1 ON THE DIRECT COSTS OF REIT SEOS Sinan Gokkaya † Ohio University Matthew D. Hill †† University of Mississippi G. Wayne Kelly ††† University of Southern Mississippi First Draft: April 9, 2012 Revision Requested: May 27, 2012 Resubmitted: July 16, 2012 Keywords: REITs; SEOs; cost of equity; investment banking fees † Sinan Gokkaya is an Assistant Professor of Finance at Ohio University. Electronic mail: [email protected]. †† Contact author Matthew D. Hill is an Assistant Professor at the University of Mississippi. Electronic mail: [email protected]. ††† G. Wayne Kelly is an Associate Professor of Finance at University of Southern Mississippi. Electronic mail: [email protected].
47
Embed
ON THE DIRECT COSTS OF REIT SEOS - Johns Hopkins …pages.jh.edu/jrer/papers/pdf/forth/accepted/ON THE DIRECT COSTS O… · ON THE DIRECT COSTS OF REIT SEOS ... that SEO proceeds
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
ON THE DIRECT COSTS OF REIT SEOS
Sinan Gokkaya†
Ohio University
Matthew D. Hill††
University of Mississippi
G. Wayne Kelly†††
University of Southern Mississippi
First Draft: April 9, 2012
Revision Requested: May 27, 2012
Resubmitted: July 16, 2012
Keywords: REITs; SEOs; cost of equity; investment banking fees
†Sinan Gokkaya is an Assistant Professor of Finance at Ohio University. Electronic mail: [email protected].
†† Contact author Matthew D. Hill is an Assistant Professor at the University of Mississippi. Electronic mail:
This table presents sample characteristics of the 460 firm-commitment REIT SEOs conducted by 144 unique REITs over the 1990-2007 period.
GrossSpread is the total dollar fees paid to investment banks divided by total proceeds, where the dollar fee is the difference between the offer price
and the price at which the underwriting syndicate buys shares from the issuer. ManagementFee is the ratio of management fees to total proceeds.
UnderwritingFee is underwriting fees scaled by total proceeds. SellingConcession is selling concessions divided by total proceeds. MktCap is market
capitalization, calculated as the number of shares multiplied by share price (SharePrice) on the day prior to SEO issuance. DLowMktCap is an indicator
variable set equal to 1 if the firm is in the lowest decile of market capitalization, 0 otherwise. Rated equals 1 if the issuer has an S&P credit rating, 0
otherwise. LOC is a binary variable set equal to 1 if the issuer has a line of credit, 0 otherwise. StockVol is the standard deviation of daily stock
returnsover the 12 months prior to the SEO issue date. PriorLend equals 1 if the underwriter has previously underwritten equity offerings for the
issuer, 0 otherwise. UPREIT equals 1 if the firm uses anoperating partnership structure, 0 otherwise. PrevSEOs is the number of SEOs in the previous
year. Transparency is 1 minus the proportion of explained variation from the expanded market model regression. Post1990 is a binary variable that
takes the value of 1 if the REIT went IPO after 1990, 0 otherwise. TotProceedsis SEO issuance proceeds. Turnover is average monthly stock volume
divided by number of shares outstanding over 12 months prior to the SEO issue date. SharePrice is the firm's stock price on the day prior to SEO
issuance. Rank represents underwriter reputation and is based on Carter and Manaster’s reputation measure. MultiBook is a binary variable equal to 1
if the issue involves multiple bookrunners, 0 otherwise. Shelf is a binary variable set equal to 1 if the offer is registered under Rule 415, 0 otherwise.
Discount is the offer price discount, defined as SharePrice minus the offer price divided by SharePrice. REITProceeds is the total SEO proceeds
raised by all REITs over the 3 months prior to the SEO.
