Good Signal or Bad Signal? Impact of CEO Fame on Takeover Outcomes ()
1. Introduction
*Corresponding author.
In the Modigliani-Miller world where information is symmetric and financial markets are frictionless, personal characteristics of CEOs do not affect the offer price of merger and acquisition (M&A). The purchase price of a deal is determined by the pre-acquisition fair value of the target and the synergy of the combined entity. The synergy is the M&A premium and is recorded as goodwill in the accounting book. As long as the corporate marriage of the acquirer and target continues, the synergy will be sustained and there will be no subsequent impairment of goodwill.
Yet in the real world, the information is asymmetric and there are frictions in financial markets. This information asymmetry creates agency problems that deviate from the principle of maximizing shareholder values (Jensen [1]). Empirical evidence finds that the personal characteristics of CEO very often play a crucial role in the negotiated price of a M&A deal in an asymmetric information real world. For example, Roll [2] developed the hubris hypothesis of managers to explain corporate takeovers. Malmendier and Tate [3] find that overconfident CEOs of acquirers often take an aggressive role to complete an overpriced deal. Overconfident CEOs may often believe that they can manage a target firm better than the incumbent CEO. The market responses to these M&A deals are a significant decline in stock prices of acquiring firms. It has been demonstrated that in 75% of mergers, where the acquiring stockholders face financial losses, the CEOs of the acquiring companies still end up in a better financial position (Harford and Li [4]).
This paper examines the impact of CEO fame on the outcomes of M&A. To our best knowledge, this paper is among the first to show that CEO’s fame is a signal to differentiate a good acquisition from a bad one under an asymmetric information environment. Based on our empirical analyses of more than 8000+ M&A deals by Chinese listed firms during the pre-COVID-19 era of 2013-2017, we find that a deal with a famous CEO of the acquiring firm is a bad signal to investors, whilst a deal with a famous CEO of the target firm is a good signal to investors.
In our research, we supplement deal and firm-level data with hand-collected information from online media data sources to identify the famous CEOs. We refine the approach of Jiang, Wan, and Zhao [5] to define a famous CEO if the number of times the CEO’s name is mentioned in the online media from year t − 3 to year t − 1 exceeds the statistical median, where year t is the year of the merger and acquisition transaction. CEO’s fame mainly stems from their track record in managing businesses including leadership qualities, innovation, and impact within the industry.
We find that a famous CEO of the acquiring firm delivers a bad signal because it paid an overpriced M&A deal with high goodwill that will eventually undergo a subsequent impairment, lowering value creation for shareholders. The market reaction to the deal is a significantly lower abnormal return after the announcement of the acquisition. A famous CEO, who has especially attained fame recently, often exhibits overconfidence in his experience and ability and is under pressure to gain the recognition from his shareholders that he could create value for them (Hayward, Rindova, and Pollock [6]). Therefore, a famous acquiring CEO is willing to be aggressive to close a deal even if the target is overpriced. In other words, he made a poorer decision that ended up with his acquiring stockholders paying a premium over the dollar value of the synergy of the acquisition. An efficient financial market senses the bad signal of the deal and reacts rationally with a lower abnormal return. Subsequent auditing verifies the financial loss of the deal and requires goodwill impairment for the post-acquisition firm.
On the other hand, we find that a famous CEO of the target firm sends a good signal to the investors. Target CEOs hold a pivotal position in M&A deals. They significantly influence their firm’s strategic decisions prior to a bid, such as determining whether to pursue a buyer or initiate acquisition discussions. In addition, when a bid is presented, they lead the firm’s efforts in responding to the bid and conducting negotiations with prospective buyers (Graham, Harvey, and Puri [7]). A famous target CEO delivered significant value for shareholders. She requested the acquirer to pay a fair value of the target, plus a fair value of goodwill that accurately reflected the underlying synergy of the acquisition, thereby creating a higher value for shareholders. The market reaction to the deal is a significantly higher abnormal return after the announcement of the acquisition. The famous target CEO fully utilized her experience and ability and is successfully gaining recognition from shareholders as she created value for them.
However, this is a tough job for the target CEO. It is well established fact that 1) the target CEO will usually lose her job after a successful acquisition and she will be replaced by the acquiring CEO, and 2) it is only rare for a departing CEO to find a senior management position in another listed firm (e.g., see Martin and McConnell [8]). This implies an M&A deal imposes large career costs on the target CEO. Therefore, an ordinary target CEO with serious career concerns has a strong private incentive to utilize her private information to mislead shareholders and to resist the bid, hampering a value-enhancing deal for shareholders.
There is a group of target CEOs who may not experience acute career-related concerns when they are close to the retirement age of 65. Jenter and Lewellen [9] find strong evidence that these target CEOs are more prepared to accept acquisition bids because of their lower personal costs after a takeover deal is closed. Their findings imply that these target CEOs’ private interests still have a significant impact on target firms’ acquisition decisions that might not have maximized the interests of target shareholders.
Yet a renowned CEO of a target company is not an ordinary CEO preoccupied with career concerns and also won’t behave like a near-retired CEO. Due to her esteemed professional reputation, she holds greater bargaining power compared to an ordinary target company CEO, both internally within the firm and externally in the executive professional market. Internally, her reputation enables her to bargain for a better compensation package including golden parachutes, special bonuses, and unscheduled equity grants during acquisition negotiations that generate additional income for her in the event the target firm is sold (Fich, Cai, and Tran [10]). Therefore, she is unlikely to resist a value-enhancing deal for shareholders.
Externally, her public image and professional reputation grant her a competitive edge, making it advantageous for her to secure a senior position at another listed company compared to an ordinary departing CEO. As a result, the reputational effect mitigates the target CEO’s career concerns and her post-acquisition loss, reducing her resistance to takeovers. A famous CEO of the target, unlike a near-retired CEO, also will not easily accept a suboptimal offer bid to get her professional reputation tarnished. Taking together, it explains why a famous CEO of a target firm can do a better job to close a deal and to create a higher value for shareholders than an ordinary young CEO or a near-retired CEO near age 65.
