THEORY AND PRACTICE OF DEBT FINANCING IN REITS
By
Huyue Zhang
M.A. The University of Toledo
B. Architecture Tsinghua University
Submitted to the Department of Urban Studies and Planning
In partial fulfillment of the requirements for the
Degree of Master of Science in Real Estate Development
ROW"M
at the
Massachusetts Institute of Technology
February, 2000
FEB 25 ZOQ
@ 2000 Huyue Zhang
All rights reserved
The author hereby grants to MIT the permission to reproduce and to distribute publicly
paper and electronic copies of this thesis document in whole or in part
Signature of Author
Huyue Zhang
Department of Urban Studies and Planning
January 7, 2000
Certified by
Blake Eagle
Chairman of The Center for Real Estate
Thesis Supervisor
Accepted by
_
V
-
14=1'
-
William C. Wheaton
Chairman, Interdepartmental Degree Program in Real Estate Development
Theory and Practice of Debt Financing In REITs
By
Huyue Zhang
Submitted to the Department of Urban Studies and Planning
On January 7, 2000 in partial fulfillment of the requirements for the
Degree of Master of Science in Real Estate Development
ABSTRACT
Miller and Modigliani's classic finance theory proposes that debt does not add to
a firm's value in the absence of corporate tax. Real Estate Investment Trusts (REITs)
have no corporate level income tax and meet M & M's perfect market assumption. In
reality, however, REIT operators use a significant amount of debt on their balance sheet.
There seems to be a contradiction between practice and theory.
This thesis developed a statistical model, which demonstrated the relationship
between leverage and the cost of equity capital in a sample of 60 equity REITs including
four property types. The model indicated a positive relationship between leverage ratio
and cost of equity. When a firm adds additional debt, the cost of equity is increased in the
form of a higher FFO yield, which is often associated with the decline of stock price. On
one hand, the increased debt level increases the required return on equity, and therefore
increases the weighted average cost of capital (WACC). On the other hand, the debt is
cheaper than the overall return on asset, and the addition of debt lowers the WACC value.
The positive and negative effects of a high debt level on WACC are then neutralized.
Finally, the model indicated that the WACC value is invariant to the changes of leverage
ratios and directionally supported the M & M theory. The model also verified that the
size of a firm has an inverse relationship with the cost of equity. The larger firms usually
have a lower cost of equity.
Thesis Supervisor: Blake Eagle
Title: Chairman of The Center for Real Estate
ACKNOWLEDGMENT
A thesis is never a piece of work for one person. Many contributors assisted me in this
work. I would like to say thank you to Mr. Blake Eagle, my thesis advisor at MIT/CRE,
Mr. Randall Newsome at Heitman Financial, and Ms. Virginia Dawson at Merrill Lynch
for their instruction, information and encouragement.
Special thanks also go to Mr. Blake Eagle for not only the direction he gave in my thesis
work, but also for the generous help in my search for a job.
I would also like to thank Dr. William Wheaton and Dr. Timothy Riddiough for their
academic instruction in my education and their support in pursuit of my career.
Last, but not least, I am grateful to my family for supporting me in completing my
education at MIT/CRE, in which I received very professional training.
TABLE OF CONTENTS
ABSTRACT
2
ACKNOWLEDGMENTS
3
TABLE OF CONTENTS
4-5
CHAPTER 1
INTRODUCTION
6-8
CHAPTER 2
HISTORY OF REIT DEVELOPMENT
History of Development
9-13
Classification of REITs
13-15
Requirements for REITs
16-17
Recent REITs Performance
17-22
CHAPTER 3
LITERATURE REVIEW OF M & M THEORY
AND REIT CAPITAL STRUCTUR
M & M Theory
23-27
REIT Related Capital Structure Theory
27-33
CHAPTER 4
TEST 1: LEVERAGE ANF REQUIRED RETURN TO EQUITY
Introduction
34-35
Data Selection
35-36
The Regression Model
36-39
Results of the regression model
40-41
Findings and Implications
41-42
CHAPTER 5
TEST 2: LEVERAGE AND COST OF CAPITAL
Introduction
43-44
Methodology
45-47
Findings
47-49
Implications
49-50
CHAPTER 6
CONCLUSIONS
51-55
BIBLIOGRAPHY
56-57
GLOSSARY OF TERMS
58-62
APPENDIX
Exhibit I - Comparative Valuation of REITs from Merrill Lynch
63-67
Exhibit 2 - Assembled Data Set
68-69
Exhibit 3 - Regression for Test 1
70-71
CHAPTER 1.
INTRODUCTION
Corporate finance theory suggests that in a perfect capital market, the value of a
firm remains unchanged when the capital structure changes. Modigliani and Miller
theorized in their landmark paper of 1958 that a firm's value is determined by the value
of its assets, not by the securities it issues (Modigliani and Miller, 1958). Thus, the M &
M Proposition I proposes that the market value of the firm is independent of its capital
structure. Modigliani and Miller further suggest (Proposition II) that while borrowing will
improve the expected return, it will also increase the required return on equity because of
an increased risk. Those changes in the discount rate will leave the value of the firm
unchanged. Under these conditions, there is no reason to assume debt. This theory was
based on the assumptions of the perfect capital market characterized by no taxes, costless trading, and borrowing and lending at a risk free rate.
In reality, most of the firms are taxed at the corporation level, but the interest
payment is considered as the production cost and is tax deductible. This provides a
rationale for borrowing. Under the above conditions, the value of the firm can be
described as the present value of its future cash flow plus the present value of its tax
shield. In an imperfect financial world (with the corporate tax), the use of a greater
amount of debt by the firm can enhance the value. However, operating at a high debt
level will incur direct or indirect costs (such as bankruptcy cost), which are known as cost
of financial distress. The cost of financial distress will offset the advantages of the tax
deductibility of the interest payment. This trade off between tax deduction and the costs
of financial distress suggests the idea of optimization of level of debt. However, during
the early part of this century when there was no income tax imposed on corporations,
corporate balance sheets had plenty of debt. This shows that the financial officers were
considering factors other than the tax shield/financial distress trade off.
In the world of real estate finance, there is an interesting case involving this
question. Real Estate Investment Trusts (REITs) are a growing financing source for
commercial real estate. A REIT eliminates the taxation of income at the entity level and
only pays tax at the shareholder level. In exchange for the benefit of the corporate tax
exemption, REITs have to meet many other conditions and requirements. As long as the
REITs adhere to the conditions imposed by the law, they pay no tax on their income at
the entity level. REITs can be described as conditionally tax exempt. Under the M & M
theory discussed above, REITs should not seek the addition of debt because this strategy
will have no impact on their value. Yet REITS do borrow money, often by considerable
amount. The phenomena of REITs to assume debt suggests that some other factors
influence REIT capital structure and leverage decision.
This thesis is an attempt to find whether theory in an academic world and
practical application in the real world of corporate finance are consistent. The focus of
the research will be equity REITs as opposed to mortgage REITs. Equity REITs invest
directly in property and comprise of more than 90 per cent of the market capitalization of
the REITs (NAREIT, 1999). The approach of this study will be to explore some of the
factors that determine REIT capital structure decisions by analyzing REIT data and
developing a model to quantify the cost of leverage to a REIT. The model will examine
the relationship between the market required return (presented as the FFO yield) and a
series of other variables, such as the REIT's size, the debt to market capitalization ratio,
the property type it specializes in, and the organization structure (UPREITs or regular
REITs) etc. The research should yield results helpful to REIT financial managers in
operating the firms.
It should be noted that the situational factors surrounding the 1992 - 1996 period
of REIT growth would not be those of the future. An environment characterized by low
interest rates, low inflation, a broadening macro-economy recovery, and improving
commercial real estate market is unprecedented in the REIT industry. These conditions
will change, and the REIT structure will be tested in new ways. The signs of these
changes first became visible in 1997. In 1998, as the commercial real estate market
moved further into a recovery period, competition for acquisition was heating up
dramatically. Early in the cycle, profits could be achieved simply by buying properties at
recession prices which produced capitalization rates of 10 to 12 per cent, packaging them
as REITs, and reselling them to the public at REIT earnings (FFO yields), which were
200-400 basis points lower. This practice became known as " positive spread investing".
With competition for properties driving up prices, these large yield spreads have been
narrowing.
CHAPTER 2.
HISTORY OF REITS
Developmental History
The concept of REITs dates back to the late 19th century when trusts were not
subject to taxation if trust income was distributed to the beneficiaries. However, during
the 1930s, a Supreme Court decision ended this treatment and required that all the
passive investment vehicles that were organized like corporations be taxed as
corporations. After World War II, the need for equity and mortgage funds to support a
national housing program promoted the extensive use of real estate investment trusts. In
1960, Congress passed legislation creating the REITs. Under the law, a REIT can be
treated as a conduit with its income distributed to its beneficiaries or shareholders.
The FirstREIT Boom
REITs were established by the Real Estate Investment Trust Act of 1960. REITs
experienced their first significant growth during the late 1960s and early 1970s. This
growth was stimulated by a strong demand for funding for construction and development,
which was not met by traditional sources. Much of this demand was met by REITs
providing long-term capital, which they in turn took from short-term paper and bank
financing. Employing this strategy, many REITs had profitable spreads until interest rates
began to rise in 1973 and 1974. In the high interest environment of 1973 and 1974, many
of these REITs had negative spreads. Also, the overbuilding in the real estate markets
contributed to the problem and REIT performance declined dramatically in the mid and
late seventies. During 1974 and 1975, the REIT share price index dropped by 50% (Han
and Lian, 1995). Significant structural changes also occurred during the late 1970s and
early 1980s. Leverage was reduced significantly. By 1984, leverage ratios in equity
REITs had declined to 55 % down from 64 % in 1972.
REITS Capitalization
160000
140000
120000
100000
80000
60000
40000
20000
na,
-I
=mli
-
AN
EM
M
AN I
-I
SEquity REIT [3 Mortgage RElT [: Hybid REI1T
Data Source: NAREIT (1999)
M
IN I M
M
M-
-M
1
The Second REIT Boom
It was not until the real estate markets improved in the early 1980s that REITs
recovered. The second REIT boom was fueled by The Economic Recovery Tax Act of
1981, which included shorter depreciation schedules for real properties and the ability to
pass tax losses to investors, who then were able to reduce their personal tax liabilities.
REITs and partnerships, together with banks, S&Ls, insurance firms, pension funds and
foreign investors put a flood of capital into real estate market. The excessive influx of
capital to the real estate market led to the biggest construction boom in American history.
The results, however, were devastating. Office vacancy levels reached all-time highs. The
number of empty apartment buildings soared.
The Tax Reform Act of 1986(TRA, 1986) also stimulated the growth of REITs.
The new law eliminated the tax advantages of many real estate limited partnerships by
increasing depreciation schedules and replacing accelerated depreciation with straight
line depreciation. In addition, the new rules decreased the deductibility of real estate tax
loss against ordinary income. These changes in the TRA 1986 changed the nature of real
estate investment. Because REITs were not a tax-advantage investment, the change in tax
law did not have the impact on REITs. On the contrary, it actually improved the
attractiveness of REITs.
The Third REIT Boom
In the real estate recession of the late 1980s and early 1990s, banks, savings and
loan institutions, and life insurance companies all exited the real estate capital markets
due to the massive overbuilding of the previous decade. As the macro economy of the
United States began to emerge from the recession in 1991, most commercial real estate
was still behind the recovery curve. By 1993, part of the industry had begun to recover
and demand for capital increased. REITs were a potential capital provider and their
growth was tremendous. By the end of 1996, the total market capitalization of publicly
traded equity REITs had grown to $78.3 billion from $4.39 billion just ten years earlier.
The total number of publicly traded equity REITs had grown to 166 from 45 during the
ten year period (NAREIT, 1996). The amount of capital to real estate from equity REITs
jumped dramatically between 1993 and 1996. During this period, over $44 billion in
initial and secondary offerings occurred compared with $6.4 billion during the previous
four years. In 1997, the market capitalization of equity REITs increased tremendously to
$ 127 billion, almost double that of 1996. The total capital raised in 1997 was $ 45.3
billion. From 1997 to present, the capitalization has been stabilized at the $125 to $130
billion level (NAREIT, 1999).
Many of the REITs established during the 1990s have very different
characteristics than those of previous years. Ken Rosen at the University of California at
Berkeley created the term "Modern Era REIT" to emphasize that the REITs that had
became public since late 1992 had very distinct investment characteristics from earlier
REITs. Their portfolios contain institutional quality real estate. They can acquire, develop
and manage their properties. Typically, they were also larger with market capitalizations
above $400 million (Rosen, 1995).
The three most common reasons for the dramatic increase of REITs' public
offerings since 1993 were access to capital, debt reduction and estate planning (Baird,
1998). Before the early 1990s, most capital sources had abandoned private real estate
companies due to the slump in the market. Insurance companies and banks were not
lending, or they were requiring low loan-to-value ratios. Private real estate companies
desperately needed access to equity capital to roll over their debt and to take advantage of
market opportunities. The second reason for going public was debt reduction. Many
loans made to real estate firms in the mid-to-late 1980s had five- to seven-year maturities.
