1 Andrea Resti Università di Bergamo [email protected]How Mergers and Acquisitions affect listed Banks* *Co-author: Lucia Galbiati Outline of this presentation • Experimental Setup – Aims and scope – Methodological Framework – Previous Literature – Sample • Results – Overall Results – Univariate Analysis of targets and bidders – Multivariate Analysis of targets – Combined (target+bidder) value creation
18
Embed
How Mergers and Acquisitions affect listed Banks* · • Zollo and Leshchinkskii (2000) – Value creation margins for the bidder’s stakeholders depend on: bidder’s ... (see next
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
• Results– Overall Results– Univariate Analysis of targets and bidders– Multivariate Analysis of targets– Combined (target+bidder) value creation
����������
Aims and Scope of the Research
• In the last 15 years a huge wave of bank M&Astook place in the main advanced countries– Did those M&As create or destroy value?– What types of M&As scored comparatively better?
• Two approaches exist to answer such questions:– Analyze efficiency/profitability ratios taken from the
financial statements• Takes time (e.g., three years after the merger)• Aggregates may not be homogeneous
– Analyze the market’s reaction (shift in stock prices) when the M&As were announced
• Did the market “believe” in those deals? And how much?• This is the approach followed in this research.
3
����������
Why to focus on Stock Prices?• Prices are the result of the behavior of many
informed, reactive, motivated investors– They incorporate an evaluation of the past, as well as
expectations on future developments• Prices are a signal, through which minority
shareholders can express their dissent• In short, prices embed the market’s opinion on the
economic viability of the discosed merger plan
The target’s shareholders
cash in the “majority
premium” paid for corporate
control
The target’s shareholders
cash in the “majority
premium” paid for corporate
control
The buyer’s shareholders may not “believe” in the value creation objectives set out by their own managers
The buyer’s shareholders may not “believe” in the value creation objectives set out by their own managers
• Results– Overall Results– Univariate Analysis of targets and bidders– Multivariate Analysis of targets– Combined (target+bidder) value creation
4
�
����������
Methodological Framework: CAR (Cumulated Abnormal Returns)
Market returns (e.g. S&P500)
Ret
runs
on B
ank
XY
+5,5%
+5%
+4%
+3%
In “normal” times, the returns on a stock reactto those fo the market, according to their “beta”
After an M&Aannouncement, returnsmay differ from those“forecast” by the beta
We can sum all redvertical segments in the days around the M&Aannouncement, to checkif this cumulatedabnormal return (CAR) isstatistically differentfrom zero.
����������
Cumulated Abnormal Returns:some inferential tests
�
car Nt1 =
scar Nt2 =
1/2N
1i
2i
N
1i
2ii
3
1/�
/�ract
��
���
��
�=
=
=ˆ
( )
B
2B
A
2A
1
N�
N�
carcar t
+
−= BA
Average CAR on N deals
Average variance of the N CARs
Mean of the standardised CARs
Weighted average CAR (weightsdecrease as variances increase)
• Results– Overall Results– Univariate Analysis of targets and bidders– Multivariate Analysis of targets– Combined (target+bidder) value creation
��
����������
A brief Literature Review:some Studies on US Banks
• Hannan and Wolken (1989)– A wealth transfer takes place between the buyer’s and the target’s
shareholders. No evidence of net wealth creation is found– The value decrease in the buyer’s stock is stronger when the target’s
size is large (suggesting a less “manageable” M&A)• Hawawini and Swary (1990)
– The buyer’s value decrease is more than offset by the target’s rise Houston and Ryngaert (1994)
– Bidders are more efficient than targets, and than the industry average– The market reacts better to M&As announced by high-performance
buyers and between banks with a strong geographic overlapping• Ferretti (2000), bidders only
– Prices do not react univocally: CARs look more negative in recent years• Zollo and Leshchinkskii (2000)
– Value creation margins for the bidder’s stakeholders depend on: bidder’s size, target’s asset quality, bidder’s management skills during the implementation phase
6
��
����������
A brief Literature Review/2:some (few!) Studies on Europe
• Structural limitations– Less observations are available– Cross-country analyses are hindered by national segmentations
• Beitel and Schiereck (2001) – Based on 54 European M&As: most have prompted a favourable
reaction in prices– This is especially true for domestic deals, while cross-border
M&As have induced a value decrease• Cybo-Ottone and Murgia (2000)
– Cross-border mergers in Europe destroy value– The best CARs are associated with domestic M&As aimed at
gaining a significant overall market share and/or at diversifying earnings. “Manageable” size deals are also rewarded by the market
Results/2: univariate analysis before and after the “price bubble”
after
before
Target’s CARsincrease (from
+10.4% to +22.3%), bidders suffer less.
