Conditional variance tests of integration between direct and indirect real estate markets
| Date | 01 August 2003 |
| Pages | 366-382 |
| Published date | 01 August 2003 |
| DOI | https://doi.org/10.1108/14635780310483647 |
| Author | Tien Foo Sing,Sook Beng Stephanie Sng |
| Subject Matter | Property management & built environment |
JPIF
21,4
366
Journal of Property Investment &
Finance
Vol. 21 No. 4, 2003
pp. 366-382
#MCB UP Limited
1463-578X
DOI 10.1108/14635780310483647
Conditional variance tests of
integration between direct and
indirect real estate markets
Tien Foo Sing and Sook Beng Stephanie Sng
Department of Real Estate, National University of Singapore, Singapore
Keywords Real estate, Volatility, Market segmentation, Integration
Abstract This study tests the hypothesis of market integration between the securitised and the
unsecuritised real estate market by examining the information contents of their respective
ex-post conditional volatility measures. The two markets are said to be integrated if the
conditional volatility terms of one market do not contain incremental information for the ex-post
conditional volatility of another market. Our empirical results showed no evidence of the ex-post
returns of the direct real estate (PPI) market incorporating the market volatility of the securitized
real estate asset. The ex-post conditional volatility of the PPI market, which contains only
information on the past shock and the past conditional volatility, is sufficient to statistically
explain the variation in the log-PPI price variations. However, there was significant evidence of
incremental information flowing from conditional volatility of the unsecuritized property market
to the securitized property market. Therefore, the securtized and unsecuritized real estate markets
are integrated, but the integration is only uni-directional. Some degree of segmentation is still
observed as the information of property market (PPI) still has significant impacts on the returns
of the property stock market.
1. Introduction
Indirect property investment vehicles, such as real estate investment trusts
(REITs) and publicly traded property stocks, have long been regarded as a
proxy for the direct real estate investment. This belief arises on the rationale
that the performance of indirect real estate assets should, in an efficient
markets condition, reflect the market value of the underlying assets of real
estate companies. If the hypothesis of integration of the two markets is not
rejected, one should then expect a high degree of substitution between the two
assets. For fund managers who seek to maintain a well-diversified portfolio, the
integration hypothesis implies that risk reduction cannot be achieved through
holding investments in the two markets. There will be no additional premium
associated with real estate market risk (Liu et al., 1990).
The integration of the direct and indirect real estate markets can be
detected if there are co-movements of their prices and yields. The study of
the relationships between securitised property vehicles like property
companies' stocks or real estate investment trusts and the direct real estate
investment has always been of great interest among real estate investors and
academics. In Singapore, studies have been made to test the relationship
between property stock and real estate prices. Chan and Sng (1991) analysed
The Emerald Research Register for this journal is available at
http://www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at
http://www.emeraldinsight.com/1463-578X.htm
The authors wish to thank the late Professor Gerald Brown and Dr Seow Eng Ong for their kind
comments on the earlier draft. The errors, if any, remain the responsibility of the authors.
Conditional
variance tests
367
property stocks and real estate returns in Singapore from 1976 to 1988 and
concluded that there were no significant differences in the returns. Similarly,
Ong (1994) found a long-term relationship between the property stock price
and the real estate price, implying that the two markets are integrated.
However, in two other independent studies by Ong (1995) and Liow (1998),
contrary results were obtained which reject the presence of a
contemporaneous long-term relationship between the property stock and the
real estate price indices. They found only a weak long-term
contemporaneous relationship in their lagged period indices. Hence, these
reveal that property stock and property markets are only partially
integrated.
It is apparent that the inconsistent results of the earlier studies failed to offer
conclusive evidence on the integration between direct and indirect real estate
markets in Singapore. Despite the differences in their findings, there are
invariably two common features found in the earlier studies. First, the studies
mainly examined the long-term contemporary effects of the returns in both real
estate markets. There is no discussion whatsoever on the issues of price
variations in both markets. Second, when the risk measures are implicitly
embedded in the earlier models as an identical and independently distributed
white noise process, it would be difficult to interpret the results on the
significant lagged effects between property stock and property prices (Ong,
1995; Liow, 1997). By looking at the graphs (Figures 1 and 2) of the log-returns
of property stock and property prices in Singapore, it is not difficult to observe
some strong, time-varying effects in the conditional volatility of the two
log-price series. What implications would these effects have on empirical
findings that implicitly assume a constant variance term?
This study hopes to contribute to the debate by offering an alternative way
of testing the integration of the two markets that looks at their risk
Figure 1.
Changes in property
price index (PPI)
To continue reading
Request your trial