# RECORD 4412b66b7af7dc2e4784f04e96dda07a5d24ef90
# TITLE: Buyers’ property asset purchase decisions: an empirical study on the high-end residential property market in Hong Kong
# YEAR: 2016.0  VENUE: nan  DOI: 10.3846/1648715x.2015.1105322
# AUTHORS: W. M. Jayantha; Jia-Ming Lau
# PAGES: 16  CANDIDATE TABLE PAGES (1-based): [10, 11, 12]
# ABSTRACT: Demand for luxury housing units from the upper and upper-middle income groups in Hong Kong has been increasing over the last few years. as the market cannot satisfy demand, some prospective buyers have turned their attention to “special” housing units. this research paper attempts to investigate buyers’ preferences for two types of “special” units, namely duplex units and adjoining flats. the study investigates the price premiums paid by the buyers, and examines the effects of these special units on property price. the study employs two hedonic price models: one measuring the buy- ers’ preference on duplex units and the other one measuring buyers’ preference on adjoining flats. The results show that buyers are willing to pay a larger premium for special residential units: HK$588/ft-super-2 more for a duplex unit and HK$934/ft-super-2 more for an adjoining property unit, respectively. furthermore, a relatively larger premium is found for adjoining flats compared to duplex units. This implies that a property unit, purchased as an adjoining flat can add more value to the property price (in terms of $s per sq. ft.) compared with being a duplex unit.


===== PAGE 1 (score 11, numeric-paren 1, table-words 0, var-words 10) =====

Copyright © 2016 Vilnius Gediminas Technical University (VGTU) Press
http://www.tandfonline.com/tspm
International Journal of Strategic Property Management
ISSN 1648-715X / eISSN 1648-9179
2016  Volume  20(1): 1–16
doi:10.3846/1648715X.2015.1105322
*	Corresponding author. E-mail: bsmjwadu@polyu.edu.hk
BUYERS’ PROPERTY ASSET PURCHASE DECISIONS:  
AN EMPIRICAL STUDY ON THE HIGH-END RESIDENTIAL  
PROPERTY MARKET IN HONG KONG
Wadu Mesthrige JAYANTHA a,*, Jia Ming LAU b
a Department of Building and Real Estate, The Hong Kong Polytechnic University, Hong Kong SAR, 
China
b Freelance author, Kowloon, Hong Kong SAR, China
Received 5 September 2014; accepted 14 January 2015
ABSTRACT. Demand for luxury housing units from the upper and upper-middle income groups in 
Hong Kong has been increasing over the last few years. As the market cannot satisfy demand, some 
prospective buyers have turned their attention to “special” housing units. This research paper attempts 
to investigate buyers’ preferences for two types of “special” units, namely duplex units and adjoining 
flats. The study investigates the price premiums paid by the buyers, and examines the effects of these 
special units on property price. The study employs two hedonic price models: one measuring the buy-
ers’ preference on duplex units and the other one measuring buyers’ preference on adjoining flats. The 
results show that buyers are willing to pay a larger premium for special residential units: HK$588/ft2 
more for a duplex unit and HK$934/ft2 more for an adjoining property unit, respectively. Furthermore, 
a relatively larger premium is found for adjoining flats compared to duplex units. This implies that a 
property unit, purchased as an adjoining flat can add more value to the property price (in terms of $s 
per sq. ft.) compared with being a duplex unit.
KEYWORDS: Property market; Duplex unit; Adjoining flat; Property price; Hong Kong
1. INTRODUCTION
Even though the housing affordability issue in 
Hong Kong has been a serious problem, especially 
among many low and medium income households, 
there has been an increasing demand for luxury 
housing units by the upper middle and high income 
households over the last few years. With the rising 
aspirations of this sector in society, along with in-
creasing national income (and per capita income), 
there is an increasing demand for so-called “spe-
cial” housing units (i.e. duplex units and adjoining 
flats). As the market cannot satisfy demand, some 
prospective buyers have turned their attention to 
these types of “special” housing units, for some as 
a way of getting into the market, for others as a 
way of increasing their investment and diversifica-
tion into this type of asset. The main reason for this 
significant increase in demand for special units is 
the existence of the wealth effect and the limited 
supply of these units in the property market (Fong 
2009). The wealth effect is defined as “…an increase 
in spending or consumption that accompanies an in-
crease in wealth, or merely a perceived increase in 
wealth.” (Pigou 1943). According to Cheung et al. 
