Takatoshi Ito, “Zero Interest Rate Policies and Quantitative Easing”（in Mathias Dewatripoint, Xavier Freixas and Richard Portes (eds.) Macroeconomic Stability and Financial Regulation: Key Issues for the G20(Centre for Economic Policy Research (CEPR) ebook), pp.67～78）
経済を刺激するために、ゼロ金利政策（ZIRP）を超えて中央銀行がなし得ることとしては一体何があるだろうか？ この問いに対して、多くの研究者から数々の非伝統的な金融政策（unconventional monetary policies） の変種（variants）が提案されてきた。これらの提案の大半は、広義の量的緩和政策（quantitative easing (QE)）―中央銀行が通常では購入しないであろうリスク資産の買い取りを通じて中央銀行のバランスシートを拡張させる手段―として分類されることになるだろう。 これら量的緩和政策の変種の各々に関しては、(1)政策の第一義的な目的、(2)政策運営上の中間目標（target instrument）とコミュニケーション戦略、(3)買い取り対象となる資産、(4)インフレ期待に影響を与える経路、の4点に基づいてその違いを指摘することができるだろう。
2001年から2006年にわたって、日本銀行はあるタイプの量的緩和政策を試みてきた。まさしく現時点において、Fedは―ヨーロッパ各国のいくつかの中央銀行の中にもFedに追随しようとの動きを見せているところがあるが―、別のタイプの量的緩和政策―Fedは、自身の政策を指して、（日本タイプの）量的緩和ではなく信用緩和（"credit easing" (CE)）と呼んでいる―を試みようとしているところである。ECB（欧州中央銀行）も近いうちにFedにしたがうことになるだろうか？ そして、日本銀行は再びゼロ金利政策/量的緩和政策に逆戻りすることになるだろうか？ もしECBや日本銀行がゼロ金利政策・量的緩和政策を採用することになるとすれば、我々はグローバルなゼロ金利・量的緩和（global ZIRP and QE）の世界―国際マクロ・国際金融の分野における未知の領域―に足を踏み入れることになる。
In postwar history, we have only a few episodes of the zero interest rate policy (ZIRP) combined with quantitative easing (QE): The Bank of Japan (BOJ) from 1999 to 2006, and the Federal Reserve Board (FRB) in recent months are prominent examples（注1）. However, several other major central banks will have to introduce ZIRP/QE in 2009 if gloomy economic predictions are realized.
The purpose of this note is to review facts of the BOJ ZIRP/QE policy, to draw lessons learnt from the BOJ ZIRP/QE, to show the difference between BOJ QE and FRB QE, and to point out some issues associated with global ZIRP/QE if BOE, ECB, and some other G20 countries adopt ZIRP/QE.
2. The Bank of Japan experience of ZIRP/QE
2.1. ZIRP, the first episode, February 1999 – August 2000
The Bank of Japan adopted the zero interest rate policy (ZIRP) in February 1999 in response to extremely weak economic conditions combined with lingering banking sector weakness. The growth rate in 1998 was negative and major banks were suffering from shortage in their capital in the process of writing off bad loans. Major banks were scheduled to receive the second capital injection in March 1999, as the first capital injection in March 1998 was apparently not enough.
The February decision was to lower the policy rate to 0.15% immediately, and to induce the rate lower as soon as possible. By mid-March, the rate was near zero, and in April, the Governor mentioned that ZIRP would continue until "the deflation concern would be dispelled"（注2）. The inflation rate measured by CPI excluding fresh food had become zero and had been declining since mid-1998. With that crisis mode, the decision of ZIRP came too late. The first lesson is that ZIRP should not be delayed when the economy is sharply declining, and the inflation rate is also fast declining toward zero. The Bank of Japan should have adopted ZIRP soon after the banking crisis of November 1997 broke out and credit crunch became evident. ZIRP of 1999 came one year too late.
The exit from ZIRP in August 2000 by the Bank of Japan turned out to be too early. The Bank had to return to ZIRP, although it was combined with QE, in March 2001, only seven months later. At the time of exit in August 2000, the inflation rate was still negative, although the degree of deflation had been moderating. However, the prospect of economic growth should have been weak, as the IT stock bubble was bursting in the United States and other major stock markets, including Tokyo. The decision of lifting the ZIRP in August 2000 was opposed by the government, as the government representative tabled a motion to delay voting by one monetary policy meeting (about one month later). The government representatives, who do not have voting power, have a right to table such a motion, according to the Bank of Japan Law. The motion of "delay" was voted down by 8 to 1, and the motion by the Governor to raise the interest rate from zero to 0.25% was approved by a 7 to 2 majority.
