ESTIMATING GLOBAL RISK PREMIUM OF PRIME DIRECT REAL ESTATE INVESTMENT - A DESMOOTHED PERSPECTIVE

Abstract

This study investigates the merits of adopting the multi-factor APT (arbitrage pricing theory) model to estimate the international direct real estate risk premiums. This paper estimates the risk premiums of direct real estate investing through a pooled-panel-data and cross-sectional approach of North Asia, South Asia and United States direct real estate total return data, utilizing the quarterly direct real estate data from the Jones Lang Lasalle Real Estate –Asia (JLL REIS-Asia) and the Russell-NCREIF Property Index (RNPI). Findings confirm the existence of appraisal smoothing for direct real estate data via adopting the Geltner and Miller (2007) 1st and 4th order autoregressive model to desmooth the direct real estate returns. This study affirms that the historical volatility can be a reasonable estimation of direct real estate risk, only when the autoregressive lag orders of the desmoothed returns are taken into account. The results from the APT model estimation reveal that the variations of the macroeconomic and real estate factors explain much more for the office and retail returns as compared to residential returns. The relatively high vacancy rate risk premium is attributable to the institutional environment, in terms of country specific institutional framework. The direct real estate risk premium for South Asia is revealed to be higher than that for North Asia, and that the risk premium for the US is lower than for Asia.