32
Table 2.Time and Property Focus Distribution of Sample of REITs
Year N GrossSpread(
%)
Management
Fee(%)
Underwriting
Fee(%)
Selling
Concession(%)
Average
TotProceeds ($M)
1990 2 5.994 1.236 1.169 3.588 97.216
1991 6 5.634 1.112 1.202 3.214 57.363
1992 9 5.450 1.029 1.049 2.990 70.836
1993 15 5.345 1.016 1.085 2.873 115.006
1994 23 5.342 1.047 1.155 2.972 117.789
1995 41 5.390 1.010 1.083 2.874 107.607
1996 52 5.002 0.967 1.004 2.674 118.835
1997 100 4.876 0.940 0.977 2.673 134.198
1998 116 3.854 0.773 0.800 2.229 62.204
1999 4 5.267 1.106 1.155 2.614 152.819
2000 5 4.372 0.849 0.849 2.547 177.113
2001 22 5.200 1.083 1.041 2.798 85.024
2002 13 4.997 0.960 0.979 2.834 104.222
2003 5 4.294 0.783 0.837 2.389 71.304
2004 14 3.875 0.754 0.733 2.229 117.593
2005 17 4.305 0.896 0.863 2.320 144.803
2006 14 3.432 0.730 0.725 1.920 192.767
2007 2 2.389 0.478 0.780 1.432 148.367
460 REIT SEOs
(144 Unique REITs) 4.647
106.564
33
Panel B: Property Focus Distribution of Sample
Property
Focus N
GrossSpread
(%)
Management
Fee(%)
Underwriting
Fee(%)
Selling
Concession(%)
Average
TotProceeds($M)
Hotel 52 4.553 0.911 0.906 2.456 128.336
Industrial 26 4.395 0.842 0.907 2.382 78.646
Office 106 4.471 0.863 0.901 2.247 145.585
Other 61 4.797 0.941 0.966 2.696 103.721
Retail 111 4.713 0.926 0.952 2.619 79.762
Residential 83 4.847 0.959 1.01 2.680 91.081
Storage 21 4.510 0.881 0.906 2.496 101.374
Total 460 REIT SEOs
Panel C: UPREIT Distribution
In an UPREIT? N Gross Spread
(%)
Management
Fee(%)
Underwriting
Fee(%)
Selling
Concession(%)
Average
TotProceeds
($M)
Yes
369
4.593
0.893
0.930
2.536
110.554
No 91 4.866 0.983 1.003 2.703 90.383
Difference in
Means (T-
statistics)
1.88*
2.38**
2.20**
1.96*
Panels A, B, and C of this table present the distribution of the sample across time, property focus, and operating structure, respectively. The sample consists of 460 firm-commitment REIT SEOs
conducted by 144 unique publicly traded equity REITs over the 1990 to 2007 period. GrossSpread is the total dollar fees paid to investment banks divided by total proceeds, where the dollar fee is the
difference between the offer price and the price at which the underwriting syndicate buys shares from the issuer. ManagementFee is the ratio of management fees to total proceeds. UnderwritingFee is
34
underwriting fees scaled by total proceeds. SellingConcession is selling concessions divided by total proceeds. TotProceeds is SEO issuance proceeds (inflation-adjusted and in millions). The 7
categories of property focus (taken from SNL) are Hotel, Industrial, Retail (Retail, Regional Mall, Shopping Center), Residential, Office, Storage, and Other (Diversified, Health Care, Manufactured
Homes, and Other). T-statistics are calculated assuming unequal variances. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively.
