This study addresses the short-term disparity between REIT returns and direct property returns, and argues that this phenomenon is due to the trading constraints in the direct property market imposed on REITs (the dealer rule). This renders REITs unable to time markets in order to realize short-term property appreciation profits, making REITs primarily a property income investment rather than a full property investment, and explains the observed disparity. Empirically, I find that REIT returns consistently reflect property income returns, but not property appreciation returns. This makes this study the first in the literature to find a consistent link between REIT returns and any portion of direct property returns at short time horizons, in the context of a linear factor model. I then set up a natural laboratory to test the trading-constraints explanation by examining the appreciation dependence of different types of REITs, which should be differently affected by the trading constraints. I find that returns to UPREITs, which are less affected by the constraints, have a stronger appreciation dependence than returns to regular REITs. I also perform a size test and find that large REITs, which are less affected by the constraints, have a stronger appreciation dependence than small REITs. When testing the effects of UPREIT and size characteristics simultaneously, I find a consistent UPREIT effect. I further find that Real Estate Operating Companies (REOCs), which are not subject to trading constraints, show short-term property appreciation dependence. These findings offer strong support for the trading-restrictions explanation.