lag of three months exists between the creation of each data series: Window guidance quotas were announced to each individual bank before the beginning of a new quarter (in late March, for instance). The actual loan figures are only available after that quarter is over (after the end of June; in fact the Bank of Japanwould publish the data with at least a two-month lag, thus in this example not before the end of August). Both series are shown here according to the quarter that they refer to, but the window guidance series is actually known a quarter earlier.
Window Guidance Used to Raise Loan Growth

We found that during the 1980s, window guidance was hardly a tool to suppress loan growth. Not many banks were keen to exceed the loan growth quotas set by the Bank of Japan. To the contrary, those quotas were often considered too high by the banks: “In the bubble period, we wanted a certain amount [of loan increases], but the BoJ wanted us to use more than that. After 1985, the BoJ said, ‘Use more!’ Normally, we would not get as much as we want to use…. Especially in 1986 and 1987, for around one year, the Bank of Japan said: ‘Use more, because we have a recession.’ Window guidance can be used not just to make borrowing smaller, but also to make it bigger. We actually thought, ‘This is a little bit much.’ But we couldn’t leave anything unused of the quota given to us. If we did, other city banks that received a similar quota might beat us. Thus, in order to keep our ranking [among banks] we had to use it all up…. Also, if we got a reputation for being weak, we would get less in the future” (bank officer 4). “Window guidance was a burden for banks, because sometimes we had to lend when we didn’t need to and at other times less than we wanted to” (bank officer 5).
Window Guidance: The Number One Policy Tool

From the interviews with BoJ officials and bank officers it quickly becomes evident that the Bank of Japan did not just sanction banks’ loan growth plans. Rather, it decided and administered credit growth quotas at its discretion. They took the same form of total control as in the pre-1980s period—indeed, the same form as during the war. Window guidance remained the main policy tool of the central bank. Interest rates were at best a supplementary tool. The Bank of Japan punished banks for overshooting or undershooting the loan growth quotas; it was also aware that competition for rank among banks meant that they would strive to fulfill and not undershoot loan growth quotas even without punishment procedures. As before 1982, qualitative window guidance continued; thus the Bank of Japan was fully aware of and “guided” the increase of loans in real estate, construction, and nonbank financial institutions, in other words, the increase in credit creation that was used for speculative purposes.31
Thanks to the BoJ’s extralegal window guidance mechanism, the BoJ could almost play the economy like a yo-yo. To accelerate growth, it raised the loan growth quota. To slow growth, it reduced it. Such nearly complete control also bred contempt for other central banks that appeared to be less firmly in the driver’s seat. In the early 1980s, for instance, the Federal Reserve apparently had great difficulties with its monetary targeting. Eventually, it capitulated: In 1982 it abandoned M1 targeting and in 1987 targeting altogether, as it felt that it could not control both the price of money (interest rates) and its quantity at the same time. Following the mainstream theories to give preference to interest rates, it seemingly abandoned quantity variables. At the time, Bank of Japan officials were reported to have made contemptuous remarks about the Fed’s problems. Unlike the Fed, the BoJ easily managed to control both the quantity of money and its price: The BoJ set interest rates and, using window guidance, at the same time controlled the creation of new purchasing power, which also determines deposit measures of the money supply, such as M2+CD.32
Window Guidance Pushed up Lending

Once the Bank of Japan had informed a bank of its loan growth quota for the next quarter, the bank would then divide it between its various branches, where in turn the branch managers would allocate loan growth targets to their various loan officers. So the entire financial system in Japan was allocating the overall loan growth quota of, for instance, a 15 percent increase into specific targets for each bank type, for each bank, for each branch, for each loan officer. Then, at the end of this credit allocation chain of command, the loan officers in the thousands of branches all over the country were faced with the task of meeting these loan growth targets.
In the beginning, their job was not so difficult. The economy had been in the endaka recession of 1985, and there was pent-up demand for money. But their loan growth targets handed down from the window guidance remained high while loan demand by the firms that used money productively had been satisfied, so loan officers increasingly targeted firms that would use money in an unproductive way. They focused on real estate lending. As it became easier for real estate companies to borrow, they began to increase real estate investments. More transactions meant more demand for land. Thus land prices began to rise. This created enormous capital gains for investors. Profits lure copycats. Other real estate firms, then construction firms, and in the end even normal manufacturing firms wanted to get in on the game by borrowing to invest in real estate. With rising land prices, this seemed a sure thing. For each individual bank, it also seemed safe: They had their loans secured by land, considered the safest collateral. The problem was visible only on an aggregate level, when it could be seen that the share of real-estate-related lending in total new loans soared dramatically. When comparing total loan growth with GDP growth, it was obvious that credit was being created and used for unproductive purposes. The land prices were not for real; they had been driven up by the excessive bank lending. A bubble had developed.
But Why Were the Credit Quotas Set So High?