This table shows Pearson correlation coefficients for many of the variables used in the analysis. The sample consists of 460 firm-commitment REIT SEOs conducted by 144
unique REITs over the 1990-2007 period. GrossSpread is the total dollar fees paid to investment banks divided by total proceeds, where the dollar fee is the difference between the
offer price and the price at which the underwriting syndicate buys shares from the issuer. Ln(MktCap) is the natural logarithm of market capitalization, calculated as the number of
shares multiplied by share price (SharePrice) on the day prior to SEO issuance. DLowMktCap is an indicator variable that is set equal to 1 if the firm is in the lowest decile of market
capitalization, 0 otherwise. Rated equals 1 if the issuer has an S&P credit rating, 0 otherwise. LOC is a binary variable that is set equal to 1 if the issuer has a line of credit, 0
otherwise. Ln(StockVol) is the natural logarithm of the standard deviation of daily stock returns over the 12 months prior to the SEO issue date. PriorLend equals 1 if the
underwriter has previously underwritten equity offerings for the issuer, 0 otherwise. UPREIT equals 1 if the firm uses an operating partnership structure, 0 otherwise. PrevSEOs is
the number of SEOs in the previous year. Transparency is 1 minus the proportion of explained variation from the expanded market model regression. Post1990 is a binary variable
that takes the value of 1 if the REIT went IPO after 1990, 0 otherwise. Ln(TotProceeds) is the natural logarithm of SEO issuance proceeds. Ln(Turnover) is the natural logarithm of
the average monthly stock volume divided by number of shares outstanding over 12 months prior to the SEO issue date. Ln(SharePrice) is the natural logarithm of the firm's stock
price on the day prior to SEO issuance. Rank represents underwriter reputation and is based on Carter and Manaster’s reputation measure. MultiBook is a binary variable that
equals 1 if the issue involves multiple bookrunners, 0 otherwise. Shelf is a binary variable set equal to 1 if the offer is registered under Rule 415, 0 otherwise. Discount is the offer
price discount, defined as SharePrice minus the offer price divided by SharePrice. Ln(REITProceeds) is the natural logarithm of total SEO proceeds raised by all REITs over the 3
months prior to the SEO. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively.
This table presents pooled OLS regressions predicting REIT investment banking fees. The sample consists of 460 firm-commitment REIT SEOs conducted by 144 unique REITs
over the 1990-2007 period. GrossSpread is the total dollar fees paid to investment banks divided by total proceeds, where the dollar fee is the difference between the offer price
and the price at which the underwriting syndicate buys shares from the issuer. ManagementFee is the ratio of management fees to total proceeds. UnderwritingFee is underwriting
fees scaled by total proceeds. SellingConcession is selling concessions divided by total proceeds. Ln(MktCap) is the natural logarithm of market capitalization, calculated as the
number of shares multiplied by share price (SharePrice) on the day prior to SEO issuance. DLowMktCap is an indicator variable set equal to 1 if the firm is in the lowest decile of
market capitalization, 0 otherwise.Rated equals 1 if the issuer has an S&P credit rating, 0 otherwise.LOC is a binary variable set equal to 1 if the issuer has a line of credit, 0
otherwise. Ln(StockVol) is the natural logarithm of the standard deviation of daily stock returns over the 12 months prior to the SEO issue date. PriorLend equals 1 if the
underwriter has previously underwritten equity offerings for the issuer, 0 otherwise. UPREIT equals 1 if the firm uses an operating partnership structure, 0 otherwise. PrevSEOs is
the number of SEOs in the previous year. Transparency is 1 minus the proportion of explained variation from the expanded market model regression. Post1990 is a binary variable
that takes the value of 1 if the REIT went IPO after 1990, 0 otherwise. Ln(TotProceeds) is the natural logarithm of SEO issuance proceeds. Ln(Turnover) is the natural logarithm of
the average monthly stock volume divided by number of shares outstanding over 12 months prior to the SEO issue date. Ln(SharePrice) is the natural logarithm of the firm's stock
price on the day prior to SEO issuance.Rank represents underwriter reputation and is based on Carter and Manaster’s reputation measure. MultiBook is a binary variable that equals
1 if the issue involves multiple bookrunners, 0 otherwise. Shelf is a binary variable set equal to 1 if the offer is registered under Rule 415, 0 otherwise. Discount is the offer price
discount, defined as SharePrice minus the offer price divided by SharePrice. Ln(REITProceeds) is the natural logarithm of total SEO proceeds raised by all REITs over the 3
months prior to the SEO. Hotel, Industrial, Office, Residential, Retail, and Storage are dummy variables representing property focus classifications. Other is the reference case.
