The Impact of Merger Announcements on Stock Prices: The Application of Event-Study Analysis to a Historical Issue
Discussion held on 26th April 2003 Economics and Business Historical Society Conference Gerhard Kling Department of Economics University of Tuebingen
I
T
R
C
Overview
Introduction
Theoretical Background Empirical Results Concluding Remarks
Event Study Merger Paradox CMR model Abnormal returns
Cumulated Effects Cross-sectional model
Pros and cons Further research
Gerhard Kling
2
I
T
R
C
Introduction
What is the Merger Paradox? Empirical finding that acquiring firms loose from mergers
Can we detect the merger paradox in the first phase of globalization? Event Study method + draw a sample (year 1908)
Gerhard Kling 3
I
T
R
C
Theoretical background
• Basic concept of event-studies – normal returns - up to six models – definition of the estimation period – construction of the event period Caution! – deriving the test statistics - two approaches! • Why do we choose the CMR model? – lacking market index – simple model - modifications easily embedded – similar results (see Warner and Brown 1980, 1985)
Gerhard Kling 4
I
T
R
C
Illustration of CMR and AR
return Rit +error term eit ARit return of stock i at t
ˆ ARit Rit i
ˆ
upper and lower bound
estimation period
event period time t event day 5
ˆ ˆ ; Var ( )
Gerhard Kling
I
T
R
C
AR and CAR Measurement
• normal range of returns • returns outside range - abnormal returns AR • accumulation of AR over firms and over time
– whole effect should be captured – which group wins? – division into subgroups (target; acquiring firm)
Gerhard Kling
6
I
T
R
C
Target and acquiring firms
5,8 4,8 3,8
2,8
1,8
0,8
-0,2
11
13
15
17
19
21
23
25
27
29
Acquiring firms
Target firms
Gerhard Kling
31
1
3
5
7
9
7
I
T
R
C
Results - What can we learn?
• • • • • • • Both groups gain from mergers Merger paradox rejected Reasons unknown - further research Adaptation takes only a few days Market is highly informationally efficient Pre-merger gains - insiders? Not enough insights - cross-sectional model
Gerhard Kling
8
I
T
R
C
Cross-sectional model
• Objective: What influences success of mergers? • Explanatory variables:
– – – – age of the firm (experience) market capitalization (firm size) growth rate of dividend payment (profitability) dummies: target, cash payment, change of management, success (approval), lines of business (banking and mining industry)
9
Gerhard Kling
I
ˆ C iindex 0 1 log capi 2 log agei 3 Successi 4 Changei 4 Cashi
T
R
C
5 DivGrowth i 6 Banki 7 Miningi 8T arg eti ui
Explanatory variable Intercept Log(capi) Log(agei) Successi Changei Cashi DivGrowthi Banki Miningi Targeti Number of Observations Adjusted R
2
Coefficients 4.5720 -0.0494 -0.4532 -0.6415 -0.4284 -0.1943 0.5855 0.4775 0.6891 -0.2004 45 0.14 1.80 (0.104)
p-value 0.000 0.592 0.009 0.266 0.916 0.937 0.089 0.098 0.074 0.435
F-Test (p-value)
White Test NR2 (p-value) 43.68 (0.177)
Gerhard Kling
10
I
T
R
C
What do I mean with causality?
direct
Success of Merger CAR
indirect
Firm Size
Estimated Normal Return
Gerhard Kling
11
I
ˆ C iindex 0 1 log capi 2 Successi 3 Changei 4 Cashi
T
R
C
5 DivGrowth i 6 Banki 7 Miningi 8 Meani ui
Meani 0 1 log capi 2 log agei 3 Successi 4Changei 5Cashi
6 DivGrowth i 7 Bank i 8 Mining i vi
Explanatory variable Equation1: Dependent Variable CAR Intercept Log(capi) Log(agei) Successi Changei Cashi DivGrowthi Banki Miningi Targeti Meani Number of Observations “Adjusted R ” F-Test (p-value)
2
Equation2: Dependent Variable Meani -0.3992 (0.002) 0.0131 (0.419) 0.0958 (0.001) 0.0781 (0.438) 0.0209 (0.771) -0.0571 (0.189) -0.0512 (0.389) -0.0154 (0.757) -0.1612 (0.017) 0.0294 (0.513) 45 0.36 2.83 (0.007)
2.6831 (0.000) 0.0127 (0.870) -0.2718 (0.545) 0.0561 (0.857) -0.2898 (0.144) 0.3434 (0.200) 0.4045 (0.068) -0.0735 (0.847) -0.0611 (0.759) -4.7322 (0.001) 45 0.47 2.96 (0.005)
Gerhard Kling
12
I
T
R
C
Concluding remarks
• Merger paradox rejected • Additional insights using simultaneous equation model
– banking industry exhibits larger gains – targets and acquiring firms do not differ significantly
• Insiders versus outsiders
– Different ways of disclosure – Does regulation protect outsiders?
Gerhard Kling 13