Ownership structure and the value of the firm: The case of REITs
Friday, H Swint
Abstract. This article examines the relation between ownership structure and firm value as proxied by market-to-book ratios for real estate investment trusts (REITs) over the period 1980 to 1994. Piecewise regression analysis reveals a nonlinear relationship between REIT market-to-book ratios and ownership structure. Low levels of inside ownership are associated with increased market-to-book ratios for equity REITs. However, as inside block ownership rises above 5%, equity REIT market-to-book ratios decline. The opposite result is true for hybrid and mortgage REITs. Similarly, higher levels of outside blockholdings have a negative impact on both equity and hybrid and mortgage market-to-book ratios.
Through the enactment of the Omnibus Budget Reconciliation Act of 1993 (OBRA, 1993), Congress relaxed the `five or fewer’ ownership constraint imposed on real estate investment trusts (REITs). This rule from the Internal Revenue Code (Sections 856-858) states that at least 100 shareholders must exist with no more than 50% of the shares outstanding held by five or fewer shareholders. A primary purpose of this section of the Internal Revenue Code for REITs was to restrict the ability of a few large block holders from using their ownership power to expropriate wealth from small outside shareholders. The intent of relaxing this constraint was to increase the presence of institutional ownership in the securitized real estate markets. Downs (1998) reported a significant positive reaction for REITs with low block ownership to the first announcement of the possible inclusion of this provision in OBRA 1993. Prior to the passage of this provision, no comprehensive academic research existed on the impact of increased block holder ownership on overall shareholder wealth for REITs. However, previous ownership research from the finance literature calls into question the wisdom of allowing a few large block holders to accumulate a large portion of outstanding REIT shares.
Widely dispersed small shareholders with limited resources may find it difficult to actively monitor a firm’s management, while, on the other hand, large outside block holders and institutional shareholders possess the resources and are better positioned to perform this task. However, it is not clear that these large investors necessarily act in the best interest of small shareholders. Barclay, Holderness and Pontiff (1993) document a positive relationship between large block holder ownership and discounts for closed-end mutual funds.’ They find that the average discount for funds with large block holders is 14% versus 4% for those funds without large block holders. The authors attribute their findings to agency costs associated with the ability of these large shareholders to receive private benefits at the expense of less well-positioned small outside shareholders.
In addition, Morck, Shleifer and Vishny (1988) report that Tobin’s Q’s tend to decline for firms as inside ownership increased from 5% to 25%. However, firm Q’s start rising again as holdings rise above 25%. The authors contend that entrenchment related agency costs explain the declining Q’s for inside ownership between five and 25%. Outside shareholders recognize these agency costs and demand higher returns as compensation for the increased costs. These higher returns manifest themselves as increased discounts. However, the authors argue that once total inside ownership levels surpass the 25% mark, additional entrenchment related agency costs are probably minimal.
The dearth of research addressing the influence of REIT ownership structure on shareholder wealth following the relaxation of the `five or fewer’ rule necessitates such a study. The implications of this legislation may have a far-reaching impact on shareholder wealth as large block holders start accumulating larger positions in REIT shares. The impact of REIT ownership structure on shareholder wealth is examined using methodology similar to that employed by Barclay, Holderness and Pontiff (1993) for closed-end investment companies and Morck, Shleifer and Vishny (1988) for a broad cross-section of firms.
This analysis provides insight into several conflicting hypotheses addressing the impact of firm ownership structure on firm value. Pound (1988) presents three hypotheses addressing the relationship between institutional ownership and firm value. Pound’s efficient monitoring hypothesis predicts that large institutional investors are superior monitors due to their greater expertise and lower marginal costs. However, his conflict of interest hypothesis contends that institutional shareholders such as banks and insurance companies sacrifice their monitoring roles for possible business relationships with the firm or management. Pound’s strategic alignment hypothesis argues that large institutional investors may find it in their best interest to cooperate with management to gain superior private information at the expense of smaller, less well positioned outside shareholders. These hypotheses can be expanded to include non-institutional large outside block holders.