Unreported standard errors are heteroskedasticity consistent and cluster by issuer. T-statistics are in parentheses. All regression models include year dummies. For brevity
intercepts are not reported. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively.
This table presents random effects regressions predicting REIT investment banking fees.The sample consists of 460 firm-commitment REIT SEOs conducted by 144 unique REITs
over the 1990-2007 period. GrossSpread is the total dollar fees paid to investment banks divided by total proceeds, where the dollar fee is the difference between the offer price
and the price at which the underwriting syndicate buys shares from the issuer. ManagementFee is the ratio of management fees to total proceeds. UnderwritingFee is underwriting
fees scaled by total proceeds. SellingConcession is selling concessions divided by total proceeds. Ln(MktCap) is the natural logarithm of market capitalization, calculated as the
number of shares multiplied by share price (SharePrice) on the day prior to SEO issuance. DLowMktCap is an indicator variable set equal to 1 if the firm is in the lowest decile of
market capitalization, 0 otherwise. Rated equals 1 if the issuer has an S&P credit rating, 0 otherwise. LOC is a binary variable set equal to 1 if the issuer has a line of credit, 0
otherwise. Ln(StockVol) is the natural logarithm of the standard deviation of daily stock returns over the 12 months prior to the SEO issue date. PriorLend equals 1 if the
underwriter has previously underwritten equity offerings for the issuer, 0 otherwise. UPREIT equals 1 if the firm uses an operating partnership structure, 0 otherwise. PrevSEOs is
the number of SEOs in the previous year. Transparency is 1 minus the proportion of explained variation from the expanded market model regression. Post1990 is a binary variable
that takes the value of 1 if the REIT went IPO after 1990, 0 otherwise. Ln(TotProceeds) is the natural logarithm of SEO issuance proceeds. Ln(Turnover) is the natural logarithm of
the average monthly stock volume divided by number of shares outstanding over 12 months prior to the SEO issue date. Ln(SharePrice) is the natural logarithm of the firm's stock
price on the day prior to SEO issuance. Rank represents underwriter reputation and is based on Carter and Manaster’s reputation measure. MultiBook is a binary variable that
equals 1 if the issue involves multiple bookrunners, 0 otherwise. Shelf is a binary variable set equal to 1 if the offer is registered under Rule 415, 0 otherwise. Discount is the offer
price discount, defined as SharePrice minus the offer price divided by SharePrice. Ln(REITProceeds) is the natural logarithm of total SEO proceeds raised by all REITs over the 3
months prior to the SEO. Hotel, Industrial, Office, Residential, Retail, and Storage are dummy variables representing property focus classifications. Other is the reference case.
Unreported standard errors are heteroskedasticity consistent and cluster by issuer. T-statistics are in parentheses. All regression models include year dummies. For brevity
intercepts are not reported. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively.
40
Table 6.Univariate Evidence on Differences in Direct SEO Costs: REITs versus non-REITs
Firm Type N Gross Spread (%) Management
Fee(%)
Underwriting Fee(%) Selling
Concession(%)
Non-REIT 460 5.219 1.076 1.103 3.040
REIT 460 4.647 0.911 0.944 2.569
Difference in
Means (T-
statistics)
6.87***
8.37***
8.42***
10.08***
This table compares the mean gross spreads of REIT SEOs with Industrial SEOs, where industrial SEOs are matched based on issue date and proceeds. GrossSpread is the total
dollar fees paid to investment banks divided by total proceeds, where the dollar fee is the difference between the offer price and the price at which the underwriting syndicate buys
shares from the issuer. ManagementFee is the ratio of management fees to total proceeds. UnderwritingFee is underwriting fees scaled by total proceeds. SellingConcession is
selling concessions divided by total proceeds. T-statistics are calculated assuming unequal variances. *** denotes statistical significance at the 1% level.