The economy deteriorated in the fall of 2000, and it became clear that economic stimulus was needed by spring of 2001. The fact of back pedaling to ZIRP, although presented as QE, in March 2001 showed that the prediction and judgment by the government (Ministry of Finance and Cabinet Office), objecting to the end of ZIRP in August 2000, was more prudent than the Bank of Japan's. Also, a suspicion that the Bank of Japan would err on the side of too early exit rather than too late exit was proven. This was a substantial damage to the reputation and credibility of the Bank of Japan. The second lesson from this episode was that a commitment to err on the side of "late" would be important in getting out of deflation（注3）.
The Bank of Japan may have proved itself in standing tall and "independent" by rejecting a transparent objection by the government, but it eventually lost credibility. The renewed fight against deflation with QE starting March 2001 was much tougher due to this lost credibility. Accountability was lacking, as Governor Hayami after reintroduction of ZIRP with QE in March 2001 did not explain what went wrong: whether the objective was wrong; whether a forecasting model was wrong; or whether judgment was wrong.
In the literature of central banking, an "inflation bias" caused by the government preferring a higher output, is always cited as a reason for central bank independence, but the incident of August 2000 shows that the Bank of Japan had a "deflation bias" （注4）.
In October 2000, the Bank of Japan initiated a semi-annual "Outlook" paper, in which Monetary Policy Board members' forecasts were shown in a range, much like the FRB's FOMC participants' forecast in semiannual Humphrey Hawkins testimony documents. In the first Outlook, the forecast horizon was just the current fiscal year (April 2000 to March 2001). Members' forecasts for CPI (excluding fresh food) ranged from -0.5 to -0.1, while the trimmed range (excluding one max and one min each similar to "central tendency" in FRB) was from -0.4 to -0.2. This is remarkable, since just two months earlier, they had voted to raise the interest rate (with no major events in between). The MPC members were not afraid to raise the interest rate in the midst of deflation. Many observers saw contradiction to the commitment of April 1999, in that ZIRP would continue until the "deflationary concern would be dispelled."
Also in October 2000, the Bank issued a document called "On Price Stability." Such a document had been promised since the spring of that year（注5）. That was expected to be an attempt to define price stability in a transparent manner, possibly with a quantitative definition. It turned out to be a disappointment. The October document defined price stability as a state that is neither inflationary nor deflationary - a tautology（注6）. (Much later in March 2006, the Bank of Japan published a document that says the "understanding" of price stability in quantitative points of all members ranging from 0.0 to 2.0.)
In sum, the too-early exit, from ZIRP, not only in hindsight but in real time, revealed that the Bank was too eager to raise the interest rate at the earliest opportunity. When ZIRP was restored with an additional feature of excess reserve target (QE), the market was still skeptical on how long the Bank would continue ZIRP/QE.
Against this criticism, the Bank of Japan came up with a new commitment: The QE/ZIRP would continue until the CPI (ex fresh food) inflation rate becomes "stably above zero." This was an indirect way of admitting the mistake of August 2000-without saying so. (Later, in 2003, "stably" was further defined as zero or above zero "for a few months" and no prospect to falling back to deflation) The definition of inflation (CPI ex fresh food) and the floor condition for an exit was finally defined. At this point, a desirable (target) rate of inflation was still not defined.
2.2. Quantitative Easing (QE)-excess reserve targeting
The current account balance, essentially the sum of required reserve and excess reserves, at the Bank of Japan became a new target, and maintaining the target balance above required reserve meant that the interbank interest rate would be automatically zero (ZIRP), given that excess reserves were not remunerated back then, and that more than enough liquidity would be provided to the interbank market. In order to induce the commercial banks to deposit to the non-remunerated current account at the BOJ, ample liquidity had to be injected by purchasing qualified securities.
This policy can be called QE-reserve targeting. The level of current account target was initially set at 5 trillion yen while the required reserve was about 4 trillion yen. Shortly after this policy was introduced, required reserves went up to 5 trillion yen, when the postal bank became a part of commercial banking system, and the QE target was raised to 6 trillion.
