Many of our Chinese friends who live in Shanghai and Beijing are telling us that the property inflation rate exceeds 50% – and 100% in some areas. Available housing for a young couple is far from downtown, closer to one-hour commuting distance, and still more than ten times the average salary. There are also reports that the state government is borrowing with “protected” forest land as collateral (Dyer 2010). However, these stories do not match the Chinese official data (National Bureau of Statistics of China 2010a). The Chinese official statistics say that the average increase (year-on-year) in property prices was 10.7% in February. The increase is accelerating from a year-on-year rise of 9.5% in January.
The official data may be significantly underestimating the real price increases, due partly to a statistical bias to be explained below. Some break down of statistics is available. The rate of price increases of newly-built residential buildings (at 90 square meters and below) in Beijing, Shanghai, and Shenzhen are 19.3%, 11.6% and, 19.6%, respectively (National Bureau of Statistics of China 2010b). In the same category, the highest property inflation was in Sanya, Hainan Island, at 57.9%. However, the location (within a city) and quality (beyond “new” vs. “used”; and “average” and “small”) of the buildings are most likely not controlled for in the official statistics. Granted, the 70-city breakdown is available, but not within the city; “new” and “used” is differentiated but no information on how old the used units are; and statistics on “smaller” units are available, but still each break down category is still coarse. Beyond “average prices,” no detailed statistical description is available.
Suppose that prices are rising uniformly at the same rate in downtown and in the suburb. Since more expensive properties are less demanded, more units in the suburb (at lower price levels) will be supplied to the market. Thus, the “average” price increase will be lower than the uniform, true rate of the price increases. More buildings are sold in the second-tier suburbs than the first-tier, and even less in the downtown area, the “average” price may be lower than the location-controlled index. The official statistics most likely underestimates the size of housing bubble.
In real estate economics, the location-adjusted, and quality-adjusted property price index can be constructed by:
- the sampling of repeated-sale (identical) properties (like the Case-Shiller index),
- by running a hedonic regression (in many academic studies for particular cities), (Ito and Hirono 1993 on Japanese housing data),or
- by using expert appraisal (like the frequently-quoted Japanese land price index). Having a location- and quality-adjusted property price index is critical for timely policy decisions.
Reminiscent of the Japanese bubble
What is happening in China now is familiar to any Japanese who lived through the bubble in the second half of the 1980s (and its subsequent burst). From the mid-1980s to 1990, Japan experienced one of the largest property bubbles in history. The land price index tripled in five years. The 6-large city price index for residential land rose from 39.2 in September 1985 to 105.8 in September 1990. The commercial land price rose from 27.9 to 104.5. At the initial stage, the land price rose in the central part of Tokyo, then spread to first-tier suburbs in Tokyo, then to other large cities and second-tier suburbs, and finally to rural land and forest. The recent US housing bubble had a similar process of moving from prime to subprime mortgages. The quality of borrowers progressively worsened both in Japan and the US. The loan-to-value ratio rose sharply toward the end of bubble.
Is this bubble now being repeated in China?
Yes, Chinese property prices increases seem to share many of the features of the Japanese property bubble in the 1980s. Of course, that does not necessarily mean that the Chinese bubble would collapse like Japanese. Chinese economic growth potential now is much higher than Japan in 1990. Chinese banks are still under heavy control of the state, so that any policy restraints on further fuelling to the bubble would be more effective in China than in Japan.
It is important to learn what caused a bubble, why it could not be stopped before it became too big, what went wrong in crisis management when it burst. In the following, I provide a short summary (see Cargill et al. 1997 and 2001 for a comprehensive analysis).
Lessons from the Japanese bubble and burst
The Japanese property price started to increase with strong fundamentals and financial deregulation in the mid-1980s. In the beginning (say, 1984 to mid-1987), property price increases seemed reasonable and well-justified with strong macro fundamentals. Real demand came from a few booming sectors, like newly deregulated financial services. Foreign investors were interested in Japanese assets including stocks and real estate. Monetary policy remained very lax (i.e., the official discount rate being 2.5%, a record low at the time) from mid-1987 to mid-1989. Both stock prices and land prices kept rising, partly helped by the low interest rate. However, some time later (mid-1987 to end-89), the process became a bubble – that is, people invest in the property only to aim at capital gains. A leverage ratio became higher in pursuit of higher yields and as banks became more confident that the value of collateral would continue rising.
