The Impacts of Land Leases on Housing Value by bvt11437

VIEWS: 7 PAGES: 16

									                        Land Leases and Housing Values




Yong Tu*                                         Helen X. H. Bao
Department of Real Estate                        Department of Management Sciences,
National University of Singapore                 Faculty of Business,
Singapore                                        City University of Hong Kong
Tel: 65 68744455                                 Kowloon, Hong Kong
Fax: 65 67748684                                 Tel: (852) 2788 7269
E-mail: rsttuy@nus.edu.sg                        Fax: (852) 2788 8560
                                                 E-mail: msxhbao@cityu.edu.hk




This is a structured abstract based on our on-going work, prepared for AREUEA 2006.




*Please forward correspondence to the first author by e-mail at rsttuy@nus.edu.sg or by
mail to The Department of Real Estate, National University of Singapore, 4 Architecture
Drive, Singapore 117566. All comments are welcome.
Abstract:


This paper attempts to reveal how and why land leases and land lease structure may
affect urban housing prices. The paper is motivated by the valuation of leasehold property
relative to freehold property. The results may have implications to efficient land uses
when leaseholds are in force. The conceptual framework developed in the paper aims at
providing a comprehensive understanding on how land leases affect housing value. The
framework draws from the approaches of properties rights as well as the options to
redevelop that is an issue arising from land tenure security. Market psychology may also
help to explain the premium fetched by freehold property when freehold land supply is
constrained.


Since Singapore and Hong Kong represent two different but comparable land lease
structures, housing transaction data in Singapore (1990-2003) and Hong Kong (1992-
2004) are used in the empirical study. The paper develops a set of comparative hedonic
analyses to test the hypotheses: first, freehold property can fetch a higher premium;
second, the premium will be enhanced during economy booms and will be different under
different land lease systems; third, constrained freehold supply may cause market
psychology leading to an even higher freehold premium.


A two order spatio-temporal autoregressive model with Bayesian heteroscedasticity
robust procedure is employed in the empirically analysis.


The following structured abstract will provide more information about this on-going
work.




                                             2
Introduction


Leasehold properties are popular in many jurisdictions, both with private vendors and
with local governments who want to retain future controls over their land uses. A
question from both vendors and purchasers has been how to value leasehold property in
relative to fee simple property (Capozza and Sick 1991). Empirically, real estate literature
has revealed that the pricing of freehold property is generally higher than that of
leasehold property. Freehold premium can be significantly different across markets with
different land lease structures. This gives rise to the questions of why freehold properties
can fetch a premium, why the premiums across international markets are different, and
will the premiums change along real estate cycles?


Asabere (2004) shows that, in Ghana, freehold property prices fetch a premium of 7.5%
in relative to leasehold, supporting the bundle of right argument. That is, everything
being equal, freehold properties should be more valuable that leasehold properties since
the bundle of rights inherent in freehold properties is typically superior to that found in
leasehold properties. He further points out that some institutional regulation that was
aimed at constraining the supply of freehold properties can enhance the premium up to
17.5%.


Capozza and Sick (1991) point out that in Canada, leasehold property’s price can be 40%
lower than the comparable freehold property. Discount cash flow cannot explain such big
price differential. An explanation from the angle of the option to redevelop is offered.
Since the option to upgrade the property is valuable and repeated redevelopment adds
more value to the assets (Williams 1991, 1997), leased property should be traded at a
significant discount because of impairment of the option.


Chau et al (2005) find that in Hong Kong, the premium is 8% and the premium will be
higher if real estate return is lower. Knight and Balihuta (2004) find evidence in Uganda
that the premium is 13.8%. However, Janssen (2003) shows that in Sweden, the effect of
freehold versus leasehold on housing price appears to be insignificant, due to well-



                                              3
founded expectations in the market that renewal will be fairly routine and that the new
lease will be similar to the old one.