The current account balance target was increased from December 2001 to January 2004 in several steps. From January 2004 to March 2006, the target level was maintained at 30-35 trillion yen, while the required reserve was at around 6 trillion yen.
The QE- reserve targeting was ended in March 2006 and the policy target became ZIRP. It took several months to mop up excess liquidity and the interest rate was raise by 0.25% in July 2006, citing that the CPI (ex fresh food) inflation rate had been above zero for several months. This was the end of the second ZIRP period.
As one of the measures to increase overall liquidity, possibly influencing the slope of the yield curve, the Bank of Japan increased the amount of monthly (gross) purchase of long-term bonds. The purchase of long bonds had been done practiced on the assumption that monetary base had to increase with the nominal GDP growth and the long-term growth of monetary base can be provided with long-term assets on the liability side. However, the increase of monthly purchase of long-term bonds during the period between August 2001 and October 2002 was for helping increase the liquidity provision and, possibly, lowering the long-term interest rate. The monthly purchase was increased form 400 billion yen to 600 billion yen in August 2001; 600 billion yen to 800 billion yen in December 2001; 800 billion yen to 1 tillion yen in February 2002; and 1 trillion yen to 1.2 trillion yen in October 2002. The rapid increase reflects implies that the Bank was less reluctant in purchasing long-term bonds-no risk if held to maturity-than other kinds of risk assets.
It is also important to note that the increase in long-term bond purchases was done under Governor Hayami, while large jumps in the target amount of QE were done under Mr. Fukui, who became Governor in March 2003. Governor Fukui's communication was clearer in carrying out large increase in QE.
When the QE-reserve targeting was wound up in 2006, the amount of monthly purchase of long-bonds did not change.
These changes in long-bonds purchase and QE were shown in Figure 1.
Figure 1 BOJ increased the purchase of JGBs and current account balances (excess reserves)
2.3. Assessment of QE-reserve targeting
It is not immediately clear from the traditional theoretical monetary policy models how QE- reserve targeting would help the economy recover from stagnation and deflation. Even if money supply increases, as monetarists would recommend in fighting deflation, if most of increases of M2 is in the central bank reserve-that is, smaller money multiplier-then the real side of economy (investment and consumption) would hardly be stimulated.
There are four different channels that QE-excess reserves contributed to restoring financial stability, helping economic recovery and fighting deflation.
First, it was expected to, and it did, have had strong effects on financial systemic stability. Since there was ample liquidity, a sudden death of a commercial bank due to liquidity shortage became extremely unlikely. So, counterparty risk in the interbank market was lessened.
Second, it was also expected, but did not happen, that credit crunch would not happen for the reason of liquidity shortage. However, bank loans were not dramatically increasing at all during the QE period, because banks were concerned with their capital adequacy ratios and deteriorating credit worthiness of corporations. When the economy was stagnant and stock prices were falling, it would be very difficult to encourage Japanese banks (or any other banks) to lend more, even with expanding QE. The task should be taken up by financial supervision policy and fiscal operations in terms of capital injection to commercial banks.
Third, QE-reserve targeting had sent a message to market participants that ZIRP would continue for some time, and there would not be an abrupt end like August 2000. This has been called the "policy duration effect" (see Oda and Ueda (2005)). As extra liquidity provided to the market, and it would take time to take out liquidity from the market, ZIRP would be guaranteed for some time. At the time of introducing QE in March 2001, there was also a commitment, in that the exit would not come until the inflation rate would be stably above zero. With high target of QE and the commitment of a clear exit condition, the medium- and long-term interest rates came down. Therefore, the yield curve would become flatter: the zero interest rate up to approximately one year, and flatter slope up to ten years. Then borrowers with investment project and potential house buyers may be encouraged to borrow and invest, provided that the expected inflation rate for a longer term stays stable (that is, not dragged down with the nominal interest rate). That would raise aggregate demand.
Fourth, with ZIRP in place with policy duration effects, commercial banks may find it safe and profitable to invest in interest bearing securities like long-term bonds, high-grade corporate bonds, equities, and low-grade corporate bonds. This is called "portfolio rebalance effects."
