Remarks by Yutaka Yamaguchi, Deputy Governor of the Bank of Japan, at a symposium sponsored by the Federal Reserve Bank of Kansas City in Jackson Hole, Wyoming between August 26 and 28, 1999.

September 2, 1999
Bank of Japan

1. I would first like to thank Federal Reserve Bank of Kansas City for giving me this opportunity. However it is not without pains for me to talk on our so called "bubble" years, for it is not yet a part of the past, but it is a part of the present. In fact, much of Japan's difficulties in the 1990s clearly has its origins in the asset market swing in the last 15 years. Its scale has been unprecedented through the modern history of Japan. The damage left on the balance sheet of firms, both financial and non-financial, has been enormous. A question haunts us to this day if monetary policy in the 1980s couldn't have been conducted to help stabilize the asset markets with the effects of leaving less painful adjustment in the 1990s. I will therefore spend most of my time in revisiting our "bubble" days.

2. The first alternative course in monetary policy I want to examine is much earlier tightening actions than the BOJ actually took in May 1989.....preemptive policy, if you will. Would it have been reasonable and what might have been the effect on asset price?

The macro-economic environment in the second half of 1980s was unique indeed. The CPI inflation stayed close to zero for three years through 1988, while real growth accelerated to 5% per year from less than 4% in early 1980s. The key elements that supported such no-inflation performance appear to have been:

1. (1) The lagged effects of 100% appreciation of the yen against the dollar after the Plaza Accord.
2. (2) The changing labor market structure that helped moderate wage pressure in the face of extremely tight market conditions. Major changes include rising market participation of women, increased dependency on part-timers, shift of labor from self-employed businesses to more modernized, and inflow of foreign workers, among others.
3. (3) Helped by easy credit and lower capital cost, businesses invested heavily, which in turn generated faster growth in productivity and lower unit labor cost. Growth accounting also suggested faster improvement in total factor productivity.

With hindsight, such excellent performance was unsustainable. The lag between rapid growth and eventual rise in inflation happened to be longer than usual but the linkage didn't disappear. As the state of high growth and near-zero inflation stubbornly persisted, however, it became increasingly difficult to distinguish cyclical upswing and trend/structural shift. The BOJ's warnings on inflation threat came to attract less public heed.

The BOJ was also constrained by the international environment. After the Black Monday crash in 1987, the policy coordination among the G-7 came to be reemphasized to ensure global stability, with by far the heaviest pressure fell on Japan's continuation of monetary relaxation. A perception emerged that Japan's interest rate would stay at historical low. It grew more deeply rooted when, in the summer of 1988, Germany and the U.S. raised the key policy rate and Japan didn't. Japan's inaction was not without reasons, as inflation was almost nil, but was read as expressing our emphasis on international cooperation. A bizarre argument gained strength that Japan as the largest creditor maintain low rate as the cornerstone for global stability and growth. Easy money was also regarded necessary to achieve a national economic goal of the day: reduction of external surplus through growth of domestic demand.

For all these reasons, policymakers felt that, for their actions to gain public support, they would need absolutely solid evidence of inflationary potential building up. When such evidence finally emerged in 1989, they quickly seized the moment to reverse the policy. Inflation picked up but peaked in 1990-91 at mid-3%. On the basis of the inflation track record, I would conclude that preemptive policy to achieve sustained price stability would probably have been desirable but must have been very hard to initiate. Professors Bernanke and Gertler indicated in their simulation results that the BOJ should have raised short term interest rates to 8% in 1988 and even higher in the following years. But all through 1988 inflation was pretty close to zero. I don't see how a central bank can increase interest to 8 or 10% when we don't have inflation at all.

3. Supposing a reversal in monetary policy at earlier timing had been implemented ideally, it would have taken away fractions of growth, resource utilization rate and inflation. It is more difficult to estimate the effects on asset price. Such effects appear to vary depending on specific circumstances. If, for instance, early rate hikes had hit the emerging sense of complacency and kept the public as vigilant as usual on potential risks ahead, the asset market should have responded accordingly. Alternatively, in the context of national confidence already growing strong enough, which might have been Japan's situation in 1988 particularly as Japan's market swiftly overcame the Black Monday shock, early but modest tightening might have invited the people to foster their perception of "eternally" sustained non-inflationary growth. In that case, the risk premium would have further fallen to offset the interest rate effect. The net result is uncertain but might possibly have been an escalation of asset price. Therefore I would remain cautious in attempting to make a hasty generalization here on the possible effect on asset price.

