Investigation the capital structure in A-REITs

I NVESTIGATING THE C APITAL S TRUCTURE OF

A-REIT S

Bwembya Chikolwa

Texas Comptroller of Public Accounts

Abstract

This paper examines the determinants of capital structure in34Australia real estate investment trusts(A-REITs)for the period2003to2008.The results show that tangibility and size positively in?uence leverage,with risk having a negative impact. No evidence is found to support A-REITs issuance of debt when interest rates are low;however,they are less likely to default on their debt obligations in a strong real estate market.Real estate sector effects are mixed and stapled management structure and international operations have signi?cant negative signs,showing that A-REITs with these features should have lower gearing levels.There is also some evidence that A-REITs follow the market timing theory of capital structure.

One of the research challenges faced by?nancial economists has been the determinants of capital structure following Modigliani and Miller’s(MM)(1958, 1963)proposition that the capital structure is‘‘irrelevant’’to the value of a company. In addition,Myers’(1984)‘‘capital structure puzzle’’has fuelled great debate on which of the two follow-on theories after MM,‘‘trade-off’’and‘‘pecking order,’’is the most relevant in determining capital structure.‘‘Market timing’’theory has also been propagated as relevant to the determination of a?rm’s capital structure(Baker and Wurgler,2002).

Real estate companies,in particular real estate investment trusts(REITs),offer some unique characteristics on which to test the theories of capital structure.REITs have a great deal of collateral that can be used to support high levels of debt and distribute nearly all of their pro?ts as dividends.

Australian REITs(A-REITs)are a major real estate investment and?nancing vehicle in Australia,with an outstanding track record and signi?cant commercial real estate assets,being available to both general and institutional investors.As at December 2008,A-REITs had assets worth AU$201billion,representing about50%of the commercial real estate investment market and55%of the securitized real estate investment market(PIR,2009).Thus,the A-REITs market was the third largest globally after the United States and France as of December2008(Newell and Peng, 2009).

This is the?rst study to examine the capital structure determinants of A-REITs.As such,the purpose of this paper is to analyze the determinants of the capital structure of A-REITs by examining the relationship between leverage and a set of explanatory variables.The study differs from previous studies as it considers capital structure theories in assessing determinants of various leverage positions on a total,long-term,

391

392J OURNAL OF R EAL E STATE L ITERATURE

short-term,and interest coverage basis.The analysis is conducted using an unbalanced panel data pertaining to34A-REITs for the period2003–2008.A total of204 observations are available for analysis.

The paper is structured as follows.The?nancing activities of A-REITs over the study period are discussed.There is then a review of the literature on capital structure, followed by a discussion of the data,methodology,and descriptive statistics.The study results and their analyses are then given.The paper closes with concluding remarks and future research directions.

F INANCIN

G A CTIVITIES OF A-REIT S

Many A-REITs used equity capital to fuel growth and expansion during the mid-1990s,but later switched to debt?nancing in1997when the Reserve Bank of Australia(RBA)cut interest rates in the second half of1996,which made debt ?nancing a cheaper option to equity capital(Kavanagh,1997).PIR(2008)states that a total AU$52billion was raised through equity raisings between1999and2008.Of the funds raised during this period,43%was used to acquire new real estate,24% for dividend re-investment plans,and16%for working capital/debt repayment.The remaining17%was used for employee remuneration,rights issue,and unit purchase price.On average,equity raisings by A-REITs show an upward trend,as shown in Exhibit1.

Debt funding has played a signi?cant role in A-REIT growth,increasing from10% in1995to35%in2007(Newell,2008).As at December2008,the A-REIT sector had total assets worth AU$201billion,comprising more than4,700institutional-grade properties in diversi?ed and sector-speci?c portfolios(PIR,2009).In structuring this debt pro?le,A-REITs have used a range of sophisticated debt products including CMBS,real estate trust bonds,and off-balance sheet?nancing(Chikolwa,2008a). Exhibit2shows details of CMBS and unsecured bond raisings by A-REITs from 2000to2008.

A total of over AU$26billion has been raised by A-REITs using these two debt funding instruments.From2007:Q3,no CMBSs or bonds were issued by A-REITs in Australia as a direct result of the credit squeeze.Chikolwa(2008b)showed that at predicted spreads of150–200bps in2008,issuance of AAA-rated CMBS tranches would have been uneconomical.Furthermore,Property Council of Australia(2008) showed the BBB?-rated bond issue by Dexus of AU$200million in February2007 priced at483bps over the three-month bank bill swap rate(BBSW),signifying a drastic change in market conditions from February2004when similar rated notes were priced at35–40bps.

Private bank debt has also been used to fund A-REIT expansion and operations. However,limited disclosure restricts information in the private debt market.Higgins (2007)estimated the value of outstanding Australian whole commercial real estate mortgages in2007to be AU$71billion.He estimated the Australian real estate investment market at AU$232billion,1of which72%was held by institutional V OLUME19,N UMBER2,2011

I NVESTIGATING THE C APITAL S TRUCTURE OF A-REIT S 393Exhibit 1

A-REIT Equity Capital Raisings:1999–2008

5

10

15

20

25

30

$0

$2,000$4,000$6,000$8,000$10,000$12,000200020012002200320042005200620072008A U $ M i l l i o n A-REIT Equity Raising Ex. IPOs (lhs)

A-REIT Bond Issuance (lhs)

No. of A-REIT Bond Issues (rhs)

Sources:Author’s compilation from PCA (2009)and Connect 4Company Prospectuses database

(1999–2006).

investors.With 55%of institutional owned real estate held by A-REITs,it can be

inferred that a sizeable share of private bank lending goes to A-REITs.

The initial response by A-REITs to the credit squeeze was to cut dividends and to

lower gearing through asset sales.It was estimated that between AU$15–20billion

worth of commercial real estate was for sale as at June 2008(Standard &Poor’s,

2008).With empirical evidence of distress asset sales leading to the depressing of the

share price,most A-REITs resorted to off-market transactions in selling their

properties.The glut of commercial real estate on the market and debt funding

dif?culties resulted in a rise in commercial real estate yields after several years of

contraction.For instance,premium and ‘A’grade of?ce yields increased by

approximately 100bps to December 2008(DTZ Research,2008).According to CB

Richard Ellis,sales of commercial real estate fell 60%to AU$3billion in the ?rst six

months of 2008(Condon,2008).Discount rates (across both core and non-core

sectors)were set to soften over the short-term (next 12–18months)to sit between

8.5%–11.5%,depending on sector and asset quality (De Francesco and Bonello,

2008).