Ugai (2007) provides a comprehensive survey of empirical work related to the policy duration effect and the portfolio rebalance effects. For policy duration effects, Baba et al. (2005), Oda and Ueda (2005) and Okina and Shiratsuka (2004) confirmed that the effect was strong in lowering the 3-year to 5-year JGB interest rate. The effect became even stronger in late 2002.
As for portfolio rebalance effects, Kimura and Small (2006) found some effects on high-grade corporate bonds (a 10 trillion excess reserve increase depressed the spread of Aa bonds by 1 to 4 basis points), but not on equities and low-grade corporate bonds.
Long bond purchases by BOJ were increased from gross purchase of 400 billion yen per month before QE to 1.2 trillion yen per month in October 2002. Oda and Ueda (2005) actually did not find the measurable impact of this policy on the JGB yield. Kimura and Small (2006) found some effects from BOJ's increased JGB purchases on high grade corporate bond yield, but not in the low-grade corporate bonds.
The Bank of Japan also purchased equities directly from commercial banks from 2002 to 2004 for the reason of financial stability, and, as the Bank of Japan emphasizes, not for monetary policy（注7）. More than 2 trillion yen worth of equities were purchased. Japanese commercial banks' holding of equities contributed to significant uncertainty over the prospect of their Basle Capital Adequacy Ratio, as 45% of valuation gains can be counted toward tier II capital, but evaluation losses had to be deducted from tier I capital. What was once a cushion in CAR became a liability by 2002. The Bank of Japan decided to avoid self-fulfilling prophecy of banks' selling equities in loss cut that would actually cause the equity price to decline, that would induce more loss cut selling.
3. Variants of QE
What the Bank of Japan did not do as QE, despite many economists were the following（注8）. (1) Purchase of foreign currency denominated government securities; (2) Outright purchase of corporate debts (except Asset Back securities); (3) Purchase of REITs, (4) Purchase of equities as monetary policy; and (5) Guarantee of interbank liabilities. Svensson (2001) recommended that BOJ would purchase foreign bonds that would amount to unsterilized intervention. This would have two benefits via probably depreciation of the yen, preventing deflation through imported inflation and stimulating exports. Svensson called this a foolproof way out of deflation.
In Japan, foreign exchange intervention is under the Ministry of Finance jurisdiction, as foreign reserves are held in the fiscal special account. (See Ito (2007) for this mechanism.) The Ministry of Finance conducted massive yen-selling intervention from January 2003 and March 2004, coincided with BOJ expansion of its balance sheet. This may be viewed as unsterilized intervention, but it was explained as a "coincidence" by then Deputy Governor of Bank of Japan. (See Ito (2004b))
Important lessons of ZIRP and QE from the experiences of the Bank of Japan are that it is important (a) to communicate the intention, objective, and target of those policies; (b) to prevent the inflation rate from falling into the negative territory for an extended period; and (c) to find ways to influence inflation expectation. BOJ failed in some of these criteria and succeeded in some.
3.2. FED, credit easing
The Federal Reserve has expanded its balance sheet in 2008, much faster than BOJ did from 2001 to 2003. Bernanke (2009) explained the action as "credit easing" to make it differentiated from BOJ experiences. The difference is best explained by his own words:
"Fedが信用市場の機能改善を支援するためにこれまで採用してきた上述のアプローチは、量的緩和（quantitative easing；QE）政策―2001年から2006年にかけて日本銀行が採用した金融政策のアプローチ―とは概念的に異なるものである。とは言いつつも、我々が現在採用しているアプローチ―「信用緩和」（"credit easing"）と呼びうるかもしれない―と量的緩和とはある1点において類似点を有している。それは中央銀行のバランスシートの拡大を意図しているという点である。しかし、この類似点を別にすれば、量的緩和政策と信用緩和政策との間には以下のような違いがある。
From these explanations, we arrive at the conclusion that QE has many forms. QE has several variations regarding (1) what is the primary purpose; (2) what would be the target instrument and communication strategy; (3) what assets to buy; and (4) how to influence inflation expectation. Therefore we need to make it clear which QE we are talking about. The Bank of Japan, in 2001-2006, went for QE-reserve targeting, hoping that providing ample liquidity to banks would change their behavior to lend more and that the yield curve would be flattened to stimulate aggregate demand. The yield curve did become less steep, but bank lending did not pick up the pace. The Federal Reserve in 2008 directly intervened in the credit and capital markets and purchased the securities and provided guarantees. This CE (credit easing) action seems to be based on a recognition that the BOJ QE-reserve targeting was rather ineffective.
