This paper examines whether the predictability of securitized real estate returns differs from that of stock returns. It also provides a crosscountry comparison of securitized real estate return redictability. In contrast to most of the literature on this issue, the analysis is not based on a multifactor asset pricing framework as we deem that a univariate time series approach is more appropriate to compare the predictability of returns between two or more asset classes. Forecasts are performed with ARMA and ARMA-GARCH models. Such forecasts are evaluated by comparing the entire empirical distributions of prediction errors, as well as with a trading strategy. The results show that the predictability of securitized real estate and stock returns differs in some countries, even though these differences are not economically significant. Similar conclusions are reached when comparing the predictability of securitized real estate returns across countries.