Most private real estate owners who had borrowed 90 %to 110 % of the cost of their
assets did not have the equity to refinance the loans. This need for debt reduction led to
the "go broke or go public" status for many real estate firms by the early 1990s. The
third reason was estate planning. A generation of developers that came of age in the
1950s and early 1960s were reaching the age considering their estate problems. Many of
them concluded that the tax bill that would go into effect once they died would require a
partial or total liquidation of their assets to meet this tax liability. Owning shares in a
public REIT allows for more liquidity without having to sell real estate assets. This
proved compelling for two reasons. First, it was tax efficient, and second, it allowed these
entities to enter a form in which a company could outlive its founder (Baird, 1998).
Classification of REITs
REITS can be classified by the nature of the assets they are holding (e.g. equity
REITs, mortgage REITs), by nature of the properties they are managing (e.g. shopping
center REITs, apartment REITs), and by the organizational structure of the REITs
themselves (e.g. UPREITS, DOWNREITs).
Equity REITs, Mortgage REITs and HybridREITs
In the REIT family, there are two types: equity REITs and mortgage REITs.
Equity REITs acquire ownership interest in properties and derive most of the income
from rents. Mortgage REITs purchase mortgages and become creditors. There are also-a
small number of hybrid REITs combining elements of both. Equity REITs comprised
about 91% of all REIT market capitalization in 1999. Mortgage REITs account for 5% of
REIT market capitalization and hybrid REITs account for 4% of REIT market
capitalization (NAREIT, 1999). Equity REITs include many forms. Blind pool trusts are
those in which specific properties have not been identified at the time of the purchase of
the shares. The fully specified trusts are those in which identified properties are described
in the prospectus. Some are close-ended, with a specified maximum amount to be raised
from investors. This investment offers protection from dilution. Some are open-ended,
selling new shares in secondary offerings.
Classification of REITs
Equity REITs
94%
E ITs
4%
Hybrid RE[Ts
2%
Data Source: NAREIT, 1999
UPREIT& DO WNREIT
In 1992, in order to address potential tax issues related to the formation of a
REIT, a new form of REIT emerged. It was called the "umbrella partnership REIT"
("UPREIT"). It proved popular in attracting capital, and since its creation, more than 75
percent of new REITs have taken that form. In the typical UPREIT, the partners of the
Existing Partnerships and a newly-formed REIT become partners in a new partnership
termed the Operating Partnership. After a period of time (often one year), the partners
may convert the Units into either cash or REIT shares (at the option of the REIT or
Operating Partnership). This conversion may result in the partners incurring the tax
deferred at the UPREIT's formation.
Existing UPREITs frequently issue additional Operating Partnership Units to
partners in existing partnerships that hold and operate commercial real estate. The new
Unit holders achieve precisely the same benefits of the tax deferral and liquidity as the
original contributors to the Operating Partnership. The REIT benefits by being able to
acquire additional assets without having to immediately tap into the capital markets.
A "DownREIT" is the same as an UPREIT, except that the REIT in a DownREIT
owes a substantial amount of property directly, whereas virtually all of an UPREIT's
holdings are in the Operating Partnership.
Requirements for REITS
Many requirements must be met for a REIT to qualify for its tax exemption status.
The conditions can be categorized as organizational requirements, asset requirements,
income requirements, and distribution requirements.
OrganizationalRequirements:
A REIT must have at least 100 persons that own its stock or interests. Each entity,
such as corporation, partnership, or pension plan, is considered as one person. No more
than 50% of the shares may be held by a group of five individuals or fewer.
Asset Requirements
At least 75% of the value of its assets must be real estate assets, cash and
government securities. Real estate assets include real properties ,debt secured by
mortgages on real properties and stocks of other qualified REITs. Not more than 25% of
the assets may consist of securities. It may not have more than 5% of its assets invested in
the securities of one issuer and it may not own more than 10% of the outstanding voting
stock of such issuer.
Income Requirements:
At least 75% of its gross income must come from rents from real property or
interest from real property secured mortgages, gains from the sale of real property,
dividends from qualified REITs, gains from the sale of qualified REIT stock, refund of
taxes on real property, or gains from the sale of foreclosed property. At least 95% of the
entity's gross income must be derived from dividends, interest, rents or gains from the
sale of certain assets (primarily real estate). Not more than 30% of the income can be
derived from sale or disposition of stock or securities held less than six months or real
property held less than four years other than property involuntarily converted or
foreclosed upon.
DistributionRequirements:
REITs must distribute 95% of taxable income(excluding net capital gains) each
taxable year.
Compliance Requirements:
Shareholders must be polled annually to determine ownership of the outstanding
shares and ascertain whether the REIT has fulfilled the requirements of the "five or
fewer" ownership requirement.
Recent REITs Performance
REITs have been effective investment vehicles in the current real estate recovery
cycle. All REITs (equity, mortgage, and hybrid) have produced total returns (dividends
and share price appreciation) of 18% in 1995, 36% in 1996, 19% in 1997, negative 19%
in 1998, and 4.94% in 1999. The Equity REITs performed similarly, achieving a total
return of 15% in 1995, 35% in 1996, 20% in 1997, negative 18% in 1998 and 4.78% in
1999.
Total Return of REITs
80.00
60.00
40.00
20.00
0.00
1988
19
199
1992
1991
1993
1994
1995
1996
-20.00
-40.00
SComposite --*Equity
Data Source: NAREIT, 1999
6
Mortgage _). Hybrid
199
1998
1999
NAREIT Indexes
350.00
300.00
250.00
>
200.00
150.00 100.00
50.00
0.00
- All
-. Equity _x Mortgage
w Hybrid -.
S&P500
Source: NAREIT (1999)
This performance before 1998 had made REIT shares particularly attractive,
especially during a period of weakness in another high-dividend equity: utility shares.
Early in 1998, it was becoming clear, however, that the conditions that sustained the
returns achieved during the previous several years would not continue forever. In the late
spring of 1998, REIT share indexes showed negative total returns. Clearly, the recovering
real estate cycle has been kind to REIT shares, but the cycle is maturing now, raising
questions about the ability of REITs to continue growing and to be awarded the share
price multiples they have received since 1995.
REIT earnings are adjusted for depreciation and other non-recurring events such
as the sales of properties, and they are reported as Funds From Operation (FFO). The
growth in FFO per share of equity REITs had been robust during the 1995-1997 period,
typically averaging 15% per year (NAREIT, 1997). Similar to other industrial firms, pershare growth rates of FFO are a primary determinant of market prices.
REITs achieve growth in FFO in three ways. First, a REIT attempts to manage its
real estate portfolio so as to increase the net operating income each year through a
combination of revenue increases and expense reductions. Early in a real estate cycle,
when most markets are in their recovery stage, revenue growth comes easily through
lease renewals at higher rental rates and leasing of vacant space, and the growth can be
dramatic. As the market moves into the equilibrium phase, growth comes down to a rate
that is close to the inflation rate. This kind of growth strategy is referred to as "internal
growth" (Marks, 1996).
REITs have two external mechanisms for FFO growth. One is to purchase
additional properties. The yields on the new properties should be greater than the REITs'
cost of capital, and should add value to the shareholders' wealth. In assessing such deals,
a REIT's weighted average cost of capital (WACC) is utilized as the discount rate to
evaluate the expected cash flows from an investment. If the internal rate of return (IRR)
of an investment is greater than the REIT's WACC, the transaction will add value to the
REIT shareholders (Goldman Sachs, 1995). Early in the recovery cycle, very little
competition existed for commercial properties and positive NPV (net present value) deals
were easy to find in most markets. As the cycle has matured and as increasing amounts of
capital have returned to the sector, the spreads between property acquisition yields and
normal first year cost of capital has narrowed and even become negative. The acquisition
may still be accretive if rents increase, vacant space is leased up, and the IRR exceeds the
REIT's cost of capital later in the holding period.
The third mechanism employed by REITs to increase FFO is the development of
new properties. After a significant downturn, such as that between 1989 to 1993, there
was little demand, substantial vacant space, and a lack of new construction. During the
early recovery period, the vacant space was filled, rents improved, and capital was
attracted back to the sector. When property values improve, it becomes increasingly
difficult to purchase property below replacement cost. Under these conditions, new
development typically begins. Development of new properties is riskier than purchasing
existing properties. The expected returns are greater too, typically 200 to 300 basis points
higher. Again, the evaluation of a development project is to see whether there can
realistically be a positive spread between the development IRR and the REIT's WACC.
An interesting question raised by the events of this cycle is whether or not REITs
will tend to add leverage to improve FFO growth as opportunities decline. The
development phase of the cycle will produce the likelihood of overbuilding. When this
occurs, markets will soften and the FFO growth in the recovery will end. Under these
conditions, some REITs will be tempted to add debt to their balance sheet to take
advantage of the positive leverage and improve the FFO growth. In the 1993 to 1998
recovery before the stock market crashed in the Fall 1998, the equity markets had been
forgiving of excessive leverage, probably because the market recognized that acquisition
and development opportunities were still rich.
This thesis will explore the implied cost to REITs that add debt on their balance
sheets. The paper will also test the M & M propositions by analyzing the data from some
sample REITs.
CHAPTER 3
LITERATURE REVIEW
The M & M Theory
Modigliani and Miller concluded that there was no value-enhancing reason to
borrow money in the ideal, tax-free world that they described. The introduction of
corporate income taxes changed their conclusion because of the tax-deductibility of
interest payments. Under those conditions, the taxed firm should use 100% debt financing
(Modigliani and Miller, 1963). In 1977, Metron Miller introduced personal taxes into the
model and suggested that if the personal tax rate on equity was less than the personal rate
on bonds (because of the capital gain tax), then leverage will affect the value of the
untaxed firm negatively. He also argued that debt financing could create a negative
impact on the value of the firm if bankruptcy costs are not trivial. These finance theorists
concluded that the use of debt financing by non-taxed firms has no effect on the value of
the firm, and that debt financing may have a negative impact on the value under the
conditions of non-trivial bankruptcy costs and a lower-than-debt personal tax rate on
equity.
M & M's arguments are often presented in the form of two propositions. The first
proposition suggests that if investors and firms can borrow or lend at the same terms, then
the firm's value is not enhanced by changes in capital structure. The first proposition is
based on the arbitrage pricing argument. Employing their argument in a real estate
investment context, the owner of a levered property (identical to an un-levered
comparison property) would be able to sell the levered property, pay off the loan,
personally borrow money to purchase the un-levered property, and realize greater
personal wealth at the same level of risk. The arbitrage pricing argument suggests that
such opportunities would soon be recognized, the price would therefore adjust, and this
advantage would eventually disappear. In the real estate context, the M & M argument
suggests that as long as investors can borrow at the same rate offered by professional
lenders on mortgage properties, the creation of leverage can not add value to the property
(Clauretie and Sirmans, 1996).
M & M's second proposition suggests that although the expected return to equity
will increase with the increases in the debt-to-equity ratio, the expected return is exactly
offset by an increase in risk (Brealey and Meyers, 1991). A number of scholars have
worked on this problem and proposed several theories to support the idea of optimal
capital structure. The theories include: bankruptcy costs, the signaling hypothesis, and
agency costs.
Bankruptcy Costs
The value of a firm in bankruptcy is reduced because payments must be made to
third parties other than bondholders or shareholders. Legal and trustee fees are examples
of costs, which reduce the asset value of a firm. This suggests the existence of an optimal
capital structure where leverage is perceived as a trade-off with the future costs of
bankruptcy (Baxter, 1967). Costs of bankruptcy include direct costs and indirect costs.
According to some studies, direct costs are in fact trivial, but also vary inversely with the
size of a firm. Indirect costs are more significant, problematic and hard to estimate
because they are essentially opportunity costs. Altman studied 19 retail and industrial
establishments that declared bankruptcy between 1970 and 1978. He concluded that
indirect bankruptcy costs were 8.1% of firm value three years prior to bankruptcy and
grew to 10.5% in the year of bankruptcy (Altman, 1984). Direct and indirect costs
associated with bankruptcy are large enough to provide credibility to the theory of an
optimal capital structure based on the trade off between tax gains from leverage and
future expected bankruptcy cost.
CapitalStructure Changes as Signals
If managers have better information than investors and if markets are not fully
efficient, then the use of capital structure changes as a signaling mechanism is plausible.
Ross(l 977) was the first to explore this concept, suggesting that the market places a value
on the perceived stream of future returns for a firm. Changes in this perception,
stimulated by changes in capital structure or in dividend policy, will alter the market's
assessment of the value of the firm. Insiders who have access to information might
choose to use changes in capital structure as a device to convey signals about the future
prospectus. Myers and Majluf (1984) also demonstrate that, under the assumption that
management acts in the best interests of passive, current shareholders, the possession of
superior information will cause the managers to prefer debt to equity when external
financing is needed.
