Analysing the Risk of Income-producing Real Estate

Abstract

Institutional investors typically allocate under 5 per cent of their portfolios to income-producing real estate, yet the results of most academic studies imply that far larger allocations are optimal. One widespread argument is that the conflict between observed behaviour and scholarly prediction is spurious in the sense that it arises due to the use of faulty appraisal-based data. While lagged appraisals undoubtedly do smooth real estate return data, we argue that two other factors help explain the low measured variance of income-producing property returns and the low covariances of appraisal-based index returns with the stock market. The first is that most appraisal-based series are unlevered while the Standard and Poor 500 firms are approximately 50 per cent levered. A synthetic leveraging (i.e. gearing) of the Frank Russell Company real property series almost doubles the variance in its returns. The second factor flows from real estate market fundamentals. Because rents basically are a fixed cost to tenant firms, rental flows on existing buildings should be more stable than are the firms' cash flows. This implies that real property returns would not co-vary strongly with the stock market.