Within Real Estate Diversification and Investment strategies

Sammanfattning: The efficient portfolios for the period 1993 – 2010 based on IPD data have a major portfolio weight in residential properties in the three largest regions Stockholm, Gothenburg and Malmo. The portfolio with the highest risk adjusted return (measured as the highest Sharpe-ratio) combines a large portfolio weight in residential properties with a small weight in industrial properties. During the time period of 2005 – 2010 a majority of the listed real estate companies held a real estate portfolio far below the efficient frontier based on the corresponding IPD data. These companies can increase their total return without taking on any more risk by using the concept of diversification. When including all available diversification categories two out of seven companies can be said to have an efficient real estate portfolio. When we excluded the outperforming residential asset class, however, none of the companies’ portfolios were in fact efficient. The real estate market is inefficient and thus results in the IPD data being less useful as it is based on transactions occurring in this inefficient market. Investors can, in this market, easily find properties with another risk and return profile than what IPD indicates is the market risk and return for a particular property type in a certain region. The inefficiency of the market, together with the IPD data being less useful, thus makes it difficult for the companies to focus on diversification in their investment strategy. Moreover, there are several reasons that explain the discrepancy between the actually held listed real estate portfolios and the optimal portfolio based on IPD data. Since each property is heterogeneous and possesses unique risks, investors are not able to accurately quantify the risk of each investment and thus rely more on their gut feeling. This also results in investors focusing on single investment opportunities rather than looking at all investments from a portfolio perspective.