The above international literature illustrates that there are price differentials between
freehold and leasehold properties. However, the understandings on the reasons behind the
price differential as well as the variations of price differential across markets are still not
comprehensive. The dynamic pattern of freehold price premium is unclear.


This paper attempts the above issues. We first develop a conceptual framework that is
aimed at providing an understanding on price differentials between freehold and
leasehold properties as well as the possible variations of the differential across markets.
Second, the paper will conduct a comparative hedonic analysis of the pricing of leasehold
versus freehold properties in the condominium and apartment markets of Singapore and
Hong Kong. The empirical analyses will provide evidence on the magnitudes and the
dynamic patterns of price differential between freehold and leasehold properties, as well
as the variations of price differential across markets under different land lease structures.
These empirical results will hopefully support our arguments raised in the conceptual
framework.


Institutional Settings in Singapore and Hong Kong


In Singapore, the earliest land lease was issued in 1826 with a lease term of 999 years.
With the ever increasing demand for land to be sold outright or leased for a term not less
than 99 years, the year of 1838 witnessed the emergence of the 99-year lease. From 1845
onwards, with agricultural prosperity and the demand for greater security of tenure,
Indentures (fee simple or freehold title) for land outside the town limits were issued to
encourage agricultural development. In 1886, the Crown Land Ordinance, which sought
to substitute for the grant in fee simple, issued a form of tenure called the ‘Statutory Land
Grant” (a grant in perpetuity, subject to quit rent and conditions). If the grant of fee
simple tenures was terminated in 1886, the Statutory Land Grant also ceased as a form of
title in 1919 (Motha and Yuen, 1999).



                                               4
Between 1990 and 2003, in Singapore private condominium market, 58.9% of
transactions are freehold properties or leasehold properties with lease term above 900
years. In this paper, we define freehold properties as those properties with perpetuity land
lease or those properties with a lease term of above 900 years. Leasehold properties are
those properties with a lease term of 99 years. There are three lease structure related
characteristics associated with this market.


First, the grant of freehold tenures was virtually terminated in the earlier 20 century. The
supply of freehold land has been constrained and in fact it has been diminishing over the
years.


Second, for the leasehold properties, when the lease term is matured, the land as well as
the buildings attached to the land will be reverted to the landowners. The fact of reversion
is known and any purchasers for value toward the close of the term will have purchased
with their eye open and adjusted their price accordingly. However, the leasehold land
purchasers can apply for the extension of land lease, subject to the approval. Since 31
July 2000, the Singapore land Authority has implemented a transparent system of
determination of differential premium for the lifting of State title restrictions involving
change of land use, extension of land leases as well as density.


Third, Singapore allows homebuyers to use their pension contributions (a percentage of
their monthly salary, matched by the contribution from employers) to pay off their
mortgage. Recently, it is also allowed to use such contributions to pay off certain portion
of down payment. However, a regulation has been in force on the use of pension
contributions to buy leasehold properties. For any leasehold property (99 years), if the
number of years to run is less than 60 years at the time of transaction, the would-be
homebuyers are not allowed to use pension contributions to buy their properties. As a
follow-up, banks have a similar term in mortgage landing. For those leasehold properties
with valid land lease less than 60 years, the bank will not grant mortgage to the new
buyers. Recently, this rule was relaxed to 50 years.




                                               5
In Hong Kong, the first land auction was held on 14 June 1841. Land lease term then was
75 years and non-renewable. The government extended the lease term to 999 years in
1849. However, it was soon realized that the lease term was too long for the government
to catch future land value appreciation effectively. Therefore in 1898 the lease term was
changed back to 75 years with a right of renewal. Leases for land in the New Territories
were normally issued for the residue of a term of 99 years less three days from 1 July,
1898, the time when the 99-years lease of the New Territories starts. Based on the Sino-
British Joint Declaration, all new land leases issued after 1997 are under the term of 50
years (non-renewable, meaning that the renewal is subject to approval and a premium is
charged). Old land leases that do not expired on 30 June 2047 will be honored.