Agency Cost Theory
Agency cost theory is very relevant to this study because it does not rely on
income taxes to support the idea of an optimal capital structure. Firms carried leverage
long before the imposition of the income tax, and there must be some other explanations
for the reason why firms use debt. Jensen and Meckling(1976) proposed that ownership
structure was a determinant of operation decision because owners' and managers'
interests are often inconsistent. For example, a levered and on-the-edge-of-bankruptcy
firm has two investment choices, one with high expected return and high risk, the other
with lower return and low risk. The firm may be tempted to take the riskier of the two
because the upside potential will insure the equity holders' benefit while the bond
holders absorb the risk of downside. The result is a transfer of wealth from bondholders
to shareholders. The cost of preventing this transfer by protective covenants and other
monitoring mechanisms may be large and vary directly with the amount of leverage
assumed.
There are also agency costs associated with equity. Jensen and Meckling suggest
that as a single owner-operator sells equity in his firm these costs increase because the
manager can maximize personal wealth at the expense of shareholders through the use of
perquisites (luxury automobiles, vacations, and personal airplane). The shareholder will
incur monitoring costs, which will vary with the amount of equity employed in the capital
structure. According to Jenson and Meckling's concept, as more leverage is assumed, the
agency cost associated with equity decline, while the agency costs of debt increase. The
result is an optimum amount of leverage.
Sheridan Titman (1984) suggested that the nature of a firm's products has
implications for capital structure. If a firm's bankruptcy has high maintenance costs for
its customers, the firm will be likely to employ lower leverage. For example, the
bankruptcy of an automobile manufacturing company will leave the customer no place to
get maintenance service or parts. If such costs are low, the firm would carry more debt.
REITs seem to belong to the second category and likely to assume higher leverage.
REIT Related Capital Structure Theory
REITs are conditionally tax-exempted, and therefor partially meet the "perfect
market" assumption of the M & M Theory. There has been some research conducted to
examine the M & M Theory in the context of REITs.
Shulman (1975)
David G. Shulman conducted research testing the M & M's theory of leverage
(Shulman, 1975). He concluded that he could not disprove M & M's two propositions.
Shulman's work tested three hypotheses: 1.that risk-adjusted returns for equity
REITs do not significantly differ from overall stock market returns; 2. that risk-adjusted
returns from high-leverage REITs are not significantly different from low-leverage
REITs; 3. that risk-adjusted returns from relatively high dividend REITs are not
significantly different from those with low dividends.
The second hypothesis is directly related to the topic of this thesis. Shulman tested
returns from a sample of nine REITs for the period between December 31, 1963 and
December 31, 1974, and he also tested an additional seven REITs for the period
December 31, 1970 to December 31, 1974. Shulman's approach was to use the Capital
Asset Pricing Model (CAPM) to develop risk-adjusted returns for this sample of REITs.
The risk adjustment is very necessary because without the risk adjustment, the absolute
difference in performance between his REIT sample and the stock market would ignore
the risk in each. He compared the absolute returns, the alpha coefficients(the intercept of
regression), and the beta coefficients (the indicator of risk) of sample equity REITs to
those in the S&P 500. The results of the initial comparison lead to the rejection of the
first hypothesis because the REIT performance was significantly poorer than the Index's.
Shulman then tested the M & M theory concerning leverage and dividends. He
utilized the risk-adjusted returns as dependent variables in a regression equation and held
the leverage policy and dividend policy as the independent variables. The model was
designed to measure the relationship between the returns and leverage and dividend
policy. His regression model demonstrated no strong relationship between leverage and
risk-adjusted returns.
Shulman thus conducted the test of M & M's theory in a tax-exempt context. He
was not able to disapprove the M & M's theory. However, his sample size was small (16)
and raised the question as to whether or not it could yield a dependable conclusion.
Maris and Elayan(1990)
In exploring the factors contributing to the leverage use in REITs, Maris and
Elayan examined the agency theory and the leverage clientele effect. They concluded in
the study that the evidence supported the leverage clientele effect. The theory of leverage
clientele effect suggests that there may be a group of specific investors who prefer to hold
shares in un-levered firms. Kim, Lewellen, and McConnell (1979) argue that investors in
a high personal income tax bracket seek investments in unlevered firms and low-bracket
investors prefer levered firms. If these preferences exist, firms will adopt certain capital
structures to attract different investor clientele, and the distribution of capital structures
should be bi-modal.
Maris and Elayan argue that both agency costs and the leverage clientele effect
are reasonable explanations for the use of leverage by REITs, and that they have a
different impact on the pattern of debt financing in the REIT industry. If the leverage
clientele effect dominates in explaining the use of leverage, then the pattern of debt
financing should be bi-modal, namely, there should be a higher frequency at both the
lower end and the higher end. Maris and Elayan did not find the uniform capital structure,
but found the bi-modal distribution. The study supported the ideal of the leverage
clientele effect as the more important rationale for leverage.
However, in research conducted in 1991, Mais and Elayan tested for tax-induced
investor clientele in REITs and did not find an inverse relationship between the personal
tax rate and the leverage of trusts which was suggested by Kim, Lewellen and
McConnell (1979). After examining their sample consisting of 94 REITs from 1982 to
1988, the authors concluded that they could not support the hypothesis of a negative
relationship between investor tax rate and financial leverage. Maris and Elayan took the
study further by dividing the sample into REITs established before the 1986 Tax Reform
Act and those formed after 1986. For mortgage REITs, there was no relationship between
leverage and tax rates, either before or after the TRA 1986. For equity REITs, the
relationship is positive for both time periods, which is the reverse of the expectations.
Maris and Elayan also explored other factors, which could explain the use of
leverage in REITs. They proposed that REIT leverage was a function of firm size (total
assets), growth rate of earnings, the standard deviation of that growth rate, and the nature
(debt or equity investment) of assets employed in the REIT portfolio. Their study sample
included 60 qualified REITs. The model regressed the leverage against the four
determinants and concluded the following: for equity REITs, the leverage ratio is
positively related to the size and volatility of cash flow and negatively related to the
growth rate.
Sohan Lee(1995)
Sohan Lee addressed the REIT capital structure issue in his doctoral dissertation.
In his research, Sohan chronicled and analyzed the REIT industry, developed a model for
REIT property valuation, and tested various finance theories using the valuation model.
His dissertation makes two significant contributions: he develops a model to determine
the net asset value of a portfolio of REITs and he employs this value to test some capital
structure theories.
Sohan's approach involved employing data from two sources to arrive at a
weighted average capitalization rate for each REIT in his sample. His sample consisted of
58 REITs. To derive the property value of each REIT, he combined REIT financial data
from annual reports with market and product data(such as the capitalization rates and
prices) obtained from The National Real Estate Index(NREI). Using the capitalization
rates data from NREI and the proportion of each REIT's various asset types, he
calculated a weighted average capitalization rate for each REIT. He then obtained the
property value of each REIT by capitalizing the REIT's net operating income with the
weighted average cap rate. Subtracting the long term and short term debt and dividing by
the number of shares outstanding yields the net asset value per share of the REIT.
Sohan conducted a number of tests with his data, which explored the relationship
between value and leverage ratio in REITs. He divided the sample into four quartiles of
leverage ratio and found that a firm's value is significantly negatively related to leverage
in the first lowest quartile of leverage. He also finds that the market awards the highest
premium to those REITs in the first quartile and awarded the negative premium to those
in the second, third, and fourth quartiles.
Goldman Sachs (1995)
The Goldman Sachs (1995) research group addressed the central question raised
by M & M Theory. By using the 111 equity REITs they monitor in the security markets,
and assigning a 12% cost of equity to each REIT, they suggested that their WACC
analysis indicates a positive relationship between higher leverage and lower cost of
capital. Goldman, however, did not estimate the cost of equity for each REIT in their
sample. The research group rather defined it as the total return required by an equity
investor (dividend yield plus the long term growth rate of funds-from-operations). They
then employed both a CAPM (Sharpe, 1964) approach and the Dividend Discount Model
(Williams, 1938) to derive a blended cost of equity of 12%. Thus Goldman suggested that
value is positively variant to leverage. But by fixing the cost of equity at 12%, the
addition of cheaper debt will bring down the Weighted Average Cost of Capital
(WACC).
Later in 1995, the Goldman research group prepared another study in which they
regressed the FFO multiple on the debt-to-market-capitalization ratio in a sample of 115
REITs. They did not find a significant relationship. However, when they divided the
sample by property type, they did find a significant relationship between these two
variables. In the group of shopping center REITs, they discovered a negative relationship
between the FFO multiple and the leverage. In the multi-family section, the relationship
was also negative. The two cases imply that increasing the amount of leverage is
associated with the higher FFO yield and the higher required returns to equity. These
findings seem to support the M & M theory: the required return to equity increases as
additional leverage is added (Goldman Sachs, 1995).
This thesis will revisit the subject of the Goldman Sachs and David Shulman
studies and test for the relationship between increased debt and the return to equity. A
model will be developed which will evaluate the change of return to equity associated
with the addition of debt. The model will also examine whether size (the dollar value of
assets under management in a REIT) will counteract the effect of additional leverage on
required return to equity. In the modem (post -1 992)REIT era, it is believed that growth
in asset is the most significant strategy for survival and prosperity. Peter Linneman
(1996) at the Wharton School wrote an article suggesting that the cheaper capital and
economy of scale associated with size will stimulate further securitization and
consolidation in the real estate industry. Linneman argues that the commercial real estate
industry has started a transfer of ownership from the private to public sectors. This shift is
similar to that experienced by the other capital-intensive industries, such as automobile,
petroleum, aerospace, steel and railroads. The model developed here will address the
relationship between size (dollar value) and FFO yield to determine whether increases in
size can reduce the required return to equity. If REITs are penalized by increasing
amount of leverage, but are rewarded for the larger size, there may be an optimum
combination of the two.
CHAPTER 4
TEST 1: LEVERAGE AND REQUIRED RETURN TO EQUITY
Introduction
Funds-from-operation yield (FFO Yield) will be the measure of return required by
equity investors. FFO is the measure of operating performance commonly used in the
REIT industry. NAREIT defines FFO as net income excluding gains or losses from the
sales of property or debt restructuring, and adding back depreciation of real estate.
We have described a theoretical background in which REIT management should
be indifferent to debt. In theory, if a REIT maintains its status, it will pay no income tax
and have no rationale for carrying leverage on the balance sheet. Yet almost all REITs do
carry a significant amount of leverage. In the sample of 60 REITs employed in this
research, the average ratio of debt to asset is 46.9%, where debt is the book value of debt,
and assets are measured as the book value of debt plus the market value of equity. This is
the preferred method of describing leverage among REITs analysts (Goldman Sachs,
1995). In theory and practice, equity investors should charge a premium for high levels of
debt. This premium can be measured as the sensitivity of FFO Yield to changes in
leverage.
The model developed for this study will test the strength and direction of the
relationship between FFO yield and leverage. The test in this chapter will explore the
relationship between a REIT's FFO yield and its leverage, size in assets, its initial
formation as an UPREIT and property type. First, it is expected that there will be a
positive relationship between a REIT's FFO yield and the amount of leverage it carries
on the balance sheet. Second, it is expected that REITs with more real estate under
management will be characterized by a lower required FFO yield. The bigger REITs with
their large shareholder base and enhanced liquidity will have access to cheaper capital
and have better growth opportunity. Third, it is theorized that REITs specializing in
certain property types might support higher levels of debt because of the nature of the
leases found in their portfolios. For example, it might be expected that the trust, which
focuses on retail properties, will have a higher capacity for leverage because of the longterm leases. Finally, the test will examine whether the formation as an UPREIT will incur
a higher debt level and consequent higher cost of equity. If a trust was formed as an
UPREIT, it might carry more leverage than those not established under this structure and
the additional leverage would require a higher return to equity. It is theorized that each of
these variables will have some impact on the FFO yield of the sample REITs.
Data Selection
The primary sample in this research is from the NAREIT's publicly traded equity
REITs. NAREIT (National Association of Real Estate Investment Trusts) is the
Washington D.C. - based nonprofit association for the REIT industry. In 1998,
NAREIT identified 9 REIT investment specialties. These included mortgage-backed
securities, retail properties, residential properties, self-storage properties, health-care
properties, diversified properties, lodging and resort properties, specialty properties, and
office and industrial properties. This research will only focus on those REITs specializing
in apartment, neighborhood shopping center, regional mall and office/industrial property.
The NAREIT data includes information about each REIT's specialty,
management, headquarters address, age, and current analytical coverage by investment
banking firms. Complete financial information was obtained from the Report of
Comparative Valuation of REITs (September, 1999) prepared by Merrill Lynch. The
integration of the Merrill Lynch data and the NAREIT data resulted in a sample of 60
REITs. The sample includes 16 apartment REITs, 11 neighborhood shopping center
REITs, 12 regional mall REITs, and 21 office/industrial REITs. Refer to Exhibit 1.
The Regression Model
In this test, linear regression is utilized to model the sensitivity of FFO yield to
changes in the amount of leverage, size in assets, formation as an UPREIT, and property
specialty.