There are two opposing hypotheses concerning the relationship between inside ownership and firm value. The convergence of interest hypothesis contends that as inside ownership increases, insiders bear a larger proportion of the costs associated with non-value maximizing behavior.2 Therefore, there exists a direct relationship between inside ownership and firm value. The entrenchment hypothesis posits an indirect relationship between inside ownership and firm value. Inside perquisite consumption increases as insiders organizational control increases and they become more difficult to replace.3
REIT Institutional Structure
REITs are essentially closed-end investment companies that provide a passive medium for investors to invest in income producing real estate properties and mortgages.4 The REIT organizational form was created through legislation in 1960 providing tax exemptions for qualified REITs. Like stock and bond closed-end funds, REITs are a passive conduit for the income they earn. According to Internal Revenue Code, to be a qualified REIT over the period 1980 to 1994, a REIT had to meet the following criteria:
1. Seventy-five percent of all assets must consist of mortgages, real estate equities or government securities
2. At least 95% of taxable income must be distributed to REIT shareholders each year; 3. At least 75% of gross income must be derived from rents, mortgages and gains from the sale of real estate;
4. Real property must not be held primarily for sale in the ordinary course of business;
5. The number of beneficial owners must typically exceed one hundred, with the five largest shareholders owning no more than 50% of the outstanding shares;
6. The trustees, directors and employees of a REIT are restricted from actively managing or operating REIT property. However, they are permitted to make property decisions related to the business of the REIT.
The National Association of Real Estate Investment Trusts (NAREIT) classifies REITs as equity, hybrid or mortgage REITs according to their respective holdings of real property or mortgage instruments. To be considered an equity REIT, at least 75% of the REIT’s investment portfolio must consist of income-producing real property. Mortgage REITs lie on the opposite end of the spectrum with at least 75% of their assets consisting of mortgage instruments, while hybrid REITs fall between equity and mortgage REITs.
The relationship between ownership structure and firm value is evaluated by determining the impact of ownership structure on REIT market-to-book ratios. For this analysis, an initial sample of REITs is obtained from the NAREIT REIT Factbooks for the years 1980 through 1994. REITs with financial data available on the Compustat AFC, AIE and AIR files, CRSP files and the Q-Data Corporation, and EDGAR SEC files are retained. The resulting group of firms is then segregated into equity, mortgage or hybrid REIT categories according to their NAREIT classification.
The need to attain comprehensive ownership information from publicly available SEC Proxy filings and key financial statistics from the CRSP and Compustat databases eliminated over one-half of the NAREIT reported population. In addition, the data collection process occurred in the 1994 and 1995 time period, limiting the number of observations available for the 1993 and 1994 period.5 The low sample sizes relative to the REIT population during these periods clearly under represents the impact of the newly minted REITs that began trading during this period.6
Once this screening process was complete, sufficient data was available from the different sources for at least one year during the period of interest for 399 equity REITs and 276 mortgage and hybrid REITs.7 Exhibit 1 contains summary statistics for the numbers of REITs examined each year, the average annual market-to-book ratios and the total block holdings for the different REIT types. Because of a smaller number of observations, the hybrid and mortgage REITs are grouped together, while the equity REITs are examined independently. Equity REIT market-to-book ratios ranged from 0.84 to 1.48, while hybrid and mortgage market-to-book ratios ranged from 0.72 to 1.32. Overall, average block ownership for the sample period ranged from 20.2% for hybrid and mortgage REITs to 23.8% for equity REITs.
A REITs market-to-book ratio is defined as the market value of the REIT’s securities divided by the replacement value of the REIT’s assets.8 That is:
Market/Book = Market Value (Equity + Debt + Preferred Stock)/Replacement Value (Real Estate + Mortgages + Other Assets)
The market-to-book ratio is a widely used measure of firm performance and value that provides insight into the perceived value of the firm’s intangible assets such as goodwill, monopoly power, growth opportunities and management competence. A high ratio for a REIT implies that the firm has some intangible asset such as a good management team or governance structure.9
Ownership of officers and directors and owners of 5% or more of a REIT’s outstanding common stock are obtained from REIT annual proxy statements. Inside block holders are classified either as insiders that are officers and directors or insiders with other affiliations. Outside block holders are categorized either as outside directors that are block holders or other outsiders. Institutional holdings are classified into two categories. The first category contains insurance companies and lending institutions, and the second category contains money management institutions such as mutual funds and pension funds.