Our study builds upon the market signaling literature pioneered by Spence [11]. It highlights the importance of a signal in distinguishing good acquisitions from bad ones, serving as a tool to mitigate information asymmetry. For example, Myers and Majluf [12] provided both theory and empirical evidence suggesting that firms prefer a stock-for-stock deal when their own stock prices are overvalued, indicating a stock payment deal a bad signal. This is because the post-acquisition share price will fall when the market realizes the overpriced deal. On the other hand, if managers have more private information than the market that the acquirer is undervalued, then they prefer a cash deal. This is because they will capture full gains from the post-acquisition share price recovery due to market adjustment. Hence, a cash payment is a good signal.
Our paper is related to Ishii and Xuan [13] who reveal that senior management’s personal connections between the acquirer and target result in weaker decision-making and diminished value creation for stockholders. It suggests that the cross-firm senior executive social ties are bad signals for M&As. Our study is also closely related to the work by Cho, Arthurs, Townsend, Miller, and Barden [14]. They provide empirical evidence that CEOs who have recently established celebrity status tend to make more M&A deals and to pay higher M&A premiums. Our study compliments and extends their work by providing a full picture of the impact of reputation on M&A from both perspectives of acquiring CEO and target CEO.
Our study is important for understanding M&A activities. Moeller, Schlingemann, and Stulz [15] discovered that just a few bad M&A deals during a four-year period of 1998-2001, which made substantial losses, had already wiped out all the cumulative gains of M&As from the previous two decades. They concluded that those big failures of M&As were consistent with the findings of Jensen [1] that if acquiring firms give their management more discretion, their management may make poor acquisition decisions if they prefer firm growth more than shareholder value.
The reminder of this paper is organized as follows. Section 2 introduces the institutional background. Section 3 develops our main signaling hypotheses. Section 4 presents empirical results. Finally, Section 5 concludes the paper.
2. Institutional Background
In the pre-COVID-19 era, governments around the world introduced various policies and regulations to stimulate economic development. Among these, China’s “New National Nine Articles” policy, officially launched in 2014, effectively spurred a wave of acquisitions and acquisitions in the Chinese capital market in the following years, providing empirical evidence that is consistent with the findings of Shleifer and Vishny [16]. Between 2013 and 2017, Chinese listed companies participated in over 26,000 M&A transactions, including more than 2000 deals each valued at over one billion yuan (about 140 million USD) and approximately 13,000 deals each valued at less than 100 million yuan (about 14 million USD).
M&A activities frequently involve substantial premiums, prominently reflected by goodwill in accounting records under the current International Financial Reporting Standards (IFRS), which has mostly been adopted by China since 2006. If goodwill is overvalued and the target company’s post-M&A performance fails to justify the premium paid, the corresponding goodwill may be subject to impairment, which could negatively impact the firm. The accumulated goodwill in China’s capital market from 2013 to 2017 reached a record of 3.6 trillion yuan (about 500 billion USD), placing significant pressure on the profitability of Chinese listed firms to sustain their goodwill.
Our research aims to examine which mergers and acquisitions (M&As) create value and which diminish it within an environment characterized by asymmetric information. As with most empirical research on M&A, quantifying post-acquisition synergy effects is often challenging due to limited disclosure. To address this, we employ accounting measures, specifically goodwill impairment loss, to infer post-acquisition synergy. This approach allows us to evaluate M&A premiums through the lens of our newly developed signaling mechanism centered on CEO fame.
3. Hypothesis Development
3.1. Impact of CEO’s Fame on Goodwill
The reputation of the acquiring CEOs typically reflects their experience of successful management capabilities and stronger management ability expectation pre- and post-M&A, that can also instill greater confidence in the post-acquisition integration of the target company and the realization of synergies (Roll [2]). Additionally, the famous CEOs are more willing to be involved in more M&A projects and, nevertheless, create the risk of overconfidence (Malmendier and Tate [17]). That may lead to higher M&A deal prices and thus increase the acquisition premium and goodwill, lowering value creation for shareholders. In addition, famous CEOs may be more eager to win the recognition of their shareholders for their potential ability. They may use their available accounting discretion to record higher goodwill in M&A deals to give the shareholders the impression that deals make positive synergy so that shareholders may benefit from high post-acquisition earnings (Shalev, Zhang, and Zhang [18]).
From the perspective of the target company, the target CEO’s fame indicates her unique vision and more negotiation skills during M&A. Thus, a famous target CEO can make more accurate estimates of the target company’s profitability and arrive at the underlying price of M&A, thereby reducing the likelihood of an underpriced deal. The CEO’s fame often stems from their operational performance and leadership, serving as a positive signal of the target firm’s sustained profitability.
Thus, we propose the following hypotheses H1a and H1b:
H1a: The impact of acquiring CEO’s fame on goodwill is positive.
H1b: The impact of target CEO’s fame on goodwill is positive.
3.2. CEO Fame as Signal and Market Reaction
Existing literature provides evidence that a famous acquiring CEO might offer high M&A premiums due to overconfidence and excessive optimism (Cho, Arthurs, Townsend, Miller, and Barden [14]). Such premiums are often difficult to be upheld by the long-term performance of the target company. Furthermore, synergies are challenging to be quantified at the time of an M&A deal due to the presence of asymmetric information. Consequently, investors might interpret the involvement of a high-profile acquiring CEO as a negative signal for the deal, prompting a less favorable market reaction.
From the perspective of the target firm, a famous CEO of the target, which signifies their management capabilities and leadership, may serve as a positive signal for the acquisition premium and an assurance for sustained profitability post-M&A. Therefore, the target CEO’s fame may lead to a positive market reaction.
Taking together, we propose the following hypotheses H2a and H2b:
H2a: The market reaction to the fame of an acquiring CEO is relatively less favorable.
H2b: The market reaction to the fame of a target CEO is relatively positive.
3.3. Impact of CEO’s Fame on the Post-M&A Impairment
As mentioned earlier, a famous acquiring CEO, due to overconfidence or pressure for recognition, is more likely to pay an excessively high M&A premium and may fail to integrate the combined entity—such as aligning organizational cultures, systems, and processes—to achieve post-M&A synergies. This implies that the post-acquisition combined entity may not meet performance expectations and has to impair, hurting the shareholders’ interests. Indeed, Harford and Li [4] reveal that in three-quarters of mergers where acquiring shareholders incur losses, the acquiring CEOs’ wealth end up better off. Their findings indicate that famous CEOs with significant bargaining power over their boards can leverage negotiation opportunities to shield their personal interests from the potential risks associated with the acquisition.