The independent variable is the FFO yield (FFOYLD). The FFO yield is
calculated by dividing its forward FFO per share by the stock price as provided in the
Merrill Lynch report. The FFO yield is the expected return on each dollar invested in the
stock. This concept is similar to the capitalization rate on a real estate investment, which
is the forward net operating income (NOI) divided by the price. The FFO yield can be
considered as the first-year cost of the equity for a REIT. It is the current return implied
by the relationship between the common stock price and expected first year earnings. Our
objective in setting the FFO yield as the dependent variable was to measure the changes
in cost associated with changes in the leverage and the other determining factors.
Mean FFO yields vary across property type within the sample. The apartment
property group has the lowest FFO yield at 11.03%. The highest mean FFO yields are
found in the regional mall group at 12.7%. Office/Industrial properties average 11.34%
and neighborhood shopping centers average at 12.1%.
All REITS
Apartment REIT
Shopping Center
Regional Mall
Office/Industrial
FFOYIELD
LEVERAGE
ASSETSIZE
Mean
11.67%
0.469
$3499.8 Million
Standard
Deviation
0.0027
0.014
399.8
Mean
11.03%
0.439
3093.6
Standard
Deviation
0.0045
0.0253
700
Mean
12.1%
0.435
1746.8
Standard
Deviation
0.0067
0.0331
337.5
Mean
12.7%
0.579
3909.9
Standard
Deviation
0.0072
0.0136
1229
Mean
11.34%
0.448
4493.3
Standard
Deviation
0.00398
0.0222
648.8
The independent variables in the model include leverage, size in assets, whether
or not the REIT was formed as an UPREIT and the property specialty. The leverage is the
measure of how much debt the REIT reflects on its balance sheet. Leverage is defined as
the ratio of the debt book value to the sum of the debt book value and the equity market
value. This is the definition commonly used by analysts at the investment banks. All data
in this model are obtained from the Merrill Lynch REIT report (September, 1999). The
mean leverage ratio of the sample is 46.9%. The standard deviation is 1.4%.
The four property type groups have significantly different leverage ratios. The
regional mall group averaged the highest leverage with 57.9%. It had a range of 14.9%
(49.6% to 64.6%) and a standard deviation of 4.7%. The apartment building REITs had
the lowest average leverage of 43.9%, a standard deviation of 2.5%, and a range of
36.9%. Office/Industrial REITs averaged 44.8% and the neighborhood shopping center
averaged 43.5%.
The size in assets (represented by SIZEASSET) is an indicator of the size of the
sample REITs in September, 1999. The mean value is $3499.8 million, and the standard
deviation is $399.8 million. The range of asset size is $15861million from $436.4 million
to $16297.4 million.
As mentioned earlier, the UPREIT format, in which former limited partnership
portfolios were rolled up and exchanged for Operating Partnership Units (OPUs),
suggests a higher propensity for leverage than found in the traditional REIT. The
increased leverage derives from tax exposure to former limited partners. It is theorized
that this potential recognition of a taxable gain resulting from debt retirement will induce
UPREITs to carry higher levels of debt on their balance sheets than non-UPREITs. To
test this idea, a dumb variable about whether formed as an UPREIT is created. The score
1 represents that it is a UPREIT, and score zero signals the non-UPREIT type.
However, the analysis of the data does not support the hypothesis of different
leverage levels between UPREITs and non-UPREITs. The mean leverage ratio for
UPREITs is 47.65%, and the standard deviation is 1.5%. The mean leverage ratio for
non-UPREIT is 46.96% and the standard deviation is 3.22%. The regression model on the
following will test whether the differences will translate into variations in the FFO yield.
There are four property-type independent dumb variables. They represent each of
the four types of commercial real estate in which the REITs in the sample specialize. The
Merrill Lynch Report has classified the sample REITs into several categories. We
examined four types of REITs provided by the report: Apartment REITs, Neighborhood
Shopping Center REITs, Regional Mall REITs, and Office/Industrial REITs. The four
variables are labeled as APART (Apartment), SHOPCENTER (Neighborhood shopping
center), REGIONALMALL (Regional mall) and OFF/IND (Office/Industrial). The dumb
variable of each REIT will be assigned a value 1or 0. For example, for apartment REIT
Amli Residential, the variable APART is assigned 1,while the variables SHOPCENTER,
REGIONALMALL and OFF/IND are assigned 0. Refer to Exhibit 3.
The above classification scheme, however, can be improved. Each REIT
specializes in one type of commercial real estate, but it does operate various properties.
For example, Mack-Cali, Inc. is classified as an OFF/IND by the Merrill Lynch report,
but it does include retail and apartment buildings in the portfolio. Therefore, the
percentage of each property type in the portfolio can be a better indicator of each REIT's
specialty. The percentage can be based on the number of assets of each type, square feet
of each type of property, or the NOI from each type. However, due to the limits of data
accessibility, we will use the dumb variables instead of the percentage indicator.
Thus, the final regression model contains the dependent variable FFO yield
(FFOYIELD) and 7 independent variables: asset size (ASSETSIZE), the ration of debt to
total market capitalization (LEVERAGE), the categorical variable UPREIT, and
property-type dumb variables: APART, SHOPCENTER, REGIONALMALL and
OFF/IND.
Results of the Regression Model
The correlation coefficient between the dependent variable and two of the 7
independent variables (ASSETSIZE and LEVERAGE) are strong. As expected,
FFOYIELD and ASSETSIZE are inversely related, implying that larger REITs benefit
from a lower cost of equity capital. FFOYIELD and LEVERAGE have the expected
positive relationship. Required FFO yield is clearly responding positively to the increase
in debt.
Correlation Coefficient Table
FFOYIELD
LEVERAGE
ASSETSIZE
UPREIT
APART
SHOP
REGIONAL
CENTER
MALL
FFOYIELD
1
LEVERAGE
.678
1
ASSETSIZE
-0.340
0.075
1
UPREIT
-0.025
0.111
0.160
1
APART
-0.188
-0.173
-0.080
0.023
1
SHOPCENTER
0.098
-0.151
-0.271
-0.591
-0.286
1
REGIONALMALL
0.250
0.513
0.067
0.207
-0.302
-0.237
1
OFF/IND
-0.114
-0.148
0.237
0.284
-0.442
-0.348
-0.367
OFF/IND
1
Our model establishes the positive and significant association between
LEVERAGE and FFOYIELD. We have seen that there is a significant difference
between the leverage ratios of each property type. Therefore, we might expect to find that
those more-highly-levered REIT types would be characterized by relatively higher
required returns to equity. Our results, however, do not support this. They show no
significant difference in FFO yields between the four property types.
We also may not reject the null hypothesis of no significant relationship between
the formation of a REIT as an UPREIT and FFO yield. The theoretical basis of this idea
was that former limited partners who became holders of Operating Partnership Units
would exert pressure on the REIT management to not pay off former partnership debts
because of the taxable gain recognition problem. The theory is not substantiated in our
data. There seems to be no significant relationship between the initial formation as an
UPREIT and the possibility to carry more debt.
Findings and Implications
The empirical tests reflected several interesting findings with implications for
REIT managers. First, the findings demonstrate a clear relationship between additional
debt and an increase of equity capital cost. The implications for REIT managers are clear.
The market will apply M & M's increased cost of equity to adjust for the increasing risk
to equity holders incurred by additional debt. The leverage-stimulated enhancement to
FFO growth is a strategy, which will cause a negative equity market reaction.
Our test also demonstrates an interesting relationship between the size of a REIT
asset base and its FFO yield. The model suggests an inverse relationship between these
two variables. The larger REITs enjoy significantly lower costs of equity than their
smaller counterparts. This finding supports Peter Linneman's strategy for REITs to grow
their asset size. As mentioned in Chapter 3, the idea that size can offset the impact of
additional leverage is proactive.
The model did not reflect any significant association between the form of REITs
(UPREITs vs conventional REITs) and required FFO yield.
CHAPTER 5
TEST 2: LEVERAGE AND COST OF CAPITAL
Introduction
David Shulman performed a test of M & M's two propositions in his research.
Shulman's test, however, involved very early REITs, the population of which was only a
fraction of its current size. This chapter, utilizing the data collected in the first test, will
perform a test of M & M's theories in a more modem context. This chapter will seek to
confirm the suggested invariance of weighted-average-cost-of-capital (WACC) under
different leverages. If M & M's propositions hold, a review of capital structure of the
sample REITs should yield relatively low variations in the WACC in the sample.
M & M's proposition I suggests firm value, the present value of future cash flows,
will not change with the changes of leverage. This argument will be sustained because
the required return to equity will increase with the increases of leverage ratios, as
suggested by Proposition II. The idea can be stated algebraically in the following
equation:
Re= Ra +-*(Ra -Rd)
E
Re: the required return to equity capital
Ra: the return on asset
D/E: the debt to equity ratio
Rd: the cost of debt capital
As more debt is added into the capital structure, the additional risk to the equity
holder is reflected in the increased required return, which offsets the impact on the
WACC implied by more less expensive debt. The net result is that the WACC remains
constant and the present value of the future cash flow of this firm will not change.
To test M & M's theory, it was necessary to obtain data of the cost of long-term
debt and to estimate the cost of long-term equity. A model constructed with the above
data can conduct a test of M & M's theory, which would supplement and update David
Shulman's work on a small number of early REITs.
The test involves obtaining liability detail on the balance sheet, estimating a longterm cost of equity capital, employing the actual REIT leverage ratios and calculating the
WACC for each REIT. These WACCs will then be compared. In Test I, we conclude that
some additional debt increases the required return on equity. However, this does not
necessarily imply that additional debt also increases the cost of capital, although equity is
part of capital. When the required return on equity is increased due to the additional debt,
the proportion of equity in capital is decreased correspondingly. As shown in the equation
on page 47, the weighted cost of capital (WACC) could still remain unchanged.
If M & M's theory is to be supported, we would not expect to find significant
variation in the WACC among a sample of REITs with various leverage ratios.
Statistically speaking, support for M & M will be indicated if the variance for the FFO
yield is found to be greater than those of the WACCs.
Methodology
To test the M & M propositions, a sub-sample from the REIT database is
required, one which has complete capital structure information. An accurate estimation
of the actual weight average cost of the long term debt is necessary. The Form 10-Q from
each REIT provides this data.
The detailed debt information about 47 REITs were obtained from Form 10-Q,
the quarterly report submitted to the Security Exchange Commission (SEC) by each
REIT. The Form 10-Q is available on the website (http//www.edgar-online.com). The
aforementioned 47 REITs were utilized as a new sample to test M & M's theories. Based
on the information from Form 1O-Q, the weighted average cost of long-term debt was
calculated for each REIT. In this process, we made an assumption that the cost of debt
was approximately equal to the coupon rate, since most of the debt was issued recently,
and interest rates have been relatively stable. The second half of the WACC calculation
involved estimation of the long-term cost of equity for each REIT. This process starts
from the regression model developed in Test I. The model in Test I provided an equation
for estimating FFO yield as a function of 7 variables: asset size, leverage, initial
formation as an UPREIT, apartment REITs, neighborhood shopping center REITs,
regional mall REITs and office/industrial REITs. In testing M & M' propositions, it was
necessary to standardize the FFO yield in the model. The standardized FFO yield was
solved by subtracting the sum of the products of each coefficient (except for
LEVERAGE) and its value from the given FFO yield. Standardization helped to isolate
the leverage and measure its impact on the FFO yield.
The FFO yield may be considered as the first year cost of equity capital for a
REIT. An estimate of the long term cost of equity capital is needed to provide a complete
test of M & M's theory. The long term cost of equity capital implies both dividend
income and price appreciation during a specified holding period. To add this additional
component of growth, the expected long-term inflation rate was employed, adding it to
the estimated standardized FFO yield.
The rationale for this procedure was that the REIT pay-out ratios tend to be
relatively high compared to traditional industrial firms because REITs are required by
law to pay out almost all (95 per cent) of their net income in exchange for their tax
exempt status. Even after considering the effects of the FFO adjustment (adding back
depreciation and gains or losses from sales), REIT payout ratios are higher than those of
most dividend-paying industrial firms (Marks & Hartung, 1996). The high payout ratios
imply that, over a longer period of time, REITs will tend to grow roughly at the long-term
inflation rate, at which the real estate values tend to grow. Following this logic, each
standardized FFO yield was increased by the expected inflation rate.
To estimate this rate, reference was made to the return difference between the tenyear US Government bond and the inflation-indexed US Government bond with
compatible maturity. This premium was 2.96% on April 27 1998 (Barron, 1998). This
extra premium in the ten-year, risk free investments approximates the anticipated
inflation rate over that holding period. For each of the sample REITs, a standardized FFO
yield is calculated to control for the effects of variables except LEVERAGE, and the
inflation adjustment of 2.96% was added to the standardized FFO yield to yield an
estimated cost of equity capital.
In the final step, each REIT's weighted cost of capital (WACC) was estimated by
using the actual leverage ratio and its weighted average cost of long term debt. Values of
debt cost, equity cost, and leverage ratio were plugged into the WACC formula:
WACC=
D
D+E
*CD
E
D+E
C
WACC: Weighted average cost of capital
D: total long term debt
E: total equity
D + E: total capitalization
CD:
cost of debt
CE:
cost of equity
After the WACC value is obtained for each REIT of the sample, it is loaded into an Excel
program, and descriptive statistics are calculated. Then the correlation between the
leverage ratio and the WACC is examined. See Exhibit 3.