The equity and hybrid-mortgage REIT samples are partitioned into three equally sized groups according to their relative market-to-book ratios. The first group consists of REITs with high relative market-to-book ratios. The second group contains REITs with average market-to-book ratios, while the third group contains REITs with low market-to-book ratios.
Financial statistics for the high, average and low market-to-book groups of the equity and hybrid-mortgage samples are contained in Exhibit 2. The results for the equity REITs contained in Panel A show that high market-to-book equity REITs trade on average at a 69% premium, while equity REITs with average market-to-book ratios trade close to par. However, the equity REITs with low market-to-book ratios trade at a 34% discount. The average asset size of over $178 million for equity REITs with high market-to-book ratios is significantly larger than the $95 million average for equity REITs with low ratios. In addition, the average debt-to-asset ratio at 37% for the high market-to-book ratio group is significantly larger than the 30% average for the low group. These results reveal that increased debt levels are positively associated with increased market-to-book ratios providing evidence that increased REIT debt levels may serve to limit agency costs. Average return on asset levels are also significantly higher for equity REITs with high market-to-book ratios. Despite increased performance levels, equity REITs with high ratios have betas similar to the other groups.
In Exhibit 2, Panel B contains the results for the hybrid and mortgage REITs with high, average and low market-to-book ratios. The hybrid-mortgage REIT group with high market-to-book ratios trade on average at a 28% premium while the average group trades at a 10% discount. The hybrid-mortgage REIT group with low marketto-book ratios trades at a 41% discount from firm asset value. No significant differences are observed between shareholder numbers, asset values and debt-to-asset ratios for the high and low market-to-book hybrid and mortgage REIT groups. The return on assets for the high ratio group is significantly larger than that of the low ratio group, while their average beta is significantly lower. Interestingly, the middle market-to-book group is significantly higher than the higher market-to-book group in every category except for return-on-assets, where it is significantly lower.
Summary statistics and difference in means tests for ownership structure characteristics are provided in Exhibits 3, Panels A and B for the equity and hybridmortgage REIT groups. Levels of total block ownership, inside director and officer block ownership, and pension and money manager block ownership are significantly higher for the low market-to-book equity REIT group. This result parallels findings by Barclay, Holderness and Pontiff (1993) that higher discounts are associated with increased block holder ownership for bond and stock closed-end mutual funds. However, outside banking and insurance block holdings are significantly lower for this group.
Similar differences are not observed for the hybrid-mortgage REIT group. The high market-to-book group has higher levels of outside director block holdings than the low market-to-book group, while having lower levels of non-director outsider block holdings. Interestingly, the middle market-to-book group has lower levels of block ownership than do its counterparts on the extremes.
To further examine the relationship between REIT market-to-book ratios and ownership structure, a piecewise analysis similar to that employed by Morck, Shleifer and Vishny (1988) is used. Ownership structure is measured as the percentage of outstanding shares held by block holders.” The knots or breakpoints of 0% to 5% ownership, 5% to 25% ownership and >25% ownership employed by Morck et al. are also employed in this analysis. The 5% threshold is the point that the SEC considers germane by requiring mandatory public reporting at and beyond this ownership level. The 25% breakpoint employed by Morck et al. was motivated by Weston (1979). Weston argued that beyond this ownership level, a hostile bid for the firm would fail.” In addition, a time-trend variable is also included in the analysis to account for the systematic variation in REIT market-to-book ratios across the sample period.
Exhibit 4 provides results for a piecewise analysis of the relation between total block ownership and REIT market-to-book ratios obtained from the following crosssectional model:lz
Interpretation of the coefficients follows that of a standard piecewise regression with knots at 5% and 25% of total block holder ownership.13 For cases where total block holder ownership is less than 5% of outstanding shares, the coefficient estimate for the variable B-OWN shows the relation between block holder ownership of less than 5% and pricing.14 The direction and magnitude of the impact of block holder ownership between 5% and 25% of outstanding shares is determined by summing the coefficient estimates for B-OWN and B-OWN>5%. For block holder ownership greater than 25%, the magnitude and direction of the relation to M/B is measured by summing all three coefficients for B-OWN, B-OWN>5% and B-OWN>25%.
