Do Public Real Estate Returns Really Lead Private Returns?

Abstract

We use sector level REIT and transaction-based direct real estate data for the period 1994-2010 to provide a clearer understanding of the dynamic relations between public and private real estate returns. We add leverage to private returns to make the private data more comparable with the REIT data. We also include economic fundamentals in the analysis to take account of the influence of fundamentals on real estate market dynamics. Moreover, we consider the influence of the 'escrow lag' in the recording of private market prices. The estimated vector error-correction and vector autoregressive models, Granger causality tests, impulse response functions and variance decompositions all provide evidence of REIT returns leading private returns in the office, retail, and apartment sectors. These lead-lag relations appear to be due to the slow reaction of private market returns to shocks in REIT returns and in some economic fundamentals. In the industrial sector, such lead-lag relation cannot be observed, however.

On the dynamic interactions between sector-level REITs, their direct real estate counterparts and the stock market in Japan

“…Recent evidence support this causal relationship in other market like Australia and Netherlands (Yong & Pham, 2015;Yunus et al, 2012). Overall, these findings indicates price discovery generally initiates in REITs, since they react faster to changes in market information over the direct real estate market (Li, Mooradian, & Yang, 2009;Ling & Naranjo, 2015;Yavas & Yildirim, 2011 (Hoesli et al, 2015;Yavas & Yildirim, 2011).…”

“…In this section, we investigate the causal relationship between one asset and another using Granger Causality Test. Consistent with Yunus et al (2012) and Hoesli et al (2015), our test on short-run causal relationship between the assets is based on the Vector Error Correction Model (VECM):…”

Does listed real estate behave like direct real estate? Updated and broader evidence

“…Germany is considerably different from the other markets considered in terms of the discrepancy in property type mix between direct and listed indexes. Based on theoretical considerations (i.e., the basic asset pricing formula) and previous empirical evidence (Hoesli and Oikarinen 2012;Hoesli, Oikarinen, and Serrano 2015), the set of economic fundamentals incorporated in the SVAR model includes GDP, the risk premium, the short-term risk-free interest rate, and consumer sentiment. GDP measures growth in the real economy that influences the demand for real estate space, and thereby the expected rental cash flows.…”

“…The sentiment component of the confidence indicators can be regarded as the component that is unrelated to prevailing economic fundamentals. Therefore, similar to, e.g., Ling, Naranjo, and Scheick (2014) and Hoesli, Oikarinen, and Serrano (2015), we regress the differenced confidence indicator on the three economic fundamentals (differenced) mentioned above and use the residual series of this OLS regression as our sentiment measure (S). For France, Germany, and the Netherlands, the GDP and S are at the country level, whereas D and R reflect the whole Euro area.…”

“…The observed discrepancy between shortand long-term correlations is not surprising given the notable frictions in direct real estate markets that tend to make direct market price adjustments sluggish. Such sluggish adjustments can cause lead-lag relations between REIT and direct returns that diminish the short-term correlations (Li, Mooradian, and Yang 2009;Oikarinen, Hoesli, and Serrano 2011;Yunus, Hansz, and Kennedy 2012;Hoesli, Oikarinen, and Serrano 2015).…”