The role of real estate in optimal portofolio choice

Abstract

The mαin issues addressed in this study are the following: Investigate the effect of real estate on the risk-return characteristics of optimal portfolios. For this purpose, there will be investigated the change in the risk-return tradeoff by including real estate in optimal portfolios consisting of bonds, stocks and Fama & French Portfolios formed on Size and on Book-to-Market. In order to assess the diversification benefits for investors of including real estate in their portfolios, there will be used a number of well-known portfolio performance and efficiency tests. These tests are known in the literature as mean-variance intersection and spanning tests. It is well known that the solution of the optimal portfolio differs according to whether the investor is allowed to sell assets short. Hence, it is important to compute optimal portfolios for both cases of short selling and no short selling. The analysis will also take into account the illiquidity characteristic of the real estate asset class. Thus, there will be also computed for portfolios where the investor is able to buy and sell all assets constantly and portfolios for which the investor is restricted in the housing asset only. This restriction implies a hedging demand for both bonds and stocks, implying that the investor will choose to change his position in real estate in order to protect his investment from the risk the real estate restriction brings in. Another interesting aspect, is following: The basic portfolio choice model of Markowitz assumes that the investor is myopic in the sense that he is only interested in a one-period investment. However, real estate is an asset that contrary to financial assets and due to its nature cannot be treated easily as a short-term investment. Households usually treat housing wealth as an investment for lifetime or at least, a long-term investment. As a result, the portfolio choice of a household, which includes housing wealth in its assets, is an intertemporal optimization problem. In this type of problems, the solution of the optimal portfolio usually consists of two parts: the standard Markowitz solution (the universal hedge portfolio) plus the hedging demand component which adjusts the universal portfolio by taking into account the effect of a change in the investment opportunity set on the optimal composition of the portfolio. So, the upcoming question is whether there is a hedging demand component for stocks and bonds induced by real estate. A hedging demand for stocks and bonds can arise in this case if for example changes in mortgage rates, which affect the refinancing costs of real estate are correlated with changes in the value of the stock, or bond component in the portfolio. If this is the case, then the fact that the investor owns a house means that he will choose a different optimal portfolio of stocks and bonds than would be the case if he would rent a house. The reason is that by owning these financial assets, the investor can hedge the risk of a change in the refinancing costs of real estate. The empirical analysis will focus on the U.S. market, since there are more reliable monthly data available on residential house prices, than for other countries.