The first lesson is that a bubble, which can be suspected from rising loan-to-value ratio, rising price-rental (dividend) ratio, increasing bank lending to the real estate sectors, and real estate developments in further-away suburbs, should be restrained by a higher interest rate and prudential regulation on banks’ behaviour. In the second half of the 1980s, monetary tightening and stronger regulatory measures should have been applied in 1987-89. The bubble might not have been completely avoided purely by monetary tightening, but the damage might have been reduced by early tightening and prudential regulations.
The great stagnation during most of the 1990s and the 2000s is rooted in the bubble and its burst. However, policy mistakes that came after mistakes during the two decades were also to blame.
- First, monetary policy was too tight when the burst bubble occurred to mitigate negative wealth effects, especially in 1991 and 1992 (see Ito et al. 2006).
- Second, fiscal stimulus was too little and too late to make large impacts as pump priming. It eventually raised the debt level without lifting the economy. Fiscal tightening in April 1997 and monetary tightening in August 2000 was clearly premature and misguided.
- Last but not least, the banking crisis was mishandled –forbearance for many years and timid capital injection – and protracted. In the 1992-94, smaller housing loan institutions became fragile, but the problem was not dealt with immediately. The problem spread to ordinary banks, and then to large banks. Early, decisive actions would have mitigated, if not prevented, the problems among very large banks in 1997-98.
US demands, yen appreciation and a bubble
Many Chinese officials tell us that they believe the origin of Japan’s 20-year stagnation after its housing bubble burst lies in its failure to stand up to US pressure for the yen to appreciate. They cite this presumption as one of the reasons why they resist US pressure to allow the renminbi to appreciate. What is the logic behind this? There are three possibilities:
- The US pressure to appreciate the yen made the bubble size bigger;
- The US pressure to appreciate the yen made economic slump from bursting bubble worse; and
- The US pressure to appreciate the yen made the Japanese export sector to lose competitiveness that caused the 20-year decline.
I examine these three possibilities, one by one.
The US pressure for yen appreciation was most prominent on two occasions. First, the Plaza Agreement of September 1985 pushed up the value of the yen sharply. Just before that agreement, the yen was at 240 yen per dollar. Three months later, it appreciated to 200 yen per dollar – a 20% appreciation. The yen continued to appreciate to 155 yen per dollar by August of 1986 – a 45% appreciation compared with the pre-Plaza level. Second, the yen sharply appreciated from 100 yen to the dollar in January 1995 to 80 yen per dollar in May 1995. This period coincides with US frustration with Japan over trade conflicts, and the market interpreted that the US pressure was threatening Japan with a choice of trade sanction or yen appreciation. However, this appreciation episode was short-lived, as the yen depreciated back to 100 yen to the dollar by September 1995.
Is the first US pressure, the Plaza Accord, the source of the 20-year stagnation? Given the timing, if the pressure resulted in making the bubble larger than otherwise, that causality may be established. However, the truth is just the opposite. First, the Plaza Accord was to correct an overvalued dollar against other major currencies – the yen was only one of them. Second, movement from 240 to 200 yen to the dollar was well in the range of correcting the overvalued dollar, and there was no objection from Japan. When the yen appreciated beyond 190 yen per dollar in the spring of 1986, interventions started to stop yen appreciation. However, the yen kept appreciating. The US agreed to stabilise the exchange rate with the Louvre Accord in February 1987, and it was widely believed that the target zone of 150-160 yen/dollar was agreed between Japan and the US. Indeed, the exporting sector suffered because of the sharp appreciation from 1985 to 1987. But the recession of 1986 was short-lived. One offsetting factor was the oil price decline in 1986. The yen tended to appreciate (or depreciate) when the oil price declined (or increased), respectively.
