1
THE IMPACT OF THE MEDIA AROUND AN ACQUISITION ON THE STOCK MARKET PERFORMANCE OF THE ACQUIRING FIRM
University of Groningen
MSc Thesis Business Administration – Strategic Innovation Management T.C. (Thomas) van Eenbergen (S2584425)
Supervisors: K.J. (Killian) McCarthy & K.R.E (Eelko) Huizingh Date: 22-06-2015
Word count: 11604
ABSTRACT
The impact of public news releases on the stock market have become a more interesting phenomenon among scholars the last decades, however no study has yet been conducted on the impact of news on an acquisition. Therefore, this study examines the impact of the tone in articles and the media coverage around an acquisition on the stock market performance of the acquiring firm. The second purpose of this study is to analyse whether the level of transparency of a firm moderates the relation between the media and the stock market performance. Based on a sample of 183 acquisitions between 2012 and 2014, this study found some remarkable results. One might expect that media pessimism would have a negative impact on the stock market performance, instead the opposite occurs. Media pessimism and media coverage both have a significant positive impact on the stock market performance of the acquiring firm. Furthermore, no significant results were found on the moderating effect of transparency.
Key words: Acquisitions, media coverage, media pessimism, transparency, stock market performance
INTRODUCTION
On 25/8/2014 Amazon.com announced the acquisition of Twitch Interactive for approximately 970 million dollars. Many newspapers responded to this acquisition with different tones. On 27/8/2014 The Times published a news article regarding this acquisition stating: “Amazon’s web servers provide a more stable connections for its members [Twitch]”. Two days later, on 29/8/2014, The New York Times published: “Investors are scratching their heads over
Amazon’s last venture into the media and entertainment world”.
2 In the last decades research regarding the
impact of public news releases have become a more interesting phenomenon among scholars. Especially the impact of information in articles about firm fundamentals (Pritamaki & Signal, 2001; Mian & Sankaraguruswamy, 2012) and involving in corporate programs such as Corporate Social Responsibility (Zyglidopoulus et al. 2012) on the stock market are widely studied.
Tetlock (2007) was one of the first who showed that especially the level of pessimism in articles, which was calculated by dividing the number of negative words by the total amount of words in an article, can predict future movements on the stock market. In the following study of Tetlock et al. (2008) they found that the level of media pessimism predicts a negative stock market performance of the firm. Besides, not only the tone of the media can predict the stock market, but also the media coverage. Fang & Peress (2009) showed that stocks that are not covered by the media earn higher returns compared to stocks that have a lot of media attention. However, high media coverage stocks still earn positive returns, which is consistent with the findings of Amen (2013). There is, by our knowledge, no research conducted regarding the impact of the media around an acquisition (McCarthy, Dolfsma & Huizingh, 2012). Therefore, the purpose of this study is to analyse the impact of the media, in specific the level of media pessimism and the amount of articles around an aquistion, on the stock market performance of the acquiring firm.
Furthermore, this study tries to make a contribution to the literature by examining
the moderating role of transparency in relationship with the media and the acquiring’s stock market performance. Transparency, and corporate disclosure, are seen as one of the most important principles of corporate governance (OECD, 2012). It is expected that a higher level of transparency, which indicates that all relevant firm fundamental information is made public, will reduce information asymmetry for investors, the target and acquiring firm and journalists. To test the assumptions made in this study a sample of 184 acquisitions are analysed during the period of 2012 and 2014. For all 184 acquisitions a total amount of 2519 articles, containing 991.598 words, are gathered from 322 different newspapers. The empirical results contradict the expectations regarding the role of media pessimism. Media pessimism has a significant positive cumulative abnormal return on the order of 1.364% for the 20 days surrounding the announcement date after controlling for important firm, deal and media characteristics. In addition, a positive significant relation is found for the impact of media coverage on the stock market performance of the acquiring firm. No support is found for the moderating role of transparency.
3 acquisitions on the corporate performance
(e.g. Healy, Palepu & Ruback, 1992; Andrade, Mitchell & Stafford, 2001), by focusing on the impact of the media on an acquisition announcement.
This paper is organized as follows. Firstly, a review of relevant literature will be given including the hypothesis. Secondly, a description of the data, methodology and the findings are reported. Thirdly, a discussion of the findings including a conclusion. Finally, the paper ends with the limitations of the study, managerial implications and suggestions for future research.
