Land Supply and Housing Price: A Case in Beijing. Jinhai Yan

Transcription

1 Land Supply and Housing Price: A Case in Beijing Jinhai Yan Department of Land and Real Estate Management of Renmin University of China, Beijing P.R.China Abstract Recently housing price in Beijing has gone up fast and restrictive Land-supply policies have been blamed for this. This article analyzes effects of government s land-supply and developers land -bank on housing price. A stock-flow model is developed in which land-supply and land-banking is incorporated. Based on data for the periods, the results show that land supply has a significant negative effect on housing price and the effect in land reduction stage is more than one in land expansion stage. We also find that more land in developers land banks, faster housing price rises. Key words Land supply, land-bank, housing price Introduction Since housing system reform was further implemented in 1998, housing market in Beijing has experienced a quick development. Floor space of selling house in 2005 is almost 7 times of one in With the market booming, housing price has continued to rise quickly in recent years. Several arguments about why housing price has risen so fast have been made by the society. One representative opinion is that the reduction in land supply since 2004 has caused the shortage of residential land and houses and then pushed the price to rise. Another argument demonstrates that developers have enormous land banks and oligopoly s nature in housing market is the source of housing prices rising. These two opinions analyze from the aspect of supply. Some scholars also study housing demand and say that the factors affecting demand such as changes of population, increasing income, speculation and redevelopment are also important reasons. In long term, because of land resource s scarcity, housing supply is relatively restricted compared with booming demand. It is meaningful both in theory and practice to analyze thoroughly the relationship between restrictive land supply, developer strategy and housing price fluctuation in order to discern the reason why housing price rises and help government regulate market 1

2 accordingly. The rest of the article is organized as follows. Section 2 reviews related documents. Section 3 describes the theoretical foundation and put forward with three propositions. Section 4 analyzes the characteristic of Beijing housing market and tests propositions. The last section concludes the article. Literature review Studies on the effect of land supply on housing market in China mainland almost only put forward with arguments. As for a few empirical study, Yangchongguang etc(2007) and Lushixiong(2007) estimate the effect and find the relationship between land supply and housing price is negative by using the series data of 1997 to 2005 in China mainland and the panel data of 1999 to 2005 in 31 provinces of China individually. Similar to land supply system in China mainland, the government in Hong Kong also controls land supply. Researches on the effect of land supply on housing market in Hong Kong are relatively thorough and also come out to two different arguments. Tam Yin Bun(1993)finds that if the supply of new land is increased by 20,000 square meters in three consecutive years, the private residential prices will drop by HK$178 per square meter. Ruijue Peng, Wheaton(1994)develops a multiequation aggregate housing model in which government land supply is incorporated in both housing demand and supply and the empirical results show that supply restrictions have caused higher housing prices through rational expectation. In contradiction with the above opinion, Raymond Y.C.Tse(1997)shows that there is no causal relationship between land supply and housing prices by using Granger causality test and developers have held quite a substantial amount of land in land banks. Lai siu fun rita(1993) Percy Ting(1998)analyzes developers role in housing market and points out that oligopoly s nature is the main reason. Neng Lai, Ko Wang (1999)examines the relationships among developers housing-supply decisions, government land-supply decisions and public housing policies and finds out that an increase in land supply may not be a solution to the perceived shortage of housing supply in Hong Kong. In all, it is clear that the relationship between land supply and housing price in Beijing has yet to be studied in-depth. This article attempts to undertake such a study. Theoretical analysis Stock-flow model is a classical model used to study housing market. This article incorporates land supply in both housing demand and supply as Ruijue Peng, Wheaton (1994). Besides this, land bank is also included in the model. In all, the inter-relationships among land supply, land bank, housing supply and price can be revealed by Figure 1: 2