All land leases in Hong Kong can be classified into two categories: renewable and non-
renewable. The renewable leases are entitled to an automatic renewal under the same
lease terms. The lessees are not required to pay a new premium upon renewal. To extend
a non-renewable land leases, the lessees must apply for a “Conditions of Regrant” from
the government as early as 20 years before the expiry date. If the application is declined,
the government could reclaim the land after the expiry date without paying any
compensation to the lessees. In practice the government usually grants the renewal
unless the land is needed for other purposes. If the application is approved, the lessee(s)
has to pay a premium and a re-valuated annual rent in order to secure a new land lease.
The premium is usually equivalent to the full market value of the land at the date of
renewal. Between1992 and 2004, 23% of transactions in Hong Kong are freehold
transactions. By definition, freehold in Hong Kong means that the land leases are more
than 99 years. Leasehold in Hong Kong means typically those properties with a land
lease term of 75 or 50 years.


Singapore and Hong Kong land lease structures are different but comparable, providing a
good basis for us to investigate how different lease structures affect housing value. First,
in terms of pricing, 999 year land lease in Hong Kong is comparable to freehold and 900
year land leases in Singapore. In both market, freehold (or equivalent) land supply is not
only constrained, but also diminishing. Second, in Singapore, the renewal of leasehold



                                             6
properties is subject to approval and a premium is typically charged. In Hong Kong,
leasehold properties are divided into two categories: automatically renewable and non
renewable, meaning that renewal will be subject to approval and premium charges.


Conceptual Framework


As addressed in Allen (2002), Demsetz (1966)’s paper on property rights contains at least
three critical ideas that have stood the test of time and have become fundamental in
subsequent literatures. First, Demsetz stresses that the term “property rights” should not
be restricted to the narrow legal notion of rights to property under law, but should be
extended to include the ability to make all manner of choices. Second, property rights
emerge when the benefits of establishing those rights exceed the cost. Third different
patterns of property rights lead to different patterns of behavior. Barzel (1997) extends
the first point as that the terms of property rights carries two distinct meanings: economic
property rights meaning the ability to enjoy the property and legal property rights
meaning what the state assigns to a person. Pejovich (1990) further identifies that the
right of ownership is a category of the general concept of property rights, containing the
following four elements: (1) the right to use an asset, (2) the right to capture benefits
from that asset, (3) the right to change its form and substance, (4) the right to transfer all
or some of the rights specified under (1-3).


Applying the above arguments to land lease, it is concluded that each type of land lease
may be associated with a bundle of rights, which may lead the land owners (or buyers)
associated with the different land lease to behave in a different ways. For example, the
buyers may like to offer a higher price for the land with a lease including more rights.
Among property rights associated with land ownership (Pejovich 1990), some rights may
affect the options of land redevelopment, for example, for leasehold property, when lease
term comes to close, the right to capture benefits from the asset, and the right to change
its form and substance may be constrained. As a result, although the land may have a
higher redevelopment value, the owner cannot capture the benefit due to the end (or
closer to the end) of land lease term. In other words, freehold property should have a



                                               7
higher value than leasehold. When economy is good, land redevelopment value may be
higher, resulting in a higher freehold premium. Since different land lease system may be
associated with different type of land rights (for example the cases in Singapore and
Hong Kong), the premium fetched by freehold properties may be different.


In both Singapore and Hong Kong, the supply of freehold land was terminated and is in
fact diminishing. Such supply constraint, together with other institutional constraints on
leasehold properties may lead to a market psychology: buyers may pay a price higher
than market price in order to obtain a tenure security generated by freehold properties.


The hypotheses derived from the framework will be tested against a set of empirical
hedonic functions. In summary, the hypotheses are, freehold property can fetch a higher
premium, and the premium will be enhanced during economy booms and will be different
under different land lease systems. Constrained freehold land supply may cause market
psychology leading to an even higher freehold premium.