LITERATURE REVIEW Mergers and Acquisitions
Mergers and Acquisitions (henceforth, acquisitions) have become an important force to survive in nowadays fast changing environment, partly due to the globalization of the economy and liberalization in the world (Gupta, 2013). It has emerged as one of the most important and effective methods for corporate structuring and therefore has become a driving force of the business strategy for firms all over the world (Mallikarjunappa & Nayak, 2007). However, acquisitions are not without risks, the evidence shows failure rates of more than 50% (Schoenberg, 2006). These high failure rates can be related to different internal and external factors. Internal factors are for instance, information asymmetry problems (Hviid & Prendergast, 1993), difficulties in resource sharing (Cartwright & Schoenberg, 2006) or cultural differences (Weber & Camarer, 2003) between the two firms. External failure factors can be related to changes in business indicators, markets or a
counterattack of a competitor (Kummer & Steger, 2008).
Especially publicly listed firms can also encounter other problems. Since their ownership is dispersed among the public, firms need to be aware of the reaction of the public, and especially investors, on the acquisition. Research has already shown that investors actively asses what kind of information appears in the public (Schijven & Hitt, 2012), which could result in an under- or overreaction of stock prices (Daniel et al. 1998) or excessive trading by investors (Odean, 1998). Therefore it is of importance to study if and how the information published in the media impacts the acquiring firm around an acquisition.
Efficient Market Hypothesis vs. Over- and under reaction Hypothesis
In the economic literature there are two fundamental theories that can explain if the stock market reacts on the media, the efficient market hypothesis (EMH) and the over- and under reaction hypothesis.
4 available public information and the strong
market efficiency states that it reflects all available public information as well as private information. Thus, EMH assumes that the media cannot have an influence on stock prices around an acquisition due to the fact that all available information is already reflected into the stock prices. Besides, EMH assumes that investors are unbiased in their reaction to information. De Bondt & Thaler (1985) found a contradicting result compared to EMH, due to bad or good news a stock can decrease or increase in price but always returns to its intrinsic value as investors realize that they have overreacted or underreacted.
Above mentioned reasoning of De Bondt & Thaler (1985) assumed that it is possible to predict future movements and obtain profits from arbitrage opportunities. One of the strategies that is based on the overreaction hypothesis is the contrarian investment strategy (Lakonishok et al. 1994). Their strategy suggests that stocks who have a worse (better) result in the past will have a more worse (better) result in the future, so buying a previous loser stock and selling a previous winner will result in a significant positive abnormal returns in a later stage. The momentum trading strategy is an opposing strategy that is based on the under reaction hypothesis. This strategy suggests that selling a loser stock and buying a winner stock will result in a positive abnormal return. Previous studies that analysed the impact of news on the stock market (e.g. Tetlock 2007; Fang & Peress, 2009), showed that the market reacts on news announcements, therefore this study assumes that the EMH will not hold.
The impact of the media on the stock market
But how does the media impact the stock market? Previous studies acknowledged that the stock market is mostly driven by news, which can be aligned to macro-economic factors, geopolitics and firm-specific factors (van de Kauter, Breesch & Hoste, 2015). Cutler et al. (1989) were the first that showed that news can predict macro-economic factors by the variation on the market indices, which accounted for almost one-third of the variance in stock returns. Cornell (2013) re-studied the findings of Cutler and found that the results have not been changed over the past two decades, despite the changes in information technology, enhanced market regulation and new financial products.
5 Not only firm fundamental news can have an
impact on the stock market performance of a firm, but also how newspapers publish about programs like Corporate Social Responsibility (CSR). The study of Zyglidopoulus et al. (2012) examined the impact of media coverage regarding CSR on the stock market performance of financial holdings listed on the Taiwan Stock Exchange. Their empirical results showed that news about positive social activities for employees trigger a negative evaluation by shareholders, whereby the stock market performance decreases. News about social activities for shareholders results in a positive stock market performance. Nowadays, news not only gets distributed via the newspaper but also via Social Media. Therefore, Bollen, Mao & Zeng (2011) investigated whether the public mood, measured from a large dataset of Tweets, can predict the Dow Jones Index. They found that based on six observed mood dimensions (calm, alert, sure vital, kind and happy), the level of calmness of the public can predict the Down Jones Index closing values with an accuracy of 87,6%.
Despite the fact that most studies show that the media has an impact on the stock market, they differ in how they measure the impact of the media. Different sources for media can be used, like focusing on financial newspapers, (Tetlock et al. 2008; Schumaker & Chen, 2009) newspapers from a certain country (Li, 2010) or studies make use of news that is available online, e.g. Google news (Yu, Duan & Cao, 2013).
The growing literature shows the significant impact of the media regarding firms’ fundamentals, the stock market and corporate governance on the stock market.
However, no study has focused on the impact of media on acquisitions. Therefore this study builds on the assumptions made in the studies of Tetlock (2007, 2008) and Fang & Peress (2009) how the media can react on an acquisition. Tetlock (2007, 2008) showed that media pessimism is a reliable proxy to measure the tone of the media on the stock market performance of an individual firm. Fang & Peress (2009) used media coverage to investigate the relation between media and the stock market, which can provide an incremental explanatory power to explain this relation as well.