(2004), property equity has dominated the wealth 
portfolio of typical middle-class Hong Kong resi-
dents for a long time. Buying “bricks and mortar”, 
i.e. property, has been viewed as one of the safest 
stores of wealth as well as investment, according to 
Chinese tradition. These special housing units, in 
particular (i.e. duplex units and adjoining flats), are 
regarded as high quality durable investment goods 
as well as consumer goods. It is clear that people 
in Hong Kong have begun, on one hand, to appre-
ciate the characteristics of these property units 
in terms of prestige value and potential for asset 
appreciation, and, on the other hand, to embrace 
green living (Jayantha, Wan 2012). Hence, these 
special housing units are in high demand in the lo-
cal property market.


===== PAGE 10 (score 42, numeric-paren 0, table-words 16, var-words 8) =====

W. M. Jayantha et al.
10
aspects of the properties, date of occupation of the 
building, etc. were obtained from Economic Prop-
erty Research Centre (EPRC) Limited and crossed 
checked with Centadata, which is established 
by Centaline  Property Agency. EPRC provides 
a database that records all the registered prop-
erty transaction in the Land Registry since 1991 
in Hong Kong. The data provided in Centadata is 
based on the registered property transactions in 
the Government Land Registry. When necessary, 
a land search was obtained from the Land Registry 
for verification and supplementary purposes. Other 
property information is obtained through various 
government departments and site investigation. 
Location, geographical and neighbourhood informa-
tion of the selected buildings are gathered from the 
GeoInfo Map powered by the Lands Department. 
Floor plans of some of the selected properties were 
collected from the Building Records Access and 
Viewing On-line (BRAVO) System established by 
the Buildings Department. The building manage-
ment information employed in this research was 
obtained from the Database of Private Buildings 
in Hong Kong provided by the Home Affairs De-
partment. Recreational facilities associated with 
the chosen residential developments were verified 
through property agents’ and developers’ websites. 
Site investigation was also conducted to identify the 
views, accessibilities, etc. of the properties.
3.4. Statistical interpretations in the HPM
Empirical results of HPM (regression model) are 
generally interpreted and analyzed by using nu-
merous simple statistical techniques (i.e. t-test, 
F-test and adjusted R2). Our HPM has eleven in-
dependent variables, and the t-test is used to test 
whether each parameter of a variable is significant 
or not. This is done by comparing the estimated 
t-value with the critical t-value. An independent 
variable is said to be significant if the absolute 
t-statistic for that coefficient of the variable is 
greater than the critical value. That implies that a 
particular variable has a significant impact on the 
dependent variable. Another important statistic in 
the regression model is the F-test, which reflects 
the overall performance of the model. The overall 
model is said to be performing well or is statisti-
cally highly significant if the estimated F-statis-
tic is larger than the critical value. That means 
the independent variables included in the model 
jointly explain the dependent variable satisfacto-
rily. Finally, the explanatory power of the model 
is evaluated through the adjusted R2 value that 
ranges from 0 to 1; and the greater the value of 
it, the better the model. High values of R2 indicate 
that the independent variables selected, together 
substantially explain the behaviour of the depend-
ent variable (meaning that no significant variables 
are missing).