Data


In Singapore, our primary database is derived from Singapore Institute of Surveyors and
Valuers (SISV) transaction database, including all condominium transactions from
January 1 1990 to June 30 2003. After deleting several hundred compulsory foreclosed
transactions, there are 74,989 observations in the database. The database includes details
of address, dwelling related hedonic factors as well as contract and transaction dates. The
condominium project and neighborhood related spatial information is added to the
database. The spatial information is mainly obtained from Singapore Street Directory. All
data is geo-coded at building level, noted that each condominium may have a few
buildings and each building corresponds to one x-y coordinate. In the appendix, Table A1
gives the definition of all variables. Table A2 gives some description of data. Table A3
gives land lease structure.




                                             8
In Hong Kong, 249,954 condominium transactions between January 1992 and February
2004 are obtained from three local sources. The Land Registry provides transaction
prices and dates. The details on neighborhood and structural characteristics are obtained
from Centaline Property Agency Ltd., one of the major players in the Hong Kong
residential property market. The geographic information for each building is from the
Lands Department. The data set covers 59 large scale housing estates from the three
districts in Hong Kong (Hong Kong Island, Kowloon and New Territories). In the
appendix, Table A4 gives the definition of variables and Table A5 gives some descriptive
statistics of data. The land lease structure in the Hong Kong dataset can be found in
Table A6.


Econometric Implementation and Further Work


Two issues must be considered in empirical hedonic function in order to ensure that the

estimated coefficients are accurate. One is the spatio-temporal autocorrelation, the other

one is heteroscedasticity. A two order spatio-temporal autoregressive model with

Bayesian heteroscedasticity robust procedure (equation 1)



Y = Xβ − W1 XβW1 − W2 XβW 2 − TXβT − STXβ ST − TSXβTS
                                                                                   (1)
        + φW1W1Y + φW2 W2Y + φT TY + φST STY + φTS TSY + u

where   βW , βW , βT , βST , βTS are the vectors of parameters indicating φW β , φW 2 β , φT β ,
          1    2                                                             1




φ ST β , φTS β separately, W1 X ,W2 X ,TX, STX,TSX are the spatio-temporal lags of

independent variables, and W1Y ,W2Y , TY , STY , TSY are the spatio-temporal lags of the

dependent variable. Instead of using OLS estimation, we use Bayesian estimation to

correct heteroscedasticity.




                                                 9
This paper will hopefully have important implications for property valuation of leasehold
versus freehold, as well as efficient land use when leases are enforced.


The on-going work includes a further development of conceptual framework, a research
design to illustrate how to test the hypotheses arising from the conceptual framework, as
well as a completed set of empirical analyses against the data in Hong Kong and
Singapore. The comparative empirical results will be used to test the hypotheses.




                                            10
References


Asabere Paul K. (2004) The Pricing of the Emergent Leasehold (Possessory) Estates of
Ghana. Real Estate Economics, V 32 4: 673-694.

Allen D. W. (2002) The Rhino’s horn: incomplete property rights and the optimal value
of an asset. Journal of Legal Studies, Vol XXXI: 339-358.

Barzel, Yoram (1997) Economic analysis of property rights. 2nd edition. Cambridge
University Press.

Besley, T. (1995) Property rights and investment incentives: theory and evidence from
Ghana. Journal of Political Economy, Vol 103(5): 903-937.

Chau KW, Wong SK and Liu CY (2005) Valuation of Leasehold Properties. Working
Paper in The University of HongKong.

Capozza D and Sick G A (1991) Valuing long tern leases: the option to redevelop.
Journal of Real Estate Finance and Economics, Vol 4 : 209-223.

Demsetz H. (1966) Towards a theory of property rights. American Economic Review,
57:347-373.

Dale-Johnson D. (2001) Long-term ground leases, the redevelopment option and contract
incentives. Real Estate Economics, Vol 29(3): pp 451-484.

Janssen Christian T.L. (2003) Estimating the effect of land leases on prices of inner-city
apartment buildings. Urban Studies Vol 40 (10): 2049-2066.