No significant relationship is observed between MIB and total block holdings for equity REITs. However, a negative coefficient is observed for B-OWN for the hybridmortgage REIT sample. This result reveals that a significant negative relationship exists between market-to-book ratios and the absence of block ownership for the hybrid-mortgage REIT group. A positive relationship is observed between MIB and ownership levels between 5% and 25%. This result indicates that as block ownership increases from 5% to 25%, REIT MIB ratios also rise. The coefficient B-OWN>25% is negative and the sum of all three coefficients is negative indicating a negative relationship exists for block ownership levels greater than 25%. This result provides evidence that the market discounts levels of block ownership as they rise above 25%.
Further analysis is performed to determine the impact of inside and outside block holders individually on REIT discounts. Exhibit 5 contains analysis of the relationship between M/B and inside director and officer blockholdings obtained from the following model:
Interpretation of this model follows that of the piecewise regression analysis previously discussed.
A significant positive relationship is observed between MIB and no inside director and officer block holdings. However, a negative relationship is observed for all inside director and officer block holdings in excess of both the 5% and 25% levels for equity REITs. The positive coefficient on B-IDO>25% indicates that a negative relationship exists because the sum of the coefficients is still negative but the rate of decline in MIB as ownership rises above 25% slows. A significant negative relationship is also observed between M/B and inside director and officer block holdings in excess of 25% for the hybrid-mortgage REIT group. This result provides support for the entrenchment hypothesis of inside director and officer block holdings. No entrenchment exists when there are no inside director and officer block holdings. As block holdings rise above 5% insider entrenchment becomes problematic. However, for equity REITs, once the 25% threshold is surpassed the negative impact of additional inside director and officer blockholdings diminishes.
Exhibit 6 contains analysis of the relation between MIB and total inside block ownership obtained from the model:
Total inside block holdings are composed of inside director and officer block holdings as well as any other block holder such as a sponsor, advisor or affiliate not represented on the board that also holds large portions of the REITs outstanding shares. The results for this analysis correspond to those reported for inside director and officer block holdings. However, they are not as strong, indicating that the impact of other insiders on M/B is minimal. Though these block holders are considered insiders because of their affiliation with the REIT, their lack of a board seat indicates reduced authority over firm operations and thus a reduced capacity to expropriate wealth from outside shareholders.
Analysis reported in Exhibit 7 shows the relationship between MIB and noninstitutional outside block holdings using the following model:
Total non-institutional outside block holdings consists of all outside director block holdings and other outside block holdings that are not classified as a bank, insurance firm or pension or money management fund.
Results for this analysis show a positive relationship between MIB and noninstitutional outside block holdings less than 5% and non-institutional block holdings in excess of 25%, while indicating a negative relation for block holdings between 5% and 25% for both the equity and hybrid-mortgage REIT groups. However, the small positive coefficient on B-NIOUT>25% for the equity REIT group indicates that the positive impact on M/B of non-institutional outside ownership as it rises above 25% is minimal.
Exhibit 8 provides analysis for the relationship between MIB and total outside block ownership for the equity and hybrid-mortgage REIT groups from the following model:
Total outside block holdings consist of all outside director block holdings, non-director outsider block holdings, banking and insurance block holdings, and pension and money manager block holdings. The results in Exhibit 8 show that a positive relationship is found for outside block holdings up to 5% for both the equity and hybrid-mortgage REIT groups. However, this relationship becomes negative for block holdings between 5% and 25% of outstanding shares. No significant impact on M/B is observed for outside block holdings in excess of 25% for either REIT group. These results provide no support for the monitoring benefits often associated with large outside block holders.
This article examines the relation between ownership structure and firm value as proxied by market-to-book ratios for REITs over the period 1980 to 1994. Difference in means tests provide evidence that increased inside director block holdings are associated with lower market-to-book ratios for equity REITs. In addition, difference in means analysis indicates that the presence of outside banking and insurance holdings are associated with increased equity REIT market-to-book ratios, while the presence of pension and money management block holdings are associated with lower market-to-book ratios.
Similar relationships are not found for a combined hybrid and mortgage REIT sample. For this group, difference in means analysis reveals that high inside director and officer block holdings are associated with either relatively high or relatively low market-tobook ratios. Outside director block holdings are associated with increased hybrid and mortgage REIT market-to-book ratios, while increased levels of non-director outside block holdings are associated with lower market-to-book ratios.