394J OURNAL OF R EAL E STATE L ITERATURE

V OLUME19,N UMBER2,2011Exhibit2

A-REIT Bond Issuance and CMBS Issuance:2000–2008

Australian CMBS Issuance A-REIT Bond Issuance

Year AU$Million No.of Issues AU$Million No.of Issues

2000$3572$1001

2001$1,3205$1,61512

2002$2,84519$1,57012

2003$2,19114$2,79228

2004$1,5137$9059

2005$2,1028$1,32012

2006$4,01311$1,65011

2007$2,5006$4903

2008$00$00

Total$16,84172$10,44288

Note:The sources are CMBS issuance,Chikolwa(2007,2008a),and author’s compilation from PCA (1999–2008).

A recent development in funding strategies by some A-REITs has been to increase equity,either by introducing a dividend reinvestment plan or seeking additional capital.A breakdown of these seasoned equity offerings issued between1999and 2008is shown in Exhibit3.

Exhibit4shows some of the equity raisings by A-REITs in October2008.GPT Group and Goodman Group raised more than AU$2.3billion,with Goodman raising AU$755 million at90cents in a unit.Buyers of these equities included some sovereign wealth funds.For example GIC of Singapore increased its stake in Mirvac and GPT Group to6%and10%,respectively.A negative share price reaction has been noted after seasoned equity issues(Brealey,Myers,and Allen,2008;Brounen and Eichholtz, 2002;Marciukaityte,Higgins,Friday,and Mason,2007).As such,these equity capital raising strategies will lead to a dilution of A-REIT base earnings.

The review of the A-REIT?nancing activity is important in investigating their capital structure,in particular the determinants of debt funding or leverage following similar empirical studies in other parts of the world.

L ITERATURE R EVIEW

Studies on capital structure have evolved from the seminar papers by MM(1958, 1963),who argued that in a world of frictionless capital markets,optimal?nancial structure is irrelevant.Subsequently,with consideration of costs of debt due to information asymmetry,the tax bene?ts of debt?nancing and the con?ict of interest between managers,shareholders,and creditors,two theories of capital structure

I NVESTIGATING THE C APITAL S TRUCTURE OF A-REIT S 395Exhibit 3

A-REIT Seasoned Equity Issues:2000–2008

0%

10%

20%30%

40%

50%

60%70%

80%90%

100%

20002001 20022003

20042005 20062007 2008 E ntitlement P lacement P riority R ights S hare Purchase Plan

Source:Author’s compilation from Connect 4Company Prospectuses database (2000–2008).

Exhibit 4

Major A-REIT Seasoned Equity Issues:October–December 2008

Date

Company Amount of Shares Amount Raised (AU$million)Oct.14

Stockland 56.6million $300.00Oct.15

CFS Retail Property Trust 162million $325.00Oct.23

GPT Group 1per 1entitlement $1,349.80Oct.28

Goodman Group 17per 50entitlement $956.20Dec.3

DEXUS Property Group 391.7million $301.60Dec.8

ING Of?ce Fund 1per 2.9entitlement $415.50Dec.12Macquarie Of?ce Trust 1per 1entitlement $508.40

Note:The sources are EPRA (2008)and Connect 4Company Prospectuses database (2000–2008).

emerged:the ‘‘trade-off theory ’’and the ‘‘pecking order hypothesis.’’A later theory

propagates that ?rms’capital structure depends on ‘‘market timing’’or conditions

(Baker and Wurgler,2002).

The trade-off theory posits that ?rms observe a target debt ratio that is de?ned by a

trade-off between tax deductibility of interest and costs of ?nancial distress.It is

observed that tax deductibility of interest payments induces ?rms to borrow to a point

396J OURNAL OF R EAL E STATE L ITERATURE

where the present value of interest tax shield is just offset by the value of loss due to the possibility of?nancial distress.Since pro?table?rms have a higher income shield,and greater free cash?ow,theory predicts higher leverage for pro?table?rms, and the opposite for?rms with investment opportunities perceived to be risky.Indirect costs substantially in?uence capital?nancing,with Castanias(1983)and Alderson and Betker(1995)?nding a signi?cant negative relationship between bankruptcy costs and leverage.High(low)-growth?rms that are more sensitive to business?uctuations in business outlook and are therefore more vulnerable to?nancial distress,choose a lower(higher)leverage ratio(Barclay,Morellec,and Smith,2006).

The pecking order hypothesis by Myers(1984)and Myers and Majluf(1984)is driven by asymmetric information and signaling problems associated with external?nancing. The hypothesis predicts that?rm’s rank?nancing choices by their sensitivity to asymmetric information,choosing to use internal rather than external?nancing,and preferring debt to equity.This hierarchical behavior is aimed at minimizing the cost of information asymmetry.Literature is contradictory on the correctness of the theory; Shyam-Sunder and Myers(1999)state that?rms facing a?nancial de?cit resort to debt,and then if needs are great enough,to sources of?nancing lower and lower down the pecking order of?nancial choice.Chirinko and Singha(2000)and Frank and Goyal(2003)doubt the correctness of the theory because?rms can never issue equities if they can issue investment-grade debt.However,the pecking order hypothesis explains some of the observations not explained by the static trade-off theory.The strong negative relationship between pro?tability and leverage is explainable(Rajan and Zingales,1995;Ghosh,Nag,and Sirmans,1997;Gaud,Jani, Hoesli,and Bender,2005).

Baker and Wurgler(2002)advocate the market timing theory on the basis that an optimal capital structure does not exist and that a?rm’s observed capital structure is nothing more than a cumulative outcome of its past attempts to time the market. Managers will issue equity when they believe that the market has overvalued their company.Recent REIT studies supporting this theory are Ooi,Ong,and Li(2007, 2010)and Boudry,Kallberg,and Liu(2010).

Real estate companies have special features that make them an interesting case study for capital structure choice.For instance,A-REITs are not required to pay taxes if they distribute100%of taxable income as dividends.As such,the two important bene?ts of debt are nulli?ed,i.e.,tax deductibility of interest is lost and with most income distributed,debt servicing is not critical in mitigating agency cost of free cash ?ow.Debt in REIT?nancing at IPO accounts for more than50%and increases to over65%in10years(Feng,Ghosh,and Sirmans,2007).