Table 2. Summary of descriptive statistics (Model 1)
Variable
Mean
STD
Min
Max
Price
9010.88
1806.72
2177.00
15472.00
FlatSize
770.01
202.62
556.00
1229.00
FlatRoof
68.50
92.34
0.00
292.00
AGE
14.56
5.30
5.73
25.24
Floor
17.14
0.88
15.00
19.00
4. EMPIRICAL FINDINGS
4.1. Results of model 1 (Duplex unit)
Table 2 presents the descriptive statistics of the 
variables used in model 1. Empirically estimated 
coefficients and the explanatory power of the mod-
el along with statistical measures are presented in 
Table 3. All the employed independent variables 
(except Orient) are found to be highly significant 
at conventional levels and also bear the antici-
pated theoretical signs of variables. The adjusted 
R2 value of 0.745 indicates that around 75 percent 
of the total variation of residential property price 
can be described by the explanatory variables, and 
this is quite a high value. The estimated F-value of 
231.95 (p-value is 0.000) also suggests that the set 
of explanatory variables chosen are jointly statisti-
cally significant and at least one of the independ-
ent variables is able to explain the variation of the 
property price in the model. Confidence interval 
levels (at 95%) were also tested for all the esti-
mates of independent variables8. All the estimated 
coefficients are within the range of 95 % confidence 
level, confirming the reliability of the estimates.
The most relevant and significant variable in 
the study is DUPLEX, and as anticipated, it is 
found to be statistically highly significant. The es-
timated t-value of 2.408 and p-value 0.016 suggest 
that the variable DUPLEX is highly significant at 
5 percent level. The coefficient of 588.54 of DU-
PLEX means a property being a duplex unit can 
add value and raise the property price by $588.54/
sq.ft. To put it differently, potential buyers are 
ready to pay $588.54 more per square feet for a 
duplex property unit than for a normal/typical 
unit. This tells us that if a residential building has 
duplex units, the value of the other units in the 
residential property increases significantly.
8	Not reported in the text, but available upon request.


===== PAGE 11 (score 53, numeric-paren 10, table-words 11, var-words 17) =====

Buyers’ property asset purchase decisions: an empirical study on the high-end residential property market…
11
A duplex unit can enhance the living standard 
of the end user due to the protection of privacy 
among users through the separation of spaces into 
two floors. In addition, the property owner can en-
joy the prestige value of the duplex unit because 
of the limited supply of these units in the market. 
The result shows that the quality of housing prop-
erty in terms of prestige and living standard seems 
to be playing an important role in buyers’ decisions 
on purchasing a property. As duplex units are now 
well recognized housing units with these attrib-
utes, potential property buyers may be willing to 
pay a premium for these special housing units that 
enhance their prestige level and living standard.
All the other explanatory variables (except Ori-
ent) incorporated into the model are also statisti-
cally significant. The coefficient of 3.229 of Flat-
Size suggests a positive relationship between the 
saleable area (excluding flat roof area) and the 
property value per square feet. Buyers are willing 
to pay a higher price as they can enjoy a larger liv-
ing space. It reveals that a property with a larger 
perimeter and larger living area naturally tend to 
appreciate more rapidly (Cebula 2009). The result 
is consistent with previous research works (e.g., 
Mok et al. 1995; So et al. 1997; Chau et al. 2003). 
Variable FlatRoof is also found to be highly sig-
nificant (t-statistics of 3.734 and p-value of 0.000) 
with the coefficient of 4.056. This indicates that 
potential property buyers value a flat roof area as 
part of the property and are willing to pay $4.056 
for every 1 square foot of flat roof area. In other 
words, every 1 sq. ft. of flat roof area causes the 
property price (per saleable area) (excluding flat 
roof area) to increase by $4.056. The positive re-
lationship is obvious since the property price per 
saleable area by definition excludes the flat roof 
area. With a bigger flat roof size, the property 
owner can enjoy a larger open space for leisure 
and recreational activities, for example barbeque, 
resting, feeding pets, etc. Hence, the purchasers 
are willing to pay a premium for the added benefit 
of a flat roof area.
The negative coefficient of age (–45.506) im-
plies a negative correlation between building age 
and property price (per square feet) with a rate of 
$45.506/sq. ft. per year. The estimate is significant 
at 5% level with t-statistics of –6.948 and p-value 
of 0.000. This finding is consistent with Mok et al. 
(1995) and Choy et al. (2007) and Coulson and Lahr 
(2005). The age of a building is expected to link 
negatively with the property value as the build-
ing structure deteriorates physically with time. 