4. Attempts to restrain asset price at later stage would face even greater difficulties. In this regard, market reaction following the BOJ tightening is illuminating. The Tokyo stock market shrugged off the first two rate hikes, continued to rise for six more months and started to fall only after the third rate increase. The lag was even longer for the real estate market to peak out. The pattern of implied forward rates indicated the market's conviction that the rising interest rate would be short-lived, presumably for the reasons I have described above.

This episode seems to suggest that, once the asset market gained momentum, it could take more than gentle strikes to break its neck, for monetary policy would have to be adjusted tight enough to match the enhanced state of expectation. A firmly imbedded market expectation of continued low interest rate could take strong actions to be corrected. But, when such expectation begins to change, perceived future risk also begins to grow and the market suddenly starts to fall. A self-feeding process tends to develop between weakening market and increasing risk premium. That is why asset market abruptly collapses when somehow turning point arrives, making "soft landing" difficult. In fact Tokyo stock market fell by 40% in one year after the peak; the BOJ's business contacts also confirmed in late 1990 an abrupt shift in the tide in the property market. In other words, if monetary policy targeted on asset price in a situation like Japan of ten years ago, the risk would be that, once the gear is reversed, the combined effects of tight money and asset market correction could well turn out to be a disproportionately large pressure on real activity.

5. I have so far discussed in the Japanese context why it is not feasible to attempt to control asset price. That said, it is also important to identify exactly how monetary policy was associated with Japan's asset "bubbles" in the late 1980s, for monetary policy may at least be able to avoid reinforcing the asset market momentum. In view of the important role of expectations in generating on asset "bubble", I would cite the following two aspects as having been more relevant than the level of interest rate or monetary growth as such.

First, monetary policy failed to effectively check the permeation of such unrealistic assumption as perpetually low interest rate. Such perception was nurtured through a series of policy actions and inactions from Plaza to Black Monday to German and U.S. rate hike. Stronger emphasis on sustained price stability as policy objective over other considerations, such as exchange rate and external surplus, would have been helpful, and more effective battles with external constraints would have been necessary.

Second, continuous and consistent warning by the BOJ on possible future risks would have been helpful. When a society somehow becomes full of confidence, it tends to grow complacent and negligent of risks ahead. Japan in the second half of 1980s was likely approaching that kind of stage, as it had just overcome the 100% appreciation of the yen with high manufacturing technology. Against that background, several years of above-trend growth and near-zero inflation probably induced a significant drop of risk premium. Thus the discount rate adjusted for risk premium, a key determinant in the asset price equation, must have declined markedly. Obviously the BOJ alone was not able to deal with the tide of national sentiment, but it appears nonetheless important for the central bank to have tried to "lean against the wind" and endeavor to keep the public alert on potential risks.

6. The last point on the Japanese episode is the vast accumulation of risk in the real estate market. One feature of the financial development in the 1980s was that, when broad money growth moderately accelerated in the second half to 10% per year from 8% in the first half, property-related bank loans increased by 20% annually. It is this credit concentration that produced the real estate "bubble" and left behind lingering balance sheet damages.

Behind such an acceleration in bank lending was the erosion of bank franchise value. The prospect of financial liberalization accentuated the bank's search for new market opportunities. Large borrowers were increasingly turning to the capital market, but the gradualist and sagmentation approach of liberalization effectively prevented bank entry into the lucrative security or investment banking businesses. Property lending was a quick way to make profit, acquire market share in small and medium sized firms with secure collateral...or so it seemed. At the initial stage the property market was probably moving on a solid ground as the abundant liquidity with low rates attracted an influx of foreign players and demand for new offices sharply rose. But the process evolved into a self-feeding cycle between credit expansion and rising property value.

My colleagues at the BOJ has found a few years ago that a simple "stress testing" incorporating a common-sense assumption...that land price-to-GDP ratio, which sharply rose in the late 1980s, return to the historical trend value...yields a rough estimate of what turned out to be the reality of bank assets in the 1990s. But one would truly have to "lean against the wind" to exploit such stress test results, for disregarding the historical pattern is the very nature of a "bubble". Furthermore, even if such results had been available, it is an open question what monetary policy could have done on the basis of the property market development alone. Obviously risk concentration can develop in other areas. When it does, it may fall, like Japan's case, in the nexus of monetary and prudential policies. I strongly feel that we need to make further progress in this area. Central bankers can play a constructive role supplementing risk management at private institutions and prudential regulations.

1. In reexamining the policy issues in the 1980s, I have benefited a great deal from discussions with Masaaki Shirakawa, Kunio Okina, Hideo Hayakawa and Masayoshi Amamiya of the Bank of Japan.