Despite several attempts to explain REIT capital structure,no consensus has been reached on the determining theory.For instance,Morri and Beretta(2008)conclude that pecking order theory is behind REIT capital structure.Morri and Cristanziani (2009)settle for the trade-off theory.Boudry,Kallberg and Liu(2010)?nd strong support for the market timing theory and some support for the trade-off theory.Other studies that advocate for the market timing theory are Ooi,Ong,and Li(2007,2010). V OLUME19,N UMBER2,2011

I NVESTIGATING THE C APITAL S TRUCTURE OF A-REIT S397

Exhibit5

Expected Signs of Leverage Determinants

Trade-off Model Pecking Order

Theory

Previous REIT

Research

Pro?tability(?)(?)(?)

Growth(?)(?)(?) Tangibility(?)(?)(?) Operating risk(?)(?)(?)

Size(?)(?)(?)

Real estate sector NA NA NA

Stapled management structure NA NA NA International operations NA NA NA

Note:The material is adapted from Morri and Beretta(2008).

Furthermore,Exhibit5shows revealed signs of the determinants of leverage according to different theories of capital structure by previous REIT research.

D ATA AND M ETHODOLOGY

Data

The?nancial data for A-REITs were collected from the Aspect Fin Analysis database for the period2003to2008.A total of34?rms found in the ASX300Index are included in the dataset,with a total of204?rm-year observations.Some?rms did not have a full six-year set of?nancial records as they had been listed after2003. Their exclusion could have resulted in an incomplete analysis as they are major market players.An unbalanced panel dataset was used in this study.

Measurement and Interpretation of Variables

Following previous studies that have used pro?tability,tangibility,size,growth opportunities,and operating risk as determinants of leverage(Rajan and Zingales, 1995),additional market timing and industry variables of real estate sector,stapled management structure,and international operations are included to fully capture the setting of A-REITs.

Leverage.A variety of ratios are used to measure leverage,with the most popular being total liability to total assets,total debt to equity,total debt to total assets,and total debt to net assets(Rajan and Zingales,1995).In addition,a?ow-like measure interest coverage ratio,i.e.,the ratio of earnings before interest and taxes(EBIT)to interest expense,is more useful when concerns are raised about the probability of a company repaying its debt.Titman and Wessels(1988)show a signi?cant correlation

398J OURNAL OF R EAL E STATE L ITERATURE

V OLUME 19,N UMBER 2,2011

between book and market values and no signi?cant differences in proxies used.In

support of the use of book values,Fama and French (2002)contend that managers

usually decide to issue debt or capital structure choices based on book value data

because market data are subject to high volatility and market values depend on a

number of factors out of the direct control of a ?rm.In this study,total liability to

total assets (TLA ),total long-term debt to total assets (LTD ),total short-term debt to

total assets (STD ),and interest coverage ratio (NIE )are used to measure leverage.

Short-term debt is included because it presents a signi?cant proportion of total debt

by real estate companies;rolling-over short-term debt to achieve longer-term debt is

a common practice in real estate companies (Brett,1990;Ooi,1999).

Pro?tability.A common measure of ?rm pro?tability is return on assets (ROA )

(Brealey,Myers,and Allen,2008).The trade-off and pecking order theories diverge

on the relationship between leverage and pro?tability.According to the trade-off

theory,high pro?tability will increase tax shields and create an incentive for higher

leverage.On the other hand,the pecking order theory posits that companies will use

internal ?nancing before external sources.Following Myers and Majluf (1984),a

negative relationship between pro?tability and leverage postulated.

Tangibility.Tangible assets are likely to have an impact on the borrowing decisions of

a ?rm because they are less subject to information asymmetries and usually have a

greater value in liquidation.Consequently,?rms with tangible assets can take on

higher leverage (Gaud,Jani,Hoesli,and Bender,2005;Morri and Beretta,2008).

From an agent-theory perspective (Jensen and Meckling,1976),lenders demand

collateral of tangible assets to avoid suboptimal investments by shareholders as a result

of the con?ict between lenders and equity owners.The trade-off theory posits that

tangible assets serve as good collateral and reduce costs of ?nancial distress.Harris

and Raviv (1991)expect that information asymmetries will be large for companies

with few tangible assets,in line with the pecking order perspective.Most studies

(Titman and Wessels,1988;Gaud,Jani,Hoesli,and Bender,2005;Morri and Beretta,

2008;Westgaard,Eidet,Frydenberg,and Grosas,2008)?nd a signi?cantly positive

relationship between tangibility and debt,while Grossman and Hart (1982)and Feng,

Ghosh,and Sirmans (2007)?nd a signi?cantly negative relationship.In this study,

the proxy for tangibility is the ratio of the book value of real estate to total assets

(PPT ).

Size.The size of a ?rm is related to the risks and costs of bankruptcy (Ngugi,2008).

Larger ?rms are usually more diversi?ed and therefore bear lower risks of ?nancial

distress (Ang,Chua,and McConnell,1982);or that their size creates less transparency

and greater need for monitoring (Ang,Chua,and McConnell,1982;Myers,1984;

Myers and Majluf,1984).The former argument suggests a positive relationship

between ?rm size and leverage in line with the trade-off theory and the latter suggests

a negative relationship in line with the pecking order theory.Bond (2006),Morri and

Beretta (2008),and Westgaard,Eidet,Frydenberg,and Grosas (2008)?nd a

signi?cantly positive relationship between ?rm size and debt,while Feng,Ghosh,and

Sirmans (2007)?nd ?rm size is non-signi?cant in determining leverage.Ooi (1999)

?nds an inverse relationship between ?rm size and leverage and further concludes

I NVESTIGATING THE C APITAL S TRUCTURE OF A-REIT S399 that smaller real estate companies may not have much choice but to rely on bank loans.Earlier studies have used size to operating revenue,number of employees,or the size of total assets as proxies of size.In this study,the proxy for size is the natural logarithm of total assets(TAA).

Growth Opportunities.Studies have had various?ndings on the in?uence of growth opportunities on leverage.According to the trade-off theory,?rms with a higher proportion of their market value due to growth opportunities will have lower debt capacity because in the case of bankruptcy the value of these opportunities is close to zero(Myers,1977).The pecking order posits that small,high-growth?rms will face large information asymmetries and will seek to issue securities that minimize such asymmetries.Therefore,high-growth?rms will issue debt(in particular short-term debt).The most popular proxy of growth opportunities is market value-to-book ratio.Other measures used are annual growth rate in total assets(Ooi,1999)and changes of operating revenue from year to year(Westgaard,Eidet,Frydenberg,and Grosas,2008).Changes of share price-to-book value of total assets from year to year are used as proxies of growth opportunities in this study(PPR).