On the other hand, the fame of the target company’s CEO built upon her experience and management skills is in a better position to negotiate the M&A deal with less career concerns. She can more effectively justify a fair value of the M&A premium that can be sustained by post-acquisition profitability, thereby preventing impairment of the associated acquisition premium and goodwill.
Therefore, we propose the following hypotheses H3a and H3b:
H3a: The impact of the acquiring CEO’s fame on goodwill impairment is positive.
H3b: The impact of the target CEO’s fame on goodwill impairment is insignificant.
4. Methodology and Measurement of Variables
4.1. Data and Sample
The data of M&A events are downloaded from Eastmoney website. We include M&A events that acquirers are publicly traded company in China A-stock markets of Shanghai and Shenzhen. The firm-level financial and accounting variables are from the CSMAR (China Stock Market and Accounting Research) database. Our final sample consists of 8056 M&A events between 2013 and 2017. We find that all the target companies in our sample are private firms. This is quite normal as mergers between two listed companies are very rare in China. After excluding samples with missing observations of some key variables and missing data for CAR (−3, 3) (acquirer’s cumulative abnormal return within 3 days before and after M&A disclosure date), we finally selected a sample of 7925 M&A events for goodwill regressions, a sample of 6144 events for CAR (−3, 3) study, and a sample of 8056 events for impairment analysis. Table 1 presents our sample selection procedures.
Table 1. Sample selection procedures.
|
Number of Acquisitions |
Acquisition events during 2013 and 2017 where the acquirers are publicly traded |
8056 |
Less samples of missing key data |
131 |
|
Less samples of missing CAR (−3, 3) data |
|
1912 |
Acquisitions remaining |
7925 |
6144 |
This table reports the sample selection procedures of M&A events from 2013 to 2017. We start with 8056 successful M&A events between 2013 and 2017.
4.2. Measurement of Variables
4.2.1. CEO’s Fame
The existing financial literature measures CEO fame by counting the frequency of a CEO’s name appearing in traditional media outlets such as newspapers and news agencies. For example, Milbourn [19] and Rajgopal, Shevlin, and Zamora [20] utilize the Lexis/Nexis database that compiles the articles from the major U.S. and global business newspapers and wire services. They construct a proxy for CEO reputation based on the number of articles mentioning the CEO’s name. Similarly, Jiang, Wan, and Zhao [5] measure the reputation of directors by counting the frequency of their names appearing within the top six official Chinese newspapers.
In our study, we have refined this measurement. Over the past two decades, the internet in China has undergone rapid development. Many traditional newspapers have been overshadowed by the rise of online social media, whose influence has grown exponentially. To reflect the current reality in China, we measure the fame of CEOs by the number of online media reports that mention their names. Specifically, we use the Baidu, the most powerful search engine in China, to conduct our online search across media platforms and websites. Since Google exited the Chinese mainland market in 2010, Baidu became the most widely used search engine in mainland China during our sample period of 2013 to 2017, holding a significant market share of 76% in the country. Baidu is tailored specifically for Chinese users, prioritizing Simplified Chinese content and indexing Chinese web pages more effectively than other search engines.
For the acquiring CEO, we extract and match the list of executives for the acquisition year from the CSMAR database. Based on the acquiring firm’s name and its CEO’s name, we use Python to collect relevant online media links from Baidu search results, totaling 200,000 entries.
Target company’s CEOs tend to have limited media exposure, primarily due to the unlisted status of their firms. The most relevant and abundant information regarding these CEOs often comes from merger and acquisition (M&A) announcement texts. Based on the initial and completion announcement dates of M&A events in our full sample, we used Python to collect and download both announcement links for each M&A event from the official disclosure website of Chinese listed companies, Juchao Zixun (www.cninfo.com.cn), resulting in approximately 16,000 announcement links. Then, we extract and organize the names of the target companies and their CEOs, totaling around 12,000 names. We use Python to collect Baidu search links from online media for these target firm names and their CEO names, yielding 50,000 entries.
We calculate the frequency of each respective acquiring and target CEO’s name appearing in online media coverage during the period from year t − 3 to year t − 1, where year t represents the year of the merger and acquisition (M&A) announcement. CEOs are classified as famous if the frequency of their names in online media coverage exceeds the sample statistical median. This process allows us to identify the famous CEOs of both acquiring and target firms for each acquisition pair in our sample, which comprises 8056 M&A events.
Our variable of CEO’s fame is ACQ_FAME (TRG_FAME), which is a dummy variable that equals to one if there is a famous CEO in the acquiring (target) firm and zero otherwise. Table 2 presents sample distribution of famous CEOs over time. Since acquirers in our sample are Chinese listed companies, there is a larger number of famous CEOs among the acquiring companies. Consequently, 40% (= 3279/8056) CEOs of the acquiring firms are famous. The proportion of the target firm’s famous CEO is only 1.4% (=103/8056), which is relatively low. This is because all target firms in our sample are privately held and relatively small in size.
Table 2. Sample distribution of famous CEOs over time.
Year |
(1) Famous
Acquiring CEO |
(2) Famous Target CEO |
(3) No Famous CEO |
(4) Total |
(5) % |
2013 |
120 |
6 |
370 |
494 |
6.13% |
2014 |
418 |
12 |
641 |
1064 |
13.21% |
2015 |
542 |
40 |
904 |
1466 |
18.19% |
2016 |
862 |
22 |
1089 |
1954 |
24.25% |
2017 |
1337 |
23 |
1733 |
3078 |
38.21% |
Total |
3279 |
103 |
4737 |
8056 |
100% |
This table reports sample distribution of famous CEOs over time. Column (1) reports the number of acquirers that have famous CEOs. Column (2) reports the number of target firms that have famous CEOs. Column (3) presents the number of acquirer-target pairs that do not have any famous CEOs. Column (4) provides number of M&A events. Column (5) provides sample distribution over time. Note that columns (1), (2), and (3) do not sum up to column (4) due to overlap famous CEOs in both acquiring and target firms in some of the M&A events.
