New research asserts that over the long term listed real estate equities are a proxy for direct real estate investment. Steffen Sebastian and Alexander Schätz present their findings

For years, experts have discussed the question of whether the performance of listed real estate is primarily driven by real estate markets or by stock markets. A recent study commissioned by EPRA and produced by IREBS International Real Estate Business School at the University of Regensburg resulted in a clear answer: the medium-term to long-term performance of listed real estate correlates significantly with the development of direct real estate markets. However, in the shorter term performance is influenced by stock market developments.

These unambiguous findings are the result of research conducted on markets in the US and in the UK. For the first time, the approach selected for the research included macroeconomic data. In addition to the clear conclusion with regard to the performance of listed real estate, the study identified serious dependencies between the development of US direct real estate markets and the development of the non-monetary US economy. The results compiled for the UK were not as pronounced as those for the US. In the UK, development of the stocks and listed real estate indices led to the conclusion that financial and real estate markets mutually influence each other.

Research on the principal behaviour of listed real estate is anything but new. Questions in this context are raised particularly by those who are looking for an investment alternative to direct real estate ownership. Because of its low correlation with other asset classes, real estate will offer stronger diversification benefits in an investment portfolio. As a tangible fixed asset, direct real estate offers a high stability of investment, and thus displays effective inflation-hedging qualities. However, lot sizes mean potential buyers need to vault high hurdles in order to enter the direct real estate market. It is clear that high investment sums and the long-term commitment (lock-up) of the equity, mean that direct real estate investment is not as fungible as stock. Moreover, the international direct real estate markets are not nearly as sophisticated as the markets for equities and bonds in terms of liquidity and transparency.

In recent years, the scope of options for indirect real estate investment has expanded significantly, and now constitutes a viable alternative to direct real estate investment. Institutional investors have the choice of a wide range of investment and diversification options for their portfolios - open and closed-end real estate funds, listed real estate operating companies (REOCs), listed real estate investments trusts (REITs) or real estate private equity funds.

Past surveys have failed, however, to draw clear conclusions about the behaviour of listed real estate. Ultimately, inconsistent data, methodology selection, sample size and market choice have all combined to hamper results. When answering the question whether or not listed real estate behaves like the direct real estate market or the equities market, we must highlight two principal presuppositions:

Under the environment of continuous trading and constant recalculation of share prices, it could be assumed that the performance of listed real estate and its subsequent risk/return profile is influenced by developments on the general stock markets; In addition, the latest economic developments are factored into the latest share prices, along with other factors such as analyst expectations and valuations.

On the one hand, listed real estate companies consequently expose themselves to market risk generated by stock market trends; on the other hand, the core business of listed real estate companies remains the long-term management of property. The question is: which hand is the strongest?

Within the framework of the study, research for the first time included macroeconomic conditions. The focus of this new analytic approach did not rest exclusively on the contexts of the three asset classes traditionally studied to address the issue - namely, real estate equities, direct real estate, and general stock. Rather, the selection of the markets investigated took into account international diversity with regard to structural conditions and parameters.

The US and the UK real estate markets are similar in having high levels of transparency and offering low-level transaction costs. In addition, the large trading volume of both markets underlines their advanced development stage and suggests a higher level of liquidity than in the real estate markets of other industrialised nations. Data and indices of the US and UK markets for direct and indirect real estate investments are deemed reliable and representative for both countries, and this was a prerequisite for the research methodology selected. That said, the results for less-developed markets indicate that the survey findings apply there as well. The more developed and transparent the market, the more likely the market is to behave along similar lines. An excellent example in this case is Australia.

To assess the degree of transparency in the US and UK markets, one may also study their existing indices. They are meant to provide an overview of price and performance figures, and to delineate trends for the markets covered. However, the direct real estate indices of other nations do not compare to US and UK indices which have the well-known and widely used NCREIF and IPD, respectively, with their comprehensive market coverage and long history. Calculation of the NCREIF Property Index (NPI) in the US started as early as 1978. In Q1 2008, it covered a total of 5,976 properties covering all types of uses with an aggregate market value of $328bn (€219bn). It is disseminated on a quarterly basis and, being a valuation-based index, measures the total return of net cash-flow return and capital growth for the mapped, predominantly commercial real estate.

The UK equivalent is the monthly adjusted direct real estate index of the Investment Property Database (IPD), which represented 3,695 properties with a combined market value of approximately £41bn (€45bn) as of August 2008. The US listed real estate market was represented by the Equity REITs index of the National Association of Real Estate Investment Trusts (NAREIT). This index also reflects the average total return of its roughly 110 constituent companies with a market capitalisation of nearly $277bn. The general stock markets in the study are represented by the S&P 500 Composite index in the US, and by the FTSE 100 index in the UK. Like the already mentioned indices, the general equity indices are weighted according to the companies' capitalisation.