Potential buyers obviously pay only lower prices 
for older buildings as structural defects are gener-
ally found in aged properties requiring significant 
maintenance, so buyers pay only lower prices.
The coefficient of Floor has a positive sign with 
a value of 187.311, which means that the higher 
the floor level of the property is located, the higher 
the property price per square feet with a rate of 
$187.311. for each successive floor level. The result 
is significant at 5% level with t-statistics of 2.843 
and p-value of 0.005. This suggests that the buy-
ers are willing to pay more ($187.31 per square 
foot) for a unit at higher level to enjoy the better 
view and the cleaner air. Generally, higher level 
units enjoy cleaner air and better view, and fewer 
mosquitoes, compared with properties located at 
lower floor levels and this is totally true in Hong 
Kong. The result is in line with previous research-
ers such as Mok et al. (1995), So et al. (1997) and 
Chau et al. (2003).
Property buyers treat ‘orientation’ as a key fac-
tor when purchasing a property in Hong Kong. The 
reason being that the West direction is normally 
considered as the worst orientation because units 
facing West would be a lot warmer in summer, 
whilst on the other hand; units facing South di-
rection would enjoy a better ventilation (airflow) 
and better natural lighting. However, the results 
do not support this view. The variable Orient is 
found to be insignificant even though it bears the 
expected negative sign. This suggests that buyers 
do not take orientation into serious considerations 
in purchasing properties. The reason may be partly 
because some blocks in selected residential estates, 
for instance Heng Fa Chuen, are surrounded by 
other blocks of building or mountain and thus have 
a less exposure to excessive sunlight. The obstruc-
tion of sunlight may somehow reduce the negative 
impact on some of the units.
Table 3. Coefficients of the model 1 (Duplex)
Variable
Coefficient
t-statistic
Sig.
Constant
3465.965*
3.111
0.002
FlatSize
3.229*
5.414
0.000
FlatRoof
4.056*
3.734
0.000
AGE
–45.506*
–6.948
0.000
Floor
187.311*
2.843
0.005
Orient
–102.268
–1.367
0.172
ViewHw
–229.866**
–2.190
0.029
ViewMount
390.095**
2.511
0.012
ViewSea
1368.619*
8.530
0.000
DUPLEX
588.540**
2.408
0.016
Adj. R2
0.745
F-value
231.950
*, ** indicate significant at 1% and 5% levels, respectively.


===== PAGE 12 (score 65, numeric-paren 14, table-words 15, var-words 14) =====

W. M. Jayantha et al.
12
The view factor is another significant element 
in Hong Kong that influences purchasing decision 
of a property (see for example, Mok et al. 1995). 
Sea view is considered as the best, followed by the 
mountain view for many property buyers whilst 
units having high-way view or obstruction with 
adjacent buildings and structures are considered 
negatively. Findings of the model support this 
view. The negative coefficient (–229.866) of Vie-
wHw indicates that the possession of a highway or 
railway view lowers the property price (per sq. ft.) 
significantly. The result is significant at 5% level 
with t-statistics of –2.190 and p-value of 0.029. It 
is obvious that traffic noise and air pollution aris-
ing from the highway or railway cause an undesir-
able living environment to the end user leading 
to discounts in the property value. This finding is 
also in line with previous works (e.g., Nelson 1982; 
Theebe 2004; Brandt, Maennig 2011). On the other 
hand, the positive coefficient of ViewMount sug-
gests that possession of a mountain view enhanc-
es property price (per square foot). It is obvious 
that the tenants would appreciate a mountain 
view (compared to a highway or buildings view) 
as it is pleasing to have a green view to the eye 
of the end user, and hence they are willing to pay 
a higher price for such properties. The most ap-
preciated view by the buyers is the sea view; and, 
this is very clearly reflected through the result. 