Table 3 presents sample distribution of famous CEOs by industry. The industry classification is based on the China Securities Regulatory Commission. We find that the top-3 famous CEO concentrated industries are 1) manufacturing (acquirer 64.7%, target 59.2%), 2) information transmission, software and information technology services (acquirer 11.6%, target 21.4%), and 3) wholesale and retail (acquirer 4.6%, target 3.9%). The manufacturing industry attracts more than 50% of renowned CEOs because it has long history and operates on a massive scale and is a hub for innovation and technology, making its leaders highly visible and influential.
4.2.2. Goodwill
The dependent variable, GOODWILL, is defined as goodwill divided by target’s fair value. In this study, we used the goodwill accounting concept to measure the M&A premium. In the subsequent goodwill impairment analysis, we can observe each impairment occurrence that corresponds to the goodwill of each original M&A event. It provides the most direct and accurate measurement for us to test post-M&A performance of the combined entity.
4.2.3. CAR (−3, 3)
CAR (−3, 3) is our second dependent variable. It is the seven-day cumulative abnormal return of the acquirer around the acquisition announcement date. Announcement dates are from the Eastmoney website. We use Shanghai and Shenzhen stock exchanges’ composite value-weighted index as the market portfolio and estimate the parameters of the market model using returns over the 250-trading-day period from event day −280 to event day −31 (day 0 is the acquisition announcement date). We regress the daily returns of the acquirer on the value-
Table 3. Sample distribution of famous CEOs by industry.
Code |
(1) Industry |
(2) Famous Acquiring CEO |
(3) Famous Target CEO |
(4) No Famous CEO |
(5) Total |
(6) % |
A |
Agriculture, forestry, animal husbandry and fishery |
31 |
0 |
100 |
131 |
1.63% |
B |
Mining industry |
39 |
2 |
89 |
130 |
1.61% |
C |
Manufacturing industry |
2,123 |
61 |
2,540 |
4,684 |
58.14% |
D |
Industry of electric power, heat, gas and water production and supply |
47 |
1 |
203 |
250 |
3.10% |
E |
Construction industry |
103 |
1 |
94 |
197 |
2.45% |
F |
Wholesale & retail industry |
150 |
4 |
243 |
394 |
4.89% |
G |
Transport, storage and postal service industry |
76 |
1 |
106 |
182 |
2.26% |
H |
Accommodation and catering industry |
7 |
2 |
12 |
20 |
0.25% |
I |
Industry of information transmission, software and information technology services |
379 |
22 |
506 |
895 |
11.11% |
J |
Finance industry |
44 |
1 |
101 |
146 |
1.81% |
K |
Real estate industry |
44 |
1 |
283 |
327 |
4.06% |
L |
Leasing & commercial service industry |
54 |
3 |
108 |
164 |
2.04% |
M |
Scientific research & technical service industry |
43 |
0 |
52 |
95 |
1.18% |
N |
Water conservancy, environment and public facility management industry |
47 |
1 |
130 |
178 |
2.21% |
P |
Education industry |
11 |
0 |
9 |
20 |
0.25% |
Q |
Health and social work |
24 |
1 |
34 |
59 |
0.73% |
R |
Industry of culture, sports and entertainment |
48 |
2 |
100 |
148 |
1.84% |
S |
Diversified industries |
9 |
0 |
27 |
36 |
0.45% |
|
Total |
3279 |
103 |
4737 |
8056 |
100% |
This table presents sample distribution of famous CEOs by industry. Column (2) reports the number of acquirers that have famous CEOs. Column (3) reports the number of target firms that have famous CEOs. Column (4) presents the number of acquirer-target pairs that do not have any famous CEO. Column (5) provides number of M&A events. Column (6) provides sample distribution by industry. Note that columns (2), (3), and (4) do not sum up to column (5) due to overlap famous CEOs in both acquiring and target firms in some of the M&A events. The industry classification is based on the China Securities Regulatory Commission.
weighted market return on the Shanghai and Shenzhen stock exchanges. The daily abnormal return is defined as the difference between the actual daily returns of the acquirer and the expected daily returns of the acquirer from the estimated market model. Following the recent literature (e.g., Krüger, Landier and Thesmar [21]), we compute seven-day cumulative abnormal returns (CARs) over the seven-day (−3, +3) event window around the acquisition announcement date to measure the market response to the M&A announcement. This dependent variable serves as a proxy for external investors’ responses to the perceived signals regarding the CEO’s fame of the acquirer and the target involved in the M&A transaction.
4.2.4. Goodwill Impairment Loss
IMPAIRMENT is our third dependent variable, which is post-acquisition 3-year accumulated goodwill impairment loss divided by its original goodwill amount. Goodwill is a proxy for acquisition premium. In addition to the newly merged assets, the accounting calculation of goodwill impairment also includes progress of assessing overall post-M&A assets (synergies). Since goodwill impairment indicates bad news to the post-acquisition combined entity, the success of an M&A can be directly assessed by examining the extent of its impairment in the following years.
4.2.5. Control Variables
A batch of control variables are constructed following the existing literature to study M&A activities (e.g., Shalev, Zhang, and Zhang [18]; Chemmanur, Ertugrul, and Krishnanwe [22]; and Li [23]). All variables are as defined in Table 4.
TRG_RD is the average of target’s R&D expenses in 2 years pre-M&A divided by the offer price. TRG_SAL_GRO is the average of target’s sales growth rate in 2 years pre-M&A divided by the offer price. TRG_ADV is the average of target advertising expenses in 2 years pre-M&A divided by the offer price. TRG_ PPE is the ratio of target’s current assets to its net fixed assets in the year prior to the announcement of M&A. ACQ_BTM is the book value divided by market value of acquirer’s net assets in the year prior to the M&A transaction. SAME_IND is dummy variable that equals one if acquirer and target are in the same industry and zero otherwise (Morck, Shleifer, and Vishny [24]).
LACK_SLACK is dummy variable that equals one if the offer price exceeds the difference between market value and book value of acquirer in the year prior to the M&A transaction, and zero otherwise (Shalev, Zhang, and Zhang [18]). LOGTV is log of offer price plus one. ACQ_TA is the natural logarithm of the acquirer’s total assets in the year prior to the M&A transaction. ACQ_TQ is Tobin’s Q (the market-to-book ratio) of the acquirer in the year prior to the M&A transaction. ACQ_ROA is the acquirer’s return on assets in the year prior to the M&A transaction. IND_RD is the average ratio of R&D investment divided by total assets at the industry level in the year prior to the M&A transaction. IND_INTANG is the average ratio of intangibles divided by total assets at the industry level in the year prior to the M&A transaction. We also include year and industry fixed effects.