3 Land price Housing demand Land supply Housing supply Housing price Land bank Figure 1. The inter-relationships among land supply, land bank and housing price In the theory we assume that: (1) land supply is monopolized by the government. Developers bid for land in public auction market; (2) developers are price-takers and make land-development decisions based on profit-maximization principle. Moreover they can t adjust floor area ratio without government s permission; (3) consumers are rational and there are no differences in cost between buying and renting houses. Demand for housing service D mainly depends on demographic characteristics H, permanent t incomey and the annual user cost of housingu. With increasing population, smaller family size and decreasing user cost, demand will increase (equation1). D = au + X ( H, Y) (1) t t t According to assumption, housing price must equal to the capitalization of user cost of houses (equation 2). U = P[ δ + i E ( Δ P, LS)] (2) t h The capitalization ratio depends on housing economic rate of depreciationδ, the prevailing market long-term interest or mortgage rate i and the expected rate of future housing price appreciation E. In reality, the formation of expected appreciation is a mix of both adaptive and h rational behavior (Ruijue Peng,Wheaton,1994). On one hand, consumers may speculate about future price appreciation based on price fluctuations ΔP in recent dates. On the other hand, the condition of land supply LS can affect consumers anticipation about future housing supply and housing price. When land supply decreases, consumers will rationally predict that housing supply in long-term will be suppressed and the price will rise. So they will adjust their myopic expectations. With the expectation of more increasing housing price appreciation in the future, 3

4 consumers will increase their current demand. The supply of housing will change over time as the existing stock depreciates at rateδ and as new construction S t is added up to the stock (equation 3). Besides housing price Pt 1 and construction costct 1, new construction is also affected by land supply LS (Tam Yin Bun,1993;Ruijue Peng,Wheaton,1994). In long-term, with fixed floor area ratio, future construction will slow when available land reduces. Then the stock of housing fails to keep pace with demand. Moreover, as many scholars(lai siu fun rita,1993;raymond Y.C.Tse,1997;Percy Ting,1998) say, developers will not develop land immediately after their land acquisition, in contrast, they will add new land into land banks. Barlow(1993) argues that in real estate industry, skills in land banking behavior and timing completions to maximize gains from price inflation tend to be more important than the ability to compete through technical innovation. So if developers add more new land to their land bank LB, the new construction at present will decrease (equation 4). TS = S + (1 δ ) TS (3) t t t 1 S ( ) t = bpt 1 Ct 1+ clst 1+ dlbt 1 Δ P (4) Based on above analysis, three propositions are put forward with: Proposition1: In long-term housing price will rise when land supply continues to decrease. Proposition2: More land in developers land banks, faster housing price rises Proposition3: During the stage of land supply s expansion, the effect of land supply s increase on housing price will be less than one in the stage of land supply s reduction. Empirical analysis The characteristic of Beijing housing market Housing supply and demand After market oriented reform was established in 1992, housing reform shifted to transforming mechanism. With government deregulating market, many developers flowed into housing market and housing investment increased. Floor space of residential buildings completed in 1995 rose by 68 percent than one in After 1992 housing market experienced a temporary adjustment period. During the time of , housing demand was dominated by group purchasing. In order to prompt consumption and quicken housing system reform, in 1998 the government decided to stop real houses distribution and establish new housing system in which consumers must buy houses directly in the market. With land, credit and taxation policies support, housing market boomed quickly. Consumer purchasing individually dominates the market and the 4

5 amount of non-registered residents purchasing increased fast also. Floor space of housing completed in 2005 was 4.8 times of one in 1998 while floor space of housing sales 5.8 times. However, in 2006 floor space of housing completed and sales all decreased a little. (Ten thousands of Square Meters) Sources: Beijing Statistical Bureau and Beijing Municipal Construction Bureau. Figure 2. Floor space of housing completed in Beijing since 1992 (Ten thousands of square meters) Sources: Beijing Statistical Bureau and Beijing Municipal Construction Bureau. Figure 3. Floor space of housing sales in Beijing since 1992 Housing price Because housing hedonic price index or repeated sale price index has not been established in Beijing, the article gets housing price by dividing sales of housing by space of housing. As figure 4 shows, housing prices also experience three different stages. From 1992 to 1997 housing prices rose fast by the average nominal speed of 28 percent or 10 percent deflated by CPI. For the period of 1998 to 2004 prices almost kept stable and only dropped a little in the total trend. However, since 2005 prices have continued to increase fastly. 5

