A-REIT BIDDER RETURNS: An Evaluation of Public and Private Targets and Method of PaymentChris RatcliffeBill Dimovski

Introduction

• M&As one of few avenues to growth for A-REITs• Australia is one of the highest securitized property

markets in the world • GFC saw market cap fall from A$135b in 2007 to

A$46b Feb 2009, as at March 2011 A$79b• Chandler (2011) suggest increase M&A activity in

future as market conditions improve

Introduction

• Investigate 56 A-REIT M&A announcements 1996-2010

• Prior US REIT studies shown mixed results for bidders of public targets• +5.78% (Allen & Sirmans, 1987) • -1.21% (Sahin, 2005)

• Private target → bidders earn CARs +1.52% (Campbell et al. 2005)

Prior literature (All REIT-REIT)

Author Study period

# sample CARs (%) Event days

Allen & Sirmans (1987) 1977-83 38 +5.78* [-1,0]

Campbell et al., (1998) 1990-98 27 -1.1 [-1,+1]

Sahin (2005) 1990-98 35 -1.21* [-1,+1]

Eichholtz & Kok (2008) 1999-2004 43 0.27 [-1,+1]

Keisers (2009) 1990-2005 93 -0.41 [-1,+1]

Campbell et al., (2009) 1997-2006 132 0.00 [-1,+1]

* Denotes statistical significance

Prior literature (Pub v Private)

Author Study period

# Sample

Type CARs (%) Event days

Campbell et al., (2001)

1994-98 4045

Pub-pubPub-private

-0.6*+1.9*

[-1,+1]

Campbell et al., (2005)

1995-2001 53 Pub-private +1.52* [-1,+1]

Keisers (2009) 1990-2005 7023

Pub-pubPub-private

-0.76*+0.66

[-1,+1]

Campbell et al., (2009)

1997-2006 7062

Pub-pubPub-private

-0.95*+1.1*

[-1,+1]

* Denotes statistical significance

Prior literature (method of payment)

Author Study period

# Sample

Type CARs (%)

Event days

Campbell et al., (2001)

1994-98 4037

Scrip (Pub-pub)Scrip (Pub-priv)

-0.6*+2.2*

[-1,+1]

Campbell et al., (2005)

1995-2001(pub-priv)

494

Scrip/comboCash

+1.58+0.76

[-1,+1]

Eichholtz & Kok (2008)

1999-2004 2815

Scrip/comboCash

+0.23+0.14

[-1,+1]

Ratcliffe et al., (2009)

1996-2007(Aust data)

2214

Scrip/comboCash

+1.55*-0.22

[-1,+1]

Campbell et al., (2009)

1997-2006 5624

Scrip/comboCash

+0.81*+0.07

[-1,+1]

* Denotes statistical significance

Event Study Method

• We employed event study methodology as described by Brown and Warner (1985)

• The market model was estimated for each company over a 120 day estimation period, OLS regression employed to determine the parameter estimations.

• The following market model is employed:

• To avoid the bias associated with the estimation of parameters using daily returns with infrequent trading we employ the Scholes and Williams (1977) adjusted beta method

titmiiti RRE ,,, )()(

Event Study Method

• The abnormal return (AR) of the common stock in the event window [-20,+20] is calculated as:

• The cumulative abnormal returns (CAR) for any interval during the event window:

]2:1[

,]2:1[tt

titt ARCAR

)( ,,, tititi RERAR

Regression Method

• Regression model was developed to examine the CARs [-1,+1] calculated above for acquirers.

• Independent variables were selected on the basis of prior literature along with variables unique to the A-REIT structure.

HHPROPBVMVPUBLICMOP

LEVRELSIZECAR

6543

21]1,1[

Regression Method

RELSIZE – ln(price paid/bidder market capitalisation) LEV – bidder financial leverage (financial debt/financial debt + equity) MOP – method of payment, dummy variable 1 if cash used, otherwise 0 PUBLIC – Type of target, dummy variable of 1 if the target is publicly listed,

0 otherwise BVMV – Book-to-market ratio calculated as book value equity/market value

equity HHPROP – measure of focus/specialisation by property type, calculated

as:i

iwHHPROP 2

Data

• Successful A-REIT M&A’s bidders were identified from the Connect 4 Takeovers Database from Jan 1996 to Dec 2010.

