The data on REIT IPOs are obtained from Thomson Financial SDC New Issues database.
Table 1 reports the number and market capitalization of REITs for the period from 1980 to 2006.
Equity REITs clearly dominate the market, accounting for 75% in terms of the number of REITs and 91.5% in terms of market capitalization in 2006. As mortgage REITs and hybrid REITs account for only a small percentage, we exclude them from our study. Due to data availability, the final sample consists of 169 equity REIT IPOs. Daily share prices, the CRSP value-weighted index, and trading volume are collected from the CRSP database. The management type and structure of the REIT are defined by SNL Real Estate Securities Quarterly.
[TABLE 1 ABOUT HERE]
Table 2 shows the distribution of IPO offer size and lockup length. About 71% (120 out of 169) of REIT IPOs include a lockup agreement, which is somewhat lower than the 80%
reported in Brau, Lambson, and McQueen (2005). The average offer size is $181 million for the whole sample, and the average offer size for REIT IPOs with lockups is almost twice that for REIT IPOs without lockups ($207 million vs. $116 million). As for the length of lockup period, the distribution of our REITs sample is quite different from those for industrial IPOs. Field and Hanka (2001), who examine 1,948 industrial IPOs for the period from 1988 to 1997, find an average lockup length of 187 days; 80% of the IPOs in their sample have a lockup period of exactly 180 days. Of 2,794 IPOs in their sample, Brav and Gompers (2003) report 64% have a lockup length of 180 days. Similarly, Bradley, Jordan, Roten, and Yi (2001) find that 75% of sample of 2,693 IPOs have a 180-day lockup. We do not find strong clustering of lockup lengths.
In our sample of equity REIT IPOs, only 37% have a lockup period of exactly 180 days, and 56% have a lockup period of more than 180 days.
Interestingly, none of the 20 REIT IPOs listed before 1988 had a lockup agreement. This may not simply be a result of the timing effect. Before the Tax Reform Act of 1986 went into effect, REITs were not allowed to be self-managed; therefore, all these 20 equity REIT IPOs were externally managed. Once REITs were allowed to self-select management types, the majority became self-managed. Of those that decided remain externally managed, most featured a lockup when they went public. These externally managed REIT IPOs exhibit an average shorter lockup period than that of their self-managed counterparts. Since 1988, 27 externally managed REIT IPOs have included a lockup agreement with an average lockup period of 258 days. By contrast, internally managed REITs IPO lockup periods averaged 348 days. These observations imply a close relation between management structure and lockup length.
While the average lockup period for our REIT IPOs (325 days) is longer than those reported in the studies of industrial IPOs, it has been declining over time. Before 1997, among the REIT IPOs with a lockup, 78% (54 out of 69) had a lockup of longer than 180 days; only 14% (10 out of 69) had a lockup period of exactly 180 days. Since 1997, the lockup periods for REIT IPOs have trended toward the 180-day mark. Among the REIT IPOs with a lockup during the period from 1997 to 2006, 68% (34 out of 50) had a 180-day lockup. A possible explanation for this trend is that a learning curve existed for REIT IPOs. That is, because investors were not familiar with the REIT industry at its early development stage, REIT IPOs offered a much longer lockup period to demonstrate the commitment of the company and ensure the confidence of public investors. As the market built its credibility and more underwriters increasingly involved in the REIT IPO market, the lockup periods slid toward the 180-day norm established by the industrial IPOs. Similar learning effect was observed by Chen and Lu (2006) on underwriting
fees of REIT IPOs.
[TABLE 2 ABOUT HERE]
Table 3 shows the distribution of institutional ownership, insider ownership, management type, and property type for REIT IPOs with a lockup for the full sample. The institutional ownership is the percentage of shares owned by institutional investors in the quarter of IPO or the following quarter if ownership data are not available. We divide the sample into two groups based on the median institutional ownership and median insider ownership. A REIT is considered as having high (low) institutional ownership if its institutional ownership is greater (less) than the median of the sample. Panel B presents that institutional investors own 53% of the REITs, 71.7%
for the high institutional ownership group and 34.8% for the low institutional ownership group.
Panel C indicates that, on average, insiders own 14.72% of the REITs, 23.68% for the high insider ownership group and 5.76% for the low insider ownership group. Panel D shows that 77% of REIT IPOs with a lockup are self-managed, with an average offer size of $210 million;
the average offer size of externally managed REIT IPOs is $155 million.
As previously mentioned, a significant difference exists between self-managed and externally managed REIT IPOs in terms of lockup length. The average lockup period for self-managed REITs is 90 days longer than that for externally managed REITs. In addition, institutional investors own a higher proportion of self-managed REITs (56.9%) than they do of externally managed REITs (40.8%). According to the NAREIT, the property types for equity REITs include retail, residential, office/industrial, self-storage facilities, health care, hotel/lodging, and specialty and diversified. Panel E reports that lodging/resort REITs exhibit
smaller offer size, shorter lockup periods, and lower institutional ownership when compared to REITs with other property types.
[TABLE 3 ABOUT HERE]
III. The Determinants of Lockup Length