6 8000 (Yuan per square meter) Housing price Housing price adjusted by CPI Sources: Beijing Statistical Bureau and Beijing Municipal Construction Bureau. Figure 4. Housing price in Beijing since 1992 Land supply Since land market-oriented reform in 1992, the amount of residential land supply in public remise market has risen irregularly while the amount of land allocation mainly used to construct houses for low-income urban residents relatively stable. Before 2004 land negotiation leasing dominates the remise market while from 2005 all land for profitable use including commercial residential land must be leased to the highest bidder through bidding or auction selling. As figure 5 shows, residential land supply can be divided into three different stages: from 1992 to 1997 at the beginning of market-establishing land supply firstly soared in 1993 then kept stable for two years while in 1996 and 1997 decreased again. Over the 1999 to 2004 period when housing market boomed land supply increased quickly, especially after hectares land averagely per year was added. In 2004 affected by the 8.31 policy, much residential land was supplied in advance which amounted to 1290 hectares statistically by Beijing Land Resource Bureau. Since 2005 after the government adopted the policy that restrictive developable land is supplied through bidding or auction selling, the area of land supply has dropped quickly (Hectares) Land remise Land allocation Sources: Beijing Land resource Bureau. Figure 5. Residential land supply in Beijing since 1992 Compared Figure 2 and Figure 4 with Figure 5, we can find that the area of residential land supply 6

7 changes in the same direction as floor space of housing completed in long-term while contrary to housing price appreciation. In the stage of land supply expansion, housing price almost kept stable while it would rise fast when land supply induces. At the same time, based on investigation by Beijing Statistical Bureau, the amount of undeveloped land in developers land banks has continued to increase quickly also since The area in 2004 was 7.6 times of one in 1999 while the amount in 2006 increased by 38 percent than in 2004 again. It is reasonable to suspect that large amount of undeveloped land caused shortage of housing supply and rising of housing price. The above two arguments maybe all can be used to explain why housing price in Beijing has increased so fast Estimation model From equation 1 to 4, we end up with the following economics model: P = α + αy + α H + α i + α Δ P + α LS + α LB + α DLS (5) t 0 1 t 2 t 3 t 4 t 5 t 6 t 7 t To reveal whether there are significant differences between effect of land supply on housing price in land supply expansion stage and reduction stage, a dumb variable D is added which equals to 1 in expansion stage and 0 in reduction stage. Definitions of variables in this model are as follows: Y : Discretionary income per capita of urban household deflated by CPI. i : Real one-year loan interest rate adjusted by inflation rate. H : Urban permanent population. Based on statistical data from Beijing Municipal Construction Bureau, non-registered residents purchasing has accounted for one out of three in total housing sales at present. These people mainly buy houses to live by themselves rather than invest. In order to represent the population factor affecting housing price correctly, urban permanent population rather than registered population is considered. LS : Area of residential land supplied by the government through both land allocation and land remise. LB : Cumulative area of undeveloped land reserved by developers. Δ P : Housing price appreciation rate computed by the formula Δ Pt = ( Pt Pt 1)/ Pt 1. All data of 1992 to 2006 come from Beijing Statistical Bureau, Beijing Municipal Construction Bureau and Beijing Land resource Bureau. Estimation results Because interest rate and price appreciation rate are not significant in the regression model, all cases in Table 1 have omitted these two variables. Case 1 and case 2 compares the effect of current 7

8 land supply with one-year lagged land supply and the results suggest that current land supply has more negative effect on housing price. Case 3 shows that the value of both R 2 and D-W statistic increase with the addition of the dumb variable. Thus, the hypothesis that the effect is different between expansion stage and reduction stage cannot be rejected. Case 4 and case 5 further examine the effects of current and one-year lagged land bank on housing price. The consistently positive correlation between land bank and housing price suggests that developers development strategy has contributed to rising housing price significantly. The final estimation equation is written as follows: P= Y H 1.83LS LB( 1) DLS (6) Table 1. Ordinary Least Square Estimates of Housing Prices (t statistics in Parentheses) Indenpent Variable Case 1 Case 2 Case 3 Case 4 Case 5 Y 0.17 (2.68) 0.19 (3.54) 0.27 (4.18) 0.21 (2.97) 0.14 (1.92) H 2.98 (5.01) 2.72 (5.12) 2.46 (4.92) 2.60 (5.41) 3.12 (6.75) LS (-2.09) (-2.33) (-2.90) (-2.68) LS(-1) -0.4 (-1.9) DLS 1.32 (1.88) 2.19 (2.52) 1.31 (2.35) LB 0.50 (1.54) LB(-1) 0.51 (1.98) R D-W From Table 1, we can see case 5 passes all tests and the goodness of fit of the model is best, at 83 percent. Land supply has negative effect on housing price. The coefficient in expansion stage is 0.52 less than 1.83 in reduction stage. That s to say, when the government increases the area of current land supply by 100 hectares in the stage of land expansion housing price will drop only by 52 yuan per square meters while 183 yuan per square meters in reduction stage. The positive 8