• Daily share price data was obtained from Bloomberg.

• Accounting data (leverage, specialisation) was collected from the Connect 4 Annual Reports collection and ASX.

• A total of 56 transactions were identified.

Announcement of 56 M&As by Year

Year # announce Year # announce

1996 2 2004 6

1997 0 2005 2

1998 1 2006 3

1999 5 2007 5

2000 9 2008 0

2001 8 2009 1

2002 2 2010 5

2003 7 Total 56

Descriptive statsAll Obs (n = 56) Mean Median Max Min S.D.Mkt Value of Bidder ($M) 2120.40 1484.49 8056.72 45.34 2077.33Value of Acquisition ($M) 697.66 378.74 8460.53 20.00 1213.28Relative Size of Acquisition 0.466 0.299 2.435 0.019 0.515

Public-Public (n = 44) Mean Median Max Min S.D.Mkt Value of Bidder ($M) 2431.86 1619.75 8056.72 184.09 2172.43Value of Acquisition ($M) 833.32 427.92 8460.53 24.98 1337.91Relative Size of Acquisition 0.435 0.290 2.362 0.019 0.461

Public-Private (n = 12) Mean Median Max Min S.D.Mkt Value of Bidder ($M) 978.38 676.72 3981.09 45.34 1147.58Value of Acquisition ($M) 200.22 177.75 475.00 20.00 139.28Relative Size of Acquisition 0.580 0.386 2.435 0.051 0.688

Means Test Diff in means p-valueMkt Value of Bidder 1453.48 (0.030)**Value of Acquisition 633.11 (0.109)Relative Size of Acquisition -0.146 (0.389)

Panel A: A-REIT Bidders

Total sample (n = 56)Cash (n = 30)_____ Combination (n = 26)

Interval CAR pValue CAR pValue CAR pValue[-2,+2] 0.880% (0.015)** 0.321% (0.622) 1.231% (0.016)**[-1,+1] 0.966% (0.001)*** 0.174% (0.707) 1.463% (0.001)***

Panel B: Public-Public

Total sample (n = 44)Cash (n = 21)_____ Combination (n = 23)

Interval CAR pValue CAR pValue CAR pValue[-2,+2] 0.326% (0.079)* -0.345% (0.840) 0.714% (0.030)**[-1,+1] 0.457% (0.017)** -0.286% (0.776) 0.947% (0.002)***

Panel C: Public-Private

Total sample (n = 12)Cash (n = 9)______ Combination (n = 3)

Interval CAR pValue CAR pValue CAR pValue[-2,+2] 2.914% (0.060)* 1.876% (0.228) 5.196% (0.272)[-1,+1] 2.834% (0.022)** 1.246% (0.262) 5.419% (0.349)

Event study results

***, **, * statistical significance at 1%, 5% & 10% level

Results – regression modelPanel A Panel B

No. of Obs   No. of Obs  Variable 56 (p-value) 54^ (p-value)Intercept 0.100 (0.049)** 0.027 (0.193)RELSIZE 0.003 (0.689) -0.006 (0.232)LEV -0.181 (0.050)** -0.079 (0.125)MOP -0.022 (0.179) -0.021 (0.034)**PUBLIC -0.043 (0.116) -0.005 (0.435)BVMV -0.029 (0.019)** -0.022 (0.042)**HHPROP 0.033 (0.088)* 0.028 (0.033)**

R2 0.259 0.266Adjusted R2 0.168 0.173White Test 46.007 (0.006) 39.260 (0.035)Jarque-Bera 122.200 (0.000) 3.358 (0.187)Ramsey Reset 25.402 (0.123) 22.166 (0.020)

Values corrected for hetroskedasticity. ^ Reported figures corrected for outliers. ***, **, * show statistical significance at the 1%, 5% and 10% level respectively.

Conclusion

• Acquiring A-REITs enjoy positive & significant CARs• Choice of payment is important • Bidding A-REITs earn higher CARs when target is

private• BVMV suggests investors penalise high BVMV A-

REITs in a M&A due to their higher risk characteristics.

• Specialisation has a positive impact on CARs