Empirical evaluation of the value of waiting to invest
| DOI | https://doi.org/10.1108/14635780110406888 |
| Date | 01 December 2001 |
| Published date | 01 December 2001 |
| Pages | 535-553 |
| Author | Tien Foo Sing,Kanak Patel |
| Subject Matter | Property management & built environment |
Academic papers:
Empirical
evaluation
535
Journal of Property Investment &
Finance, Vol. 19 No. 6, 2001,
pp. 535-553. #MCB University
Press, 1463-578X
ACADEMIC PAPERS
Empirical evaluation of the
value of waiting to invest
Tien Foo Sing
Department of Real Estate, School of Design & Environment,
National University of Singapore, Singapore, and
Kanak Patel
Department of Land Economy, University of Cambridge, UK
Keywords Property markets, Empirical study, Investment, Land use
Abstract The lack of transaction data has been identified as one of the major obstacles for the
empirical evaluation of real option. Quigg's study in 1993 was one of the first to empirically
estimate the premium for the option of waiting to develop using data from 2,700 land sales in
Seattle. This study modified Quigg's methodology and applied it to estimate the premium for the
option of waiting to develop based on a sample of data from 2,286 property transactions in the
UK collected over a 14-year sample period from 1984 to 1997. Based on a one-factor contingent
claim valuation model, we found that the average premiums for the timing options were 28.78
percent for office sector, 25.75 percent for industrial sector and 16.06 percent for retail sector.
We also tested the robustness of the theoretical-based land value estimates in explaining the
market-based land values. The regression results showed a statistically significant relationship in
logarithm form between the market-based residual land value and the model-based land values
(with embedded timing option), with R
2
of 0.75, 0.79 and 0.82 for office, industrial and the retail
sectors respectively.
1. Introduction
In the conventional residual approach of valuation, land value is simply
estimated as the difference between the market value of completed building and
the gross development costs that areinclusive of developer's profits and holding
costs. However, in the standard land option model, which assumes that land is a
call option on its underlying asset, which is an optimal scaled building that
would potentially be built on that land, land value should thus conceptually
reflect a premium for the options to select an optimal time and scale of a
development. Quigg (1993) and Williams (1991) have developed theoretical real
option models with two stochastic variables, namely development costs and
project value, to explicitly evaluate the optimal timing option premiums. Based
on the proposed two-factor model, Quigg (1993) went on to empirically estimate
the option premium for waiting to develop land in the Seattle area, which was
found to be approximately 6 per cent of the land price. In the UK, however, there
is no comparable empirical evidence on premiums for the option of waiting to
The research register for this journal is available at
http://www.mcbup.com/research_registers
The current issue and full text archive of this journal is available at
http://www.emerald-library.com/ft
The authors wish to thank Professor D.A. Paxson, Manchester Business School, Professor G.
Brown and Dr S.E. Ong, National University of Singapore, and Dr Ozcan Sezer, University of
Connecticut, for their kind comments. The authors remain solely responsible for any errors.
Your comments are much welcome.
JPIF
19,6
536
develop lands. This study is therefore undertaken to empirically estimate the
values of optimal timing options based on the UK commercial property
transactiondata over a sample period from 1984 to 1997.
The paper is organised into six sections. Section 2 reviews empirical
literature that tested the validity of real option models and/or estimated real
option premiums. Section 3 outlines the empirical methodology we use to
estimate the premiums of timing option in the UK and explains how it differs
from Quigg's methodology. Results of option premium estimation are analysed
in Section 4. In Section 5, results of the least squares regressions that test the
robustness of the estimated intrinsic and the option values in relation to the
residual land values are presented. Section 6 concludes the findings and
highlights the limitations of the study.
2. Previous research
There are basically two approaches by which real option pricing models are
empirically tested in the literature. The first approach involves tests of the
validity of the theory of irreversible investment, which postulates a positive
relationship between uncertainty and hurdle price of an undeveloped land
(Capozza and Schwann, 1989, 1990; Patel and Sing, 2001). The second approach
jointly tests the predictability of real option pricing models and the efficiency of
market (Shilling et al., 1985, 1990; Quigg, 1993).
Capozza and Schwann (1989, 1990) used the census data for the Canadian
Metropolitan Areas from 1969 to 1985 to test the theoretical relationship
between uncertainty and land price within an urban land development
framework. Two variables in the theoretical urban land development model
were dropped from their empirical analysis due to non-availability of data.
They recognized that the regression results are sensitive to the assumptions
made in deriving the input variables. When regressing the hurdle price of
undeveloped land on the rental variance, opposite results were found in the two
independent studies by Capozza and Schwann (1989, 1990). Capozza and
Schwann (1989) obtained a negative, albeit statistically insignificant,
coefficient for the hurdle price variable, when the rental variance was computed
exogenously from the housing completion rate. When the rental variance
variable was represented by the changes in three endogenous factors:
household size, apartment vacancy rate and housing start, the coefficient of the
risk variable was found to be significantly positive, which contradicted their
earlier findings (Capozza and Schwann, 1990). They, however, concluded that
the hurdle price that triggers a decision to convert an undeveloped land was
positively related to the levels of unsystematic risk.
Patel and Sing (2001) empirically tested the irreversibility theory that
proposes a negative relationship between property investment and market
uncertainties using macroeconomic and property market data in the UK. Based
on the proposed regression models that include various market shock terms on
the right hand side of the models, the coefficients for the market shock terms
were found to be significantly negative at the 5 per cent level for the models
To continue reading
Request your trial