An assessment of the risk and return of residential real estate
y Dale Domian, Rob Wolf, Hsiao-Fen Yang
Format:
Article
Publication year:
2015
|
Peer-reviewed
| No other editions or formats
Journal:
· Managerial Finance v41 n XXXXXXXXXX): XXXXXXXXXX
– The home is a substantial investment for most individual investors but the assessment of risk and return of residential real estate has not been well explored yet. The existing real estate pricing literature using a CAPM-based model generally suggests very low risk and unexplained excess returns. However, many academics suggest the residential real estate market is unique and standard asset pricing models may not fully capture the risk associated with the housing market. The purpose of this paper is to extend the asset pricing literature on residential real estate by providing improved CAPM estimates of risk and required return. – The improvements include the use of a levered β which captures the leverage risk and Lin and Vandell XXXXXXXXXXTime on Market risk premium which captures the additional liquidity risk of residential real estate. – In addition to presenting palatable risk and return estimates for a national real estate index, the results of this paper suggest the risk and return characteristics of multiple cities tracked by the Case Shiller Home Price Index are distinct. – The results show higher estimates of risk and required return levels than previous research, which is more consistent with the academic expectation that housing performs between stocks and bonds. In contrast to most previous studies, the authors find residential real estate underperforms based on risk, using standard financial models.
– The home is a substantial investment for most individual investors but the assessment of risk and return of residential real estate has not been well explored yet. The existing real estate pricing literature using a CAPM-based model generally suggests very low risk and unexplained excess returns. However, many academics suggest the residential real estate market is unique and standard asset pricing models may not fully capture the risk associated with the housing market. The purpose of this paper is to extend the asset pricing literature on residential real estate by providing improved CAPM estimates of risk and required return. – The improvements include the use of a levered β which captures the leverage risk and Lin and Vandell XXXXXXXXXXTime on Market risk premium which captures the additional liquidity risk of residential real estate. – In addition to presenting palatable risk and return estimates for a national real estate index, the results of this paper suggest the risk and return characteristics of multiple cities tracked by the Case Shiller Home Price Index are distinct. – The results show higher estimates of risk and required return levels than previous research, which is more consistent with the academic expectation that housing performs between stocks and bonds. In contrast to most previous studies, the authors find residential real estate underperforms based on risk, using standard financial models. Read Less
An assessment of the risk and
eturn of residential real estate
Dale Domian
School of Administrative Studies, York University, Toronto, Canada, and
Rob Wolf and Hsiao-Fen Yang
Department Finance, University of Wisconsin at La Crosse, La Crosse,
Wisconsin, USA
Abstract
Purpose – The home is a substantial investment for most individual investors but the assessment of
isk and return of residential real estate has not been well explored yet. The existing real estate pricing
literature using a CAPM-based model generally suggests very low risk and unexplained excess
eturns. However, many academics suggest the residential real estate market is unique and standard
asset pricing models may not fully capture the risk associated with the housing market. The purpose of
this paper is to extend the asset pricing literature on residential real estate by providing improved
CAPM estimates of risk and required return.
Design/methodology/approach – The improvements include the use of a levered β which captures
the leverage risk and Lin and Vandell XXXXXXXXXXTime on Market risk premium which captures the
additional liquidity risk of residential real estate.
Findings – In addition to presenting palatable risk and return estimates for a national real estate
index, the results of this paper suggest the risk and return characteristics of multiple cities tracked by
the Case Shiller Home Price Index are distinct.
Originality/value – The results show higher estimates of risk and required return levels than
previous research, which is more consistent with the academic expectation that housing performs
etween stocks and bonds. In contrast to most previous studies, the authors find residential real estate
underperforms based on risk, using standard financial models.
Keywords Housing, Asset pricing, Real estate
Paper type Research pape
I. Introduction
Residential real estate is perhaps the most commonly held investment asset in the
world. Practitioners and some academic research suggest residential real estate has
high returns with low risk[1]. Nevertheless, due to the unique characteristics of
esidential real estate, many authors (e.g. Francis and I
otson, 2009) argue the risk of
esidential real estate investment is higher than often suggested. The recent fallout
of the housing market supports the latter view. Regardless, improved understanding of
esidential real estate performance is long overdue.
Goetzman XXXXXXXXXXuses a mean-variance framework to estimate the effects of
including a single family home in an investment portfolio. The results suggest home
ownership reduces portfolio risk and diversified real estate ownership adds further risk
eduction suggesting real estate portfolios are comparatively stable investments.