Target’s CARsincrease (from
+10.4% to +22.3%), bidders suffer less.
��
����������
���
�
��
� �
���
� �
���
" �
"��
� �
�� �
5�2�!�(�(!��!�4�!������)�����!��2�
���
'�&�� ����2�/�
'�&�� �����2�/�
µ#σ$√�
µ�σ$√�
µ#σ$√�
µ�σ$√�
Results/2: univariate analysis targets before/after the “bubble”
After the bubble targets continue to earn long after the announcement: the rally is not only driven by the bid price but also by an evalutation of the value creation prospects for the target?
After the bubble targets continue to earn long after the announcement: the rally is not only driven by the bid price but also by an evalutation of the value creation prospects for the target?
14
��
����������
�!�
�(�
���
���
�
��
��
(�
!�
� �
���
�� �
5�2�!�(�(!��!�4�!������)�����!��2�
���
3�,,� ����
2�/�
3�,,� �����2�/�
µ#σ$√�
µ�σ$√�
µ#σ$√�
µ�σ$√�
Results/2: univariate analysis bidders before/after the “bubble”
During the bubble, CARs are markedly and persistently below zero. After the bubble, the
decrease is limited and quickly compensated in the following 20 days.
Are deals perceived as more credible in a “poor”
market environment?
Did banks improve their investor-relation skills over the latest years?
During the bubble, CARs are markedly and persistently below zero. After the bubble, the
decrease is limited and quickly compensated in the following 20 days.
Are deals perceived as more credible in a “poor”
market environment?
Did banks improve their investor-relation skills over the latest years?
Payments-in-cash, correlation (=diversification) and cross-border weresubsequently eliminated from the model, not being significant at the 5% level(blue-shadowed cells). Profitability, efficiency, relative size, capital tightenessand market conditions (“post2000”) seem to explain some 40% of total variance.
• Results– Overall Results– Univariate Analysis of targets and bidders– Multivariate Analysis of targets– Combined (target+bidder) value creation
17
��
����������
Results/4: the combined effect
• Targets (smaller) enjoy a considerable rise in stock prices. Bidders (larger) suffer from a limited decrease.– Does this imply a zero-sum game?– Is the combined effect (and the net value creation) significantly different from
zero?• To check this, we simulated the investment in a stock portfolio including both
bidder and target (“combined”)– Both companies must be listed (69 observations)– Weights are given by market capitalizations
Estimation window Event window
EventTime
Combined Estimate betaCompute CAR
�
����������
-60%
-40%
-20%
0%
20%
40%
60%
80%
100%
120%
140%
-40% -30% -20% -10% 0% 10% 20%
Car bidder
Car
targ
et
ρρρρ = +12%
Results/4: the combined effect
Group 1 Group 2 test p-valueDomestic (60) vs. cross-border (9) -1.10% -2.60% 0.5 59.1%Before (38) vs. after (31) March 2000 -3.20% 1.20% -2.1 3.5%Target TA / bidder TA < 20% (30) or > 20% (24) -2.10% -1.00% -0.5 62.5%Italian (13) vs. other European (17) -1.60% -0.20% -0.4 65.8%European (30) vs. US (39) -0.80% -1.60% 0.4 69.6%
CARs Equal means?
•The average combined CAR is -1.3%, not significantly different from zero.•However, no negative correlation isobserved between target and bidderCARs on an individual basis
•This is not a zero-sum game!•Univariate segmentations (see below) do not look significant, the onlyexception being the March 2000 watershed
18
�
����������
Summary of the main findings
• Target and bidder CARs are significantly different from zero– Bidders suffer less when targets are small
• Bidders’ CARs revert to zero in the long term– Targets earn more when bought from abroad, by large
banks and when the market trend is gloomy• Multivariate analysis confirms this, while showing that target’s
profitability and efficiency, as well as bidder’s capitalizationlevels, also matter.
• Combined CARs are not, on average, nonzero– However, no systematic “value transfer” emerges– Some deals actually destroy value (e.g., before 2000),