Piecewise regression analysis indicates a non-linear relationship between REIT market-to-book ratios and ownership structure. Low levels of inside ownership are associated with increased market-to-book ratios for equity REITs. However, as inside block ownership increases above 5%, REIT market-to-book ratios decline. This result provides support for the entrenchment hypothesis concerning increased inside ownership. As inside ownership increases, the insiders become more entrenched and are able to expropriate outside shareholder wealth with less fear of being ousted from their inside positions via the market for corporate control.
The opposite result is true for hybrid and mortgage REITs. A negative relationship is observed low inside ownership. However, a positive relationship occurs as ownership increases from five to 25% and then turns negative again as ownership increases above 25%. This result provides support for the alignment hypothesis associated with increased inside ownership. As inside ownership increases, insiders’ interest in maximizing shareholder wealth increases. However, as inside ownership crosses the 25% threshold, concerns about entrenchment and wealth expropriation resume causing a decline in market-to-book ratios.
The conflicting results for inside block holdings observed for equity and hybridmortgage REITs may result from their respective asset structures. Transactions involving real property assets held by equity REITs may be more opaque.
Equity REIT assets are not homogenous products that are readily priced in the marketplace but are often an amalgamation of heterogeneous properties from across the United States. To determine current asset values, REIT management often employ professional property appraisers. However, the inside management team and directors decide if and when a property will be appraised and who will perform the appraisal. In addition, Damodaran and Liu (1993) find evidence that insiders utilize these appraisals to expropriate wealth from outside shareholders by trading on the appraisal information prior to public release. To obtain unbiased and independent estimates of value for equity REIT properties, small outside shareholders must obtain accurate data on reasonable property appreciation rates, rental rates, expense rates and other property-specific information. Therefore, the presence of large inside block holders in equity REITs increases the likelihood that outside shareholders may be trading against insiders with superior information. Consequently, outside shareholders discount the REIT’s stock to account for this possibility.
On the other hand, mortgage REIT transactions should be more transparent than equity REIT real property transactions. Mortgages generally are considered more liquid than real property holdings. In addition, outside shareholders may find it easier to determine a risk-adjusted required rate of return for mortgages allowing mortgage REIT assets to be more readily priced in the market. Therefore, there is less opportunity for mortgage REIT inside block holders to expropriate wealth from outside shareholders through asset acquisition and disposition transactions. Consequently, actions by mortgage REIT inside block holders become more aligned with outside shareholders as they strive to increase profits for the REIT in general, rather than line their own pockets through complex inside deals.
Outside blockholdings appear to have a negative impact on both equity and hybrid and mortgage market-to-book ratios. As levels of outside block holdings rise above 5%, market-to-book ratios for these REIT groups starts to decline. This result contradicts potential monitoring benefits associated with increased outside block holder ownership. Rather, the opposite is true providing support for findings reported by Barclay, Holderness and Pontiff (1993) that outside block holders may utilize their position to gain an advantage over less informed small outside shareholders.
Overall, these results reveal that the presence of large block holders is not necessarily in the best interest of shareholders. Increased block holdings is associated with lower market-book-ratios for all REIT types. However, this negative relationship between block holdings and REIT market-to-book ratios is much stronger for equity REITs indicating that asset type may influence the market’s response to the presence of large block holdings. As firm asset valuations become more opaque the presence of large block holders, especially inside block holders, is viewed with increased suspicion by less well positioned outside block holders. The presence of large block holders increases the likelihood that these outside shareholders will be trading against more informed insiders. Consequently, these outside shareholders discount the price of these firms relative to their asset values to account for the presence of this increased informational asymmetry.
1 Demsetz (1983) contends that there is no relationship between ownership structure and profitability. He argues that ownership structure is based on an equilibrium level arrived at through competitive selection in which different cost advantages and disadvantages are balanced.
2 Jensen and Meckling ( 1976) argue that as inside ownership increases, insiders interests converge with those of outside shareholders and their propensity to pursue non-value maximizing behavior declines.
3 Fama and Jensen (1983) point out that managers with significant holdings can engage in nonvalue maximizing behavior with less fear of dismissal.