Findings
The results of this analysis support M & M's theory. The REIT sample has
leverage variations ranging from 25.94% to 67.5%, averaging 47.3% with a standard
deviation of 1.5%. However, the derived WACC values range from 9.5% to 12.75%,
averaging 10.9% with a standard deviation of 0.12%. The correlation between the
leverage ratios and the WACC values is as low as 0.0379, which indicated that WACC
value is invariant to the leverage changes. Refer to the table and the plot for detailed
information.
_LEVERAGE WACC
Mean
47.3%
10.9%
Standard
Deviation
Range
1.56%
0.12%
41.56%
3.2%
Min.
25.9%
9.51%
Max.
67.5%
12.75%
LEVERAGE & WACC
0.18
0.16
0.14
0.12
*WACC
C.)
01
Linear (WACC)
........
0.1
0.08
0.06
0.04
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Leverage Ratio
Implications
The results provide evidence indicating that a change of leverage ratio does not
change the weighted cost of capital if all the other conditions remain the same. However,
this test is not trying to demonstrate absolute support of M & M theory. To provide such
a test, a set of benchmark data for a larger sample of REITs is required. This research
provided directional support of M & M theory. The distribution of the WACC values is
well clustered around the mean in this test, implying that the WACC is relatively constant
despite significant variations in the leverage ratios and the FFO yields.
These findings suggest that REIT managers, who are concerned about stock
prices, are keeping the debt level moderate. The result of the first research question
implied a penalty for the FFO yield for additions to the leverage. The result of the second
research question supporting M & M propositions that debt policy is irrelevant in
determining firm value if there is no corporate income tax.
CHAPTER 6
CONCLUSIONS
This thesis has described empirical research exploring the subject of leverage in
Real Estate Investment Trusts. Interest in this topic was initially stimulated by the
apparent contradiction between M & M theory and the REIT operation. M & M suggests
that, in the absence of corporate income tax, firms have no value-improving reason to
assume debt on the balance sheet. However, REITs clearly carry some leverage. This
research is an exploration of the relationships between REIT debt levels and other factors.
The research also provided an opportunity to test the M & M propositions, using modern
REIT data.
J. P. Morgan securities analyst Lee Schalop expressed his belief that during the
last five years, growth of REIT funds-from-operation (FFO) has been the product of a
combination of both secular and cyclical trends. The secular change is the shift from
private to public ownership of real estate. The profit potential from this shift is enormous
and includes the potential for scale economy and lower costs of capital. The cyclical
changes refer to the supply/demand imbalance in commercial real estate markets in the
last several years. This imbalance, the product of a deep real estate recession occurring
between 1989 and 1993 and a macro-economy recovery, has lent the real estate operators
the opportunity to raise rents and occupancy. Schalop concludes that the profit
opportunities continue to exist only as the product of this secular change; but the cyclical
recovery is over.
If Schalop is correct, the findings presented in this thesis have some significance
for REIT operators. This research began with the collection of data from a sample of
equity REITs. The data collected was utilized as the basis for the creation of an empirical
model designed to test the association between a REIT's financial leverage and its cost of
equity capital, or its FFO yield. The relationship found here supported the M & M theory
that an increased required return to equity is associated with increasing the amount of
debt.
The second research question explored in this thesis involved a test of M & M
theory. In this test, a sample of REITs was utilized to test the idea of value-invariant to
debt changes. The research result clearly indicated the directional support; the
distribution of weighted average cost of capital (WACCs) was clustered around the mean
of the sample REITs. This narrow distribution of WACCs was in a contrast with the
broad distribution of leverage ratios of the same sample REITs.
The findings of this research seem to suggest to the REIT operators that a strategy
of increasing leverage as a mechanism to generate higher property return and better FFO
growth will not be viable. This thesis has attempted to find the consistency between
theory and practice. The theory will provide useful guidance to the REIT operators with
respect to an important question. The theory suggests that the strategy of increasing
leverage to stimulate FFO growth would encounter resistance on Wall Street, indicated
by a further decline in multiples.
If the REIT operators are not convinced by the theoretical arguments for low
leverage, additional rationale is available. Mike Kirby at Green Street advisors, a
California-based, buy-side research firm, suggests that most real estate operators would
think that "M & M refers to the popular candy." Kirby presents and defends the M & M
propositions in a discussion about REITs. He points out that there are times when the
equity market is receptive to REITs and times when it is closed. Kirby suggests that low
leverage may be a mechanism by which to maintain the ability to raise capital. With
mandatory high payout ratios, a high-leverage REIT has relatively few alternatives to
raise capital when the equity market is not available. Under these conditions, it is advised
to maintain low leverage so as to keep the flexibility to take advantage of opportunities
for which significant amounts of capital are required. This strategy is known as "keeping
dry powder" and reflects the desire for financial flexibility. Thus, relatively low leverage
may reflect a desire for financial flexibility.
Suggestions for Future Research
The research in this thesis is generally supportive of the work done by M & M on
value invariance to leverage change. David Shulman attempted to test M & M theory in
his work in 1975. A formal update of Shulman's work would be valuable because his
work involved a small population of early REITs, and that new research would use a
greater number of modern-era REITs.
Another area of research that might prove noteworthy is that of an exploration of
the question of the relationship between leverage and the degree of institutional
ownership of the issued shares of a REIT. Then, it might be seen that regressing FFO
yield on this independent variable was circular because of the confusion in identifying
causality. High institutional ownership associated with low leverage could be derived
from either an institutional bias towards low-levered REITs in their portfolios, or from
REIT management's keeping leverage low to meet the institutions' bias toward low
leverage. An interesting research question might involve an assessment of those factors
that seem to attract institutional investors to own the shares of particular REITs. Level of
debt could be one factor, as might property-type specialty, geographic diversity and asset
size. The power of institutional ownership to move share prices suggests that this might
be an important area for further study.
Reference was made in this research as to the inverse relationship between firm
size and leverage. The idea that an optimum combination of these two variables might
exist would also merit further study. Theoretically, a large REIT might have a lower
WACC than its smaller counterpart, while both have the same leverage. M & M theory of
value invariance to leverage does not mention the effects of size. The idea of an optimum
combination of the two variables is implicit in the data and further research would be
helpful in determining whether another variation on the optimal capital structure might be
present.
Another question that might deserve additional research would be the effect of
UPREIT form on the REIT operation. During the period between 1993 and 1998, most
markets, particularly those in which institutional investors had exhibited interests, had
been in a recovery stage. They have now gained equilibrium and will begin the transition
into over-capacity and decline. During the equilibrium and saturation phases, a REIT's
portfolio management practices will become critical in FFO growth. These practices
include selective disposition of assets in the portfolio. These sales will, in many cases,
trigger the tax payment for the holders of Operating Partnership Units. The interesting
question will involve how REIT management groups reconcile the OP unit holders'
reluctance to sell with appropriate portfolio management strategy. REIT regulation rules
allow active participation of outside board members. But there are maybe sometimes
when the conflict is intense due to the size of triggered tax. Another study, focused on the
UPREITs, which examines the property disposition over the years, would reveal how the
REIT operators deal with this challenge during downside cycle of real estate market.
Finally, additional research work exploring the relationship between FFO yield
and property-type specialty would be fruitful. The result in Test I did not reveal any
property-type effects on FFO yield. A different approach might give more satisfying
results. As mentioned in the expectation before Test I, property type should matter
because lenders are concerned about the leases associated with the properties to secure
their loans. When leases are significantly longer, like those of regional malls with long
term anchor tenants, the premium required for the equity return should be less. More
research in this area would be useful for operators and lenders.
As mentioned above, this thesis attempted to reconcile theory with practice. The
apparent paradox between M & M theory and the practices in REIT capital structure led
to this research. After analyzing the results of the two tests, we can see that theory and
practice do merge. Public REITs, without corporate-level income tax, do carry some debt.
But they only carry a fraction of what their private real estate counterparts do.
Their disinclination to high leverage reflects the discipline of the watchful equity market.
BIBLIOGRAPHY
Altman, E, A further empirical investigation of the bankruptcy cost question. Journal of
Finance, 1067-1089, 1984
Baird, W. Blake, Going Public, The Formation of a REIT, Real Estate Investment
Trusts Capital Structure, Analysis, and Strategy, McGraw-Hill, New York, 1998,
Richard T. Garrigan, John F. C. Parsons, Editors.
Barron's, The Dow Jones Business and Financial Weekly. Market Week. Chicopee,
Mass., April, 1998.
Baxter, N., Leverage, risk of ruin and the cost of capital., Journal of Finance, 396-403,
September 1967
Black, F., & Scholes, M., The pricing of options and corporate liabilities. Journal of
Political Economy, 637-654, May 1973
Brealey, A & Myers, S, Principles of Corporate Finance (4 th Edition), New York:
McGraw Hill. 1991
Clauretie, T. & Sirmans, G., Real Estate Finance: Theory and practice (2nd Edition).
Upper Saddle River, New Jersey: Prentice Hall
Frequently Asked Questions about REITS (1998), Washington D.C. A publication of
the National Association of Real Estate Trust
Han, J. & Liang, Y. (1995), The Historical Performance of Real Estate Investment
Trusts. The Journal of Real Estate Research, 10 (3), 235-262
Jensen, M., & Meckling, W. (1996). Theory of the Firm: Managerial Behavior, Agency
Costs, and Ownership Structure., Journal of Financial Economics, 305-360.
Kim, E. H., Lewellen, W. G., & McConnell, J., (1979). Financial Leverage Clienteles:
Theory and Evidence. Journal of Financial Economics, 7 (1), 83-109
Kirby, Mike (1998), Why don't REITs Borrow More Aggressively? Property 98, A
Publication of Dow Jones Financial Publishing Corp. Shrrewsbury, N.J. 72
Lee, Sohan., (1995), Wall Street vs. Main Street: Valuing Securitized Assets, University
of Michigan
Linneman, Peter, (1996), The Forces Changing the Real Estate Industry Forever.
Wharton Real Estate Center. University of Pennsylvania
Maris, B. A., & Elayan, F. A., (1996), Capital Structure and the Cost of Capital for
untaxed Firms: the Case of REITs. AREUEA Journal, 18 (1), 22-3 9
Comparative Evaluation of REITs (1999), Merrill Lynch Report
Miller, J. D., (1997), Capital Flows. Emerging Trends in Real Estate: 1998. A
Publication of Equitable Real Estate Investment Management, Inc. and Real Estate
Research Corp. 14-23
Miller, M. H., & Rock, K., (1985), Dividend Policy Under Asymmetric Information.
Journal of Finance, 40, 1031-1051
Modigliani, F., & Miller, (1958), The Cost of Capital, Corporation Finance, and the
Theory of Investment. American Economic Review, 48 (3), 261-297
Myers, S. & Majluf, N, (1984), Corporate Financing and Investment Decisions When
Firms Have Information That Investors Do Not Have. Journal of Financial Economics,
13, 187-221
REIT Handbook, (1996), Washington D.C. A Publication of the National Association
of Real Estate Investment Trusts.
Rosen, K. T., (1995), REITs: Stocks, Bonds, or Real Estate? Fisher Center for Real
Estate and Urban Economics. Berkeley, California
Ross, S., (1977), The Determination of Financial Structure: The Incentive Signaling
Approach. Bell Journal of Economics, 8, 23-40
Schalop, L., (1998), Industry Update: NAREIT Conference Notes: Denial is not a River
in Egypt. J. P. Morgan Securities, Inc.
Sharp, W.F., (1964), Capital Asset Prices: A Theory of Market Equilibrium under
Conditions of Risk. Journal of Finance, 19, 425-442.
Shulman, David. (1975), Leverage, Dividends and Rates of Return on Equity Real
Estate Investment Trusts. University of California at Los Angeles
Titman, S. (1984) The Effect of Capital Structure on a Firm's Liquidation Decision
Journal of Financial Economics. 13
Williams, J.B. (1938), The Theory of Investment value. Harvard University Press
http://nareit.com
http://finance.yahoo.com
http://www.edogar-online.com
GLOSSARY OF TERMS
Adjusted Funds From Operations (AFFO)
This term refers to a computation made by analysts and investors to measure a real
estate company's cash that is available for distribution to shareholders. AFFO is
usually calculated by subtracting from Funds from Operations (FFO) both (1)
normalized recurring expenditures that are capitalized by the REIT and then
amortized, but which are necessary to maintain a REIT's properties and its revenue
stream (e.g., new carpeting and drapes in apartment units, leasing expenses and tenant
improvement allowances) and (2) "straight-lining" of rents. This calculation also is
called Cash Available for Distribution (CAD) or Funds Available for Distribution
(FAD)
Capitalization Rate
The capitalization rate (or "cap" rate) for a property is determined by dividing the
property's net operating income by its purchase price. Generally, high cap rates
indicate higher returns and greater perceived risk.
