REAL ESTATE RETURNS AND FINANCIAL ASSETS IN EXTREME MARKETS

The benefits of diversification from including real estate in a mixed-asset portfolio are typically established on
the basis of low average correlations with other asset classes, in line with modern portfolio theory. However, the
existing literature suggests that correlations may vary through time. Correlations tend to be higher in periods of
increased uncertainty and during bear markets.
This evidence suggests that returns from real estate investments exhibit ‘asymmetric dependence’ with respect
to equities and bonds. Tail dependence is defined as a disproportionately high probability of jointly occuring high
or low returns. Asymmetric dependence refers to the situation where this occurs in one tail of the joint return
distribution only, typically the negative tail. That might imply that the benefits of diversification commonly ascribed
to real estate may vanish when they are in fact most needed to protect portfolio values from eroding in downturns.
A statistically robust assessment of asymmetric dependence is therefore crucial for determining the benefits of
diversification associated with including real estate in mixed-asset portfolios. However, analysing asymmetric
dependence is a complex, multi-dimensional problem. Existing studies of asymmetric dependence often rely solely
on correlations, which may be an incomplete description of dependence, or employ methodologies that are unable
to distinguish between different dimensions of dependence. They may, therefore, produce misleading results.
Using Monte Carlo simulations, the research team identifies the most suitable metric of asymmetric dependence
out of a range of methods commonly employed in real estate finance: the adjusted-J statistic. This metric is used in
combination with the CAPM beta to provide a comprehensive assessment of dependence between the returns on
real estate investments and other asset classes.
This set of statistics is employed to empirically examine the dependence structures in a large sample of return data
for direct and indirect UK real estate investments, and a range of benchmark assets over the period 1990–2010.
When we control for linear dependence and focus on the strength, direction and statistical significance of higher-
order, asymmetric dependence, the benefits of diversification offered by real estate appear stronger than sometimes
reported in recent studies. There are asymmetric dependence relationships between real estate and equities.
However, these seem to be a high-frequency phenomenon – relevant only to securities funds that are frequently
rebalanced – and the effects reduce rapidly with lower frequency.
By contrast, dependence between bonds and real estate appears to be more complex than previously assumed. The
relationship re-emphasises the importance of gearing and credit conditions in affecting property performance.
These finding highlights the importance of employing a statistically robust metric of asymmetric dependence. It
may be possible to develop portfolio strategies based on such measures that can limit the downside risk while not
adversely affecting the upside. As yet, though, such models are confined to listed securities and would be hard to
implement in private real estate