Copyright © 2012 Vilnius Gediminas Technical University (VGTU) Press Technika
http://www.tandfonline.com/TSPM
INTERNATIONAL JOURNAL OF STRATEGIC PROPERTY MANAGEMENT
ISSN 1648-715X print / ISSN 1648-9179 online
2012 Volume 16(1): 71–84
doi:10.3846/1648715X.2011.602756
RELATIONSHIP BETWEEN RESIDENTIAL PROPERTY PRICE INDEX
AND MACROECONOMIC INDICATORS IN DUBAI HOUSING MARKET
1
and Metin VATANSEVER
2
1
Istanbul University, Faculty of Business Administration, Department of Finance, Turkey
E-mail: alihepsen@yahoo.com
2
Yıldız Technical University, Faculty of Arts and Science, Department of Mathematics and
Statistics, Turkey
E-mail: vatansevermetin@gmail.com
Received 25 February 2011; accepted 30 June 2011
ABSTRACT. The main purpose of this study is to investigate whether there is a long-run
relationship between macroeconomic indicators and property price index in Dubai. This paper
uses the monthly data for the eight year period from January 2003 to December 2010. In order
to identify long term equilibrium between property price index and macroeconomic indicators,
cointegration analyses are utilized for the study. The results of the empirical analyses show
that there is a long term positive equilibrium relationship not only between REIDIN.com
Dubai Residential Property Price Index (DRPPI) and gold prices; but also between DRPPI and
volume of total direct foreign trade. On the other hand, there is a negative long-run relation-
ship between DRPPI and the number of completed residential units. In addition, there is a
signicant positive relation between DRPPI and the rst lag of DRPPI and also the rst lag of
error term. Our paper is the rst academic study that identies this relationship in Dubai.
KEYWORDS: Dubai; Real estate; Residential Property Price Index; Macroeconomic indica-
tors; Cointegration
1. INTRODUCTION
Over a period of half a century the city state
of Dubai has progressed from pre-industrial to
industrial to post-industrial status. Change is
evident in the economic, social and cultural
characteristics of the city and, most visibly, in
the scale, pace and nature of real estate de-
velopment. Dubai has also changed from an
economy based on oil and trade to a centre for
the service industries of tourism, trade, pro-
fessional and nancial services. This transi-
tion has included a rise in property construc-
tion and a massive proliferation of large-scale
residential building developments in the past
years. Dubai has recreated itself in the image
of the cutting edge nancial sector, creating
buildings appealing to wealthy indigenous hab-
itants as well as the wealthy expatriate, and
offshore population. Development of the city
has been energized by a synergistic relation-
ship between global and local forces embedded
within a particular historical and geographical
context. Central to the planned urban growth,
and as part of the city’s strategy to establish
itself as the region’s hub for commerce, serv-
ices and leisure, is the construction of a se-
ries of, cities within the city, mega-projects.
72
A. Hepşen and M. Vatansever
Principal among these are Burj Al Arab, Burj
Khalifa (the world’s tallest building), Dubai
Mall (the largest shopping mall in the world),
Dubai Marina, Palm Jumeirah, Dubai Inter-
national Financial Center, Internet City, Fes-
tival City, Media City and International City.
As seen real estate has been a key driver of
growth and has been a steady and robust per-
former over the years. For instance, the real
estate sector contributed 14.73% of Dubai’s
total Gross Domestic Product (GDP) in 2008
(Dubai Statistics Center, 2010). Legal factors
that have been also critical to the strong devel-
opment of this sector including property rights,
transaction costs and capital gains taxes. With
the opening up of the market to allow freehold
ownership of properties to foreigners in Du-
bai in May 2002, international investors have
driven a huge demand for properties. At that
point, monitoring the evolution of property
prices and the market trend over time is obvi-
ously a necessary requirement in Dubai. The
REIDIN.com Dubai Residential Property Price
Index (DRPPI) comes out of this necessity.