41
Table 7 Multivariate Evidence (OLS) on Differences in Direct SEO Costs: REITs versus non-REITs
This table presents pooled OLS regressions testing for differences in investment banking fees across REIT and non-REIT SEOs. Industrial SEOs are matched based on issue date
and proceeds. Note that the sample is not the same as in Table 1 because matching non-REITs are included here. GrossSpread is the total dollar fees paid to investment banks
divided by total proceeds, where the dollar fee is the difference between the offer price and the price at which the underwriting syndicate buys shares from the issuer.
ManagementFee is the ratio of management fees to total proceeds. UnderwritingFee is underwriting fees scaled by total proceeds. SellingConcession is selling concessions divided
by total proceeds. Ln(MktCap) is the natural logarithm of market capitalization, calculated as the number of shares multiplied by share price (SharePrice) on the day prior to SEO
issuance. DLowMktCap is an indicator variable set equal to 1 if the firm is in the lowest decile of market capitalization, 0 otherwise. Rated equals 1 if the issuer has an S&P credit
rating, 0 otherwise. Ln(StockVol) is the natural logarithm of the standard deviation of daily stock returns over the 12 months prior to the SEO issue date. PriorLend equals 1 if the
underwriter has previously underwritten equity offerings for the issuer, 0 otherwise. PrevSEOs is the number of SEOs in the previous year. Transparency is 1 minus the proportion
of explained variation from the expanded market model regression. Ln(TotProceeds) is the natural logarithm of SEO issuance proceeds. Ln(Turnover) is the natural logarithm of
the average monthly stock volume divided by number of shares outstanding over 12 months prior to the SEO issue date. Ln(SharePrice) is the natural logarithm of the firm's stock
price on the day prior to SEO issuance.Rank represents underwriter reputation and is based on Carter and Manaster’s reputation measure. MultiBook is a binary variable that equals
1 if the issue involves multiple bookrunners, 0 otherwise. Shelf is a binary variable set equal to 1 if the offer is registered under Rule 415, 0 otherwise.Discount is the offer price
discount, defined as SharePrice minus the offer price divided by SharePrice. Ln(MktProceeds) is the total SEO proceeds raised by all REITs and Industrials over the 3 months
prior to the SEO in question. REIT is an indicator variable that equals 1 if the observation is a REIT, 0 otherwise. Unreported standard errors are heteroskedasticity consistent and
clustered by issuer. T-statistics are in parentheses. All regression models include year dummies. For brevity intercepts are not reported. ***, **, and * denote statistical significance
at the 1%, 5%, and 10% levels, respectively.
43
Table 8.Multivariate Evidence (Random Effects) on Differences in Direct SEO Costs: REITs versus non-REITs
This table presents random effects regressions testing for differences in investment banking fees across REIT and non-REIT SEOs. Industrial SEOs are matched based on issue
date and proceeds. GrossSpread is the total dollar fees paid to investment banks divided by total proceeds, where the dollar fee is the difference between the offer price and the
price at which the underwriting syndicate buys shares from the issuer. ManagementFee is the ratio of management fees to total proceeds. UnderwritingFee is underwriting fees
scaled by total proceeds. SellingConcession is selling concessions divided by total proceeds. Ln(MktCap) is the natural logarithm of market capitalization, calculated as the number
of shares multiplied by share price (SharePrice) on the day prior to SEO issuance. DLowMktCap is an indicator variable that is set equal to 1 if the firm is in the lowest decile of market
capitalization, 0 otherwise. Rated equals 1 if the issuer has an S&P credit rating, 0 otherwise. Ln(StockVol) is the natural logarithm of the standard deviation of daily stock returns
over the 12 months prior to the SEO issue date. PriorLend equals 1 if the underwriter has previously underwritten equity offerings for the issuer, 0 otherwise. PrevSEOs is the
number of SEOs in the previous year. Transparency is 1 minus the proportion of explained variation from the expanded market model regression. Ln(TotProceeds) is the natural
logarithm of SEO issuance proceeds.Ln(Turnover) is the natural logarithm of the average monthly stock volume divided by number of shares outstanding over 12 months prior to
the SEO issue date. Ln(SharePrice) is the natural logarithm of the firm's stock price on the day prior to SEO issuance.Rank represents underwriter reputation and is based on Carter
and Manaster’s reputation measure. MultiBook is a binary variable that equals 1 if the issue involves multiple bookrunners, 0 otherwise. Shelf is a binary variable set equal to 1 if
the offer is registered under Rule 415, 0 otherwise. Discount is the offer price discount, defined as SharePrice minus the offer price divided by SharePrice. Ln(MktProceeds) is the
total SEO proceeds raised by all REITs and Industrials over the 3 months prior to the SEO in question. REIT is an indicator variable that equals 1 if the observation is a REIT, 0
otherwise. Unreported standard errors are heteroskedasticity consistent and clustered by issuer. T-statistics are in parentheses. All regression models include year dummies. For
brevity intercepts are not reported. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively.