Cash (or Funds) Available for Distribution
Cash (or Funds) available for distribution (CAD or FAD) is another measure of a
REIT's ability to generate cash and to distribute dividends to its shareholders. In
addition to subtracting from FFO normalized recurring real estate-related
expenditures and other items to obtain AFFO, CAD (or FAD) is usually derived by
also subtracting nonrecurring expenditures.
Cash Flow
With reference to a property (or group of properties), the owner's rental revenues
from the property less all property operating expenses. The term ignores depreciation
and amortization expenses, as well as interest on loans incurred to finance the
property. Sometimes referred to as "EBITDA" (Earnings before interest, taxes,
depreciation and amortization).
Cost of Capital
The cost to a company, such as a REIT, of raising capital in the form of equity
(common or preferred stock) or debt. The cost of equity capital generally is
considered to include both the dividend rate as well as the expected equity growth
either by higher dividends or growth in stock prices. The cost of debt capital is merely
the interest expense on the debt incurred.
DownREIT
A DownREIT is structured much like an UPREIT, but the REIT owns and operates
properties other than its interest in a controlled partnership that owns and operates
separate properties.
EBITDA
Earnings before interest, taxes, depreciation and amortization.
Equitization
The process by which the economic benefits of ownership of a tangible asset, such as
real estate, are divided among numerous investors and represented in the form of
publicly-traded securities.
Equity Market Cap
The market value of all outstanding common stock of a company.
Equity REIT
A REIT which owns, or has an "equity interest" in, rental real estate (rather than
making loans secured by real estate collateral).
Funds From Operations (FFO)
The most commonly accepted and reported measure of REIT operating performance.
Equal to a REIT's net income, excluding gains or losses from sales of property or debt
restructuring, and adding back real estate depreciation.
Hybrid REIT
A REIT that combines the investment strategies of both equity REITs and mortgage
REITs.
Implied Equity Market Cap
The market value of all outstanding common stock of a company plus the value of all
UPREIT partnership units as if they were converted into the REIT's stock. It excludes
convertible preferred stock, convertible debentures and warrants even though these
securities have similar conversion features.
Leverage
Debt.
Mortgage REIT
A REIT that makes or owns loans and other obligations that are secured by real estate
collateral.
Net Asset Value (NAV)
The net "market value" of all a company's assets, including but not limited to its
properties, after subtracting all its liabilities and obligations.
Positive Spread Investing (PSI)
The ability to raise funds (both equity and debt) at a cost significantly less than the
initial returns that can be obtained on real estate transactions.
Real Estate Investment Trust Act of 1960
The federal law that authorized REITs. Its purpose was to allow small investors to
pool their investments in real estate in order to get the same benefits as might be
obtained by direct ownership, while also diversifying their risks and obtaining
professional management.
Real Estate Investment Trust (REIT)
A REIT is essentially a corporation or business trust that combines the capital of
many investors to acquire or provide financing for all forms of real estate. A REIT
generally is not required to pay corporate income tax if it distributes at least 95% of
its taxable income to shareholders each year.
Securitization
Securitization is the process of financing a pool of similar but unrelated financial
assets (usually loans or other debt instruments) by issuing to investors security
interests representing claims against the cash flow and other economic benefits
generated by the pool of assets.
Straightlining
Real estate companies such as REITs "straight line" rents because generally accepted
accounting principles require it. Straight lining averages the tenant's rent payments
over the life of the lease.
Tax Reform Act of 1986
Federal law that substantially altered the real estate investment landscape by
permitting REITs not only to own, but also to operate and manage, most types of
income-producing commercial properties. It also stopped real estate "tax shelters" that
had attracted capital from investors based on the amount of losses that could be
created.
Total Market Cap
The total market value of a REIT's (or other company's) outstanding common stock
and indebtedness.
Total Return
A stock's dividend income plus capital appreciation, before taxes and commissions.
UPREIT
In the typical UPREIT, the partners of the Existing Partnerships and a newly-formed
REIT become partners in a new partnership termed the Operating Partnership. For
their respective interests in the Operating Partnership ("Units"), the partners
contribute the properties from the Existing Partnership and the REIT contributes the
cash proceeds from its public offering. The REIT typically is the general partner and
the majority owner of the Operating Partnership Units.
After a period of time (often one year), the partners may enjoy the same liquidity of
the REIT shareholders by tendering their Units for either cash or REIT shares (at the
option of the REIT or Operating Partnership). This conversion may result in the
partners incurring the tax deferred at the UPREIT's formation. The Unitholders may
tender their Units over a period of time, thereby spreading out such tax. In addition,
when a partner holds the Units until death, the estate tax rules operate in a such a way
as to provide that the beneficiaries may tender the Units for cash or REIT shares
without paying income taxes.
Exhibit 1
Comparative Valuation of REITs
from Merrill Lynch
Table 4: Dividend Yield, FFO Yields, Payout Ratios, and Return of Capital
Table 5: Capital Structure Statistics
Merrill Lyrnch
Comparative Valuation Of REITs - 9 September 1999
Table 4: Dividend Yield, FFO Yields, Payout Ratios, and Return of Capital (1)
Apartment REITs
Amli Residential Properties
Pnce
9/3/99
52 Week
High
Low
$20.81 $23.31 $18.44
Current Dividend
Payout
1998 (2) Ratio (3)
Adjusted
Adjusted
Payout 1999E FF0 1999E FFO
Ratio (4)
Yield (5)
Yield (6)
Yield
8.6%
1999-2003
N/A
7%
70%
81%
12.3%
10.7%
$30.00
$17.88
$9.00
$30.50
$21.50
$29.00 $23.88
$35.75 $28.13
$48.38 $3469
$35.50 $25.69
$27.00 $21.69
$27.25 $20.88
$42.13 $35.00
$20.63 $16.00
$12.38
$9.06
$23.94 $16.00
$2.50
$1.48
$1.86
$2.04
$1.56
$2.08
$2.14
$2.84
$2.20
$2.04
$2.30
$2.80
$1.67
$1.06
$1.93
6.2%
7.0%
20.1%
5.7%
6.2%
7.5%
6.2%
6.4%
6.3%
8.5%
10.2%
6.9%
8.4%
9.4%
10.3%
8.4%
10.0%
7.0%
N/A
9.0%
8.4%
5.1%
N/A
8.1%
8.7%
N/A
4.8%
9.0%
6.3%
3.5%
4.5%
7.0%
59%
10%
48%
19%
11%
0%
20%
0%
0%
32%
25%
13%
25%
12%
50%
61%
74%
132%
63%
67%
65%
66%
64%
68%
71%
80%
76%
76%
75%
78%
74%
68%
77%
141%
66%
71%
72%
73%
72%
73%
81%
101%
81%
85%
91%
97%
83%
10.1%
9.5%
15.2%
9.1%
9.3%
11.6%
9.3%
10.1%
9.3%
11.9%
12.7%
9.1%
11.1%
12.5%
13.3%
11.0%
9.1%
9.1%
14.3%
8.7%
8.8%
10.5%
8.4%
8.9%
8.7%
10.5%
10.1%
8.5%
9.9%
10.4%
10.6%
9.8%
$9.16
$13.63
$20.00
$8.63
$18.63
$33.75
$11.25
$17.81
$18.75
$38.25
$9.94
$1.44
$1.40
$1.76
$0.94
$1.58
$2.40
$1.92
$1.63
$1.84
$2.84
$1.12
12.9%
9.4%
7.4%
9.7%
8.0%
6.5%
15.7%
8.7%
8.8%
7.1%
10.2%
9.5%
N/A
7.6%
3.7%
N/A
9.7%
8.2%
4.4%
6.4%
N/A
8.3%
0.0%
6.0%
0%
96%
100%
0%
8%
75%
78%
85%
67%
73%
102%
12.9%
12.8%
8.6%
11.6%
10.0%
9.0%
15.3%
11.0%
74%
77%
79%
84%
13.5%
14.0%
9.9%
12.4%
10.4%
9.8%
17.1%
11.9%
11.8%
9.8%
12.5%
12.1%
7.7%
11.6%
6.0%
12.6%
9.7%
7.9%
10.5%
5.1%
7.9%
7.4%
7.3%
9.6%
8.6%
N/A
N/A
9.3%
N/A
4.8%
4.1%
5.5%
6.5%
4.0%
3.8%
4.9%
4.8%
5.3%
18%
100%
65%
65%
51%
73%
74%
72%
83%
40%
66%
75%
68%
84%
68.0%
77%
79%
61%
85%
82%
80%
95%
45%
77%
93%
77%
99%
79.1%
11.9%
18.0%
11.7%
17.3%
13.1%
10.9%
12.7%
12.8%
12.1%
9.9%
10.6%
11.5%
12.7%
10.0%
14.7%
9.8%
15.0%
11.7%
9.9%
11.0%
11.4%
10.3%
8.0%
9.4%
9.8%
10.9%
8.6%
15.7%
9.9%
11.4%
8.6%
N/A
3.7%
6.2%
0%
0%
0%
70%
78%
69%
72%
75%
79%
74%
76%
12.3%
20.3%
14.3%
15.6%
11.5%
19.8%
13.4%
14.9%
18.7%
10.9%
20.7%
11.6%
15.5%
2.3%
5.3%
Rvw
4.2%
3.9%
0%
0%
Rvw
0%
84%
87%
Rvw
83%
85%
90%
88%
Rwi
83%
87%
22.1%
12.6%
Rvw
13.9%
16.2%
20.8%
12.4%
Rv
13.9%
15.7%
$44.06
$23.50
$18.94
$37.00
$26.38
Neighborhood Shopping Center REITs
Center Trust
Deelopers Diversified'
Federal Realty *
IRT Property Co.
JON Realty Corporation*
Kinco Realty *
Kranzco Realty Trust'
New Plan Excel Realty'
Regency Realty Corp
Weingarten*
Western Investment
Neighborhood Shopping Center Average
$11.13
$14.88
$23.94
$9.69
$19.69
$36.69
$12.25
$18.75
$21.00
$4.00
$11.00
$13.50
$19.63
$24.88
$10.81
$23.38
$40.75
$17.13
$23.75
$24.00
$46.88
$13.44
Regional Mall REITs
C8L & Associates Properties
Crown American Realty Trust
General Growth Properties, Inc.*
Glimcher Reatty Trust
JP Realty
Macench'
Mills Corporation
Rouse Company'
Simon Property Group'
Taubman Centers, Inc.'
Urban Shopping Centers'
Westfield America. Inc.*
Regional Mall Average
$25.31
$7.06
$32.88
$15.19
$19.25
$24.50
$19.13
$23.44
$25.44
$12.94
$30.88
$15.06
$21.88
$6.13
$31.13
$13.31
$17.44
$21.25
$16.13
$21.13
$23.88
$11.50
$27.56
$14.38
$1.95
$0.82
$1.96
$1.92
$1.86
$1.94
Factory Outlet REITs
Chelsea GCA Realty*
Prime Retail
Tanger Factory Outlet *
Factory Outlet Average
$33.44 $39.38 $27.88
$7.50 $11.19
$7.38
$24.44 $26.75 $18.69
$2.88
Health Care REITs
ElderTrust'
Health Care Property Investors *
Meditrust'
Nationwide Health Properties'
Health Care Average
$7.81 $14.75
$7.50
$25.63 $35.88 $24.06
$8.88 $20.25 $8.69
$15.56 $23.25 $14.94
$1.46
$2.80
$1.84
$1.80
Note: Footnotes can be found at the end of the report.
Return of
Capital
Dividend
$1.80
$40.25
$21.13
$9.25
$35.56
$25.06
$27.63
$34.69
$44.19
$34.69
$23.94
$22.56
$40.63
$19.81
$11.25
$18.69
Apartment Investment & Management Co.
Archstone'
Associated Estates Realty
Avalon Bay Communities*
BRE Properties inc*
Camden Property Trust
Charles E. Smith Residential
Equity Residential Properties Trust*
Essex Property Trust *
Gables Residential Trust*
Mid-America Apartment Comm.'
Post Properties, Inc. *
Summit Properties Inc.*
United Dominion Realty Trust'
Walden Residential*
Apartment Average
Dividend
Growth Rate
$27.00
$9.50
$38.63
$18.63
$22.63
$28.63
$25.00
$28.88
$31.00
$14.19
$33.44
$18.25
$2.01
$1.20
$0.96
$2.24
$1.45
$1.18
$2.42
0%
40%
27%
2%
0%
17%
40%
0%
0%
10%
84%
11.4%
9.0%
10.8%
11.1%
Merrill Lynch
Comparative Valuation Of REITs - 9 September 1999
Table 4: Dividend Yield, FFO Yields, Payout Ratios, and Return of Capital (1)(condnued)
Officellndustrial REITs
AMB Property Corporation*
Arden Realty
Boston Properties'
Brandywine Realty Trust*
Brookfield Properties Corporation*
CarrAmenca Reelty"
Centerpont Properties
Cornerstone Properties*
Crescent Real Estate'
Duke Realty investments*
Equity Office*
First Industrial Realty Trust*
Highwoods Properties*
Kilroy Realty Corp.*
Liberty Property Trust
Mack-Cali Realty Corporation'
Prentiss Properties
ProLogis*
Reckson Associates Realty'
Spieker Properties*
TrizecHahn*
Office/industrial Average
Self-Storage REJTs
Shurgard Storage Centers*
Sovran*
Public Storage, Inc.'