Statistics on residential property prices
have a number of important uses. They are
used as a macro-economic indicator of ina-
tion, as a measurement of wealth, as a deator
for national accounts, as an input into an indi-
vidual citizen’s decision making on whether to
invest a residential property and as an input
into other price indices, most particularly the
Consumer Price Index for use, amongst other
things for wage bargaining or indexation (Fen-
wick, 2009). Additionally, as the International
Monetary Fund mentioned, property price in-
dices are direct input into an analysis of expo-
sure to risk of default (International Monetary
Fund, 2006): “During an upswing in real estate
prices, real estate may be used as collateral for
extensions of credit for further purchases. But
once conditions begin to reverse, such exposure
can cause the downturns in economic activity,
credit, and real estate prices to become mutu-
ally reinforcing.”
The main purpose of this study is to inves-
tigate whether there is a long-run relationship
between Dubai’s property market and macro-
economic factors. The rest of this article is or-
ganized in four sections. The next section gives
brief information for the REIDIN.com Dubai
residential property price index methodology.
Section 3 begins by reviewing some of the ex-
isting studies on dynamic interaction between
property prices and macroeconomic indicators.
Section 4 explains the data and theoretical
framework adopted in this study, and section
5 describes the empirical results. Finally, sec-
tion 6 offers concluding comments.
2. BRIEF INFORMATION FOR THE
REIDIN.COM DUBAI RESIDENTIAL
PROPERTY PRICE INDEX
METHODOLOGY
The objective of the residential property
price index (RPPI) is to provide an accurate
measure of the contemporary rate of change
in the prices of the properties. The methods
used to compile residential property price in-
dices vary considerably both between countries
and between alternative sources within indi-
vidual countries. The differences between the
available property price indices cover almost
every aspect of price index construction: the
conceptual basis of index; data sources (land
registry transactions, tax records, mortgage
applications and completions, estate agents,
newspaper advertisements); market cover-
age (geographical coverage, type of property,
mortgage/cash transactions); weighting (stock
or transaction weighted). The problems caused
by these different factors can be exacerbated
by the fact that housing markets can be highly
heterogeneous.
There are also four distinct approaches for
constructing property price indices. The rst
and the easiest approach to construct indexes
is the unit method. That is taking an average of
all property prices observed in a period-usually
73
Relationship between Residential Property Price Index and Macroeconomic Indicators ...
a mean or median. In general, given the heter-
ogeneity of properties, the median is preferred
to the mean. The mean method considers the
values of all sales activities regardless of ex-
tremely high or low values, whereas median
series considers all sales activities but is not
affected by extremely high or low values. The
benet of this method is median house prices
have been used by several researchers (i.e.,
Mark and Goldberg, 1984; Crone and Voith,
1992; Gatzlaff and Ling, 1994; Wang and Zorn,
1997; Prasad and Richards, 2006; Hoffmann
and Lorenz, 2006; Olczyk and Neideck, 2007;
Prasad and Richards, 2008; McDonald and
Smith, 2009); however, medians also form the
basis of publicly available house price indexes
(Bourassa et al., 2006). Most major property
price indices (including those of The U.S. Na-
tional Association of Realtors, the U.S. Census
Bureau, the Real Estate Institute of Australia,
the Real Estate Institute of New Zealand, and
the Standard Bank in South Africa) all pub-
lish house price data based on median meas-
ures (Prasad and Richards, 2006; Nhleko and
Tlatsana, 2009). A second approach-the repeat
sales methodology-originally developed by Bai-
ley et al. (1963), and focuses on houses that
have sold more than one. The straightforward
idea motivating the repeat sales methodology
is that a property’s quality approximately the
same over time; hence there is no need to in-
clude the properties’ attributes in the model.
