Using REIT Follow-on Equity Offerings to Pay down Credit Line Balances: Bank Certification or Monitored Financial Flexibility?

AbstractUsing hand collected data from offering prospectuses and other corporate filings, we examine the market response to real estate investment trust (REIT) follow-on stock offerings’ stated uses of proceeds. We also track REIT banking relationships over time. Consistent with the idea of bank certification, we show that markets react relatively favorably to REIT equity offers where issuers have lending relationships with affiliates of the underwriters. However, we also find that reactions are most favorable where REIT issuers intend to repay their bank’s line of credit, regardless of the bank’s affiliation with the underwriters. This pattern is particularly strong among smaller firms with lower institutional ownership. We posit that credit line repayments preserve benefits of bank monitoring while enhancing financial flexibility. Further examination reveals that this monitored financial flexibility is the dominant effect.Keywords: REITsequity offeringscredit linesbanking relationshipscertificationmonitored financial flexibility AcknowledgmentsWe thank Nevin Boparai, Paul Brockman, John Cobb, Sandeep Dahiya, Chitru Fernando, Ioannis Floros, Kathleen Weiss Hanley, Bill Hardin, David Harrison, Masaki Mori, Christo Pirinsky, Victoria Rostow, Calvin Schnure, Paul Schultz, Qinghai Wang, Ke Yang, two anonymous reviewers, and participants at the American Real Estate Society Conference and European Real Estate Society Conference for helpful discussions and comments. Natalya Bikmetova, Debanjana Dey, Xin Fang, and Sulei Han provided invaluable research assistance. We remain responsible for any errors. Gatchev and Singh are grateful for the financial support provided by the SunTrust Endowment; Nayar appreciates support from the Hans Julius Bär Endowed Chair; Price acknowledges support from the Collins-Goodman Endowed Chair.Disclosure StatementNo potential conflict of interest was reported by the author(s).Notes1 Indeed, the increase of REITs’ commercial bank borrowings, such as lines of credit and term loans, in recent years has attracted the attention of investors, credit rating agencies, and the press. See, for example, https://www.wealthmanagement.com/reits/are-reits-maxing-out-bank-borrowing.2 Consistent with Puri (Citation1996), we also use the term investment houses interchangeably with investment bankers (or underwriters) to distinguish them from pure commercial banks. To denote the dual underwriting and commercial banking relations, we use the term “underwriting-banking relations” through the rest of the paper.3 The general conclusion across prior studies is that the certification benefits, net of any conflicts of interest costs, stem from information advantages gained through the lending channeland are most, and sometimes only, evident for junior, information sensitive securities. For example, Duarte-Silva (Citation2010) finds that announcement returns to equity offers are less negative when underwriters also have prior lending relationships with the issuer. See also Ang and Richardson (Citation1994), Kroszner and Rajan (Citation1994), Puri (Citation1996), Gande, Puri, Saunders, and Walter (Citation1997), Schenone (Citation2004), and Drucker and Puri (Citation2005). Accordingly, we focus our study exclusively on REIT equity offerings where we expect the certification effect to be most prominent.4 Hardin and Wu (Citation2010) and Chang et al. (Citation2021) examine banking relationships in the context of REIT debt issuance. Although they do not focus on it, Hardin and Wu (Citation2010) acknowledge certification as a possibility.5 The borrower pays a commitment fee on the untapped amount of the credit line.6 Hardin and Hill (Citation2011) find that more than 95% of REITs have credit facility access. Ott et al. (Citation2005) report that only 7 percent of REIT investment is financed with retained earnings, compared to 70 percent of industrial firm investment being funded from retained earnings as reported in Fama and French (Citation1999). Ooi et al. (Citation2012) indicate that lines of credit represent 73.8% of total liquidity available to REITs.7 Myers and Majluf (Citation1984) argue that adverse selection costs are most severe for equity issues, compared to issuance of other securities.8 The certification hypothesis does not make any specific prediction related to offerings where the proceeds are used to repay credit facilities owed to the investment house.9 See Letdin et al. (Citation2019) for a recent overview of how leverage affects REIT returns.10 Further, there is a rich literature that discusses the importance of banks as delegated monitors, e.g., Diamond (Citation1984), Fama (Citation1985), Rajan (Citation1992), James (Citation1987), and Petersen and Rajan (Citation1994, Citation1995).11 Also, see Acharya et al. (Citation2020) on the impact of violations of covenants on lines of credit.12 See Appendix A for a tabulation of follow-on equity offerings. The average firm conducts roughly four offerings covering the period from 