4.3. Descriptive Statistics
Table 5 reports the number of observations, mean, minimum, first quartile, median, third quartile, maximum, and standard deviation of all the variables. All continuous variables are winsorized at the 1% and 99% levels to limit the effect of outliers.
Table 4. Variable definitions.
Variable |
Definition |
Dependent variables |
GOODWILL |
Goodwill generated by an acquisition divided by target’s fair value. |
CAR (−3, 3) |
Acquirer’s seven-day cumulative abnormal return calculated using a market model estimated over the period [−280, −31] relative to the M&A announcement date (day 0). |
IMPAIRMENT |
Post-acquisition 3-year accumulated goodwill impairment loss divided by its original goodwill amount. |
Famous CEO dummies |
ACQ_FAME |
Dummy variable that equals one if the number of times the acquiring CEO’s name is mentioned in the online media from year t − 3 to year t − 1 exceeds the sample median, and zero otherwise. |
TRG_FAME |
Dummy variable that equals one if the number of times the target CEO’s name is mentioned in the online media from year t − 3 to year t − 1 exceeds the sample median, and zero otherwise. |
Target characteristics |
TRG_RD |
The average of target’s R&D expenses in 2 years pre-M&A divided by the offer price. |
TRG_SAL_GRO |
The average of target’s sales growth rate in 2 years pre-M&A divided by the offer price. |
TRG_ADV |
The average of target advertising expenses in 2 years pre-M&A divided by the offer price. |
TRG_PPE |
The ratio of target’s current assets to its net fixed assets in the year prior to the announcement of M&A. |
Acquirer characteristics |
ACQ_BTM |
The book value divided by market value of acquirer’s net assets in the year prior to the M&A transaction. |
LACK_SLACK |
Dummy variable that equals one if the transaction price exceeds the difference between market value and book value of acquirer in the year prior to the M&A transaction, and zero otherwise. |
ACQ_TA |
Log of acquirer assets in the year prior to the M&A transaction. |
ACQ_TQ |
Tobin’s Q, acquirer’s market-to-book ratio in the year prior to the M&A transaction. |
ACQ_ROA |
Acquirer’s return on assets in the year prior to the M&A transaction. |
M&A characteristics |
LOGTV |
Log of M&A’s offer price plus one. |
SAME_IND |
Dummy variable that equals one if acquirer and target are in the same industry and zero otherwise. |
Industry characteristics |
IND_RD |
Average ratio of R&D investment divided by total assets at the industry level in the year prior to the M&A transaction. |
IND_INTANG |
Average ratio of intangibles divided by total assets at the industry level in the year prior to the M&A transaction. |
For the dependent variable GOODWILL, which is the average ratio of goodwill divided by M&A offer price, is 8.5% out of 8056 M&A transactions. Among these, 1177 transactions generated goodwill, while 85% of the transactions did not generate goodwill. However, the highest ratio of goodwill divided by M&A offer price reached 92.7%.
Table 5. Descriptive statistics.
Variable |
obs |
mean |
min |
p25 |
median |
p75 |
max |
sd |
Dependent variables |
|
|
|
|
|
|
|
|
GOODWILL |
7925 |
0.086 |
0 |
0 |
0 |
0 |
0.927 |
0.235 |
CAR (−3, 3) |
6144 |
0.042 |
−0.317 |
−0.036 |
0.010 |
0.090 |
0.489 |
0.147 |
IMPAIRMENT |
8056 |
0.007 |
0 |
0 |
0 |
0 |
0.457 |
0.051 |
Famous CEO dummies |
|
|
|
|
|
|
|
|
ACQ_FAME |
8056 |
0.407 |
0 |
0 |
0 |
1 |
1 |
0.491 |
TRG_FAME |
8056 |
0.013 |
0 |
0 |
0 |
0 |
1 |
0.112 |
Target characteristics |
|
|
|
|
|
|
|
|
TRG_RD |
7925 |
0.018 |
0 |
0.003 |
0.014 |
0.025 |
0.093 |
0.018 |
TRG_SAL_GRO |
7925 |
0.260 |
−0.387 |
0.019 |
0.149 |
0.320 |
3.359 |
0.513 |
TRG_ADV |
7925 |
0.001 |
0 |
0 |
0 |
0.001 |
0.034 |
0.005 |
TRG_PPE |
7925 |
17.572 |
0.218 |
1.874 |
3.930 |
9.242 |
380.67 |
51.352 |
Acquirer characteristics |
|
|
|
|
|
|
|
|
ACQ_BTM |
7925 |
0.336 |
0.047 |
0.183 |
0.282 |
0.428 |
1.128 |
0.217 |
LACK_SLACK |
7925 |
0.028 |
0 |
0 |
0 |
0 |
1 |
0.165 |
ACQ_TA |
8056 |
22.989 |
21.089 |
22.375 |
22.916 |
23.538 |
25.428 |
0.871 |
ACQ_TQ |
6144 |
4.336 |
0.883 |
2.286 |
3.461 |
5.137 |
20.130 |
3.289 |
ACQ_ROA |
6144 |
6.701 |
−8.850 |
3.540 |
6.230 |
9.250 |
24.820 |
5.316 |
M&A characteristics |
|
|
|
|
|
|
|
|
LOGTV |
6144 |
18.207 |
12.368 |
16.811 |
18.315 |
19.714 |
22.633 |
2.036 |
SAME_IND |
8056 |
0.289 |
0 |
0 |
0 |
1 |
1 |
0.453 |
Industry characteristics |
|
|
|
|
|
|
|
|
IND_RD |
6144 |
0.015 |
0.000 |
0.007 |
0.019 |
0.020 |
0.025 |
0.008 |
IND_INTANG |
6144 |
0.044 |
0.001 |
0.040 |
0.043 |
0.043 |
0.159 |
0.021 |
This table reports the number of observations (obs), mean, min, p25, median, p75, max, and standard deviation (sd) of variables. Definitions of all variables are provided in Table 4.