Flavin and Yamashita XXXXXXXXXXuse a mean-variance analysis to determine the impact of
home ownership on asset allocation. They argue home ownership has life-cycle
implications, as young homeowners are burdened with high mortgage debt and must
alance their risk with increased ownership in bonds, while older homeowners have
less home ownership debt and can benefit from increased equity ownership. Also,
Waggle and Johnson (2003, 2009) use a mean-variance utility function to estimate the
Managerial Finance
Vol. 41 No. 6, 2015
pp XXXXXXXXXX
©Emerald Group Publishing Limited
XXXXXXXXXX
DOI XXXXXXXXXX/MF XXXXXXXXXX
Received 25 July 2013
Revised 2 July 2014
24 December 2014
Accepted 4 January 2015
The cu
ent issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/ XXXXXXXXXXhtm
591
Return of
esidential
eal estate
impact of the home ownership decision on optimal asset allocation. They find most
investors should have a higher allocation of risky stock, when including the impact of
home ownership. Assessing the impact of home ownership on a scenario analysis of
portfolio performance is useful, however it is also important to price asset risk and
eturn with some theoretically based models.
CAPM is the standard asset pricing model for financial assets, but the unique
market characteristics of residential real estate such as low liquidity, high leverage,
high information asymmetry, etc., suggest the traditional CAPM which assumes the
market is perfect and complete may not fully capture the risks in the residential real
estate market (See footnote 1). How to adjust the one-factor CAPM model to the unique
characteristics of the residential real estate market is an issue of interest to academe
and to practitioners.
This research presents a CAPM-based asset pricing model with improved
assessment of asset risk. Specifically, we incorporate two important risks, associated
with the housing investment, into the traditional CAPM: liquidity risk and leverage
isk. Due to the heterogeneity of housing assets, the residential real estate market is
very illiquid. Investors incur additional costs, including transaction costs, search
costs, financing costs, and maintenance costs, when their houses cannot be sold
quickly. The liquidity risk taken by housing investors is not trivial. In this paper, we
consider the liquidity risk explicitly and use Lin and Vandell’s XXXXXXXXXXTime on
Market (TOM) multiplier to proxy for the liquidity risk. Also included is a premium
to proxy for the higher leverage of home ownership. Residential real estate is usually
the largest investment in a household. To finance the initial investment, most
families bo
ow a substantial percent of the home value. The high leverage
magnifies the risk of the investment and increases the possibility of foreclosure
during an economic downturn. To capture the increased risk, we use a levered β to
estimate the systematic risk and the required housing return. Our results show that
oth liquidity and leverage risk premiums capture unique risks associated with
esidential real estate. The result is to increase the estimated risk measures including
standard deviation, β and required return.
The paper proceeds with a literature review in Section II, a review of the model, data
and results in section III, and the conclusions presented in Section IV.
II. Literature review
The literature has taken at least two distinct approaches to developing an asset pricing
model for real estate. The first approach is CAPM (Sharpe 1964; Lintner 1965), the
standard asset pricing model. Due to its strong theoretical, intuitive, and practical
considerations, CAPM is the most commonly used model in the industry. Using
portfolio theory or CAPM, Geltner (1989), Cascio and Clutter (2008), MacKinnon (2008),
Francis and I
otson (2009), and Cotter and Roll XXXXXXXXXXsuggest, that compared to
equity investments, residential real estate has often provided high levered returns with
low risk and low co
elation to the market. Cotter and Roll XXXXXXXXXXfind the risk of REITs
less than that of equity markets. MacKinnon XXXXXXXXXXfinds residential real estate to have
the fourth highest Sharpe ratio of twelve asset classes. Consistent with Goetzman
(1993), MacKinnon concludes housing’s contribution to an investment portfolio is more
isk reduction than return maximization. Francis and I
otson XXXXXXXXXXfind the average
eal estate investor accumulates wealth slowly with the additional benefits of high
leverage, tax deductibility of interest and valuable diversification opportunities.
592
MF
41,6
In addition to the one-factor CAPM, numerous multifactor models have been
developed for real estate, often justified by the Fama and French XXXXXXXXXXthree facto
model. In contrast to Fama and French, the multifactor models for real estate are
designed for and use factors specific to the real estate industry. Cannon et al. (2006)
develop an asset pricing model for the housing market from 1996 to 2003 finding a
positive relationship between asset volatility and return as well as finding stock market
isk priced directionally in the housing market. Beracha and Skiba XXXXXXXXXXdevelop
a cross-sectional pricing model for housing returns including the factors for the
US real estate market, the local market income growth, land supply elasticity, and
a momentum factor. Over the 25 year sample period, the model factors provide
economically significant explanations of housing market returns that are robust ove
various market segmentations. Case et al XXXXXXXXXXdevelop a multifactor asset pricing
model for housing including a real estate market factor, an idiosyncratic risk factor, a
momentum factor, and a MSA size effects factor. Their model is robust after controlling
for socioeconomic variables. They find the market factor significant suggesting
a clear risk and return framework for housing returns using real estate systematic
isk. Additionally, they find the model robust over different MSAs, but with varying
significance. However, Pai and Geltner XXXXXXXXXXdevelop a three factor model, simila
to the Fama