Sunday, September 2, 2012

Chinese Housing: A Reply

This post is meant to be a reply to Scott Sumner and Matt Yglesias' thoughtful points on the Chinese housing market and what China could be building besides housing. Reading their responses, it seems that a large part of the debate is about whether China needs more housing right now, and if the current explosion in housing construction is in-line with long term fundamentals.

Yet this debate about fundamentals seems to neglect the most important part about housing in China: the role of housing as a savings vehicle. What I think is sometimes forgotten is that Chinese citizens face a severe shortage of effective ways to save money to 保值, or preserve value. Bank deposit rates consistently run below the rate of inflation. When better investment opportunities come along, those deposit rates don't necessarily rise. There are securities companies, but it's very difficult for the people to trust them. Moreover, given the recent explosion in these companies, it's not hard to imagine that any risk-adjusted excess returns on securities should quickly go to zero. The stock market is seen as a capricious creature, and given past stock crazes involving nannies and farmers, people are right to be suspicious of investing in stocks as a way to secure wealth.

Therefore, housing is special because it is much more concrete than other investments. It depreciates relatively slowly and is easy to verify; this is unlike many other investment opportunities such as equities or securities. It is useful, because you can immediately start to live in it or rent it out. Also, because of rising incomes across China, it seems very reasonable that demand for housing will increase and therefore building a house will have a high return. Urbanization and demographics are sometimes used by Shanghai residents to justify that housing prices “will never fall,” or that if they do, the drop will be minor. As a result, Shanghai residents view houses as an incredibly important way to save money. This past summer, I had the opportunity to work with a Chinese instructor on my language skills. She was a middle-class Chinese citizen and often mentioned that, although housing wealth was rising, she and her family didn't feel much richer. This was because, for them, the houses were seen as a savings vehicle, much too important to be sold on a whim.

Under such conditions, large scale housing investment is not necessarily a neutral “revealed preference”. Rather it can be a dangerous “constrained preference”. Given the option between negative real rates in a bank account or housing, Chinese citizens choose housing. But if they had the option of higher deposit rates, we might have seen a shift away from housing towards regular bank deposits. This is very similar to the problem developmental economists face when discussing whether people in poor countries have too many children. While parents might prefer to take care of fewer children, they choose to have more children so that enough of them survive to take care of their elderly parents. So similarly, poor people choose to have a lot of children given their constraints, but with enough financial innovation they would consider having fewer children.

Applied back to housing, these dynamics result in a world where the statements “there are too many houses” and “housing prices are too high” can both be true. Chinese people want housing, but they would prefer an alternative savings mechanism. Therefore, houses are built because there is the demand for the savings vehicle. But there are “too many” houses because the demand doesn't have to be this high. The high demand pushes up prices, pushing those who need the a roof over their heads into cold or other cramped conditions.

From an empirical standpoint, this story should manifest itself in a rising price to rent ratio as well as a lower housing yield. This is because if the house is seen mostly as a savings vehicle, those who own houses would be willing to take a lower rental rate, as their income growth is not dependent upon the flow of rental fees but rather on the growth in the value of the underlying asset. And as seen on page 12 of a 2010 IMF working paper, housing markets in cities such as Shanghai, Beijing, and Senzhen were “overvalued” in 2010 by this metric. More recent data on rental yields also supports this hypothesis, as rental yields in Beijing have fallen to around 2.2% while yields in Shanghai are around 2.8%, much reduced from above 9% and around 6% when the data was first collected.

This has serious implications for policy. First, housing controls such as one-house per family or severe taxes against house-flipping may not be very effective in stanching the speculative demand. The speculative demand is not from, as it was in the US, large banks trying to manufacture subprime mortgages to sell CDS. Rather, Chinese housing demand is from everyday people desperately trying to stretch their savings so that they have something solid to own many years from now. These “mom and pop” speculators are sometimes even willing to go through sham divorces to get around housing restrictions to make this saving method work. As the ant colonies of everyday workers get richer, they too will try to invest in housing for the purpose of saving money, only compounding this problem. This is all a natural outgrowth, and not some nefarious plot by one financial firm. In some ways, larger scale speculation may even be beneficial as it would encourage firms with enough capital to point out the lack of fundamentals in certain areas and help moderate prices.