A third approach-the hedonic regression ap-
proach-uses regression techniques to control
for compositional and quality change. It treats
a property as a bundle of attributes, each with
its own price that changes over time. However
it is a widely used technique, it requires a tre-
mendous amount of data on property attributes
(Rappaport, 2007). The nal approach is the ap-
praisal-based indexes-the Sales Price Appraisal
Ratio (SPAR) approach. The method combines
sale prices and valuations or appraisals from
an earlier period to compute price relatives
or sale price appraisal ratio’s and thus con-
trols for quality mix changes (Bourassa et al.,
2006). Property price indices for a number
of cities in New Zealand based on the SPAR
method are published by Quotable Value, a
property valuation company. The SPAR meth-
od is also being used by the statistical agencies
of Sweden and the Netherlands. The method
can of course only be used in countries where
appraisals of sufcient quality are available
(de Vries et al., 2009).
The REIDIN.com DRPPI employs the unit
method-the median approach and uses sales
transaction data made available exclusively
through “the Government of Dubai Land De-
partment
1
”. The main advantages of that ap-
proach are the simplicity of their methodology
and the data compiling stage is low cost. An-
other advantage is to include the new period
data to the model instantly and the application
methods are relatively easy to understand by
the users. On the other hand, property price
indices that are based on median measures
can be constructed for different types and lo-
cations of housing (Diewert, 2006). But city-
wide median price measures do not control for
the location of houses. Cities have areas where
houses tend to be more expensive and other
areas where they tend to be less expensive. If
the mix of houses between these groups varies
signicantly, it would have negative effect on
median price measures. As a result, changes in
median prices may contain substantial noise
from regional composition change and provide
poor estimates of true price changes. To control
1
The Government of Dubai Land Department is
the real estate registry for the Emirate. It is re-
sponsible for registering built and unbuilt land,
villas, apartments, commercial and residential
buildings, leases and mortgages throughout Du-
bai including the designated freehold areas. The
Department records and ofcially authorizes as
legitimate transactions involving Dubai real es-
tate and transfers of ownership, whether these
are between buyer and seller, donations, gifts or
inheritance. It is Dubai’s ofcial registry, valuer,
auctioneer, regulator, information provider and
property overseer (http://www.dubailand.gov.ae).
74
A. Hepşen and M. Vatansever
for compositional change, the REIDIN.com
DRPPI employs arithmetic average of the me-
dian prices of districts for constructing index
series. All indices are also calculated by using
a moving average algorithm. A moving aver-
age is commonly used with time series data to
smooth out short-term uctuations and high-
light longer-term trends or cycles. Outliers
and extreme values (as a result of incomplete,
inconsistent or erroneous data) are excluded
by the outlier detection procedure of the inter-
quartile range (IQR) based on the calculated
price per square meter of each property. This
commonly used methodology considers any
data that is more than 1.5 times the IQR from
the upper or lower quartile to be an outlier
(Tukey, 1977). The REIDIN.com DRPPIs are
calculated by using the Dutot price index for-
mula. The formula can be written as:
0
0
1
100 100
1
t
t
DUTOT
p
p
n
I
p
p
n
×
= ×=×
×
∑
∑
∑
∑
, (1)
where: p
t
is the median of price/sqm, t months
after the base period; p
o
is the median of price/
sqm during the base period; n
is the number
of districts.
Dutot price index formula dened as the ra-
tio of the unweighted arithmetic average of the
prices in the current period to the unweighted
arithmetic average of the prices in the base
period 0. The following table shows the name
of indices and their features.
The Figure 1 shows the time series of REI-
DIN.com Dubai Residential Property Price
Index (DRPPI) for the period January 2003-
December 2010.