3.3. Balance sheet concern
A common concern in expanding central bank balance sheet, either by BOJ style or FRB style, is the possible losses from holding risky assets, like corporate bonds. In fact, the concern was one of the reasons that the Bank of Japan did not take bolder actions.
Here it becomes crucial that the central bank and the fiscal authority must cooperate. The central bank has to put the health of the economy ahead of the health of its balance sheet, and the fiscal authority has to find an implicit or explicit mechanism to fill losses from the central bank action of holding risk assets for the sake of economic recovery. The fiscal authority can take reduced payment of seigniorage from the central bank over the years, directly inject capital (if losses wipes out capital), or reimburse the losses in risky securities holding in joint facilities, such as Maiden Lane that would assume possible liabilities of Bear Stearns assets.
The trust between the central bank and the fiscal authority is a key. The Bank of Japan in 2001-2006 did not have such a trust, possibly due to a crash over lifting the first ZIRP in August 2000. The FRB and the Treasury seem to enjoy such a trust. It would be a challenge for ECB to build such trust with 16 different fiscal authorities of the euro area.
4. Implications of global ZIRP
When several major economies simultaneously fall into ZIRP, the effectiveness of QE is diminished as a stimulus via exchange rate depreciation may be limited.
When Japan was suffering from stagnation while other major economies were booming like the late 1990s, the correct policy prescriptions for Japan included currency depreciation. Currency depreciation is a potent weapon for getting out of deflation because it works directly via imported inflation and indirectly through stimulation of output among the export sectors. Japan needed this help in 2001-2006, and other major countries that were growing at healthy pace did have room to absorb exports from Japan. Japan did maintain the zero interest rate, and as other countries raised interest rates, the widening interest rate gap induced the yen to depreciate in
2004 to 2006.
Moreover, Svensson (2001) recommended non-sterilized intervention as a fool-proof way of getting out of deflation. Whether it is a result of natural consequences of monetary base expansion or a result of outright non-sterilized intervention, the currency depreciation should be considered as a part of the prescription. However, currency depreciation becomes difficult when ZIRP is practiced by many countries simultaneously. Some currency has to appreciate to help others. The burden of appreciation should be shouldered by a country with large trade surpluses with relatively sound financial system and growth prospects.
5. Inflation targeting
Flexible inflation targeting has been popular among many OECD and emerging market central banks. It has been argued that successful inflation targeting has effects of stabilizing inflation targeting. In the current context, anchoring inflation expectation, say, at around 1-2 % has the following benefits. Even if actual inflation rate becomes negative, the public continues to believe that the inflation rate would return to 1-2% in the medium term (say, two to three years). The investment and consumption decision would be based on negative real interest rate (ZIRP minus positive nominal expected inflation) that is stimulative, rather than a positive interest rate, that is contractionary. Thus, credible, flexible inflation targeting would be beneficial to managing the economy during a deflationary period.
QE-reserve targeting by the BOJ in 2001-2006 had "policy duration effects" that flatten the yield curve by lowering the long-term interest rate. The lower long-term interest rate can be a result of low inflation rate expectation, or low real interest rate expectation, or both. The credible inflation targeting may lower expected real interest rate, but not the nominal rate. An introduction of credible inflation targeting amidst deflation may be counterproductive by raising the inflation expectation and in turn raising the nominal long-term interest rate. This logic was also used as one of the objections to inflation targeting proposal to the Bank of Japan. Thus, it is important to have credible inflation targeting regime during a normal period with positive inflation rate, and to use it to prevent the inflation expectation falling into deflation territory in the beginning phase of deflation. As the central bank shows resolve to get out of deflation in a short period, inflation targeting may be a useful tool.
The Federal Reserve of the United States, the ECB (Eurosystem) of the Euro Area and the Bank of Japan have not adopted inflation targeting. Would it be better for them to adopt inflation targeting at this juncture? Although the Group of Big Three (G3) central banks have not declared inflation targeting, some of them have improved on many of the pre-conditions for inflation targeting, such as independence, transparency, communication strategy, and history of inflation rates, with some deficiency remaining.
The Bank of Japan has been quite negative on adopting inflation targeting during the ZIRP period and even after the exit from ZIRP. Ito (2004) discussed the political economy of the inflation targeting debate in Japan during ZIRP period. He summarized the Bank of Japan's opposition as follows.
