Can we predict the property cycle' A study of securitized property market

摘要

Academia takes interest in cyclicality of real estate market. Compared to various findings on housing cycles, no literature takes insight into the cycles of securitized property markets. To address the issue, a nonlinear model is developed to probe into the characteristics of cycles in global markets (US, UK, Australia, Japan, Singapore and Hong Kong) over the last 23 years. The findings suggest that (a) cointegrating relationships influence the six markets in the long term and become stronger during bullish markets. (b) The short-term dynamics of each market is more likely to have a regime-switching structure. (c) The cyclical pattern shows differences between securitized property and housing markets, as well as between securitized property and general stock markets. Meanwhile, the cyclical pattern in developed markets is also different from that in developing markets. (d) The duration dependence shows a weak effect of the boom on predicting the occurrence of the upcoming bust. Instead, the magnitude of boom growth plays a significant role in predicting the duration of following bust. (e) The asymmetric analysis brings forward the "paralleling effect" which indicates that the asymmetry in returns is parallel with the movements of r. The methodology shall serve in providing detailed implications on the characters of cycle and duration forecast in securitized property markets for investors and governments.