45
Endnotes
1Prior to 2001, the dividend payout mandate was 95% of taxable income.
2 Despite the large payout requirement, most REITs choose to pay additional discretionary dividends, further reducing their ability to accumulate
capital internally (Wang, Erickson, and Gau (1993)). The minimum dividend is not binding due to depreciation expense, which coupled with capital
gains on property sales, allows REITs to pay dividends in excess of taxable income, as shown by Bradley, Capozza, and Seguin (1998), Ghosh and
Sirmans (2006), Feng, Ghosh, and Sirmans (2007b), and Hardin and Hill (2008). Although, REITs have some discretion when choosing their degree
of financial constraint, the mandatory dividend precludes these firms from pursuing meaningful growth given the capital intensity of the industry.
3Recent research by Riddiough and Wu (2009), Hardin and Wu (2010), and Hardin and Hill (2011) examine the interaction between short-term debt
offerings and long-term sources of financing. Francis, Lys and Vincent (2004) show that REITs access capital markets more frequently than
industrial firms.
4 Reduced agency problems are attributable to the industry's high payouts to shareholders, while the latter point results from asset and income
restrictions that limit REITs to invest primarily in real estate.
5 In addition, REITs may provide an improved identification strategy by mitigating the potential endogenous nature of adverse selection problems.
Previous studies generally conclude that information asymmetry drives a substantial portion of SEO gross spreads but asymmetric information is not
directly observable, hence proxies vary across studies. Furthermore, these proxies may not be exogenous from omitted variables nor cross-industry
shocks (Lee and Masulis (2009). This endogeneity problem should be most severe in industries investing heavily in intangible assets and for firms
with greater free cash flow where both characteristics aggravate the adverse selection problem clouding causal inferences with respect to existing
gross spread determinants. Structural differences between equity REITs and non-REITs may attenuate the endogeneity concern. Later in the text, it is
noted that the relative REIT transparency argument is unresolved.
6Hardin and Hill (2011) examine the determinants of REIT line of credit rating and use and find that most publicly-traded equity REITs have access
to credit lines.
7Chou, Hardin, Hill, and Kelly (2011) find evidence of market value implications of the partnership structure for REITs. The market values of
discretionary dividends paid by REITs using the partnership structure, relative to non-OP REITs.
8The expanded market model is defined as: r i,k,t= α i,t + β1* r m,t + β2* r j,t+ β3* rk,t+ ε i,t,
where r i,k,t is the monthly return of REIT I in property k in year t, r m,t is the monthly CRSP value-weighted market return in year t, r j,t is the
monthly value-weighted return of the equity REIT’s industry in year t, and rk,t is the monthly value-weighted return of the equity REITs in property
k in year t.
46
9Only 51 unique lead investment banks underwrote REIT SEOs between 1990 and 2007.