During the war and early postwar era, window guidance allocated credit to priority, productive areas. It also helped to suppress “unproductive” and hence inflationary credit creation. That is how Japan’s economy grew so fast in the 1950s and 1960s. In the 1960s and 1970s, we saw that the Bank of Japan warned of the dangers of excessive, unproductive real estate lending, due to the temporary relaxation of the window guidance ceilings. That fueled a real estate lending boom and produced rocketing asset prices. The BoJ responded by tightening its loan growth quotas. When the excess credit creation was stopped, it turned into bad debts, which produced a bank-centered recession.
Given its experience, it is not surprising to find that the BoJ kept a close watch on real estate-related lending in the 1980s. The Bank of Japan window guidance had become so detailed that not just sectoral loan aggregates were monitored and controlled by the BoJ, but even the names of the main borrowers. The increased creation and allocation of credit for nonproductive and speculative use was therefore known and sanctioned by the BoJ. Yet, instead of reducing its window guidance after 1986, it kept it at very high levels: over 12 percent YoY growth for most years. With GDP growth and hence demand for productive credit being much lower, the BoJ knew that the only way for banks to fulfill these high quotas was to expand nonproductive credit.
Not all BoJ officials agreed with this window guidance policy. The Nikkei reported in 1990 that some BoJ officials were complaining that the window guidance system, which punished banks for not using up their loan quotas by reducing them next time, was one reason for the “unnecessarily strong loan growth” in the 1980s, as banks would always use up their loan quota, even when unnecessary. “Put bluntly, one side effect of window guidance was that it spurred the expansion of bank lending during the bubble period,” said one Bank of Japan executive.33 Another BoJ official we interviewed stated that “the BoJ has promoted loan expansion in the bubble period” (BoJ official 5). “If there is no demand for credit [from low-risk borrowers] and we want to use up the quota, the risk gets worse. Thus during the bubble, bad loans rose. The quota became a target for a period of time during the late 1980s” (bank officer 3).
The Plaza Excuse