"In fact, inflation targeting became somewhat a symbol of additional steps that the Bank of Japan should or should not take, since in order to achieve a positive inflation rate target, some of the unconventional measures have to be taken. Most of the Board members and staff economists of the Bank spoke out negatively about inflation targeting. Several reasons were mentioned. First, inflation targeting was characterized as a simple-minded reflation policy and rejected. Second, no country had adopted inflation targeting to increase an inflation rate from the state of deflation. Third, there was no policy measures, given that the interest rate was zero, are available to lift the inflation rate to the positive territory so that announcement of inflation targeting, without tools to achieve, would damage the credibility of the Bank. Fourth, a mere announcement of an inflation target would not change expectation. Fifth, if the public happens to believe in the inflation targeting, the long-term interest rate would increase and it would damage the economy."
6. Conclusion: Policy Recommendations
Based on Japan's experiences of ZIRP from February 1999 to August 2000, and ZIRP and QE from March 2001 to March 2006, and FRB's experience of credit easing from mid-2008 to now, with some forecasts on what will be coming in Europe, the following policy recommendation seems to emerge.
(1) Determination and communication of a do-everything attitude to avoid a prolonged deflation is critically important. Arriving at ZIRP should not be delayed until the inflation rate reaches zero. Inflation targeting-such as aiming at returning to 1-2 percent inflation rate in three years-may be helpful as a clear, transparent message.
(2) The authorities should not hesitate to expand the central bank's balance sheet by buying risk assets that would most effectively restore financial stability. Many variations of QE should be considered and attempted. In this respect, the FRB actions to purchase wide range of securities are remarkable and commendable. It is firmly believed at the FRB that restoring the economic health is more important than keeping the central bank's balance sheet clean and riskfree.
(3) It is important for ensuring the end of deflation to have a commitment to err on the side of late than early for the timing of exit and an implementation of easier-than-usual monetary policy after the exit of ZIRP.
(4) Global ZIRP would shut down one channel of recovery, that is, stimulus via currency depreciation. Under a global ZIRP environment, a country with large trade surpluses with positive GDP growth should refrain from large scale intervention since intervention to prevent appreciation would be equivalent to a beggar-thy-neighbor policy. The currency of a relatively
strong economy should appreciate (at least in the real effective exchange rate) in the globally deflationary environment.
(5) The trust between the central bank and the fiscal authority is important to induce the central bank to take bold actions in QE and/or CE.
(6) Globally-coordinated fiscal stimulus is desirable, but room for sustainable fiscal deficits is quite different from a country to another. The higher the debt ratio is, the more constrained is the country in implementing fiscal stimulus. Even in these countries, current fiscal spending associated with structural reform removing inefficiency and with subsidies to green innovations, with a promise to raise a less-distortionary tax in the medium run may be prudent fiscal policy under the circumstance.
（注1）The ZIRP does not necessarily mean literal 0.00% policy rate, although that was the case in the BOJ experience, 2001-2006, when the excess reserve as well as required reserve was not remunerated. In 2008, the Bank of Japan has lowered the policy rate to 0.1%, and excess reserves are subject to remunerated deposit rate at the Bank of Japan. This is effectively ZIRP. The FRB lowered the policy rate to the range 0.00 - 0.25%, with the remunerated deposit rate at 0.1%. This is another case of ZIRP.
（注2）Some may argue that the ZIRP had not been adopted firmly until April 1999. But for simplicity, I date the Japanese first ZIRP started in February 1999. See Okina and Shiratsuka (2004), Ito (2004a) and Ito and Mishkin (2006) for details of developments and statements of policy decisions.
（注3）Theoretical underpinning of this lesson can be given by modeling inflation expectations depending on the future policy commitment (Krugman (1998)) and by modified Taylor rule argument, taking into account the zero bound of the nominal interest rate (Oda and Ueda (2005).
（注4）This was almost expected as Governor Hayami was quite negative on calls for additional measures, including quantitative easing in his speech in his speech in March 2000. He cautions against the danger of inflation and is against the policy to call for aiming at inflation rate of 2 to 3 percent since inflation once occurred would be difficult to control. He sounded like fighting inflation in the midst of deflation. See Hayami (2000).
