Constraining optimal portfolios and the effect on real estate’s allocation
| Date | 01 August 2000 |
| Published date | 01 August 2000 |
| Pages | 488-506 |
| DOI | https://doi.org/10.1108/14635780010345445 |
| Author | Simon Stevenson |
| Subject Matter | Property management & built environment |
JPIF
18,4
488
Journal of Property Investment &
Finance, Vol. 18 No. 4, 2000,
pp. 488-506. #MCB University
Press, 1463-578X
Received 29 March 1999
Revised 27 March 2000
ACADEMIC PAPERS
Constraining optimal
portfolios and the effect on
real estate's allocation
Simon Stevenson
Graduate School of Business, University College Dublin, Ireland
Keywords Property portfolio, Real estate, Republic of Ireland
Abstract This paper re-examines the role of real estate within mixed asset portfolios from the
perspective of an Irish portfolio manager. The nature of the investment market in the Republic of
Ireland leads to the study extending the existing literature by expanding the universe of assets
beyond a solely domestic setting and by imposing constraints on the optimal portfolios. Irish funds
generally hold proportionately more in international equities than in the domestic market due to
the small and illiquid nature of the Irish market; therefore, unconstrained tests do not adequately
model the behaviour of Irish portfolio managers. The study finds that while real estate plays an
important role in both the domestic and international unconstrained portfolios, it exits the
optimal portfolios at relatively low return levels. Additionally, the real estate series adjusted for
smoothing fails to enter any of the optimal portfolios. However, the use of 20 per cent band
constraints leads to an increase in the diversification role real estate can play in a mixed asset
portfolio, with the asset maintaining a presence up to more acceptable return levels.
Introduction
The role of commercial real estate in mixed asset portfolio has been the subject
of extensive research over recent years, with many studies advocating large-
scale allocations in the sector. However, despite the large number of empirical
studies that have addressed the issue, a number of key issues remain largely
unresolved. Using data on the Irish market, this study intends to examine two
areas that have been the subject of relatively little debate in the literature,
namely the capital market coverage of the data sets used and the imposition of
constraints on the optimal portfolios.
The benefits of international diversification have been widely accepted since
the work of Grubel (1968), and Solnik (1974); however, the majority of real
estate asset allocation studies have generally taken place in a domestic context,
with few studies extending the universe of assets into an international arena[1].
In addition, those studies that have examined the issue in an international
context have generally been focused on issues relating to international real
estate markets. This has had the consequence that even when foreign capital
markets have been examined they have been generally limited to those nations
whose real estate markets are being analysed[2]. The growth in international
The research register for this journal is available at
http://www.mcbup.com/research_registers/jpif.asp
The current issue and full text archive of this journal is available at
http://www.emerald-library.com
The author would like to extend his appreciation to the comments of two anonymous referees
and to participants at the 1998 RICS Cutting Edge conference for comments on an earlier
version of this paper.
Academic papers:
Constraining
optimal portfolios
489
investment means that the majority of portfolio managers are not solely
concerned with a domestic asset set, therefore it would be reasonable to extend
the universe of assets to include foreign capital market securities.
The inclusion of international assets is also an issue in terms of the
imposition of constraints. Irish funds are in the unusual situation of holding
proportionately more in international equities than in the domestic market.
This reverse home bias is largely due to the small size of the Irish market, both
in value terms and in the number of actively traded stocks. Therefore, unlike
the common home bias that occurs in most markets, the allocation decision
faced by Irish funds is constrained by practical issues[3]. Because of this
situation it is felt reasonable to constrain the optimal portfolios in a similar
fashion to the policies generally adopted by Irish fund managers. The primary
objective of this study is therefore to examine the role of real estate in a scenario
that explicitly takes account of the policies that portfolio managers pursue.
Methodological framework
The methodological framework adopted in this study is primarily based on the
Mean Variance Analysis approach of Markowitz (1952, 1959), The short selling
of assets is not allowed for two primary reasons. First, the characteristics of
real estate mean that the short selling of the asset is impossible, and secondly,
most institutions are constrained in terms of the short selling activity that they
can partake in. Therefore, as the aim of the study is to provide a more accurate
assessment of the policies pursued by portfolio managers, short selling is
prohibited throughout the tests conducted[4].
Initially unconstrained optimal portfolios were constructed, first on a
domestic only basis and secondly with the introduction of international capital
market assets. This analysis should allow an examination of whether the
introduction of such assets has an impact on the contribution that real estate
could play in mixed asset portfolios. With regard to the treatment of foreign
exchange risk, the tests are conducted twice, first using local returns and
secondly using the appropriate spot rate to convert the local returns to Irish
pounds. The first strategy is equivalent to assuming that fund managers have
perfect hedging ability. The estimated allocations in these scenarios are
unconstrained with one exception, that being the treatment of cash assets. A
number of studies, including Byrne and Lee (1995), have found that the
inclusion of unconstrained cash assets leads to extremely high estimated
allocations for cash in the optimal portfolios. As fund managers would not
consider such allocations the asset is constrained to a maximum allocation of
10 per cent.
The tests are then re-examined under weight constraints. This is designed to
see whether the imposition of the constraints that many fund managers would
naturally impose, has an impact on the allocations placed in property. Table I
displays the average pension fund allocations for Ireland, and for comparative
purposes the UK and USA. While the allocations for US and UK funds are
broadly similar, the Irish funds show a number of marked differences, in
To continue reading
Request your trial