4 For a description of the qualifying provisions for a REIT, see Internal Revenue Code, Sections 856-858.
5 Publicly available proxy filing data from the SEC Q-Data Files or EDGAR files was often not available for newly minted REITs for at least a year after the initial public offering and often much longer. Similar delays were also observed for existing REITs.
6 It is our contention that the fundamental relationship between REIT value and ownership structure that existed for REITs from 1980 through 1992 should persist for REITs into the future. Shareholder perceptions of agency costs should be consistent across time periods, barring major legislation to limit these costs.
7 It is important to note that most REITs appear more than once in the sample.
8 As Damodaran and Liu (1993) and Cappoza and Lee (1995) point out, the estimation of the replacement costs of REIT assets is both difficult and tricky. However, barring the ability to observe readily traded prices in a highly liquid market such as the one that exists for assets of closed-end investment companies, one must employ the values reported to shareholders and the Securities and Exchange Commission via financial statements and assume that any errors from the true underlying value are randomly distributed around the actual value for the observed REITs.
9 The market-to-book ratio is clearly a noisy indication of ownership influence. However, a superior measure to examine the cross-sectional relationship between ownership structure and the markets perceived value of this structure has yet to be presented in the literature.
10 As a diligent reviewer pointed out, both the percentage block holder ownership and the numbers of block holders are important measures of ownership structure. Increased percentage blockholdings combined with a low total number of block holders indicates a high concentration of ownership in a few hands while the reverse indicates a lower concentration. The different ownership structures may have different impacts on firm value. However, our analysis indicates a very high positive correlation between the percentage of blockholdings and the numbers of block holders.
11 Morck, Shleifer and Vishny (1988) experimented with other breakpoints but found the five and 25% breakpoints have a stronger theoretical underpinning and produced the lowest sum of squared errors.
12 White’s adjustment is employed to correct for the presence of heteroskedasticity in all of the models for this analysis.
13 For an expanded explanation of the piecewise regression methodology employed in this analysis, the reader is referred to Mendenhall and Sincich (1989) or a similar text on regression analysis.
14 Block ownership is either 0% or greater than 5%. Consequently, the coefficient for B-OWN provides the relationship between MIB and no block ownership when block ownership is less than 5%.
Barclay, M. J., C. G. Holderness and J. Pontiff, Private Benefits from Block Ownership and Discounts on Closed-End Funds, Journal of Financial Economics, 1993, 33, 263-91.
Capozza, D. R. and S. Lee, Property Type, Size and REIT Value, Journal of Real Estate Research., 1995, 10, 363-80.
Damodaran, A. and C. H. Liu, Insider Trading as a Signal of Private Information, The Review of Financial Studies, 1993, 6, 79-119.
Demsetz, H., The structure of Ownership and the Theory of the Firm, Journal of Law and Economics, 1983, 26, 375-90.
Downs, D. H., The Value in Targeting Institutional Investors: Evidence from the Five-or-Fewer Rule Change, Real Estate Economics, 1998, 26, 613-49. Fama, E. F. and M. C. Jensen, Separation of Ownership and Control, Journal of Law and Economics, 1983, 26, 301-25.
Jensen, M. C. and W. H. Meckling, Theory of the Firm, Managerial Behavior, Agency Costs and Ownership Structure, Journal of Financial Economics, 1976, 3, 305-60.
Mendenhall, W. and T. Sincich, A Second Course in Business Statistics: Regression Analysis, Second Edition, San Francisco: Dellen Publishing Company, 1989.
Morck, R., A. Shleifer and R. Vishny, Management Ownership and Market Valuation: An Empirical Analysis, Journal of Financial Economics, 1988, 20, 293-316.
Pound, J., Proxy Contests and the Efficiency of Shareholder Oversight, Journal of Financial Economics, 1988, 20, 237-65.
Weston, J. F., The Tender Takeover, Mergers and Acquisitions, 1979, 74-82.
H. Swint Friday* G. Stacy Sirmans** C. Mitchell Conover***
*University of South Alabama. Mobile, AL 36688-0002 or firstname.lastname@example.org. **Florida State University, Tallahassee, FL 32306-1042 or email@example.com. ***University of North Carolina-Wilmington, Wilmington, Wilmington, NC 28403 or firstname.lastname@example.org.
Copyright American Real Estate Society 1999
Provided by ProQuest Information and Learning Company. All rights Reserved