Monetary policy was relaxed from 1986 to 1987, and the record-low discount rate (at that point) of 2.5% was maintained from February 1987 to May 1989, with a hope that low interest rate would stop or moderate the speed of yen appreciation. Hence, it was not caving in to yen appreciation demand but resistance to US pressure which made monetary policy too lax and contributed to bubble enlargement. The logic is just the opposite of what Chinese officials may believe.
How about the second episode? Trade conflicts over automobiles in 1994 to 1995 led to US frustration. However, the yen appreciation pressure from 1994 to 1995 was more informal than during the first episode. Investors appreciated the yen whenever Japan was reported to have resisted US pressure for numerical targets of “voluntary import expansion.” From macroeconomic reasons, no factors were supporting a sudden appreciation from 100 to 80 yen per dollar in five months. So, decisive interventions directed by Mr. Sakakibara, then Director General of the International Finance Bureau, brought the yen back to 100 yen per dollar. It is also evidence that the appreciation had no fundamental basis. Since the appreciation was so short a period, it is doubtful that exporting sectors suffered a permanent damage.
In either episode, US demand for yen appreciation has little relevance for the 20-year stagnation of Japan. What is to be blamed is a series of policy mistakes: not preventing a bubble from becoming too large, not addressing the bursting bubble early enough, and not dealing with the banking problem decisively in the early stages.
In failing to prevent a bubble from getting bigger and bigger, the Chinese authorities are making a mistake – low interest rate policy and resisting currency appreciation – similar to Japan’s in the 1980s. China’s reported inflation rate does not show rampant CPI inflation, but there was no CPI inflation in Japan in the 1980s. The property bubble is a clear sign of overheating. Trade surpluses were large both in Japan in the 1980s and in China now. The first best policy is interest rate hikes to restrain the property boom, resulting as well in an appreciation of the Chinese renminbi that prevents hot money inflows. The Chinese authorities are already taking some measures to restrain banks from increasing credits to the property sector – tightening regulations of loan-to-value ratios and increasing the reserve requirement and this should continue. If the renminbi does appreciate, overheating of China’s export sectors will be slowed, while standards of living will improve with higher purchasing power.
China did allow the renminbi to rise from July 2005 to 2008 by 20%. Yet, productivity increases in China’s export sector overcame this appreciation, and trade surpluses grew. The Chinese renminbi was re-pegged to the dollar in the summer of 2008, and remains so until now. It is high-time China lets its currency appreciate once again. It should not fail to do so just to confront the US. China should maximise its own welfare whether or not that is what the US wants. Renminbi appreciation is clearly good for macro management. It gives more policy tools to the People’s Bank of China and is beneficial to the standard of living of the general public. China should learn the right lessons from the Japanese experience.
Editors’ note: A shorter version of this piece was published in the Financial Times on 17 March 2010.
Cargill, Thomas, Michael Hutchison, and Takatoshi Ito (1997), The Political Economy of Japanese Monetary Policy, MIT Press.
Cargill, Thomas, Michael Hutchison, and Takatoshi Ito (2001), Financial Policy and Central Banking in Japan, MIT Press.
Dyer, Geoff (2010), “China warned of growing ‘land loan’ Threat”, Financial Times, 28 March.
Ito, Takatoshi and Frederic S Mishkin (2006), “Two Decades of Japanese Monetary Policy and the Deflation Problem,” in Takatoshi Ito and Andrew Rose (eds.), Monetary Policy with Very Low Inflation in the Pacific Rim, NBER-University of Chicago Press:131-193.
Ito, Takatoshi and Keiko N. Hirono (1993), "The Efficiency of the Tokyo Housing Market," Bank of Japan, Monetary and Economic Studies, 11(1):1-32.
National Bureau of Statistics of China (2010a), “Sales Price Indices of Buildings in 70 Medium-Large Sized Cities (2010.02)”, www.stats.gov.cn, February.
National Bureau of Statistics of China (2010b) “Sales Price Indices of Buildings in 70 Medium-Large Sized Cities(90 sq.meters and below) (2010.02)”, www.stats.gov.cn, February.