Could it be that the low interest rate that had been ordered by MoF policies left the BoJ no choice? During and after the negotiations that led up to the Plaza and Louvre Agreements, U.S. Treasury officials put significant pressure on the Japanese officials from MoF’s International Finance Bureau to reduce interest rates and, once they had sunk to a record low 2.5 percent, to keep them there for years. When MoF complied with U.S. wishes, the Bank of Japan had little choice but to implement this interest rate policy. Does this not mean that the Ministry of Finance is responsible for the creation of the bubble economy?
To the contrary. During most of the postwar era, interest rates in Japan have been kept artificially low in order to subsidize industry, encourage investment, and increase noninflationary economic growth. In order to prevent inflation and overheating, the Bank of Japan simply engaged in quantity control and allocation of the available funds. With interest rates below those rates that the banks would have set, the credit market was in a constant state of excess demand—even more so than it usually is. The BoJ then monitored, in every detail, which sectors of the economy were receiving purchasing power and to what use the money was put.
In other words, window guidance is the ideal mechanism to prevent high loan growth and slow the economy when it is politically difficult to raise interest rates. In the 1950s and 1960s the BoJ had pursued a policy of artificially low interest rates to support industry, with economic growth kept in check by restrictive credit policies. Reducing window guidance quotas in the face of low interest rates would have been possible; indeed, it had been standard practice in the past. So even if external pressure by the United States and other G7 countries had forced MoF to ask the BoJ to lower interest rates, the BoJ could still have tightened loan growth quotas and thus prevent a bubble. But it raised loan growth quotas, thus forcing banks to lend more, even to unsound projects. Banks felt safe; “even when banks were lending wildly [nohōzu], they said that it was okay, because the BoJ said it was okay.”34 This puts the blame for the creation of the bubble squarely on the Bank of Japan.
In their defense, Bank of Japan officials have publicly claimed that they tried to prevent the creation of the bubble by raising interest rates, but they failed to push through a rate rise politically.35 Indeed, it would have been difficult for the BoJ to raise interest rates, since it was not independently in control of this monetary policy tool. But low interest rates did not create the bubble. It was the window guidance lending quotas that did. And the BoJ could have easily reduced those without any resistance—banks would have preferred lower quotas—and thus it could have prevented the bubble.
The BoJ officials I talked to fully confirmed this: “The Bank of Japan promoted loan expansion in the bubble period…. When one looks back now [to the 1980s], then it was a mistake. My own opinion is: When one reduced the interest rate and reduced the window guidance loan growth limit, then with this policy mix, the bubble would not have developed. But in reality, interest rates were reduced and window guidance was very relaxed. Thus the money supply rose by 10 percent, up to 13 percent. The question why they didn’t close the window guidance more is extremely puzzling…. All the banks tried to use their loan growth quota to the maximum and did all they could to give out loans. But the loans did not go to normal corporations, such as steel, automobiles, but instead to construction, non-bank financial institutions [which engaged in real estate speculation]. This became the bubble” (Bank of Japan official 5).
Bank officers took the same view: “I worried a lot about the policy of the authorities…. If they had taken a little tighter policy in the window guidance, this kind of phenomenon, the bubble, would have been prevented…. If the Bank of Japan had wanted to tighten, it would have achieved a lot” (bank officer 2).
It is indeed puzzling why the Bank of Japan, for almost five consecutive years, in the face of already exploding real estate prices, and despite its public protestations that it was worried about a bubble, continued to set outlandishly high loan growth quotas for Japan’s banking system. Not only did the BoJ have the benefit of the experience of the 1970s bubble, but in the 1960s, researchers had already identified the Bank of Japan’s window guidance as the source of excessive credit growth and argued that this could not simply be blamed on the banks: Placing the blame on “the ‘excessive’ competition of city banks in extending loans resulting in an overly rapid expansion of credit is a misplacing of the responsibility for credit control. It is the duty of the Bank of Japan to control the amount of credit creation by commercial banks;… The excess lending [of the 1950s] was the fault not of the banks trying to maximize profits, but of the Bank of Japan for allowing such credit expansion to take place.”36
Clearly, pleading ignorance is implausible for those at the Bank of Japan who decided the credit quotas and could look back on decades of experience with them. It is impossible to avoid concluding that these decision makers must have been aware they were creating a bubble.
Hiding the Smoking Gun?

Window guidance was suddenly abolished in July 1991—apparently in a rush. Bank of Japan officials said that window guidance was abolished at such short notice that it took the window guidance officers themselves by surprise. Hearings still took place in June, and Bank of Japan officials were preparing the window guidance quotas as usual, when the abolition was announced and they suddenly ceased their work. No reasons were given to Bank of Japan window guidance officers (BoJ official 5).
Bankers were equally surprised, indeed often left almost helpless: Soon after the abolition of window guidance, bankers complained that they did not know how to make their lending plans anymore: “So far, first the total credit amount was decided by the BoJ credit allocation quota, then we decided how to divide that allocation.” Until then, “when a certain branch said they would like to lend more, we could respond that the window guidance quota was used up—now we won’t be able to do that anymore.”37
The official reason given to the public for the abolition of window guidance was that it had become increasingly ineffective due to financial deregulation. But none of the interviewees felt that window guidance was abolished because it did not work. It worked very well. Neither bankers nor BoJ officials really knew the real reasons for its abrupt abolition.38

Perhaps there is a different explanation. Window guidance created the bubble and then pricked it, and hence ushered in the recession of the 1990s. Indeed, over the entire postwar era, the window guidance loan growth quota was the single most important economic variable. It is the smoking gun, the possession of which would make it easier to identify the perpetrator. We therefore need to find out just who was holding the gun and pulled the trigger.