Institutional involvement and the REIT January effect over time

The Revenue Reconciliation Act of 1993, implemented since 1 January 1994, facilitates the increase of institutional investment in the real estate investment trust (REIT) market. Utilizing this particular feature in the market, we examine the time‐series effect of institutional holdings to distinguish the tax‐loss‐selling hypothesis and the window‐dressing hypothesis for REITs. Consistent with the tax‐loss‐selling hypothesis, we have evidence that the January premiums decreased with the level of institutional involvement for REITs. Furthermore the January premiums declined significantly for equity REITs only that attracted more institutional investors than mortgage REITs. On the other hand, the January premiums did not decrease significantly for mortgage REITs. Overall the results suggest that trading strategies to profit the higher January returns may work only when institutional investors exit and leave the market.