10
Consistent with this, Choe, Masulis and Nanda (1993) find that price reactions to industrial equity offerings are lower during high financing activity
in the market. Furthermore, Goodwin (2011) finds a negative relation between underpricing of REIT SEOs and primary market activity.
11
Discounting measures the difference between closing price on the day prior to the SEO issue date and offer price. On the other hand, underpricing
is defined as the difference between the closing price on the issue date and offer price. Discounting and underpricing are analogous, and represent the
indirect costs associated with SEOs (Altinkilic and Hansen (2003).
12
Lease, Masulis, and Page (1991) shows that stated offer dates are often not accurate since most offers actually take place after the trading is closed.
Following Safieddine and Wilhelm (1996), corrections to the offer dates provided in Securities Data Corporation (SDC) New Issues Database are as
follows. If the trading volume on day following the reported offer date is (1) more than twice the trading volume on the offer data on SDC and (2)
more than twice the average daily trading volume over the previous 250 trading days, then the trading day after the reported offer day on SDC is
selected as the corrected offer date. This methodology results in offer date corrections for 54% of the sample.
13
Goodwin (2011) argues that discounting is a more appropriate measure of the indirect cost of issuance when analyzing SEOs. Thus, discounting is
used as a proxy to capture indirect cost of SEOs. However, the results are qualitatively similar when the models are run again with the underpricing
variable taking the place of the discounting. An anonymous referee provided this point.
14
SEOs are typically underwritten by syndicates of managing, underwriting, and selling groups. The managing group is responsible for structuring
the syndicate, and typically receives 20% of gross spread for its services. Along with the managing group, underwriting group make an underwriting
commitment for an agreed upon number of shares, and conduct the bookbuilding process. For these services, underwriters also generally receive 20%
of the gross spread. Finally, managing underwriters also assemble a selling group, and typically gets 60% of the gross spread for selling the shares to
retail as well as institutional customers. While 20/20/60 division is perceived as widely standard in the industrial literature, Torstilla (2001) finds
significant deviations from 20/20/60 split for industrial IPOs. Similarly, untabulated results show that only 103 out of 460 REIT SEOs provide a
20/20/60 split.
15
The specific source used here is REIT Watch, NAREIT’s monthly statistical report that lists firm name, ticker, investment focus, and property
focus for equity, mortgage, and hybrid REITs comprising their REIT return indices.
16
In a firm commitment contract, a syndicate of investment banks guarantees to buy an issuer’s equity offering at an offer price discount. The
discount is compensation for bearing the price risk associated with reselling the shares to the public following the agreement. This compensation or
discount is called underwriter gross spread, and represents a substantial portion of total flotation costs (Lee and Masulis (2009)).
47
17
It should be noted that SNL does not provide historical data on REITs' use of the UPREIT structure, as SNL backfills the UPREIT field. Thus,
when one pulls the UPREIT variable for a time-series, the researcher will note that the variable exhibits no within-REIT variation. An anonymous
reviewer indicated this to the authors.
18
Goodwin (2011) shows a significantly lower level of discounting for REIT SEOs during periods of high financing activity in the overall market.
19
Data on the breakdown of investment bank fees is also retrieved from the Securities Data Corporation (SDC) New Issues Database.
20
Because of the tradeoff between bias and efficiency, emphasis is given to the result from the full model. That is, since the primary concern is
omitted variables bias, relative to less efficient standard errors, results shown across columns 1 and 4 receive emphasis. This approach parallels
Chan, Stohs and Wang (2001) find no statistical difference between REITs and non-REITs for a sample of REIT IPOs listed in Hong Kong Stock
Exchange.
22
Results are qualitatively and quantitatively similar when excluding industrial SEOs from the computation of total SEO proceeds.
23
Specifically, Transparency for industrials is defined as ri,,t= α i,t + β1* r m,t+ ε i,t where r i,k,t is the monthly return of the non-REIT issuer in year t,
and rm,t is the monthly CRSP value-weighted market return.
24
This suggestion was provided by an anonymous reviewer.