The selection of macroeconomic factors is rooted in theoretical assumptions, integrating the key drivers of the macroeconomic environment without overloading the model with parameters. The three factors under review were economic growth, inflation and the influence of the money market. The benchmark used to reflect economic growth in the surveyed US and UK markets was the respective gross domestic product (GDP). The consumer price indices (CPI) of the two countries provided the determinants for the respective inflation rate, enabling the researchers to appraise to what extent property does hedge inflation. Interbank rates were used in turn to gauge the role played by the money market. These interest rates permit inference of the resulting loan costs, which affect the investment climate.

The evaluation started in 1992, once the US data for the years between 1978 and 2008 along with data from the UK from 1988 through 2008 had been screened for structural breaks. Such breaks could qualify trend assumptions in the analysed time series and eventually lead to misinterpretations. In both countries, the records suggested just such a break in 1992, and it was explained by the foregoing recessive cycle. The recovery of both national countries was boosted through a characteristic cut in the key interest rates by the respective central banks. While the key interest rate in the US dropped from 9.75% to 3% between 1989 and 1993, the expansive monetary policy pursued in the UK bottomed out at 5.25% in early 1994. One needs to remember that the Bank of England's key lending rate was 15% as late as the end of 1990.

With the observation periods selected, a complex Vector Error Correction Model (VECM) was used to evaluate the data. This econometric procedure helps to evaluate time series such as stock quotes/prices. The variables taken into consideration are part of a meaningful yet - unlike with simple linear regression models - initially unknown context. Whenever they mutually influence each other, they are called co-integrated. These co-integration models are particularly well suited for the study of long time horizons with fewer data points widely spaced along the time axis. After all, a key objective of the study was to avoid distortions possibly caused by the specific characteristics of the selected time series. Indices for the general stock market and real estate equities are continuously calculated on a daily basis.

By contrast, macroeconomic data are published at best once a month or once a quarter - as is the case with GDP. Moreover, data on the national economy are often revised after their publication. The indices for direct real estate markets are compiled even less frequently because they are based on the valuations of individual properties. Economic developments or fluctuations that might affect real estate prices thus do not enter into these indices except with a time lag. Obviously, this hardly constitutes a sensible basis for a monthly analysis of direct real estate price trends. The study ultimately used quarterly data so as to take the peculiarities of the direct real estate market into account.

The large number of factors posed yet another problem for the analyses. The more factors are fed into a co-integration model, the higher the likelihood that the findings become too unstable to derive meaningful conclusions from them. However, the data proved to be very consistent, remarkably meaningful, and clear. Stable findings demonstrated the suitability of the selected macroeconomic data. Moreover, they confirmed the selection of the time increments between the data. The conclusion derived from the long-term survey, though, were:

The performance of real estate equities in both countries is significantly affected by developments on the underlying direct real estate markets; The longer the time period under consideration, the stronger the influence of the direct real estate market. This escalates to the point where you can deduce with reasonable certainty that the performance of listed real estate over a very long investment horizon will ultimately match the performance of direct real estate ownership; While a short-term study of listed real estate reveals its susceptibility to the trends of the general stock market, they are definitely driven in the longer run by the performance of the actual or underlying real estate held in the respective portfolio; The study also shows how strong these dependencies are over differing time horizons;

Aside from having profiled the characteristics of listed real estate, the study confirmed the following economic-theory assumptions for both economies:

Rising quotes/prices on the general stock market will in turn prompt a positive performance for direct real estate investments; Negative performance, by contrast, is explained by an increase in the inter-bank rate, as real estate tends to be financed with a large share of debt capital. Whenever loan costs go up, the investment climate deteriorates and demand for direct real estate investment declines.

That said, there are manifest differences between the countries studied. In the US, the development of the real estate market is more closely intertwined with the macroeconomic development than is the case in the UK. For instance, the findings suggest strong reciprocal relationships between GDP and interest levels in the US. Over the entire observation period, elevated growth rates of the overall economy coincided with low interest rates. The latter encouraged additional, debt-financed investments, and thereby precipitated an increase in real estate prices.

In the UK, the observed development of the indices for stocks and listed property companies suggested that financial and real estate markets mainly influence each other. The US figures revealed a positive influence of the CPI on the development of listed real estate, which thus benefited from rising inflation rates. The figures for the UK, by contrast, failed to suggest either a positive or a negative influence in the same context. Any statement on effective inflation hedging of real estate investments needs to take the economic environment and its linkage to the real estate sector into account.

Real estate investments are particularly suitable for investors with multi-asset portfolios because of their low correlation with other asset classes. Direct real estate investments, however, are constrained by entry barriers such as high transaction costs, transparency gaps and poor liquidity. Assuming that listed real estate serves as an adequate medium-term to long-term substitute, or proxy for direct real estate investments, investors with an extended investment horizon can profit from the advantages of both asset classes - from the liquidity, transparency, and management of listed real estate, on one hand, and from the diversification qualities and the risk/return profile of direct real estate, on the other hand.

The study shows that listed real estate can not only act as a proxy for direct real estate investment, but also illustrates how this investment approach pans out over various investment horizons. Anyone wishing to invest long term in real estate, and having a sufficient degree of flexibility, will find listed real estate a sound alternative to direct real estate ownership.

Steffen Sebastian is professor of real estate finance at IREBS
Alexander Schätz, department of real estate finance at IREBS