Storage USA*
Self-Storage Average
Manufactured Home REITs
Chateau Communities'
Manufactured Home'
Sun Communities, Inc.*
Manufactured Home Average
Hotel REITs
Feicor Lodging Trust*
Hospitality Property Trust*
Host Mariot*r
Meristar Hospitality Corporation*
Starwood Hotels & Resorts *
Hotel REIT Average
Mixed Use and Miscellaneous REITs
Colonial Properties Trust*
Cousins Properties*
Forest City Enterpnses'
Franchise Finance Corp. of America'
HRPT Properties Trust*
National Golf Properties*
Pennsylvania REIT*
Realty Income'
US Restaurant Properties'
Vomado *
Washington REIT*
Mixed Use and Miscell. REITs Average
Price
9/3/99
$21.13
$23.25
$33.13
$17.75
$12.44
$22.63
$33.25
$15.50
S20.19
$21.88
$25.31
$25.25
$24.25
$23.00
$24.00
S28.06
$22.38
$19.50
S20.50
$37.94
$19.94
Didend
52 Week
Current Dividend Growth Rate
High
Low
Dividend
Yield 1999-2003
$26.00 $20.31
$1.40
6.6%
5.9%
$27.19 $19.75
$1.78
7.7%
N/A
$37.50 $23.44
$1.70
5.1%
8.8%
$20.44 $15.75
$1.56
8.8%
5.6%
$14.63 $11.50
$0.25
2.0%
16.3%
$26.75 $19.00
$1.85
8.2%
9.2%
$38.56 $30.56
$1.90
5.7%
N/A
$17.00 $13.25
$1.20
7.7%
7.0%
$28.56 $19.50
$2.20
10.9%
5.6%
$24.38 $19.94
$1.56
9.0%
7.1%
$29.38 S20.19
$1.48
5.8%
9.7%
6.7%
$2.40
$28.25 $21.63
9.5%
$29.13 $22.13
S2.22
8.0%
9.2%
7.0%
S26.50 $18.50
$1.68
7.3%
N/A
$25.94 S20.13
$1.80
7.5%
S2.20
8.0%
7.8%
$33.63 $26.13
$1.76
N/A
S24.38 $18.13
7.9%
$24.38 $18.63
$1.30
6.7%
8.5%
$26.75 $12.06
$1.49
7.3%
8.9%
$41.56 $31.00
6.4%
9.3%
$2.44
8.5%
SO.35
1.8%
$22.81 $16.13
7.0%
8.4%
$25.31 S27.88
$22.94 $27.06
S26.13 $29.38
S29.56 $35.38
Return of
Capital
Payout
1998(2) Ratio (3)
0%
67%
0%
68%
0%
59%
N/A
64%
0%
15%
8%
65%
10%
65%
25%
76%
N/A
76%
4%
73%
4%
59%
35%
73%
12%
64%
0%
67%
0%
63%
0%
66%
19%
62%
0%
65%
0%
66%
5%
72%
0%
17%
62%
$23.75
$21.25
S22.88
S27.75
$2.00
S2.24
SO.88
$2.68
7.9%
9.8%
3.4%
9.1%
7.5%
9.4%
5.4%
9.0%
8.5%
8.1%
7%
0%
0%
0%
$29.25
$31.00 $25.88
S24.44 $27.00 $21.81
$35.44 $37.13 $29.88
$1.94
$1.55
$2.04
6.6%
6.3%
5.8%
6.2%
5.0%
5.0%
5.0%
5.0%
$18.63 S26.13 $16.69
$25.50 $31.50 $23.81
$9.63 $15.44
$9.25
$16.75 S24.31 $12.25
$24.13 $37.75 $18.75
$2.20
11.8%
10.8%
8.7%
12.1%
2.5%
9.2%
$26.94
$2.32
8.6%
4.0%
0.8%
8.9%
11.5%
7.9%
9.5%
9.0%
11.4%
5.2%
7.3%
7.6%
$35.88
S25.31
$22.13
$13.19
$22.31
$19.81
$23.38
$16.00
$33.75
$16.00
Note: Footnotes can be found at the end of the report.
$29.00
$38.25
$28.75
$27.81
$17.50
S30.00
S21.69
S25.94
$25.81
S40.00
$18.75
S24.44
$24.19
$17.75
$20.13
$12.88
$21.19
$18.56
$20.31
$15.69
$26.00
$15.06
$2.76
$0.84
$2.02
$0.60
$1.44
SO.20
$1.96
$1.52
$1.76
$1.88
S2.10
$1.82
$1.76
$1.17
73%
76%
Adjusted
Adjusted
Payout 1999E FFO 1999E FFO
Ratio (4)
Yield (5)
Yield (6)
76%
9.9%
8.7%
76%
11.3%
10.1%
68%
8.7%
7.6%
81%
13.7%
10.9%
21%
13.7%
9.4%
78%
12.6%
10.5%
69%
8.8%
8.3%
86%
10.2%
9.0%
88%
14.4%
12.4%
81%
9.7%
8.8%
10.0%
68%
8.6%
83%
11.4%
13.1%
76%
14.4%
12.0%
75%
9.7%
10.8%
67%
11.8%
11.2%
78%
10.0%
11.8%
70%
11.3%
12.8%
75%
10.3%
8.9%
75%
11.1%
9.7%
82%
8.9%
7.9%
23%
7.8%
10.3%
71%
9.7%
11.3%
33%
81%
66%
78%
82%
35%
85%
70%
10.8%
12.9%
10.1%
11.2%
11.3%
10.1%
11.9%
9.7%
10.7%
8.5%
35%
21%
25%
79%
70%
68%
72%
86%
81%
74%
80%
8.4%
9.0%
8.5%
8.6%
7.7%
7.8%
7.8%
7.8%
0.0%
3.7%
N/A
N/A
14.0%
5.9%
N/A
25%
N/A
N/A
0%
58%
72%
47%
52%
16%
49%
81%
90%
81%
79%
27%
72%
20.5%
15.0%
18.5%
23.2%
15.7%
18.6%
14.5%
12.1%
10.8%
15.2%
9.2%
12.4%
7.0%
12.4%
14.7%
5.2%
2.5%
8.9%
3.8%
4.1%
7.0%
8.0%
4.0%
7.1%
26%
0%
N/A
8%
6%
0%
0%
6%
36%
17%
2%
72%
59%
5%
76%
84%
63%
72%
85%
76%
57%
78%
66%
79%
61%
6%
77%
89%
65%
89%
86%
79%
73%
90%
72%
12.0%
6.8%
16.9%
11.7%
13.8%
12.5%
13.2%
10.6%
15.0%
9.2%
9.4%
11.9%
10.9%
6.5%
14.1%
11.4%
12.9%
12.1%
10.6%
10.5%
14.4%
7.1%
8.1%
10.8%
Merrill Lynch
Comparative Valuation Of REITs - 9 September 1999
Table 5: Capital Structure Statistics
($millions)
Apartment REITs
Amli Residential Properties
Apartment Investment & Management Co.'
Archstone*
Associated Estates Reatty
'Awton Bay Communities*
,iRE Properties Inc*
Camden Property Trust
Charles E. Smith Residential
Equity Residential Properties Trust
Essex Property Trust *
Gables Residential Trust *
Mid-America Apartment Comm.*
Post Properties, Inc. *
Summit Properties Inc.
United Dominion Reatty Trust
Walden Residential *
Apartment Average
Neighborhood Shopping Center REITs
Center Trust
Developers Diversified*
Federal Realty *
IRT Property Co.
JDN Realty Corporation
Kimco Reatty *
Kranzco Realty Trust
New Plan Excel Realty
Regency Reatty Corp
Weingarten *
Western Investment
Neighborhood Shopping Center Average
Regional Mail REtTs
CBL & Associates Properties
Crown American Reatty Trust
General Growth Properties, Inc.*
Glicher Realty Trust
JP Reaty
Macerich
Mills Corporation*
Rouse Company*
Simon Property Group
Taubman Centers, Inc. *
Urban Shopping Centers
Westfield Arnerica, Inc.*
Regional Mail Average
Factory Outlet REITs
Chelsea GCA Realty
Prime Retail
Tanger Factory Outlet
Factory Outlet Average
Health Care REITs
EklerTrustHeafth Care Property investors
Meditrust'
Nationwide Heatth Properties*
Health Care Average
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
UPREIT
Structure
Yes
Yes
No
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Shares
OP units
24.4
75.5
148.4
22.6
65.2
47.8
47.2
36.0
147.3
20.5
32.5
21.9
43.7
32.5
124.8
31.6
Floating Floating Dt/
Total
Total Non-conv.
Debt Pref. Stock Mkt Cap () Rate Debt Dt+Pref. Stk
42.2%
$947.8
$185.3
$0.0
$439.2
1.5%
$31.7
$518.5 $5,124.5
$3,037.3 $1,568.7
.0%
$5,621.5
'
$155.D
$3,135.2 $2,331.3
40.2%
$816.3
$244.3
$209.0
$551.1
-- $56.3
$57.3
2.8%
$458.1
$4.379.5
$2,319.1 $1,602.3
34.6%
$50.0 $1,968.9 $267.1
$1,197.5
$721.4
$39.0
3.2%
$1,305.1 S1,127.5
$100.0
$2,532.5
$1,249.5
$863.5
$50.0
$2,163.0
$0.0
0.0%
8.4%
$459.0
$775.5
$11,991.7
$6,508.2 $4,708.0
12.6%
$384.8
$105.0
$1,199.4
$61.6
$709.6
$169.7
17.6%
$1,744.4
$778.5
$796.4
$169.5
$154.3
16.9%
$1,408.9
$741.5
$173.5
$493.9
$350.5
34.4%
11,774.9
58P9.9
$150.0
$2,794.8
$94.0
12.7%.
$1,380.4
$6573
$85.0
$643.0
$3,833.3
$388.9
16.0%
$255.0
$1,403.8 $2,174.5
18.2%
$1,590.6
$181.9
$150.0
$591.0
$849.6
16.3%
-~'
tt3,093.6
Equity
Mkt Cap
$508.7
Floating Det/ Cash On
Equity Mk Cap
Hand
$4.5
36.4%
$105.9
1.0%
0.0%
$29.5
116.9%
$20.2
2.5%
$18.6
$14.7
22.3%
$9.7
3.0%
0.0%
$5.0
7.1%
$95.5
8.7%
$17.5
21.8%
$22.9
31.2%
$14.2
19.7%
$31.0
14.6%
$25.4
27.7%
$21.9
$7.1
30.8%
21.5%
Yes
No
Yes
No
No
No
No
No
No
No
No
39.8
66.0
45.3
36.1
33.2
60.6
12.3
101.1
63.3
27.1
19.6
$472.9
$443.1
$981.1 $1,457.0
$749.2
$1,084.3
$349.9 $259.1
$503.4
$654.5
$2,223.1 $1,449.7
$355.6
$150.3
$801.5
$1,895.6
$987.6
$1,328.5
$1,083.1
$489.3
$220.5
$215.9
$946.3
$0.0
$303.8
$0.0
$0.0
$50.0
$225.0
$45.0
$75.0
$80.0
$280.0
$0.0
$916.0
$2,741.9
$1,833.5
$609.1
$1,207.9
$3.897.8
$550.9
$2,772.1
$2,396.1
$1,852.4
$436.4
$1,746.8
$181.3
$210.0
$291.9
$8.0
$0.0
$92.6
$49.1
$0.0
$216.2
$43.6
$85.7
38.3%
11.9%
39.0%
3.1%
0.0%
5.5%
12.3%
0.0%
20.2%
5.7%
38.9%
15.9%
40.9%
21.4%
26.9%
2.3%
0.0%
4.2%
32.7%
0.0%
16.3%
4.0%
39.7%
17.1%
$6.8
$2.8
$12.5
$1.2
$0.0
$40.8
$2.9
$34.6
$32.4
$3.9
$1.5
(21)
(22)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
36.5
36.2
59.9
24.5
21.3
55.7
40.0
82.8
252.5
84.4
31.7
99.1
$922.9
$255.4
$1,970.4
$372.2
$410.4
$1,364.4
$764.7
$2,166.3
$6,422.4
$1,092.5
$978.2
$1,492.2
$1,517.7
$1,204.9
$658.7
$3,486.7
$623.1
$469.3
$1,981.9
$1,022.4
$3,461.3
$9,525.0
$1,274.9
$1,054.9
$2,722.5
$71.9
$125.0
$337.5
$0.0
$0.0
$0.0
$0.0
$137.0
$350.0
$200.0
$0.0
$0.0
$2,199.7
$1,039.1
$5,794.6
$995.3
$879.7
$3,346.3
$1,787.1
$5,764.6
$16,297.4
$2,567.4
$2,033.1
$4,214.7
$3,909.9
$495.5
$60.1
$1,292.7
$155.8
$88.1
$289.0
$117.7
$0.0
$1,927.6
$232.1
$50.2
38.8%
7.7%
33.8%
25.0%
18.8%
14.6%
11.5%
0.0%
19.5%
15.7%
4.8%
9.9%
16.7%
53.7%
23.5%
65.6%
41.9%
21.5%
21.2%
15.4%
0.0%
30.0%
21.2%
5.1%
18.1%
26.4%
$6.8
$27.7
$14.5
$7.6
$9.0
$23.9
$4.1
$29.5
$145.0
$12.5
$2.2
$27.6
(23)
(24)
Yes
Yes
Yes
20.4
67.8
11.8
$379.6
$681.1
$508.3 $1,211.6
$287.2 $305.2
$492.2
$50.0
$57.5
$0.0
$1,110.7
$1,777.4
$592.4
$1,160.2
$70.0
$0.0
$79.4
16.3%
0.0%
26.0%
141%
10.3%
0.0%
27.6%
12.6%
$17.9
$5.8
$0.2
(25)
(26)
(27)
No
No
No
No
7.7
34.4
Rvw
46.2
$60.4
$881.6
Rvw
$719.2
$553.7
$267.8
$679.9
Rvw
$735.1
$0.0
$187.8
Rvw
$100.0
$328.1
$1,749.3
Row
$1,554.3
$1,210.6
$95.8
$100.5
Rwv
$61.5
35.8%
11.6%
Rvw
7.4%
18.2%
158.7%
11.4%
Rvw
8.6%
59.6%
$9.4
$4.3
Rv'
$14.4
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
Note: Footnotes can be found at the end of the report.