（注5）See Hayami (2000) for mentioning the initiation of the study in March 2000 as a part of enhancing transparency.
（注6）See Ito (2004a) for in-depth analysis on why the Bank of Japan was reluctant to define price stability inthe quantitative manner.
（注7）See Ito (2004a) for political economy over the debate of purchasing equities by the Bank of Japan. The purchase was decided in the "regular" meeting of monetary policy board, and not in the "monetary policy" meeting of the same board.
（注8）See Meltzer (1999, 2001), Svensson (2001), and others summarized in Ito and Mishkin (2006).
〇Baba, Naohiko (2006), "Financial Market Functioning and Monetary Policy: Japan's Experience," Monetary and Economic Studies, vol. 21, Special Edition, December: 39-71.
〇Baba, Naohiko; S. Nishioka; N. Oda, M. Shirakawa, K. Ueda, and H. Ugai (2005), "Japan's Deflation, Problems in the Financial System and Monetary Policy," Monetary and Economic Studies, vol. 23, no. 1: 47-111.
〇Bernanke, Ben S. (2009). "The Crisis and the Policy Response," At the Stamp Lecture, London School of Economics, London, England, January 13, 2009.
〇Bernanke, Ben S. (2003). "Some Thoughts on Monetary Policy in Japan," Before the Japan Society of Monetary Economics, Tokyo, Japan, May 31, 2003.
〇Hayami, Masaru (2000). "Price Stability and Monetary Policy" Speech given to the Research Institute of Japan, Tokyo, March 21.
〇Ito, Takatoshi (2007). "Myths and reality of foreign exchange interventions: An application to Japan", International Journal of Finance & Economics, Volume 12, Issue 2, April 2007: 133-154.
〇Ito, Takatoshi (2004a). "Inflation Targeting and Japan: Why has the Bank of Japan not adopted Inflation Targeting?," in Christopher Kent and Simon Guttmann(eds.)The Future of Inflation Targeting, August 2004: 220-267.
〇Ito, Takatoshi (2004b). "The Yen and the Japanese Economy, 2004." in C. F. Bergsten and J. Williamson (eds.) Dollar Adjustment: How Far? Against What?, ch.8, Washington, D.C.: Institute for International Economics: 171-196.
〇Ito, Takatoshi (2009). "How Close Are They to Inflation Targeting: BOJ, FRB, and ECB?" Presented at the AEA meeting January 3, 2009.
〇Ito, Takatoshi and Frederic S. Mishkin (2006). "Two Decades of Japanese Monetary Policy and the Deflation Problem," in T. Ito and A. Rose (eds.) Monetary Policy with Very Low Inflation in the Pacific Rim, NBER-University of Chicago Press: 131-193.
〇Kimura T. and D. Small (2006). "Quantitative Monetary Easing and Risk in Financial Asset Markets," The B.E. Journal of Macroeconomics, vol. 6, no. 1.
〇Krugman, Paul (1998), "It's Baack! Japan's Slump and the Return of the Liquidity Trap," Brookings Papers on Economic Activity 1998, vol. 2, pp. 137-187.
〇Meltzer, Allan H (1999). "Comments: What More Can the Bank of Japan Do?" Monetary and Economic Studies, vol 17, December: 189-191.
〇Meltzer, Allan H (2001), "Monetary Transmission at Low Inflation: Some Clues from Japan," Monetary and Economic Studies 19, no. S-1, Bank of Japan, February: 13-34
〇Oda, Nobuyuki and Kazuo Ueda (2005). "The Effects of the Bank of Japan's Zero Interest Rate Commitment and Quantitative Monetary Easing on the Yield Curve: A Macro-Finance Approach," CARF Discussion Paper, No. F-013, University of Tokyo, 2005.
〇Okina, Kunio and Shigenori Shiratsuka (2004). "Policy Commitment and Expectation Formation: Japan's Experience under Zero Interest Rates," North American Journal of Economics and Finance, vol. 15: 75-100.
〇Svensson, L.O. (2001), "The Zero Bound in an Open Economy: A Foolproof Way of Escaping from a Liquidity Trap," Bank of Japan, Monetary and Economic Studies, vol.19, no. S-1, February: 277-312.
〇Ugai, Hiroshi (2007), "Effects of the Quantitative Easing Policy: A Survey of Empirical Analyses," Monetary and Economic Studies, vol. 25, March: 1-47.