The mean of the CAR (−3, 3), which is the cumulative abnormal return during the 3 days before and after the M&A disclosure date, is 4.2%. The minimum CAR (−3, 3) is −31.7%, and the maximum is 48.9%. This indicates that investors hold divergent outlooks across the M&A deals related to ongoing operational capabilities of listed companies.
The average goodwill impairment (IMPAIRMENT) was 0.7%, with a third quartile of zero and the maximum reached 45.7%. This indicates that within the entire sample, instances of impairment are relatively few. However, when impairment does occur, it can amount to nearly half of the goodwill.
The mean of the dummy variable for the acquiring and target companies being from the same industry (SAME_IND) is 29%, indicating that about 71% of the samples in our study are cross-industry acquisitions.
4.4. Regression Model and Research Design
Following Shalev, Zhang, and Zhang [18], Chemmanur, Ertugrul, and Krishnanwe [22], and Li [23], we construct the following regression model to test our hypotheses 1 to 3:
(4)
The dependent variables are GOODWILL, CAR (−3, 3), and IMPAIRMENT, respectively. The key explanatory variables are the famous CEO of acquirer (ACQ_FAME) and target (TRG_FAME).
and
are coefficients. We include a set of control variables, as well as year and industry fixed effects. ε is the error term.
5. Methodology
5.1. The Effect of CEO’s Fame on Goodwill
In this section, we examine the impact of CEO’s fame on goodwill using multivariate ordinary least squares (OLS) regressions. The dependent variable is goodwill, a proxy for M&A premium. It is goodwill generated by an acquisition divided by targets fair value. The key explanatory variables are the fame of acquiring CEO and target CEO. Following Shalev, Zhang, and Zhang [18], our control variables include a batch of target pre-acquisition characteristics such as R&D intensity, sales growth, advertising expenditures, and ratio of current assets to net property, plant equipment (PPE).
For acquirer pre-acquisition characteristics, we control for acquirer’s book-to-market ratio of net assets (ACQ_BTM). We also include a dummy variable LACK_SLACK for the amount of pre-acquisition unrecognized assets of acquirer, indicating acquiring firm’s flexibility to avoid future goodwill impairment loss. If its value is one, it indicates that the acquirer has limited unrecognized assets prior to the acquisition and a higher risk of future goodwill impairment (Shalev, Zhang, and Zhang [18]). As a result, the excessive M&A premium becomes costlier, which leads to the expected negative relationship between LACK_SLACK and goodwill. We further include an indicator variable of acquirer-target industry proximity (SAME_IND), which captures the amount of synergy created by the acquisition. Finally, both industry and year fixed effects are also controlled.
Table 6 displays the regression results. In column (1), the coefficient of ACQ_FAME (the dummy variable for famous acquiring CEOs) is 0.208, which is statistically significant at the 1% level. This indicates that a famous acquiring CEO tends to increase the amount of goodwill due to confidence in their own capabilities. In column (2), the coefficient of TRG_FAME (the dummy variable for
Table 6. Effect of CEOs’ fame on goodwill (acquisition premium).
|
Dependent variable = GOODWILL |
|
(1) |
(2) |
ACQ_FAME |
0.208*** |
|
|
(42.64) |
|
TRG_FAME |
|
0.186*** |
|
|
(8.15) |
TRG_RD |
−1.035*** |
−0.618*** |
|
(−6.44) |
(−3.49) |
TRG_SAL_GRO |
0.003 |
−0.008 |
|
(0.67) |
(−1.45) |
TRG_ADV |
−0.918* |
−0.037 |
|
(−1.86) |
(−0.07) |
TRG_PPE |
0.000 |
0.00002 |
|
(1.58) |
(0.36) |
ACQ_BTM |
−0.023 |
−0.030* |
|
(−1.58) |
(−1.84) |
LACK_SLACK |
−0.006 |
0.001 |
|
(−0.34) |
(0.04) |
SAME_IND |
0.032*** |
0.036*** |
|
(5.90) |
(6.14) |
Industry fixed effects |
Yes |
Yes |
Year fixed effects |
Yes |
Yes |
Number of observations |
7925 |
7925 |
R2 |
0.227 |
0.057 |
This table presents OLS regression results on the impact of CEOs’ fame on goodwill (acquisition premium). Goodwill is measured as target’s deal price minus its fair value, scaled by the deal price. ACQ_FAME (TRG_FAME) is a dummy variable that equals one if the acquiring (target) firm has a famous CEO, and zero otherwise. Definitions of all variables are provided in Table 4. t-statistics are presented in parentheses below the coefficients and are corrected for cross-sectional correlation at the firm year levels. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.
famous target CEOs) is 0.186, which is also significantly different from zero at the 1% level. This finding suggests that a famous CEO of a target firm can secure a higher price and generate a greater amount of goodwill by leveraging her management expertise. The empirical evidence strongly supports our two hypotheses, H1a and H1b, which propose that the fame of both acquiring and target CEOs positively influences the goodwill in M&A transactions.
Across column (1) and (2), both estimated coefficients on the target company’s R&D intensity variable (TRG_RD) are negative and significant at the 1% level, indicating that a R&D intensive target is expensive, and the M&A premium cannot be excessively high. We find that the dummy variable LACK_SLACK, a proxy for the amount of pre-acquisition unrecognized assets of acquirer, is not significant at the 10% level. It suggests that acquiring CEOs in China exhibit overconfidence in their management ability and are less concerned about firm’s flexibility to avoid future goodwill impairment loss. The coefficient on the dummy variable indicating whether the acquirer and the target are from the same industry (SAME_IND) is 0.032 is significant at the 1% level, which captures synergy effect created by the acquisition and leads to higher goodwill. Furthermore, it shows that industry insiders have a better understanding of the M&A targets within the same industry, mitigating information asymmetry.
5.2. CEO’s Fame as Signal and Market Reaction
In this section, we run OLS regressions to examine the market reaction to CEOs’ fame using the dependent variable of CAR (−3, 3), which is the cumulative abnormal return over 3 days before and after the M&A announcement date. The key explanatory variables are the fame of acquiring CEO and target CEO. Following Chemmanur, Ertugrul, and Krishnanwe [22], our control variables include a set of acquirer pre-acquisition characteristics such as total assets (ACQ_TA), Tobin’s Q (ACQ_TQ), and ROA (ACQ_ROA). We further include an indicator variable of acquirer-target industry proximity (SAME_IND), which is trying capture the post-acquisition market competitiveness. We also control M&A deal size (LOGTV), industry level R&D intensity (IND_RD), industry average ratio of intangible assets to total assets (IND_INTANG), as well as industry and year fixed effects.