Operating Risk.Operating risk or business risk is de?ned as variability of expected earnings.High variation in earnings increases the probability of default.Leverage increases the volatility of net pro?t.Both the pecking order and the trade-off theory predict a negative relationship between operating risk or volatility and the degree of leverage.From a pecking order perspective,?rms with highly volatile results try to accumulate cash during good years,to avoid under investment issues in the future. The trade-off theory,on the other hand,posits that?rms that have high operating risk can lower the volatility of the net pro?t by reducing the level of debt.This results in the reduction of bankruptcy risk and an increase in the probability of fully bene?ting from tax shields.Some of the commonly-used proxies for operating risk are operating income to interest expense,percentage change in EBIT,standard deviation of EBIT scaled by total assets,standard deviation of EBIT to net sales,and systematic risk represented by?rm beta.In this study,standard deviation of EBIT scaled by total assets(SDE)for each?rm over the entire period covered and percentage change in EBIT are used as a proxies for operating risk.

Macroeconomic Factors.Real estate companies are more likely to default in their debt commitments in a weak real estate market,with costly contracting theories positing an inverse relationship between secured debt and market condition(Ooi,1999). Similar to Ooi(2000),the Property Council of Australia(PCA)/Investment Property Databank(IPD)Total Return Index for all real estate types over the preceding year is used as a proxy for real estate market condition(MKT).

Previous real estate research has shown that?rms are more likely to use debt when the cost of borrowing is low and to use equity?nancing when interest rates are high (Ooi,1999;Boudry,Kallberg,and Liu,2010;Ooi,Ong,and Li,2010).Higher interest rates also increase the probability of?nancial distress.Therefore,debt ratio is hypothesized to be inversely related to the prevailing term structure.The term structure of interest rates is measured by the year-end yield difference between three-month bank accepted bills and ten-year federal government bond yield(TRM).

400J OURNAL OF R EAL E STATE L ITERATURE

V OLUME 19,N UMBER 2,2011Other Firm Attributes.The study also includes other ?rm-speci?c attributes such as the

real estate sector the A-REIT operates in,whether a stapled management structure is

adopted,and whether it operates internationally.Ooi (1999)shows that the nature of

activities and the asset structure of real estate companies have a signi?cant impact on

their debt funding activities.Furthermore,Giambona,Harding,and Sirmans (2008)

show that ?rms with less liquid assets will choose to use lower leverage and shorter

maturity.Therefore,it is hypothesized that the underlying real estate base of A-REITs

will in?uence their debt funding activities.In Australia,the stability of cash ?ows

and asset values of the major real estate types,ranked in order from lowest to highest

volatility,is as follows:retail (SER ),industrial (SEI ),of?ce (SOF ),hotel (SEO )

(Moody’s Investor Service,2003).As discussed earlier,this volatility impacts

operating risk.

Adoption of the stapled management structure (SSS )by A-REITs was meant to booster

returns (Newell,2006;Tan,2004a).Real estate development carries higher risk than

passive real estate holdings for investment purposes and uses more short-term debt

(Ooi,1999).

The internationalization of A-REITs was a response to increased geographical

diversi?cation and returns and because of a dearth of quality local commercial real

estate to invest in (Tan,2004b).Debt has played a major role in this international

expansion.Newell (2006)states that debt levels in some A-REITs with 100%

international real estate have debt levels in excess of 50%,compared to the average

debt level of 35%.Despite the ?nding that A-REITs with international properties in

their real estate portfolio have not increased their risk pro?le,market sentiment is that

such A-REITs are risky investments resulting in a fall in their share price.Therefore,

the study investigates the relationship between A-REIT internationalization of

operations (INT )and leverage.

The coef?cient direction for stapled management structure and international operations

is a matter of empirical investigation since in times of prosperity they booster an A-

REIT’s earnings (Tan,2004a,2004b;Newell,2006)and in periods of economic slump

contribute to the underperformance of an A-REIT (Newell and Peng,2009).

These ?rm attributes are presented as dummy variables in the models.For example,

a dummy variable is equal to 1if the A-REIT has international operations,and is

equal to 0otherwise.

D ESCRIPTIV

E S TATISTICS

Descriptive statistics regarding the sample are provided in Exhibit 6.Exhibit 7shows

a correlation matrix for the dataset.Most of the variables are not highly correlated.

Following a rule of the thumb,multicolinearity becomes an issue if the R 2is higher

than 80%and pair-wise correlations among regressors are in excess of 0.8.

Model Speci?cation

Apart from the standard panel data estimation techniques (Ooi,1999;Gaud,Jani,

Hoesli,and Bender,2005;Westgaard,Eidet,Frydenberg,and Grosas,2008),other

I NVESTIGATING THE C APITAL S TRUCTURE OF A-REIT S401

Exhibit6

Descriptive Statistics

Variables Mean Median Min.Max.Std.Dev. Dependent

TLA0.3670.3770.000 1.0000.233 LTD0.2230.2210.0000.6110.155 STD0.0530.0030.0000.4960.096 NIE13.144 3.825?18.4101805.070126.182 Independent

ROA0.0500.054?0.1180.2200.041 PPR0.946 1.0300.000 3.3300.634 PPT0.4580.5280.0000.9930.384 SDE0.0240.0160.0000.1120.025 EBT0.6040.080?15.10040.808 3.668 TAA7.5899.0620.00010.706 3.618 TRM?0.003?0.002?0.0110.0030.004 MKT0.5880.6190.000 1.0160.357 Notes:The summary statistics are based on the?nal sample of204?rm-year observations.The dependent variables for the models are the ratio of total liability to total assets(TLA),ratio of long-term debt to total assets(LTD),ratio of short-term debt to total assets(STD),and the interest coverage ratio(NIE).Independent variables for the models are pro?tability:return on assets(ROA); growth opportunities:changes of share price-to-book value of total assets from year to year(PPR); tangibility:ratio of book value of property to total assets(PPT);operating risk:standard deviation of EBIT scaled by total assets over the entire period covered(SDE),changes in EBIT from year to year(EBT);size:natural logarithm of total assets(TAA);time-variant variables:term structure of interest rates measured by year-end yield difference between3-month bank accepted bills and10-year federal government bond yield(TRM),market condition measured by Property Council of Austral(PCA)/Investment Property Databank(IPD)total return index for all property(MKT). techniques have been used to analyze capital?nancing behavior in real estate studies: simultaneous equation model(Giambona,Harding,and Sirmans,2008),linear regression(Morri and Beretta,2008)and logistic models(Ooi,Ong,and Li2007; Boudry,Kallberg,and Liu2010;Ooi,Ong,and Li,2010).In this study,the panel data estimation technique is adopted.Panel data estimation techniques improve the ef?ciency of the estimates by providing a larger number of data points and blending characteristics of both cross-sectional and time series data.