Highly significant positive coefficient (1358.619) 
of ViewSea indicates that the possession of a sea 
view would increase the property price dramati-
cally, by $1358.619/sq.ft. Potential buyers would 
be willing to pay a larger premium if the unit pos-
sesses a sea view as the sea view can yield more 
satisfaction to the end users with a better living 
environment. The estimate is significant result at 
1% level with -statistics of 8.530 and p-value of 
0.000. This finding is also consistent with previous 
studies (e.g., Mok et al. 1995; Choy et al. 2007).
All these physical and locational attributes 
have been well documented in the literature with 
regard to the housing price. For example, floor 
area of the unit (Cebula 2009; Neill et al. 2007; 
Sirmans et  al. 2005; Coulson, Lahr 2005; Chau 
et al. 2003; So et al. 1997; Mok et al. 1995), age 
(Choy et al. 2007; Neill et al. 2007; Coulson, Lahr 
2005; Sirmans et al. 2005; Mok et al. 1995), floor 
(Jayantha, Wan 2012; Chau et al. 2003; So et al. 
1997; Mok et al. 1995), orientation (Jayantha, Wan 
2012), view (Brandt, Maennig 2011; Theebe 2004; 
Mok et al. 1995; Nelson 1982).
4.2. Results of model 2 (Adjoining flat)
Descriptive statistics of the data series for the 
model are summarized in Table 4. The empiri-
cal estimates of variables of the model 2, includ-
ing t-statistics and other measures are shown in 
Table 5. The adjusted R2 (0.81) suggest that 81 
percent of the total variation of housing price (for 
saleable area) can be explained by the attributes 
of the model, meaning that is has a very high ex-
planatory power. Again, as in model 1, all the ex-
planatory variables (excluding Orient) are found 
be significant at conventional levels. The F-statis-
tic value of 266.09 indicates that the overall model 
performs very well and all the selected explanatory 
variables are jointly statistically significant.
The main emphasis of the model is on ADJOIN, 
which represents the special housing units (adjoin-
ing flats). As in model 1, it is found that variable 
ADJOIN (special unit) is highly significant and 
bears the expected positive sign. The estimated 
coefficient of ADJOIN is 934.079, which is much 
larger than (estimates for DUPLEX) that of in 
model 1. The t-statistic (2.513) along with 0.012 of 
corresponding p-value indicates that the ADJOIN 
variable is highly significant. Thus this variable 
Table 4. Summary of descriptive statistics (Model 2)
Variable
Mean
STD
Min
Max
Price
8852.52
2555.75
3608.00
21426.00
FlatSize
852.41
295.64
518.00
1654.00
FlatRoof
100.01
177.25
0.00
541.00
AGE
14.17
8.55
0.00
32.58
AgeSq
273.88
269.01
0.00
1061.51
Floor
32.85
6.57
0.25
42.00
Table 5. Coefficients of the model 1 (Adjoining)
Variable
Coefficient
t-statistic
Sig.
Constant
1786.530*
1.854
0.064
FlatSize
2.690*
5.376
0.000
FlatRoof
4.469*
4.442
0.000
AGE
–56.700*
–2.833
0.005
AgeSq
1.538*
2.618
0.009
Floor
110.119*
4.763
0.000
Orient
–101.477
–0.999
0.318
ViewOpen
596.154*
3.226
0.001
ViewMount
338.198**
1.977
0.048
ViewSea
1040.848*
6.207
0.000
ADJOIN
934.079*
2.513
0.012
Adj. R2
0.810
F-value
266.086
*, ** indicate significant at 1% and 5% level, respec-
tively.