Table 7 displays the regression results. The findings strongly support our both hypotheses H2a and H2b that a famous acquiring CEO delivers a bad signal, and a famous target CEO conveys a good signal respectively to the market on the acquisition announcement date. In column (1), the estimated coefficient of the dummy for the famous acquiring CEO (ACQ_FAME) is -0.011, which is significant at the 1% level. This indicates that M&A involving a famous CEO from the acquiring company is perceived as a bad signal by investors. It suggests that investors are concerned about the potential issues related to the overconfidence of high-profile executives in mergers and acquisitions and the possible realization of synergies post-M&A.
In column (2), the estimated coefficient of the dummy for the famous target CEO (TRG_FAME) is 0.039, which is significant at the 1% level. This indicates that the target CEO’s fame has a positive impact on market reactions, suggesting that investors view the fame of the target company’s CEO as a good signal for the performance of the post-M&A combined entity.
In addition, the regression results for the control variables in both column (1)
Table 7. CEOs’ fame and acquirer CAR (market reaction on M&A).
|
Dependent variable = CAR (−3, 3) |
|
(1) |
(2) |
ACQ_FAME |
−0.011*** |
|
|
(−2.94) |
|
TRG_FAME |
|
0.039*** |
|
|
(2.68) |
LOGTV |
0.015*** |
0.015*** |
|
(16.25) |
(15.76) |
ACQ_TA |
−0.003 |
−0.002 |
|
(−1.53) |
(−1.30) |
ACQ_TQ |
0.002*** |
0.002*** |
|
(2.95) |
(3.08) |
ACQ_ROA |
0.001*** |
0.001*** |
|
(3.76) |
(3.68) |
IND_RD |
−14.033*** |
−13.806*** |
|
(−5.80) |
(−5.71) |
IND_INTANG |
1.320*** |
1.289*** |
|
(3.16) |
(3.08) |
SAME_IND |
−0.010** |
−0.010** |
|
(−2.52) |
(−2.48) |
Industry fixed effects |
Yes |
Yes |
Year fixed effects |
Yes |
Yes |
Number of observations |
6144 |
6144 |
R2 |
0.145 |
0.145 |
This table presents OLS regression results on the effect of CEOs’ fame on CAR (−3, 3). CAR (−3, −3) is calculated by market-adjusted return model, which refers to the cumulative abnormal return over the 3 days before and after the M&A event date. ACQ_FAME (TRG_FAME) is a dummy variable that equals one if the acquiring (target) firm has a famous CEO, and zero otherwise. Definitions of all variables are provided in Table 4. t-statistics are presented in parentheses below the coefficients and are corrected for cross-sectional correlation at the firm year levels. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.
and (2) are consistent. The coefficient on M&A deal size (LOGTV) is 0.015 across both columns and is statistically significant at the 1% level. This suggests that larger M&A deals lead to higher investor expectations for sustained profitability and a more positive market response. Similarly, the coefficient on the acquirer Tobin’s Q (ACQ_TQ) is 0.002 in both columns, also significant at the 1% level, indicating that higher acquirer valuations are associated with better investor expectations for post-M&A profitability and a favorable market reaction. The coefficient on the acquirer’s return on assets (ACQ_ROA) is 0.001 in both columns and significant at the 1% level, implying that a higher ACQ_ROA enhances investor expectations for sustained profitability after the M&A, contributing to a more positive market response.
In Table 7, The coefficient on the dummy variable indicating whether the acquisition is within the same industry (SAME_IND) is -0.01 in two columns, both are significant at the 1% level. A takeover within the same industry may eliminate potential competition and create more market power and future profits. However, on the acquisition announcement date, investors may be more concerned that the reduced competition may trigger regulatory scrutiny or legal challenges that could delay or block the acquisition. Even if the deal is approved, the price increase as a result of weakened competition may face backlash from consumers or regulators, potentially impacting future profitability. The market reaction is negative due to these uncertainties on the announcement date.
The coefficients on the average ratio of R&D investment to the total assets at the industry level (IND_RD) are −14.033 and −13.806 in the two columns respectively, and both are significant at the 1% level. It implies that the higher the industry average R&D investment, the worse the investors’ expectations for the sustained profitability of the acquisition, and the more negative the market reaction is. This is because the market takes the view that the higher industry level R&D intensity, the fiercer competition pressure on the post-acquisition combined entity.
The coefficients on the average ratio of intangible assets to total assets at the industry level (IND_INTANG) are 1.32 and 1.289 in column (1) and (2) respectively, both are significant at the 1% level. Intangible assets such as intellectual property, brand value, R&D, and software are key drivers of innovation and future growth at the industry level. Our empirical finding shows that the higher the proportion of intangible assets at the industry level, the more likely there are spillover effects among the firms within the industry. It boosts confidence among investors for the sustained profitability of the post-M&A combined entity, and the market reaction is favorable.
5.3. The Effect of CEO’s Fame on the Post-M&A Impairment
In this section, we examine the impact of CEO’s fame on post-acquisition goodwill impairment loss by running OLS regressions. The dependent variable is IMPAIRMENT, which is post-acquisition 3-year accumulated goodwill impairment loss divided by its original goodwill amount. The key explanatory variables are the fame of acquiring CEO and target CEO. Following Li [23], our control variables include natural logarithm of pre-acquisition total assets of the acquirer (ACQ_AT), an indicator variable of acquirer-target industry proximity (SAME_IND), as well as both industry and year fixed effects.
Table 8 displays the regression results. In column (1), the coefficient of the
Table 8. Effect of CEOs’ fame on goodwill impairment (post-M&A performance).