$269.5
Merrill Lynch
Comparative Valuation Of REITs - 9 September 1999
Table 5: Capital Structure Statistics (continued)
($millions)
OceflndustrialREITs
AMB Property Caprabkn*
Arden Realty
Bcston Properties'
Brandyine Realty Trust'
Brookield Properties Corpcration
CarrAmencaRes*
Centerpcint Properties
Comerstone Prcperties*
Crescent Res Estate'
Duke Realty Investments'
EqutyOffice
First Industrial Reslty Trust
HigtwAods Properties
Kilroy Reatty Corp.*
Uberty Property Trust
MackCali Reetty Corporatim'
Prentiss Properties
ProLogis'
Reckso Associates Reetty
Spiew Properties*
TrizecHahn*
OfKenndustrial Average
(28)
(29)
(30)
(31)
(32)
(33)
(34)
(35)
(36)
(37)
(38)
(39)
(40)
(41)
(42)
(43)
Self-Storage REITs
Shurgard Storage Centers*
Pubi: Stoage, Inc.*
StorageUSA*
Self-Storage Aerage
Manufactured Home REITs
Ctateau Cornmunities*
Manufactured Horne'
Sun Canrriunities, Inc.*
Manufactured Horne Averge
Hotel RBTs
Felcor Lodging Trust*
Hospitalty Property Trust*
Host Marnctt*
Meristar Hospitality Corporation*
Starwood Hotels & Resorts
Hotel REIT Avrage
(44)
(45)
Mixed Use and Miscellaneous REITs
Colonial Properties Trust*
Cousins Properties*
Forest City Enterprises*
Franchise Firnace Corp. d Amenca'
HRPT Properties Trust*
Nationa Got Properties
Pennsylvania REIT*
Reatty Incorne*
US Restaurant Properties*
Vornado
Washington REIT*
Mixed Use and Misceit. REITs Avrage
Note: Footnotes can be found
a the end df
(46)
(47)
(48)
the report
UPREIT
Struture
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
No
Shares
OP units
90.5
65.6
104.7
44.1
155.7
74.0
20.2
152.7
137.5
102.2
297.0
45.3
70.6
32.3
76.1
73.8
45.5
176.2
67.5
75.0
153.0
Equity
M14Cap
$1,911.2
$1,524.2
$3,468.6
$666.9
$1,936.4
$1,673.3
$670.9
$2,366.6
$2,776.4
$2,236.5
$7,518.2
$1,142.6
$1,712.4
$742.9
$1,826.7
S2,071.9
S1.018.1
S3.61Q2
$1,434.3
$2,844.4
$3,050.4
$2,200.2
Total Non-corw.
Total
Floating Rcating/
FoctVng Debt/ Cash Cn
Debt Pref. Stock Mld Cap (1) Rate Debt Dt+Prer. Stk Equity Mk Cap
Hand
$1,486.4
$275.0
$3,672.6
$316.0
17.9%
16.5%
$29.2
$840.5
$0.0
$2,364.7 $296.5
35.3%
19.4%
S4.6
$2,960.3
$0.0
$6,448.9
$142.0
4.8%
4.1%
$32.2
$1,036.8
$0.0
$1,703.7
$531.3
51.2%
79.7%
$57.5
$5,607.6
$246.8
$7,790.8
$823.2
14.1%
42.5%
$132.8
$1,676.9
5400.0
$3,750.2
$417.5
20.1%
25.0%
$111.1
$2 13
S75.0
$1,007.2
$80.7
24.0%
12.0%
$0.6
$2,004.0
$0.0
S4,370.6
$661.4
33.0%
27.9%
$32.8
$2,890.0
$0.0
$5,666.4 $1,376.4
47.6%
49.6%
$105.4
$1,113.6
$325.0
$3,675.1
$00
0.0%
0.0%
$35.0
$6,183.7
$315.0 $14,017.0
$650.5
10.0%
8.7%
$35.9
$1,235.5
S391.3
$2,769.3 $158.1
9.7%
13.8%
$46.0
$1,779,9
$397.5
$3,889.8
$293.3
13.5%
17.1%
$104.0
$459.8
$108.2
$1,311.0
$56.0
9.9%
7.5%
$2.5
$1,416.2
$125.0
$3,367.8
$144.4
9.4%
7.9%
$14.7
$1,495.0
$0.0
$3,566.9
$228.8
15.3%
11.0%
$11.5
$970.0
$95.0
$2,031
$169.4
15.9%
16.6%
$11.8
$3,374.7
$535.0
$7,519.9
$684.6
17.5%
19.0%
$60.4
$1,197.2
$0.0
$2,631.5
$255.1
21.3%
17.8%
$13.7
$1,975.3
S431.3
$5,250.9
$28.0
1.2%
1.0%
$20.5
S4,451.4
$0.0
$7,501.8 $1,325.3
29.8%
43.4%
$244.2
$4,493.3
19.1%
21.0%
No
No
No
Yes
29.0
13.3
129.3
31.7
$735.0
$305.1
S3,378.2
$936.8
$1,338.8
$531.2
$196.3
$172.6
$809.4
$100.0
$30.0
$1,098.9
$65.0
$1,366.2
$531.3
$4,649.6
$1,811.1
$2,089.6
$114.0
$116.0
$0.0
$67.2
18.1%
51.3%
0.0%
7.7%
19.3%
15.5%
38.0%
0.0%
7.2%
15.2%
Yes
Yes
Yes
31.5
32.0
19.9
$922.1
$412.0
$733.7
$350.8
$75.0
$0.0
$0.0
$1,409.1
$1,516.6
$1,092.0
$1,339.3
$45.0
$136.0
$0.0
9.2%
18.5%
0.0%
9.3%
4.9%
17.4%
0.0%
7.4%
$0.3
$6.6
$2.0
Yes
No
Yes
Yes
Yes
75.9
56.4
343.4
56.5
213.6
$1,414,0
$1,767.0
$1,439.3
$414.8
$3,305.2 $5,136.0
$946.9 $1,429.9
$5,153.7 $7,431.0
$2,451.8
$143.5
$75.0
$0.0
$0.0
$0.0
S3,324.5
$1,929.0
$8,441.2
$2,376.8
$12,584.7
$5,731.3
$643.7
$0.0
$1,428.0
$1,212.8
$2,057.0
33.7%
0.0%
27.8%
84.8%
27.7%
34.8%
45.5%
0.0%
43.2%
128.1%
39.9%
51.3%
$27.0
$94.0
$436.0
Yes
35.7
32.1
30.0
55.8
143.3
21.4
14.6
26.8
19.8
109.0
35.8
$961.2
$889.5
$1,152.7
$403.2
$758.9 $2,433.0
$804.6
$1,234.0
$1,739.5 $1,231.5
$469.3
$476.9
$288.9
$420.6
$627.0
$323.3
$317.6
S404.1
S3,678.4 $3,035.0
$572.7 $292.8
$1,073.4
$125.0
$0.0
$0.0
$0.0
$0.0
$75.0
$0.0
$103.5
$0.0
$200.0
$0.0
$1,975.7
$207.7
$20.8
$633.0
$304.0
$500.0
$270.3
$139.4
$93.3
$195.0
$1,439.7
$54.0
20.5%
5.2%
26.0%
37.8%
40.6%
49.7%
33.2%
21.9%
48.3%
44.5%
18.4%
31.4%
21.6%
1.8%
83.4%
24.6%
28.7%
56.7%
48.3%
14.9%
61.4%
39.1%
9.4%
35.5%
No
No
No
No
Yes
Yes
No
Yes
Yes
No
$782.9
$741.2
$815.4
$1,555.9
$3,191.9
$2,038.6
$2,971.0
$1,021.3
$709.5
$1,053.8
$721.6
$6,913.4
$865.5
S2,092.6
$10.3
$3.1
$68.0
$117.4
$4.2
$17.9
$3.7
$5.4
$79.0
$3.7
$27.0
$1.3
$12.4
$16.8
$0.2
$80.1
$23.8
Exhibit 2
Assembled Data Set
60
Observations
ANOVA
SS
df
Regression
Residual
Total
7
52
59
Coefficients
Intercept
LEVERAGE
ASSETSIZE
UPREIT
APART
SHOPCENTER
REGIONALMALL
OFF/IND
0.778161821
0.147892156
-1.85407E-06
0.000875137
-0.72769707
-0.718436125
-0.730329411
-0.723366674
0.014938517
0.010957666
0.025896183
Standard Error
355718.6378
0.020382038
6.457E-07
0.005354274
355718.6378
355718.6378
355718.6378
355718.6378
MS
F
0.002134074 10.12732423
0.000210724
t Stat
2.18758E-06
7.256004175
-2.871403575
0.163446499
-2.04571 E-06
-2.01968E-06
-2.05311 E-06
-2.03354E-06
P-value
0.999998264
1.93335E-09
0.005899389
0.870800489
0.999998375
0.999998397
0.99999837
0.999998386
Significance F
6.41398E-08
Lower 95%
Upper 95%
ower 95.0 pper 95.0
-713800.4063 713801.9626 -713800
713802
0.106992631
0.18879168 0.106993 0.188792
-3.14976E-06 -5.58375E-07 -3.1E-06 -5.6E-07
-0.009868992 0.011619266 -0.00987 0.011619
-713801.9121 713800.4567 -713802 713800.5
713800.466 -713802 713800.5
-713801.9029
-713801.9148 713800.4541 -713802 713800.5
-713801.9078 713800.4611
-713802 713800.5
Exhibit 3
Regression For Test 1
0
0i0
00~~~~~~
-
0
-
0
!D
>
z
*
-
0
-
-
0
-
0
.HJ0*C*I
~
0
.
$-P.
-4)
0
0
-
e
~
W
Wme
--
00000000000000--
0
0.
"-"
0
O0
ss
0
-
-
0 0000C
0
0
0
h
00OO
Ms
$
w'
au
!-a
et -4-4&',1W
-4 -4 -4 -9
-.-
N
a
T
Ws
T.
PPOO
0
A,)
a(a
-o
-".-
Q0C0000
s
&W000
-
0Q0QQ0Q00Q0
e
2-
C--4
R
-a
.o.
-
00
2
W i 00
WsQDCO
W CA.
0
W~ K) -4 0
o
0
LA W
0
a
000C
5
C 3QQQ
Q
(
i4-40
a,
>
4P
9
0
-
9
-
App
W
7
00
e
0 0
~
.
eO
-M -.-
00
0
-
sP
0 F
W-"
0
12
-4
OP)p5ApC
a
000
0
a
-. 4-
't,
-.-.
00
.4)
AG@(.4
-4
CD
-
W
Wo
0
000
C
--
COO0.000
~
a-
C.)
2
-4
oV
C0
~0
0000000000000000
3
000
-0
A.
bU-.-
0
1
0000000000000000000000
-
00
a
0A.
W-
0C
S-2
3 03aMM
2cI ,- 0 m
J3
00
00
AaCeC"
WJ
0
999
a-00
go
- 0.sAas""- -.
0
0 0
:r
z c:a
00
O0
00
.- "e CD
o We
0 0
a
W -
0W000-
-000000000
000
0
QQQQ00000
Q00
0
00
q CDa0 (Asi:ss1s,
-N -40
~
-- ta
#.
iM(
90
00
0A.
.
no
sammmmasasamssaa
-
000
-4
00
a- e
1C
"
0J
~i
0
a
0
V
b
4;
Ok.
I
a
A-.4
00000
--
as
-
o
e
4C
86,
0.
n
8
S
0
-
-4
-w
C
C3
s:
0
W
a
CL
4
pPi#
6W(