The explanatory variables in the measurement and interpretation of variables section are used to proxy for the determinants of A-REIT leverage.It is posited that leverage can be explained by the following variables:

Leverage?f(profitability,growth opportunities,tangibles,operational risk,size, market timing,stapled structure,international operations).(1)

The least square dummy variable(LSDV)model similar to Ooi(1999)is used.The LSDV model may be speci?ed as:

402J OURNAL OF R EAL E STATE L ITERATURE

V OLUME 19,N UMBER 2,2011Exhibit 7

Correlation Matrix

ROA PPR PPT SDE EBT TAA TRM MKT SER SOF SEI SEO SSS INT

ROA 1.000

PPR 0.492 1.000

PPT 0.3160.187 1.000

SDE 0.2530.2610.124 1.000

EBT 0.1460.032?0.0100.099 1.000

TAA 0.5520.6970.5840.4120.069 1.000

TRM ?0.355?0.422?0.181?0.264?0.120?0.406 1.000

MKT 0.5890.7350.3750.3800.1200.777?0.778 1.000

SER 0.082?0.0030.2570.240?0.0200.241?0.1110.175 1.000

SOF 0.0770.0020.302?0.005?0.0150.172?0.0380.117?0.153 1.000

SEI 0.0280.0930.1550.037?0.0260.136?0.0860.105?0.115?0.091 1.000

SEO 0.1950.364?0.1890.2540.0730.197?0.0570.177?0.230?0.182?0.137 1.000

SSS 0.4140.4330.0620.2030.0450.435?0.1570.347?0.059?0.131?0.0800.041 1.000

INT 0.2200.2250.3320.3540.0330.498?0.2270.3750.2930.1680.269?0.1280.086 1.000Notes:The pairwise correlation coef?cients are based on the ?nal sample of 204?rm-year observations.The regressors are pro?tability:return on assets (ROA );growth opportunities:changes of share price-to-book value of total assets from year to year (PPR );tangibility:ratio of book value of property to total assets (PPT );operating risk:standard deviation of EBIT scaled by total assets over the entire period covered (SDE ),changes in EBIT from year to year (EBIT );size:natural logarithm of total assets (TAA );time-variant variables:term structure of interest rates measured by year-end yield difference between 3-month bank accepted bills and 10-year federal government bond yield (TRM ),market condition measured by Property Council of Australia (PCA)/Investment Property Databank (IPD )total return index for all property (MKT );real estate sector ?:retail sector dummy variable of 1or 0otherwise (SER ),of?ce sector dummy variable of 1or 0otherwise (SOF ),industrial sector dummy variable of 1or 0otherwise (SEI ),other property sectors,e.g.,hotel,dummy variable of 1or 0otherwise (SEO );stapled management structure dummy variable of 1or 0otherwise (SSS );and international operations dummy variable of 1or 0otherwise (INT ).

?The diversi?ed property sector has been excluded.

I NVESTIGATING THE C APITAL S TRUCTURE OF A-REIT S403

y????X?v,(2) it i it it

where y

it represents the dependent variable,subscript i denotes the cross-sectional

dimension,and t is the time-series dimension.?is a scalar;X

it contains the set of

explanatory variables in the estimation mode;?is a11?1column matrix of the

partial regression coef?cients;and v

it represents the remaining disturbances in the

regression,which varies with individual?rms and time.

Four proxies are used to measure leverage:total liability to total assets(TLA),total long-term debt to total assets(LTD),total short-term debt to total assets(STD),and interest coverage ratio(NIE).The?rst three measure the determinants of various debt positions and the last measures the ability to meet interest payments on debts.The use of several debt ratios also acts as a robustness test.

Various models are constructed to investigate the determinants of the various leverage positions of total liabilities,long-term debt,short-term debt,and ability to cover interest payments.Model1with TLA as the dependant variable and traditional independent variables found in extant literature is the base case.Market-timing factors, such as real estate market return,term structure of interest rates,and other?rm attributes such a real estate sector(PTY),stapled management structure(SSS),and international operations(INT)are added to the base case for models2to8using TLA, LTD,STD,and NIE as dependant variables.

The unbalanced panel regressions were where carried out in EViews?version6(QMS, 1997).

E MPIRICAL R ESULTS AND A NALYSIS

Eight separate,single equation models were estimated by LSDV.The results are given in Exhibit8.The models explain between56%and89%of the within-sample variance in the dependent variables and the F-statistics show that the models are,overall, strongly signi?cant.

In order to maintain brevity,results of the base model with long-term leverage(LTD), short-term leverage(STD),and interest coverage ratio(NIE)as dependent variables are not shown as their full models produced better results.Except for the coef?cients for tangibility and market timing factors,all the signi?cant regressors have signs that are consistent with a priori expectations.

The signi?cant negative impact of pro?tability(ROA)on total leverage(TLA)is in line with the pecking order theory,which stipulates that more pro?table?rms have a reduced external need for?nancing need.Studies that have come to this conclusion are Rajan and Zingales(1995),Gaud,Jani,Hoesli,and Bender(2005),Bond and Scott(2006),and Morri and Beretta(2008).Hammes and Chen(2004)and Westgaard, Eidet,Frydenberg,and Grosas(2008)?nd a positive relationship,while Ooi(1999)?nds it to be entirely insigni?cant.LTD is signi?cant at1%,though with an anomalous positive sign.STD and NIE have correct signs but are statistically insigni?cant.As pointed out earlier,because of the regulatory requirement for A-REITs to distribute

404J OURNAL OF R EAL E STATE L ITERATURE

V OLUME 19,

N UMBER 2,

2011Exhibit 8

LSDV Regression Results

Variable Expected

Sign Regression 1[1]TLA Regression 2[2]TLA Regression 3[3]TLA Regression 4[4]TLA Regression 5[5]TLA Regression 6[6]LTD Regression 7[7]STD Regression 8[8]NIE ROA (?)?0.581?0.589?0.534?0.566?0.512 2.686?1.276?0.223