07/25/1996 to 12/12/2018.13 Although the Glass-Steagall Act was fully dismantled in 1999, the Federal Reserve had relaxed the rules even by 1996 such that bank holding companies were permitted to earn up to 25% of their revenues from “ineligible” investment banking activity through their Section 20 subsidiaries (Rodelli, Citation1998).14 Firms do not receive any proceeds from the sale of secondary shares.15 We do not separately tabulate boilerplate terms such as the stated uses of proceeds for “possible future acquisitions” or “investment in securities”. The term “possible future acquisitions” is typically accompanied by “general corporate purposes”. Moreover, REITs frequently employ the following generic terminology: “until such uses, we will invest in securities consistent with our REIT status.” These terms essentially contain little additional information.16 For around 80% of the SEOs in our sample, the announcement and the pricing dates are on the same day or a day apart. We verify that our findings are not sensitive to controls for differences between the pricing and the announcement dates.17 We compare the last reported trading price with data on the daily CRSP files and correct the few stale quotes in the final prospectus.18 We examine alternate measures of whether the firm employs the services of new banks, including measures based on the set of book-runners, and also based on the entire underwriting syndicate. Our main findings are robust to alternate measures of whether the firm establishes new underwriting relations.19 We compute this measure using data on all IPOs and follow-on SEOs by our sample of 89 equity REITs. For 5 of the 379 SEOs, the underwriter does not participate in any offerings in the past three years. In these cases, we use this underwriter’s sample average over all years to compute the underwriter relative activity and dominance measures. Our findings remain similar if instead we use reputations of 0 for these underwriters or exclude these offerings altogether.20 https://site.warrington.ufl.edu/ritter/ipo-data/.21 https://fred.stlouisfed.org/series/GDPDEF.22 Corwin (Citation2003) reports that the relative offer size, defined as offered shares divided by pre-issue shares outstanding, averages around 16.0 percent on the NYSE relative to 26.8 percent on Nasdaq.23 Feng et al. (Citation2007) examine the propensity of REITs to finance growth opportunities with debt.24 Our REIT estimate is less negative compared to the -2.7 percent reported for general firms in Duarte-Silva (Citation2010).25 Howe and Shilling (1988) report -1.897 percent mean adjusted returns for days (-1,0), while Ghosh et al. (Citation1999) report -1.05 percent average standardized abnormal returns for days (0,1).26 We examine alternate measures of whether the firm employs the services of new banks, including measures based on the set of book-runners, and also based on the entire underwriting syndicate. Our main findings are robust to alternate measures of whether the firm establishes new underwriting relations.27 In Table 3, the point estimate measuring the relative increase in stock return to the joint underwriting-banking relationship ranges from 0.69% to 0.88%, which is higher than the point estimate reported by Duarte-Silva (Citation2010) of 0.39%.28 All multivariate regressions include the residual discount as an explanatory variable. The residual discount is estimated based on each model’s explanatory variables as shown in Appendix C.29 Even when the offering’s explicitly stated use of proceeds is to refinance existing underwriter-affiliated bank debt, Gande et al. (Citation1997) do not find a significant difference between yield spreads on similar debt issues underwritten by banks and investment houses.30 Figure 2 provides a categorization of the sample based on the status of (i) credit facility repayment, and (ii) the underwriting-banking relationship.31 Residual discount estimation regressions are shown in Appendix C.32 It could be argued that reducing outstanding balances on lines of credit reduces firm indebtedness and thus bankruptcy costs, which in turn, could cause the favorable market reaction to those offerings. We contend that the reduction in bankruptcy costs is not likely to explain our results. We justify this view by pointing to the negative coefficient on the indicator variable for repayment of debt other than credit facility repayment in models (3) and (4) of Table 6, Panel A. If reduction in bankruptcy costs was the driver of our results, this coefficient would have been positive.33 For brevity, we do not tabulate the estimates from this specification. All estimates are available upon request.34 Given the discussion above on the relevance of the intertwined relation between lines of credit and underwriting activities, we explore how underwriting by investment houses affects future lending relationships. The findings, presented in Appendix D, provide further support for the complementarity of the two functions.35 Additionally, diversifying funding sources, along with additional monitoring, should reduce REIT funding risk.36 https://www.directedgar.com/