Second, the question of housing investment is intimately related to efficient resource allocation. Because state-owned enterprises depend on low deposit yields to make a profit, they are a critical structural driver of continued housing price growth. We can't just write off billions of dollars of malinvestments and non-performing loans just because China is quickly growing. The continued support of enterprises that make those kinds of investments is a large reason why there is such excessive growth in housing. Moreover, these malinvestments prevent the necessary structural adjustment that Matt talks about. So unless we can address the problem of providing average Chinese citizens with more savings vehicles, we're not going to be able to address excessive growth in housing construction.

This is the reason why I believe that, besides housing, there are many things China could be building right now. I don't mean to deny the fundamentals story advocated by Scott and Matt, but while China does need a lot of housing to keep up with rapid urbanization, it does not mean that all the housing investment that is occurring at the margin is beneficial. In fact, the same IMF paper does point out that a change in the interest rate would have dramatic effect on price to rent ratios, highlighting the point that the Chinese could be pouring their resources into a lot of other places. Yet, if these places are not immediately obvious to us, it does not mean the market cannot use the higher price of credit to create more productive enterprise.

Third, demand for housing can be highly capricious, creating a dangerous situation for private companies and government alike. To play with expectations is to play with fire, and warning flags are popping up that investment demand for housingis running out. The worst case scenario would be some massive loss of confidence as people pull out of housing as it suddenly has lower yield than other wealth-preserving investments such as gold or jade. Such a drop in prices could shift expectations, driving yields lower. While the probability is uncertain, the fragility is certainly there. When I read stories about entire credit chains dependent on house sales, I have to wonder how severe the problem is. The fear is not about demographics or the fundamentals, it's the possibility of a market that, all of a sudden, falls apart. Perhaps this is yet another reason why there's too much housing; it fragilizes society to be dependent on houses for credit, revenue, and cash flow. And when the music finally stops, we all fall down.


  1. Even if the investment question is somehow fixed, I suspect it'll take awhile to deflate the bubble. Even Chinese outside of China arguably overinvest in residential real estate, and elite-level Chinese that have the ability to get money out of China also invest (or more properly, "park") it in property.

    This has inflated a RE bubble in Vancouver, BC, and propped up prices in areas of the US popular with Chinese, such as the SF Bay Area and the "Chinese" parts of the LA area.

    I think there's also cultural forces at work: China is a very low-trust investment environment, and has always been so, so investing in people or entities you don't know as well as family is a very steep cultural hill to climb. This will probably mean the RE bubble in China will deflate slowly versus having a sudden "pop".

    The thing that could cause a sudden pop is any sort of regime uncertainty, since RE in China is almost all leaseholds (quite sensitive to the whims of politicians) versus fee-simple titles (much harder to play political games with).

  2. I like your word "fragilize" but, with respect to Chinese housing, believe that there is a home ownership number below which the economy is not more fragile.
    That number is maybe about 50% -- and I have little data except the US 65% - 69% (at peak), and a very low (don't know number) West EU ownership percentage; plus in Slovakia (Central / Eastern EU) a strong desire to own.

    As long as China has fewer than 50% of its households as homeowners, it is highly unlikely to have "too many houses".

    It is still possible that too many expensive houses are built for the homebuyers -- based on builders not being able to sell them at a profit.

    So the empirical question is: how many Chines homebuilders are going bankrupt -- or stopping to build?

    tiny typo 3rd from last paragragh, missing "related to":
    the question of housing investment is intimately to efficient resource allocation

  3. I'm not sure many people who have spent significant time in recently-built Chinese housing would agree with the assessment that it "depreciates relatively slowly."

    Most residential buildings I've seen have experienced rapid physical deterioration. Also, new construction is much better today than it was even 5 years ago, and I would expect it to make similar leaps in the next decade. Taken together, I expect a building constructed in 2005 to be considered virtually unlivable by 2020.

    And when redevelopment is necessary, it becomes a question of unit owners' rights in the redevelopment process. This can be a thorny issue in the US. I can't imagine it will be much smoother in China.

  4. James, I've seen firsthand how poor some Chinese housing is. The way they deal with water damage is ridiculous, and there are always persistent problems with appliances breaking down. But even so, buildings do depreciate slower than say, farm machines, and the fact that they give a concrete service today probably makes them more attractive of an investment.

    You are probably correct in terms of the future difficulties, which seem to compound the issues we see in Chinese housing even more.