- We analyse the drivers of real estate returns, and evaluate real estate exposures in the context of a diversified equity and fixed income portfolio. We focus on US real estate, analysing both listed and unlisted assets.
- A common approach to analysing real estate is to regress its returns on equity and fixed income returns. This ’spanning’ regression approach explains listed real estate returns relatively well, but only captures a small proportion of the variation in de-smoothed unlisted real estate returns. We argue, however, that the apparent divergence between listed and unlisted real estate is overstated by this regression approach.
- The differences between listed and unlisted real estate appear to reduce over the longer term, where the correlation of their returns increases with horizon. In addition, the correlations with the broader equity market are lower at longer horizons for both real estate segments. These correlation patterns suggest that differences between listed and unlisted real estate returns are short-term and largely driven by transitory factors.
- We estimate the exposures of real estate to fundamental return drivers: expected cash flows, inflation, real interest rates and risk premiums. We find that both segments of real estate hedge inflation risk more than the aggregate equity market, and that listed real estate has a high exposure to transitory risk premium shocks. The inflation and risk premium exposures help to reconcile the spanning regression results and patterns of real estate correlations across different horizons.
- We estimate exposures and correlations for other equity sectors, and find patterns that support our analysis for real estate. In addition, when allowing for inflation exposures in spanning regression models, we better explain unlisted real estate returns, accounting for levels of variation comparable to listed real estate.