Evidence from REITS

Brent W. Ambrose (The Pennsylvania State University),Shaun Bond (University of Cincinnati), & Joseph Ooi (National University of Singapore)

1. Introduction2. Research

Question 3. Data & Design 4. Empirical Results 5. Conclusions

Current crisis in the credit market raises a number of interesting questions regarding the use of debt by corporations. Agency theory - the role of leverage as a

mechanism to mitigate conflict of interests between managers and shareholders.

Relatively few studies have examined the implications that arise from the use of secured and unsecured debt.

▪ Secured and unsecured debt holders have different incentives to engage in costly monitoring activities.

“Does the use of unsecured debt result in a performance difference between firms that employ a higher proportion of unsecured

debt versus firms that utilize secured debt?”

------------------------------------------------------ We answer this question using data on secured

and unsecured debt holdings of REITs. Why REITs? In contrast to general firms, REITs face a real choice

with respect to issuing secured versus unsecured debt (because their asset holdings are suitable for collateralization).

All equity REITs having information in the SNL database and security prices Center of Research in Security Prices (CRSP) database. Final sample: 114 publicly traded REITs

spread over 63 quarters (1990Q1 to 2007Q4). Secured debt ratio (SecDebt) is defined as:

We measure the quarterly change in SecDebt as:

1 tt SecDebtSecDebtSecDebt

tititi DebtTotalDebtSecuredSecDebt ,,, /

1,,, tititi SecDebtSecDebtSecDebt

3. Data & Research Design

SecDebt is on a downward trend, in contrasts with the debt ratio, which has generally been increasing over time.

Cross-Sectional Mean of Secured Debt Ratio and Debt Asset Ratio

0

10

20

30

40

50

60

70

80

90

Time

(%)

Scatter Plot of Within Firm Secured Debt Ratio and Volatility of the Secured Debt Ratio

0

5

10

15

20

25

30

35

40

45

50

0 20 40 60 80 100 120

Average Firm Secured Debt Ratio

Sta

nd

ard

Dev

iati

on

of

Sec

ure

d D

ebt

Rat

io

A formal analysis of the impact of changes in the use of secured debt on firm stock performance.

itiii

itit

ititit

ititi

l

llijt

eSelfAdvSelfMgtUPREIT

recessionagebeta

ratiodebtROAFFOprice

sizeSecDebtDUMr

,11109

8,7,6

,5,4,3

,2,1

11

1,

_

)ln(

[1]

What is the relationship between and firm stock performance? The overall pattern in Table 2 suggests that REITs in the

first category tend to pose below average returns, whilst REITs in the last category tend to register above average returns.

HPR   ALL

< 0 = 0 > 0

    Mean Median Mean Median Mean Median Mean Median

Ret1m   0.94% 0.97% 0.83% 1.06% 1.02% 0.78% 1.03% 1.10%

Ret3m   4.43% 4.35% 4.57% 4.71% 4.96% 3.92% 3.81% 4.22%

Ret1y   16.85% 15.91% 15.82% 15.54% 19.38% 15.97% 16.36% 16.71%

Ret3y   65.99% 59.20% 60.51% 55.39% 69.85% 62.32% 70.46% 63.83%

100% 44.6% 31.2% 24.2%

SecDebt SecDebtSecDebt

4. Empirical Results

SecDebt

We find evidence of a positive relationship between

and subsequent performance at the one-month holding period horizon. 1% increase in the use of secured debt corresponds

to a 5.2 basis point increase in the 1-month HPR. But the coefficients are not statistically significant for

3-month, 1-year and 3-year holding periods.

Holding period return

Variable 1 month 3 month 1 year 3 year

0.00052* 0.00027 0.00027 -0.00068

(0.00028) (0.00046) (0.00046) (0.00180) itSecDebt ,

itSecDebt ,

Debt ratio (+): significantly positive over each horizon period, consistent with the additional risk associated with higher leverage.

Systematic risk (+): weakly significant in the 1-month holding period return regression, consistent with the risk-return relation.

Return On Assets (+): Accounting profit and stock performance are related positively and the relationship is statistically significant in two of the cases (3-mth & 1-yr HPR).

Full results in Table 4

Firm age is positively related to stock performance.

Firm size is negatively related with future stock returns up to one year holding period. Beyond the one year period, firm size and future stock returns are positively related.

UPREITs and externally-managed REITs tend to perform poorly.

The lack of significance for the price-to-FFO variable implies that stock returns are not correlated with current period valuations.

More Determinants of Performance

Since the use of secured debt may be endogenous to the factors associated with future stock performance, we also estimate a 2-stage regression.

itiii

itit

ititit

ititi

l

llijt

eSelfAdvSelfMgtUPREIT

cessionAgeBeta

ratiodebtROAFFOprice

sizeSecDebtDUMr

,11109

8,7,6

,5,4,3

,2,1

11

1,

Re

_

)ln(

iti

itit

iti

l

ltit

eXSelfAdvSelfMgt

UPREITRatioDebtROA

sizeDUMSecDebt

,87

6,5,14

,121

11

1,

_

)ln(

Estimation results in Table 6

[3]

[2]

How the results fit in with existing theories? The positive relation between firm utilization of secured debt and future stock performance is consistent with: Adverse selection models of secured debt

▪ Borrowers use collateral to signal quality▪ High quality (low risk) firms use more secured debt

Moral hazard models of secured debt ▪ Collateral improves the incentives for borrowers to

work hard to repay debt▪ Low quality (high risk) firms post collateral to commit

to higher effort.Although these models have conflicting views on who uses secured debt,

they are, nevertheless, consistent in predicting a positive relation between change in secured debt and subsequent corporate performance.

In the robustness test, we address the potential endogeneity problem with respect to who actually uses secured debt. The positive relation between changes in SecDebt

and stock performance is robust. Interestingly, we observe that large firms with low

leverage are more likely to increase utilization of secured debt. On the basis that they are less risky (high quality) borrowers, the results suggest that:

the adverse selection models are more relevant in explaining who and why REITs utilize secured debt.