===== PAGE 13 (score 30, numeric-paren 0, table-words 5, var-words 19) =====

Buyers’ property asset purchase decisions: an empirical study on the high-end residential property market…
13
exerts a strong positive impact on residential prop-
erty price. The coefficient of 934.079 suggests that 
a property being an adjoining flat can add val-
ue and raise the property price by $934.079 per 
square feet. Put differently, potential buyers are 
willing to pay $934.079 more per square feet if it 
is an adjoining property unit. This means if a resi-
dential building is with adjoining flats, the value of 
the other normal units in the residential property 
also increases significantly. Adjoining flats by their 
nature improve the living standard of the end user 
due to a wider view and extensive enlargement of 
floor spaces. In addition, great prestige feeling val-
ue that potential buyers have living in such a unit 
encourages them to pay extra money for the prop-
erty being an adjoining flat. This finding is in line 
with the results obtained from model 1. The find-
ings of both models indicate that potential prop-
erty purchasers value special units (Duplex units 
and Adjoining flats) and are ready to pay an extra 
amount of money for them. This is very much in 
line with findings of some of the previous stud-
ies. For example, Cebula (2009), Coulson and Lahr 
(2005) and Sirmans et al. (2005) find out that de-
sirable internal luxury housing attributes such as 
courtyards, number of bedrooms and bathrooms, 
stories in structure, garage car spaces, square 
footage, the presence of a deck, a pool/hot-tub, etc. 
are positively related with the real housing price. 
These special units (Duplex units and Adjoining 
flats) possess all these housing attributes. Similar 
to model 1, all the other variables (except Orient), 
are also found to be highly significant, implying 
the fitness of the model.
4.3. Difference in magnitudes of price effect 
between Model 1 and Model 2
In an attempt to compare these two models, the 
magnitudes of price effects between these two 
cases are analyzed in this section. Models 1 and 
2 clearly indicate that potential buyers value spe-
cial units (Duplex units and Adjoining flats) and 
are ready to pay a premium for them. Both models 
also show that special units exert a clear positive 
impact on property price.
The empirical results summarized in Table 6 
clearly show that the coefficients of special units in 
both models are highly significant. This suggests 
that potential property buyers are happy to pay 
an extra amount over and above typical units (a 
premium) for special units: duplex units and ad-
joining flats. Potential buyers normally tend to 
appreciate and value a better living environment, 
and hence are willing to pay a higher premium for 
these units. However, the magnitude of coefficient 
of ADJOIN is much larger than that in DUPLEX: 
588.054 and 934.079, respectively. That means po-
tential buyers are willing to pay a higher premium 
for adjoining flats compared to duplex units: a pre-
mium of HK$934/ft2 compared to HK$588/ft2. In 
other words, a property being an adjoining flat can 
add more value (i.e. HK$346/ft2) to the property 
price compared with being a duplex unit. The re-
sult indicates that buyers prefer an adjoining flat 
than duplex unit. It can be therefore concluded 
from the empirical results that an adjoining flat 
has a greater positive impact on the property price 
and is more preferred by homeowners in Hong 
Kong compared with duplex unit. This situation 
could be explained this way.
Though duplex units enhance the prestige im-
age and provide a house-like feeling for the own-
ers, it also has some disadvantages for the users. 
Duplex units normally comprise of two stories 
with a stair case, and staircase occupied some 
floor spaces of the unit. However, adjoining flats 
do not have this disadvantage. Also, older people 
may not feel safe in climbing stairs, so adjoining 
floor space would be safer and preferable. On the 
other hand, adjoining flats enhance the floor space 
usage and also possess a broader view. Thus, in-
vestors specially consider adjoining flats as a good 
investment.
5. SUMMARY AND CONCLUSIONS
The study aimed to investigate the buyers’ prefer-
ence for special housing units (duplex units and 
adjoining flats), and to examine the effects of these 
special housing units on property prices. This re-
search issue was investigated by adopting two 
models (cases). The empirical results suggest that 
special units can add value and raise the property 
price significantly. Potential buyers are willing to 
pay HK$588/ft2 and HK$934/ft2 more for a duplex 
unit and an adjoining property unit, respectively. 
In other words, the findings suggest that poten-
tial buyers are willing to pay a larger premium 
for special housing units. For example, one may 
be willing to pay HK$ 588,000 (model 1) and HK$ 
934,000 (model 2) more on a duplex unit and 
Table 6. Comparison of coefficients of special units 
(Duplex and Adjoining)
Coefficient
t-statistic
Sig.
Model 1 (Duplex)
588.540
2.408
0.016
Model 2 (Adjoining)
934.079
2.513
0.012