(?5.821)***(?4.302)***(?4.123)***(?4.472)***?4.168*** 2.395**?0.766?0.418

PPR (?)?0.072?0.039?0.039?0.037?0.060?0.0170.003

(?4.057)***(?1.345)(?1.465)?1.441?4.085***?0.4940.529

PPT (?)?0.027?0.025?0.044?0.039?0.0440.052?0.0160.023

(?1.551)(?1.723)*(?2.447)**(?2.467)**?2.840** 2.093**?0.302 3.221**

SDE (?)?2.113?2.088?1.8570.001?0.0010.001

(?3.499)***(?3.564)***?3.409*** 1.239?0.9860.949

EBT 0.0040.003

(?1.497)(?1.228)

TAA (?)0.0530.0950.0930.0930.1010.026?0.0520.009

(?16.374)***(?7.690)***(?9.423)***(?9.851)***14.776*** 1.035**?1.996**0.820

TRM (?) 5.202?1.585 2.648 1.067?0.0210.147?0.004

(?1.745)(?0.486)(?0.715)(?0.393)?0.488 1.623*?0.395

MKT (?)0.222?0.0190.1230.0810.0000.0000.000

(?3.107)**(?0.194)(?1.310)(?1.278)?6.388**0.960?0.636

SER 0.1450.1960.2130.1910.052?0.156?0.024

(?1.916)*(?2.938)**(?3.273) 2.882**0.977**?2.030**?2.519**

SOF ?0.161?0.035?0.040?0.050?0.063?0.0900.035

(?1.360)(?0.281)(?0.343)?0.434?2.509**?1.606 3.157***

SEI 0.0700.0850.0800.08811.538 2.880?5.518

(?3.632)***(?3.056)**(?4.105) 3.151** 2.598* 1.676*?1.792*

SEO 0.0760.1420.1460.1410.3260.266?0.092

(?13.256)***(?12.016)***?10.49810.917*** 2.698** 2.179**?1.300

I NVESTIGATING THE C APITAL S TRUCTURE OF A-REIT S

405Exhibit 8(continued)

LSDV Regression Results

Variable Expected

Sign Regression 1[1]TLA Regression 2[2]TLA Regression 3[3]TLA Regression 4[4]TLA Regression 5[5]TLA Regression 6[6]LTD Regression 7[7]STD Regression 8

[8]NIE SSS (?)?0.287?0.292?0.313?0.3050.023?0.309?0.006

(?3.051)**(?3.716)***?4.426?3.995***0.990?3.303***?0.460

INT (?)?0.205?0.195?0.201?0.204?0.4490.234?0.080

(?4.198)***(?4.760)***?4.991?5.170***?4.282*** 1.399?1.461

R 20.8710.8940.8960.8940.8950.7020.7720.568F-Stat.23.80025.07125.58425.70426.854 4.533 6.518 2.527

Notes:Estimation results of LSDV regression on 204?rm-year observations.The dependent variables for the models are the ratio of total liability to total assets (TLA ),ratio of long-term debt to total assets (LTD ),ratio of short-term debt to total assets (STD ),and the interest coverage ratio (NIE ).The regressors are pro?tability:return on assets (ROA );growth opportunities:changes of share price-to-book value of total assets from year to year (PPR );tangibility:ratio of book value of property to total assets (PPT );operating risk:standard deviation of EBIT scaled by total assets over the entire period covered (SDE ),changes in EBIT from year to year (EBT );size:natural logarithm of total assets (TAA );time-variant variables:term structure of interest rates measured by year-end yield difference between 3-month bank accepted bills and 10-year federal government bond yield (TRM ),market condition measured by Property Council of Australia (PCA)/Investment Property Databank (IPD )total return index for all property (MKT );real estate sector ?:retail sector dummy variable of 1or 0otherwise (SER ),of?ce sector dummy variable of 1or 0otherwise (SOF ),industrial sector dummy variable of 1or 0otherwise (SEI ),other property sectors,e.g.,hotel,dummy variable of 1or 0otherwise (SEO );stapled management structure dummy variable of 1or 0otherwise (SSS );and international operations dummy variable of 1or 0otherwise (INT ).Intercepts are not reported.Absolute values of White’s t-statistics are given in parentheses below coef?cient estimates.

*Signi?cant at the 0.10level.

**Signi?cant at the 0.05level.

***Signi?cant at the 0.01level.

?The diversi?ed property sector has been excluded.

406J OURNAL OF R EAL E STATE L ITERATURE

most of their earnings,a pro?table?rm does not necessarily lead to more?nancial slack.

Contrary to the pecking theory,the regression results show a signi?cant negative relationship between TLA and LTD and growth opportunities(PPR)at the1%and 5%levels of signi?cance,respectively.Growth opportunities are statistically insigni?cant determinants of STD and NIE.Ooi(1999)explained the negative relationship on the basis that?rms with a larger proportion of their value accounted for by growth opportunities employ less debt.Fama and French(2002)attribute the signi?cant negative relationship between growth prospects and leverage to?rms having higher agency costs as a result of the underinvestment and asset substitution problems.Furthermore,Rajan(1995)asserts that the negative relationship holds because of the tendency by?rms to issue equity when the stock price is relatively high and from the theoretical evidence that?nancial distress costs are higher for high-growth?rms.Other researchers?nd a positive relationship(Morri and Beretta,2008), and explain this on the basis that growth opportunities add value to a?rm,increase its debt capacity,and allow a large amount of debt to be taken on(Titman and Wessels, 1988).As pointed out by Morri and Cristanziani(2009),the effect of higher or lower growth prospects on leverage is low since the regulatory environment does not allow REITs to retain a larger amount of earnings.This is con?rmed by the omission of PPR in model4,which results in a minimal change in R2of0.2%between models3 and4.

Results show tangibility(PPT)is a signi?cant determinant of leverage,though with an anomalous positive sign for the full models of TLA.LTA and NIE show positive signi?cant signs,with STD showing a negative insigni?cant result.These results are inconsistent with Feng,Ghosh,and Sirmans(2007),who argue that there should be no relationship between tangible assets and leverage ratios on the basis that REITs are expected to have mostly tangible assets,which results in not much variability in the data.The result in the current study supports the proposition that with the adoption of stapled management structures,A-REITs not relying heavily on earnings from real estate development and funds management activities should?nd having a higher tangibility ratio advantageous for leveraging in the long-run.