Dependent variable = IMPAIRMENT |
|
(1) |
(2) |
ACQ_FAME |
0.016*** |
|
|
(13.83) |
|
TRG_FAME |
|
0.008 |
|
|
(1.57) |
ACQ_TA |
−0.001* |
−0.002** |
|
(−1.81) |
(−2.44) |
SAME_IND |
0.001 |
0.001 |
|
(0.85) |
(1.11) |
Industry fixed effects |
Yes |
Yes |
Year fixed effects |
Yes |
Yes |
Number of observations |
8056 |
8056 |
R2 |
0.042 |
0.020 |
This table presents OLS regression results on the effect of CEOs’ fame on impairment. ACQ_FAME (TRG_FAME) is a dummy variable that equals one if the acquiring (target) firm has a famous CEO, and zero otherwise. Definitions of all variables are provided in Table 4. t-statistics are presented in parentheses below the coefficients and are corrected for cross-sectional correlation at the firm year levels. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.
acquiring CEO fame dummy (ACQ_FAME) is 0.016, which is significant at the 1% level. This indicates that M&As involving famous CEOs of acquiring companies are more likely to result in goodwill impairment loss. In column (2), the coefficient of TRG_FAME (the variable of target CEO’s fame) is 0.008, which is not significant at the 10% level. This indicates that an M&A involving a famous CEO on the target side is unlikely to incur post-acquisition goodwill impairment loss. The findings strongly support both of our hypotheses H3a and H3b that famous acquiring CEOs are more likely to generate impairment loss whilst famous target CEOs are unlikely to make post-acquisition impairment.
Furthermore, the coefficients of ACQ_TA (the natural logarithm of the acquirer’s pre-M&A total assets) are both negative and significant at the 10% and the 5% level in column (1) and (2), respectively. This suggests that the larger the site of the acquiring company, the lower the likelihood of goodwill impairment post-M&A. This is because larger companies often have stronger financial positions and have more experience and resources to integrate acquisitions effectively, minimizing the risk of goodwill impairment.
6. Conclusions
This paper discovers a new signal channel in the M&A events. We examine the heterogeneous effects of the fame of acquiring and target CEOs on M&A outcomes. Our findings reveal that the fame of both CEOs of the acquisition pair significantly elevates the M&A premium and goodwill. However, a renowned acquiring CEO is more prone to generating an excessive premium due to overconfidence and the pressure for recognition, which often leads to post-M&A goodwill impairment losses. In contrast, the fame of the target company’s CEO is more effective in assessing the target’s sustainable profitability, thereby justifying the associated M&A premium and reducing the likelihood of impairment losses.
Furthermore, we find that market reactions at the M&A announcement date highlight that investors can discern the differential signals emanating from the fame of CEOs on either side of the deal. If the acquirer has a famous CEO managing the M&A transaction, it is perceived as a bad signal for the acquisition, and the market responds less favorably. Conversely, if the target firm’s CEO is famous, it sends a good signal, prompting positive market responses.
In this study, we adopt a proxy for famous CEOs based on the frequency of their names appearing in online media coverage. Instead, Malmendier and Tate [25] focus on an alternative proxy for famous CEOs: superstar CEOs. They are recipients of prestigious CEO awards granted by leading organizations, such as Ernst & Young, and prominent publications, including Business Week and Financial World in the United States. A superstar CEO is typically described as a high-profile, charismatic leader who manages the company and is also a public figure. These CEOs often dominate headlines, are seen as visionaries or innovators, and are sometimes credited with being pivotal to their company’s success. They can represent a blend of exceptional leadership and a strong personal brand that captures public attention. In China, there are also several prestigious awards and recognitions for CEOs such as China Business Leaders Awards, China Entrepreneur Awards, China Top 100 CEOs, Forbes China Best CEO List, Fortune China CEO Rankings, and Hurun Rich List CEO Rankings. In future research, it would be useful to consider broadening the fame metric to incorporate qualitative aspects such as executive awards, industry nominations, and leadership recognitions to enhance its comprehensiveness.
Another limitation of this study is that our explanation for the observed effects of CEO fame such as overconfidence for acquiring CEO and bargaining power for target CEO may be overly simplistic. Further research could consider agency variables, including the acquiring CEO’s long-term holding of unexercised vested options, the CEO’s initiation of multiple M&A deals during their tenure, and analysis of the market’s timely reactions to CEO statements, to enrich our understanding of and their impact on corporate decisions. These analyses could help deepen our understanding of the behavior of famous acquiring CEOs and their impact on the corporate decisions. To understand the bargaining power of the target CEO, future research may apply textual analysis to extract relevant metrics from M&A disclosures. These metrics include customer concentration, pricing power, customer loyalty, market share, industry entry barriers, and vertical integration. Such an approach could further illuminate the influence of the famous target CEO’s bargaining power on the results observed in this study.
Finally, the main findings of this study are based on an empirical analysis of Chinese M&A events. It is questionable if the findings from the Chinese financial market may be generalizable to other countries with different regulatory environments and corporate ownership structures. On the one hand, the performance of Chinese economy and financial markets are unique in the world. The Chinese economy has an extraordinary growth in the past four decades, lifting millions out of poverty. The manufacturing sector is strong, which is the growth engine of the economy, generating persistent trade surplus with the rest of the world. Yet the Chinese service sector is weak, especially the financial sector, as evidenced by the continued service trade deficits and prolonged capital control. Furthermore, it is puzzling that mainland listed Chinese A-share companies have experienced lower returns than oversea listed Chinese firms, OECD listed corporates, and emerging economy listed firms over the past two decades (Allen, Qian, Shan, and Zhu [26]).
On the other hand, a growing strand of literature shows that some of the apparent anomalies in Chinese financial markets can be explained by standard financial models. For example, Liu, Stambaugh, and Yuan [27] find that the adjusted Fama-French three-factor model can explain most reported anomalies such as profitability and volatility anomalies in the Chinese A-share market. Carpenter, Lu, and Whitelaw [28] find that both price informativeness and investment responsiveness in the A share market have reached levels similar to those in the US. Therefore, the question of whether our newly discovered signal mechanism, which works in China, can be applied effectively in other economies remains a challenging avenue for future research.
Acknowledgements
We thank an anonymous referee, Hanwen Chen, Yubo Li, Nian Liu, Xiongwen Lu, Changjiang Lyu, Yaxin Ming, Mingjie Rui, Yu Yang Tony To, Houmin Yan, Haibin Yang, Liu Zheng, and Workshops participants at Fudan University and City University of Hong Kong for their helpful comments. We are responsible for any remaining errors.