Kim A.M. (2004) A market without the right property rights: Ho Chi Minh City.
Economics of Transition, Vol 12(2):275-305.

Knight J. R., Herrin W. and Balihuta A. M. (2004) Housing prices and maturing real
estate markets: evidence from Uganda. Journal of Real Estate Finance and Economics.
28:1, 5-18.

Motha, P and Yuen, B. K. P. (1999) Singapore Real Property Guide. 4th edition.
Singapore University Press.

Payne G. (2001) Urban land tenure policy options: titles or rights. Habitat International,
25: 415-429.

Pejovich, S. (1990) The economics of property rights: towards a comparative systems. 1st
edition.




                                            11
Williams J. T. (1991) Real Estate Development as an Option. Journal of Real Estate
Finance and Economics. 4:1991-208.

Williams J. T. (1997) Redevelopment of real assets. Real Estate Economics. V25 (3):387-
407.




                                          12
                                              Appendix


 Table A1 Variable Definition


Variable     Definition (measurement)
             The dependent variable measured by log of transacted price, used in the second step to
Ln_price     predict housing price
                                       Independent variables
Area         Built-area measured in sq.m which indicates the size of the apartment
Level        Floor level of the apartment
Freehold     Freehd_1=1, when a property is “freehold”. Or freehd_1=0
Age          Age of the property (number of quarters)
Leaserun     Number of lease years to run for leasehold properties
Bbq          Dummy variable with 1 indicating that the condo has a Barbecue area, otherwise 0
Carpark      Dummy variable with 1 indicating that the condo has a Covered car park, otherwise 0
GYM          Dummy variable with 1 indicating the condo has a Gymnasium, otherwise 0
Jacuzzi      Dummy variable with 1 indicating the condo has a Jacuzzi, otherwise 0
Fitness      Dummy variable with 1 indicating the condo has a Fitness area/jogging track, otherwise 0
Minimart     Dummy variable with 1 indicating the condo has a Minimart, other wise 0
MPH          Dummy variable with 1 indicating the condo has a Multi-purpose hall, other wise 0
Plargrou     Dummy variable with 1 indicating the condo has a Playground, otherwise 0
Sauna        Dummy variable with 1 indicating the condo has a Sauna, otherwise 0
Squash       Dummy variable with 1 indicating the condo has a Squash court, otherwise 0
Swimming     Dummy variable with 1 indicating the condo has a Swimming pool, otherwise 0
Tennis       Dummy variable with 1 indicating the condo has a Tennis court, otherwise 0
Wading       Dummy variable with 1 indicating the condo has a Wading pool, otherwise 0
Security     Dummy variables with 1 indicating the condo has a 24-hrs security system, otherwise 0
Totaluni     Total number of units in a condo
MRT          Refers to linear distance to the nearest MRT Station (km)
CBD          Refers to the linear distance to CBD (km)
School       Distance to the nearest popular primary school (top 30 primary schools)
Location-1   Dummy variable with 1 indicating unit location City and South west, otherwise 0
Location-2   Dummy variable with 1 indicating unit location Orchard/Tanglin/Holland, otherwise 0
Location-3   Dummy variable with 1 indicating unit location Newton/Bt. Timah/Clementi, otherwise 0
Location-4   Dummy variable with 1 indicating unit location Balestier/MacPherson/Geylang, otherwise 0

Location-5   Dummy variable with 1 indicating unit location East Coast, otherwise 0




                                                   13
Location-6      Dummy variable with 1 indicating unit location Changi/Pasir Ris, otherwise 0
Location-7      Dummy variable with 1 indicating unit location Serangoon/Thouson, otherwise 0
Location-8      Dummy variable with 1 indicating unit location West, otherwise 0
Location-9      Dummy variable with 1 indicating unit location North, otherwise 0