Although the risk of A-REITs is not constant over the study period and the use of alternative risk measures such as EBT(changes in EBIT from year-to-year)on a rolling window could have been more appropriate,the results of models with EBT were not very different to those in which SDE(standard deviation of EBIT scaled by total assets over the entire period covered)was used.SDE was found to be a signi?cant negative determinant of leverage at1%on a TLA basis.Westgaard,Eidet,Frydenberg, and Grosas(2008)also arrived at a similar result.The result con?rms Morri and Beretta’s(2008)assertion that leverage ratios are very sensitive to changes in the degree of operating risk borne by REITs.Also con?rmed is the assertion by Morri and Cristanziani(2009)that?rms with a high level of operating risk have lower leverage,due to maintaining a moderate total risk pro?le.Leverage increases the volatility of net pro?t.The trade-off theory and the pecking order perspective both recognize this inverse relationship.

V OLUME19,N UMBER2,2011

I NVESTIGATING THE C APITAL S TRUCTURE OF A-REIT S407 The signi?cant positive coef?cient between size(TAA)to TLA and LTD at1%and 5%,respectively,is in line with both static trade-off theory and pecking order theory when size is interpreted as a proxy for cost of?nancial distress((Rajan and Zingales, 1995)and as a proxy for information asymmetries(Rajan and Zingales,1995;Fama and French,2002;Feng,Ghosh,and Sirmans,2007).Rajan and Zingales(1995) contend that the probability of bankruptcy is lower for larger?rms than it is for smaller ?rms and that a positive correlation with leverage should be expected.Furthermore, Rajan and Zingales(1995)and Fama and French(2002)claim that larger?rms disclose more information to external investors than smaller ones do.On a TLA and LTD basis,the results show that larger A-REITs should take on more debt.

As for market timing factors,the results show that only the Property Council of Australia(PCA)/Investment Property Databank(IPD)Total Return Index(MKT)is signi?cant at5%,with a coef?cient of?6.39on a long-term basis.It appears that A-REIT leveraging is not in?uenced on short-term basis by MKT,which is con?rmed by the fact that most A-REIT investments are on a long-term basis and long-term borrowing is preferred to short-term borrowing.The term structure of interest rates variable(TRM)is statistically insigni?cant in all the models.

The results of including real estate sector dummy variables are mixed.Only the retail, industrial,and other real estate sectors are positively signi?cant.The of?ce real estate sector had negative insigni?cant results.A partial explanation in line with trade-off theory that the productivity of a?rm’s assets may also play a role in?nancing choice, those assets that yield high levels of predictable cash?ows,such as retail and industrial assets,are more likely to support higher levels of debt.On a TLA basis, inclusion of real estate sector variables improved the explanatory power of the determinants of A-REITs leverage from87.1%(model2)to89.4%(model2).Bond and Scott(2006)found real estate type effects to be insigni?cant.

Stapled management structure(SSS)is found to be signi?cantly negatively related to TLA and STD at mainly1%,with international operations(INT)being signi?cantly negatively related to TLA and LTD at5%,respectively.These results imply that A-REITs with these features should have lower gearing levels.This result is puzzling as anecdotal evidence has shown that A-REITs with stapled management structures and international operations are much more geared than those with single unit management structures and local operations.However,the?nding by Newell and Peng (2009)of underperformance by A-REITs with these two features during the global ?nancial crisis may partially explain the result.

C ONCLUSION

Research on the determinants of a?rm’s capital structure is an area that?nancial economists have not reached consensus on after MM’s proposition.As for research on REIT capital structure,results of earlier studies were inclined towards the perking order theory,with more recent studies advocating for market timing theory.These studies are predominantly on the U.S.REIT market.This study investigates the determinants of A-REIT capital structure using data on34A-REITs found in the

408J OURNAL OF R EAL E STATE L ITERATURE

V OLUME 19,N UMBER 2,2011

ASX300Index.The signi?cance of the study is shown by A-REITs being the second

largest REIT market globally until December 2008and particularly the major role

that debt funding has played in their expansion.

The results can be summarized as follows:

ⅢThe impact of both pro?tability and growth prospects on REIT leverage is low as the regulatory environment does not allow REITs to retain a

bigger amount of earnings.A pro?table ?rm does not necessarily lead

to more ?nancial slack.The standard rule that debt is supposed to grow

when investments exceed retained earnings and to fall when vice versa,

does not hold due to the payout requirement.

ⅢA-REITs not relying heavily on earnings from real estate development and funds management activities should ?nd having a higher tangibility

ratio advantageous for leveraging in the long run.

ⅢRisk is a signi?cant negative determinant of A-REIT leverage.A-REITs with a high level of operating risk should have lower leverage,due to

the purpose of maintaining a moderate total risk pro?le.The trade-off

theory and the pecking order perspective both recognize this inverse

relationship.

ⅢThe signi?cantly positive relationship between size and leverage is in line with both static trade-off theory and pecking order theory when size

is interpreted as a proxy for the cost of ?nancial distress,as well as

information asymmetries.Intuitively,larger A-REITs should take on

more debt.

ⅢNo evidence is found to support the notion that A-REITs are more likely to use debt when the cost of borrowing is low or to issue equity when

interest rates are high.The term structure of interest rates variable (TRM )

was found to be a statistically insigni?cant determinant of leverage.

However,evidence was found of A-REITs being less likely to default

on their debt commitments in a strong real estate market.

ⅢThe results of the impact of real estate sector on leverage are mixed.Retail,industrial,and other real estate sectors are positively signi?cant.

The of?ce real estate sector had negative insigni?cant results.A partial

explanation,in line with trade-off theory that a productivity ?rm’s assets

may also play a role in ?nancing choice,is that those assets that yield

high levels of predictable cash ?ows,such as retail and industrial assets,

are more likely to support higher levels of debt.

ⅢStapled management structure and international operations have a signi?cant negative impact on leverage,showing that A-REITs with

these features should have lower gearing levels.The result is puzzling

as it has been shown that in a bull market A-REITs with these features

perform strongly.The inverse applies in a bear market.

The study,though,can bene?t from further investigation of the roles played by both

private and public debt funding in A-REITs.This has not been possible in this study

I NVESTIGATING THE C APITAL S TRUCTURE OF A-REIT S409 due to lack of data on private debt.With the passage of time and the growth of the A-REITs market,more data will be available for a much more detailed the investigation of effects of different real estate sectors on A-REIT capital structure.

E NDNOTES

1.Stapled management structure involves funds management and real estate development,in

addition to the traditional passive real estate holding for investment.

2.De Francisco and Bonello(2008)reviewed the value of the Australian commercial real estate

to AU$300billion as at June2008,with A-REITs representing about50%of the market.