Table 1. The REIDIN.com Dubai Residential Property Price Index series and prices*
Index name Property type USD/SQM prices Index start date Update interval
Dubai – Residential Residential 2,193 Jan 2003=100 Monthly
Dubai – Apartment Apartment 2,400 Jan 2003=100 Monthly
Dubai – Villa Villa 2,018 Jan 2003=100 Monthly
Dubai (51SQM and Less) Apartment 2,724 Jan 2006=100 Monthly
Dubai (51SQM-100SQM) Apartment 2,214 Jan 2006=100 Monthly
Dubai (101SQM-150SQM) Apartment 2,549 Jan 2006=100 Monthly
Dubai (151SQM and More) Apartment 2,572 Jan 2006=100 Monthly
Arabian Ranches Villa 1,478 2005Q1=100 Quarterly
Business Bay Apartment 3,886 2006Q1=100 Quarterly
Discovery Gardens Apartment 2,076 2006Q2=100 Quarterly
Downtown Dubai Apartment 3,995 2006Q1=100 Quarterly
Dubai Marina Apartment 2,654 2003Q1=100 Quarterly
Dubai Sports City Apartment 2,372 2007Q1=100 Quarterly
Emirates Hills First Apartment 2,231 2004Q1=100 Quarterly
International City Apartment 1,413 2005Q2=100 Quarterly
Jumeirah Beach Residences Apartment 3,029 2007Q1=100 Quarterly
Jumeirah Lake Towers Apartment 2,225 2004Q1=100 Quarterly
Mirdiff Villa 1,136 2005Q1=100 Quarterly
Old Town Apartment 4,019 2006Q1=100 Quarterly
Palm Jumeirah Apartment 3,005 2007Q1=100 Quarterly
Palm Fronds Villa 2,502 2007Q1=100 Quarterly
The Greens Apartment 2,620 2003Q1=100 Quarterly
The Springs&The Meadows Villa 1,684 2003Q1=100 Quarterly
*USD/SQM values as of December 2010 and 4
th
Quarter of 2010.
75
Relationship between Residential Property Price Index and Macroeconomic Indicators ...
3. LITERATURE REVIEW
In the context of real estate research, the
academic studies primarily focus on the rela-
tionship between the housing sector and a com-
mon set of macroeconomic variables. Nellis and
Longbottom (1981) nd that the determinants
of housing price in the United Kingdom are real
disposable income, loan interest rates, and total
loans. Reichert (1990) identies that regional
housing prices react uniformly to certain nation-
al economic factors, such as mortgage rates.
On the other hand, local factors such as
population shifts, employment, and income
trends often have a unique impact on housing
prices. Clapp and Giaccotto (1994) study the
inuence of economic variables on local house
price dynamics and nd that some variables
(including population, employment and income)
have considerable forecasting ability for hous-
ing price. To identify the factors inuencing
yearly urban housing prices, Potepan (1996)
uses a number of indicators including the pri-
vately owned dwelling price index, monthly
rent based on the Hedonic Model, land price,
medium income, population, quality of public
services, crime rate, air pollution, non-dwelling
consumable price, mortgage rate, construction
cost, farm land price, land restriction and so
on. Baffoe-Bonnie (1998) uses a nonstructural
estimation technique (a vector autoregression
(VAR) model) to analyze the dynamic effects of
four key macroeconomic variables on housing
prices and the stock of houses sold in US sub
regions. The impulse response functions derived
from the VAR suggest that the housing market
is very sensitive to shocks in the employment
growth and mortgage rate at both national and
regional levels. Case (2000) discusses global
macroeconomic effects on US house prices. He
notes that the impact of fundamentals depends
on the openness of the different states.
By applying the vector error correction mod-
el (VECM), Kasparova and White (2001) exam-
ine the housing markets in selected European
countries, including Germany and the UK.
Granger tests reveal that the effect of house
prices on GDP is signicantly greater than the
effect of GDP on house prices. Tsatsaronis and
Zhu (2004) look at the importance of a number
of macroeconomic factors affecting the dynam-
ics of residential property prices. The authors
examine the determinants of house prices for
17 industrialized economies between 1970 and
2003 by employing a vector autoregression
(VAR) model. The model includes ve endog-
enous variables besides house price growth: the
growth rate of GDP, the rate of ination in con-
sumer prices, the real short-term interest rate,
the term spread, and the growth rate in ina-
tion adjusted bank credit. The main nding of
their study is that economic growth, ination
and interest rates, bank lending, and equity
prices have signicant explanatory power on
Figure 1. REIDIN.com Dubai Residential Property Price Index (DRPPI) (January 2003=100)