 Table A2 Data Description

 Variables                          Average                            Std
 Transaction Price (S$)             881188.56                          544243.344
 Area (sqm)                         137.12                             57.729
 Level                              7.36                                5.81
 Age (Years)                        3.23                                5.11
 Leaserun (Years)                   92.94                               6.082
 Totaluni                           375.04                              300.90
 MRT (km)                           1.37                                0.83
 CBD (km)                           8.89                                4.29
 School (km)                        1.5457                              1.2389
                                             Frequency variables
 Freehold                           58.9%
 Bbq                                68.3%
 Carpark                            84.3%
 GYM                                63%
 Jacuzzi                            21.3%
 Fitness                            43.5%
 Minimart                           9%
 MPH                                42.8%
 Plargrou                           81%
 Sauna                              56.1%
 Squash                             62.6%
 Swimming                           95.1%
 Tennis                             77.2%
 Wading                             76.1%
 Security                           91.2%
 Location-1                         9.5%
 Location-2                         20%
 Location-3                         15.4%
 Location-4                         7.7%




                                                     14
Location-5                         17.5%
Location-6                         6.3%
Location-7                         7.7%
Location-8                         10.6%
Location-9                         5.4%



Table A3 Lease Structure in Singapore Dataset

Types of leases          Frequency                   Average prices               Std
Freehold                 37,049 (49.4%)              1,015,875.2                  668,825.945
999 year lease           6,554 (8.7%)                934,592.80                   391,852.802
956 year lease           116 (0.2%)                  672,227.59                   157,696.190
947 year lease           45     (0.1%)               1,013,548.1                  430,217.791
946 year lease           59     (0.1%)               765,190.73                   156,506.184
929 year lease           373 (0.5%)                  839,607.37                   278,673.447
99 year lease            30,793 (41.1%)              709,091.61                   311,159.944
All                      74,989                      881,188.56                   544,243.344

Table A4 Variable Description for Hong Kong

       Variable                                               Description
 L_price                Log transformed selling price per square foot
 Floor                  Floor level of the unit
 Garea                  Gross sellable area of the unit
 Leaserun               Number of lease years to run
 HK                     =1 if the unit is located in Hong Kong Island and 0 otherwise
 KL                     =1 if the unit is located in Kowloon District and 0 otherwise
 Age                    Age of the unit at the time of sales
 Freehold               =1 if the unit has more than 99 year to run in the land lease and 0 otherwise
 Seav                   =1 if the unit has a sea view and 0 otherwise
 Transport              =1 if the unit has good access to public transportation and 0 otherwise
 Y1993 – y2004          Time dummy variables


Table A5 Descriptive Statistics for Hong Kong

             Variable              Mean              Std Dev           Minimum            Maximum
 L_price                            8.22               0.41              6.22              10.58
 Floor                             15.68               9.68                0                67
 Garea                             734.62            303.05              136               7098
 Leaserun                          128.49            222.98               30                908
 HK                                 0.29               0.46               0                  1
 KL                                 0.27               0.44               0                  1
 Age*                               9.15               7.54               -2                35
 Freehold                           0.23               0.42               0                  1
 Seav                               0.36               0.48               0                  1
 Transport                          0.29               0.45               0                  1


                                                  15
 Y1993                                0.12               0.32               0      1
 Y1994                                0.09               0.29               0      1
 Y1995                                0.08               0.28               0      1
 Y1996                                0.13               0.34               0      1
 Y1997                                0.14               0.35               0      1
 Y1998                                0.07               0.25               0      1
 Y1999                                0.05               0.22               0      1
 Y2000                                0.04               0.20               0      1
 Y2001                                0.05               0.21               0      1
 Y2002                                0.04               0.19               0      1
 Y2003                                0.04               0.19               0      1
 Y2004                                0.05               0.22               0      1
* Transactions with negative age values are presales.


Table A6 Lease Structure in Hong Kong Dataset

    Types of leases               Frequency                Average prices         Std
      Freehold                 58113 (23.25%)                4908.76            1857.25
      Leasehold                191841 (76.75%)               3766.16            1687.56
         All                       249954                    4033.24             1819




                                                    16

								
To top