R EFERENCES

Alderson,M.J.,and B.L.Betker.Liquidation Costs and Capital Structure.Journal of Finance, 1995,39,39–69.

Ang,J.S.,J.H.Chua,and J.McConnell.The Administrative Cost of Corporate Bankruptcy:A Note.Journal of Finance,1982,37:1,219–26.

Baker,M.and J.Wurgler.Market Timing and Capital Structure.Journal of Finance,2002,57: 1,1–32.

Barclay,M.J.,E.Morellec,and C.W.J.Smith.On the Debt Capacity of Growth Options.Journal of Business,2006,79,37–59.

Bond,S.A.and P.Scott.The Capital Structure Decision for Listed Real Estate Companies. SSRN paper,2006.

Boudry,W.I.,J.G.Kallberg,and C.H.Liu.An Analysis of REIT Security Issuance Decisions. Real Estate Economics,2010,38:1,91–120.

Brealey,R.A.,S.C.Myers,and F.Allen.Principles of Corporate Finance.New York:McGraw-Hill,2008.

Brett,M.Property and Money.London:Estate Gazette,1990.

Brounen,D.and P.M.A.Eichholtz.Capital Structure Theory:Evidence from European Property Companies’Capital Offerings.Real Estate Economics,2002,29:4,615–32.

Castanias,R.Bankruptcy Risk and Optimal Capital Structure.Journal of Finance,1983,38:5, 1617–35.

Chikolwa,B.Development of Australian Commercial Mortgage-Backed Securities.Paci?c Rim Property Research Journal,2007,13:4,398–422.

——.Determinants of Listed Property Trust Bond Ratings:Australian Evidence.Paci?c Rim Property Research,2008a,14:2,123–49.

——.Structuring Issues for Australian Commercial Mortgage-Backed Securities.Paci?c Rim Property Research,2008b,14:3,334–55.

Chirinko,R.S.and A.R.Singha.Testing Static Tradeoff Against Pecking Order Models of Capital Structure:A Critical Comment.Journal of Financial Economics,2000,58:3,417–25. Condon,https://www.360docs.net/doc/fc9950921.html,mercial Sales Plunge.The Australian.Sydney,2008.

De Francesco,A.J.,and M.Bonello.The Pricing of the Australian Commercial Property Market: An Update.Property Research.Sydney:Colonial First Estate Global Asset Management,2008. DTZ Research.Australian Investment Horizon.Sydney:DTZ,2008.

European Public Real Estate Association.FTSE EPRA/NAREIT Global Real Estate Index: October2008.Amsterdam:European Public Real Estate Association,2008.

410J OURNAL OF R EAL E STATE L ITERATURE

V OLUME 19,N UMBER 2,2011

Fama,E.F.and R.K.French.Testing Trade-off Theory and Pecking Order Predictions About Dividends and Debt.Review of Financial Studies ,2002,15:1,1–33.

Feng,Z.,C.Ghosh,and C.F.Sirmans.On Capital Structure of Real Estate Investment Trusts (REITs).Journal of Real Estate Finance and Economics ,2007,34,81–105.

Frank,M.Z.and V .K.Goyal.Testing the Pecking Order Theory of Capital Structure.Journal of Financial Economics ,2003,67:2,217–48.

Gaud,P.,E.Jani,M.Hoesli,and A.Bender.The Capital Structure of Swiss Companies:An Empirical Analysis Using Dynamic Panel Data.European Financial Management ,2005,11:1,51–69.

Ghosh,C.,R.Nag,and C.F.Sirmans.Financing Choice by Equity REITs in the 1990s.Real Estate Finance ,1997,Fall,41–50.

Giambona, E.,J.P.Harding,and C.F.Sirmans.Explaining the Variation in REIT Capital Structure:The Role of Asset Liquidation Value.Real Estate Economics ,2008,36:1,111–37.

Grossman,S.J.and O.Hart.Corporate Financial Structure and Managerial Incentive.In:The Economics of Information and Uncertainty .J.McCall (ed.).Chicago,IL:University of Chicago Press,1982.

Hammes,K.and Y .Chen.Performance of the Swedish Real Estate Sector 1998–2002.SSRN paper,2004.

Harris,M.and A.Raviv.The Theory of Capital Structure.Journal of Finance ,1991,46:1,207–355.

Higgins,D.M.Placing Commercial Property in the Australian Capital Market.RICS Research Papers .London:RICS,2007.

Jensen,M.C.and W.Meckling.Theory of The Firm:Managerial Behaviour,Agency Costs and Capital Structure.Journal of Financial Economics ,1976,3,305–60.

Kavanagh,J.Trust Changing Gear.Property Australia ,1997,June,14–15.

Marciukaityte,D.,E.J.Higgins,H.S.Friday,and J.R.Mason.Positive Performance and Private Equity Placements:Outside Monitoring or Inside Expertise?Journal of Real Estate Portfolio Management ,2007,13:4,389–99.

Modigliani,F.and https://www.360docs.net/doc/fc9950921.html,ler.Corporate Income Taxes and The Cost of Capital.American Economic Review ,1963,53:3,433–43.

——.The Cost of Capital,Corporation Finance and Theory of Investment.American Economic Review ,1958,48:3,261–97.

Moody’s Investor Service.CMBS:Moody’s Approach to Rating Australian CMBS.International Structured Finance:Rating Methodology .Sydney:Moody’s Investor Service,2003.

Morri,G.and C.Beretta.The Capital Structure Determinants of REITs.Is it a Peculiar Industry?Journal of European Real Estate Research ,2008,1:1,6–57.

Morri,G.and F.Cristanziani.What Determines the Capital Structure of Real Estate Companies?An Analysis of the EPRA/NAREIT Europe Index.Journal of Property Investment &Finance ,2009,27:4,318–72.

Myers,S.C.Determinants of Corporate Borrowings.The Journal of Financial Economics ,1977,5:147–175.

——.The Capital Structure Puzzle.Journal of Finance ,1984,39:3,575–92.

Myers,S.C.and N.S.Majluf.Corporate Financing and Investment Decisions When Firms Have Information That Investors Do Not Have.Journal of Financial Economics ,1984,13:2,187–222.

Newell,G.The Changing Risk Pro?le of Listed Property Trusts.Australian Property Journal ,2006,39:3,172–80.

——.Implications for the Strategic Development of UK REITs From The Experience of LPTs in Australia.Research Programme 2006–2009.London:Investment Property Forum,2008.