Exposure to real estate is usually incurred by acquiring properties. There are, however, other less direct ways of investing in real estate markets, such as purchasing shares in listed real estate companies or investing in a non-listed real estate fund. These two alternative investment paths allow investors to employ much smaller amounts of capital, certainly less than is required for direct investments. What’s more, with each of these approaches a good level of diversification within the real estate market can be achieved with limited capital. But it remains of utmost importance to investors whether these two investment solutions behave in the same way as the underlying real estate and more specifically whether they react to various risk factors in the same way as direct investments themselves.
There has been much research into the risk factors of direct and listed real estate, and it suggests that listed real estate investments are akin to stock investments in the short run but behave more like the underlying real estate if the time horizon exceeds three or four years. Studies that deal with the risk factors of non-listed investments are far rarer, though, and hence less is known about how such investments compare with other types of real estate exposure, this knowledge gap existing predominantly due to the heterogeneity of fund characteristics and the lack of publicly available information. But understanding the determinants of non-listed fund performance is of great importance, notably to investors engaged in making the relevant allocation decisions.
Our research seeks to identify the risk factors affecting the performance of non-listed real estate funds. Given the aforementioned heterogeneity of funds, we stress the importance of taking their characteristics into account. The results of our research for non-listed funds are then compared with those pertaining to direct and listed real estate investments. We interrogate non-listed real estate fund data sourced from the European Association for Investors in Non-Listed Real Estate Vehicles (INREV) and MSCI/IPD data for direct investments and Thomson Reuters Datastream data for listed investments. Our analyses are based on yearly data for the period 2001‒14, and we focus on funds invested in France, Germany, Italy, the Netherlands and the United Kingdom.
Several important fund characteristics, including sector, investment style, vehicle structure, size and gearing, are considered. For sectors, we consider industrial, office, residential and retail. Residential funds, for instance, should be less profitable and also less risky than industrial funds. Next, the investment style indicates whether a fund has a more or less conservative, or aggressive, strategy. We distinguish between core and value-added funds, with the former targeting stable returns and the latter strong capital growth. One would expect core funds to display lower but more stable returns than do value-added funds. For vehicle structure, we consider open- and closed-end funds. An investment in an open-end fund should be more liquid than one in a closed-end fund. Size is also taken into account to test its impact on fund returns and to assess whether an optimal fund size exists. Finally, gearing should impact a fund’s return and risk characteristics.
Our results show that there is no difference in returns across sectors for non-listed real estate when controlling for other variables. For vehicle structure, open-end funds performed better than closed-end funds during the subprime crisis, reflecting the greater flexibility of capital allocation allowed by an open-end structure. Regarding investment style, value-added funds performed worse than core funds after the crisis. No doubt the uncertain context that characterized the post-subprime period, notably involving the European debt crisis, was not ideal for riskier investments such as value-added ones. Interestingly, during the subprime crisis itself there were no differences in returns across investment styles, suggesting that price drops were largely irrespective of investment style.
A more complex relationship is found between size and fund returns—it seems that as a fund grows, its return increases but that there is a turning point beyond which the return decreases. This implies that there is an optimal fund size. Our model indicates an optimal size of €2.3 billion; but from a statistical point of view the confidence interval is rather large. The median fund size in our sample is only €350 million with no more than 10 percent of observations displaying a size above €1.2 billion and the largest fund having a size of €5.1 billion. Thus, most of the funds in our sample would improve their performance if they were larger. More specifically, the model indicates that a fund with this optimal size would have a yearly return 2.7 percent higher than that of a fund with the median size. A possible explanation for this is that, given the large unit value of properties, only very large funds can acquire a sufficient number of assets to achieve a good level of diversification. Large funds are also able to seize opportunities and invest in properties in prime locations, a strategy that should also yield higher returns. Economies of scale are yet another factor that could explain these results.
For gearing, we find a similar relationship as for size, but only for value-added funds, for which the optimal long-run gearing level over the whole business cycle is 28.7 percent. This leads to an increase of 6.3 percent in the yearly return, compared with a fund that does not rely on debt at all. At 49.7 percent, the median gearing level of value-added funds in our sample is much higher than the optimum and does not lead to any significant increase or decrease in return. Further computations suggest that beyond a gearing level of 60 percent, the impact on the return becomes significantly negative. An optimal gearing thus exists, and too high a gearing is detrimental to a fund’s performance.
We also investigated the optimal level of gearing by market phase of the underlying real estate market—that is, by separating periods during which the direct market return was positive from periods during which it was negative. Due to data limitations, however, it was not possible in this case to differentiate funds according to investment style. The optimal gearing level during market-up phases is 24.5 percent, which allows for an additional return of 3.8 percent compared to the performance of a fund with no debt. This level is very close to the median gearing of core funds (22.9 percent), but substantially below that of value-added funds. In down markets, gearing should be at 12.8 percent, which would provide a 1.7-percent positive contribution to returns. A possible explanation for the somewhat surprising result of the positive impact of debt on returns in a down market is that some debt provides for flexibility, making it possible to seize investment opportunities that occur in bearish markets. However, beyond a level of 12.8 percent, debt becomes detrimental to performance, as is the case, for instance, for value-added funds in our sample. In sum, our results emphasize the need to manage gearing as it can help generate substantial additional returns if set at the proper level.
Our results also suggest a strong relationship between fund returns and several macroeconomic variables. First, a positive link to real gross domestic product (GDP) growth is found: fund returns should increase on average by 2.9 percent following a 1 percent increase in real GDP growth. This impact is approximately twice as large for listed companies, suggesting that their returns are more sensitive to economic growth than those of non-listed funds. For direct real estate, the effect of GDP growth is lower as a 1 percent increase in real GDP growth should lead to an additional 1 percent of return. Inflation components are also found to impact real estate returns, with notable differences across the types of real estate exposure. While non-listed and direct returns respond positively to inflation and negatively to unexpected inflation, these relationships are not as clear-cut for listed investments.
An increase in the long-term real interest rate has a negative influence on returns for the three types of real estate exposure studied, but that influence is more muted for direct and non-listed investments than it is for listed ones. The negative effect of interest rate increases on returns reflects the negative effect of a higher cost of capital on real estate values. We also report a positive relationship with the real money-supply growth for non-listed and direct real estate, and to a lesser extent for listed companies. This reflects investors’ easier access to credit under such circumstances, which should lead to higher demand and hence to higher real estate prices.
Both non-listed and listed real estate returns are positively related to stock market returns. A 1 percent increase in real stock returns leads to a positive impact of 0.28 percent for non-listed funds and of 0.5 percent for listed real estate companies. These figures can be seen as the usual market betas widely used by financial analysts. The greater impact of the stock market on listed returns is not surprising given that real estate securities trade on an exchange, which should lead to strong linkages between stocks and listed real estate, in the short-term anyway. Direct real estate returns appear unrelated to stock returns.
In general, with respect to the results reported for the responses of real estate returns to macroeconomic risk factors, it appears that non-listed funds are more similar to direct investments than they are to listed real estate companies. The same conclusion is reached when examining the linkages between the returns of the three types of real estate exposure. Hence, non-listed funds are a valid alternative to direct investments for an investor wishing to allocate part of his or her portfolio to real estate.
Delfim, J.-C. and Hoesli, M., 2016, “Risk Factors of European Non-Listed Real Estate Fund Returns”, Journal of Property Research, Vol. 33(3), pp. 190-213.