9 coefficient between land bank and housing price indicates that if this 100 hectares land is absorbed by developers into their land banks, the price will rise by 51 yuan per square meters. The existence of enormous areas of undeveloped residential land in reality suggests that housing price s soaring relates to developers land reserve very much. Based on statistical data of Beijing Land Resource Bureau, among the total 512 plots, 5800 hectares land leased from January 2004 to June 2006 by the government, nearly 3000 hectares land has not been constructed at the end of June There are three main reasons: firstly, because land market was still being established before 2004, when land was released through negotiation leasing many developers absorbed it into their land banks in order to develop or transfer to others in the future to get high price appreciation with the anticipation that land price would rise fast after land bid or auction policy strictly adopted after Secondly, since 2005 the government has regulated and controlled land and housing market several times. This uncertainty about market and policy has made developers postpone and reduce current building construction and wait for some days before policy classified and market better by reserving land as real option. Thirdly, because negotiated leased land before 2005 was mainly raw land of large areas, it would take 4-5 years to complete land appropriation and household-removal with higher acquisition cost and complex land use rights. Under such current condition that lack of land holding taxation, low loan cost, more scarce land and increasingly rising land and housing price, developers would not be willing to quicken raw land s development or transfer land even they don t have the ability to develop until loan cost and taxation fees overtook net profit gained from price appreciation anticipation. Conclusions and suggestions The article analyzes the relationship between land supply and housing price in Beijing. The results show that land supply has negative effect on housing price and the effect in reduction stage is more than in expanded land supply stage. In the current condition that housing price is rising fast, the government should determine when and how much land will be put on the market scientifically after investigating all undeveloped land s situations. In long-term, on the background that land becomes scarcer in Beijing in the future, the government should make a comprehensive land supply plan based on City Plan and Land Use Plan and publish it to the public to direct and stable market participations anticipations. The empirical results also suggest that more undeveloped land in land banks faster housing price rises. Faced with the face that large areas of undeveloped land exit, it is not feasible to take administrative measures such that land use right will be cancelled if land hasn t been developed in two years because these measures not only can t been well implemented technically with developers many counteractions but also they don t accord with market efficiency and fair. Some 9

10 economic measures such as land holding taxation and expensive loan cost should be taken to establish efficient incentive-constraint mechanism to prompt developers develop land in time rather than absorb land in their land banks. At the same time the policy of supplying developable land rather than raw land in bid or auction market should be implemented seriously to reduce the cycle time among land acquisition, land development and building completion. References Yangchongguang, Fenglei, 2007, Research on applying land policy to regulate macro economy, Chinese Geological Publishing House. Lushixiong, 2007, study on the effect of land supply on housing market in China, Master dissertation of Renmin University of China. Tam yin bun, influence of land supply on rising residential prices, University of Hong Kong Ruijue Peng, Wheaton, 1994, effects of restrictive land supply on housing in Hong Kong, Journal of housing research, Volume 5, Issue 2 Raymond Y.C.Tse, 1998, Housing price, land supply and revenue from land sales, Urban studies, Vol.35, No.8 Lai siu fun rita, 1993, Housing price and government land policies, University of Hong Kong Percy ting, 1998, Property prices is independent of the amount of land supply, University of Hong Kong Neng lai, Ko wang, 1999, Land-supply restrictions, developer strategies and housing policies, International real estate review, Vol.2 No.1 10