Tone of the media
6 In the following study of Tetlock et al. (2008)
the scope of the research is refined from the stock market to the individual firm. Their data was not only limited to the column, but they also measured the media pessimism in all articles related to the individual S&P 500 firms in the Wall Street Journal and the Dow Jones News Services from 1980 to 2004. Two positive advantages were mentioned by Tetlock et al. (2008) for using media pessimism as predictor to examine the impact of media news on individual stock returns. The first advantage is that analyzing the language gives access to focus on a variety of events, whereas most studies focus on one particular event. The other advantage is that news can be an interesting source of firm’s fundamental information, it can provide an incremental explanatory power to predict a firms’ future stock market performance, beside the traditional fundamentals such as analyst forecasts. Their main results suggest that negative words in news articles forecast low firm earnings and confirms that media is a reliable source for a firm’s fundamental information. In addition, the prediction of stock returns becomes especially important when the news presents information regarding firm fundamentals.
Based on Tetlock (2007, 2008) it can be assumed that how pessimistic the media reacts on an acquisitions can predict how the stock prices will react. More specifically, a higher level of media pessimism will drive the stock price down. This research will specifically focus on the impact of media pessimism in articles around an acquisition on the stock market performance of the acquiring firm. Due to the aforementioned reasoning it can be hypothesized that:
H1: High media pessimism in news articles will result in a negative stock market performance to an acquisition announcement for the acquiring firm.
Media Coverage
The studies of Tetlock et al. (2007, 2008) focused on the tone of the media to evaluate how the stock market reacts, as opposed to Fang & Peress (2009) who used the amount of articles in the Wall Street Journal during the period 1993 - 2002 to explain this relation. They found that firms with no media coverage outperform firms with high media coverage for over 0,20% per month, even after controlling for market, size, book-to-market, momentum and liquidity. Despite the fact that stocks without media coverage outperform stocks with media coverage, high-coverage stock groups still obtain a positive monthly return. When the media coverage is high, there will be more attention from the public and especially investors. The results of Fang & Peress (2009) are consistent with the study of Engelberg and Parsons (2011), they examined whether media coverage can have an effect on the investor response to financial events. Their empirical results showed that the presence of media coverage has a strong positive impact on the trading activity in financial markets. Additionally, Amen (2013) provided the same evidence from a sample of firms in Japan, large media coverage stimulate large reactions in the stock market to corporate news.
7 media coverage around an acquisition has a
positive impact on the stock market performance of the acquiring firm. The following hypothesis will be tested:
H2: High media coverage will result in a positive stock market performance to an acquisition announcement for the acquiring firm.
Besides the examination of the impact of media on the stock market performance, this paper also seeks to contribute to the literature by investigating whether transparency moderates the relation between the media and the stock market.
The moderating effect of Transparency
News can contain fundamental firm specific information that helps the acquiring firm, target firm and the investor to resolve the information asymmetry (Tetlock, 2010). Previous research on moderating variables, which can have implication on the media are very limited. This paper takes the transparency of a firm as moderating role to explain whether the level of transparency will resolve the information asymmetry problem in the media. The definition of transparency differs a lot from which perspective it is defined. In the paper of Florini et al. (2000) a broad definition of transparency is given: “the release of information by institutions that is relevant to evaluate those institutions” (p. 168). From a public accountability perspective Kopits and Craig (1998) define transparency as “openness toward the public at large about government structure, functions, fiscal policy intentions, public sector accounts and projections.” (p. 2). In this paper the definition of Stiglitz (2000) will be used, because it specifically focuses on the relation
between transparency and information asymmetry. Stiglitz (2000) defines transparency as a way of minimizing information asymmetries in the market. Transparency is often related to the broader concept of corporate governance, or more in specific regarding the corporate disclosure a firm reveals (Kothari, 2000). The OECD (Organisation for Economic Co-operation and Development) is one of the initiators of corporate governance and divided the concept in five principles, whereas transparency and disclosure are regarded as an extremely important factor. The study of Brown & Caylor (2004) showed that a proper corporate governance structure will most likely lead to more profit, value, profit out of dividends and less volatility. It not only improves the firm’s financial performance, but also the productivity. Tian & Twite (2011) showed in their empirical findings that internal corporate governance mechanisms, such as more efficient boards and CEO-stock compensation, are effective instruments for improving the productivity of a firm.
8 These problems may have an impact on a
firm’s valuation (Chi, 2009). By increasing the transparency of a firm, it can reduce information asymmetry among inside and outside investors (Chiyachantana et al. 2013). Besides, it can enable shareholders and investors to effectively analyse the management decisions a firm makes and it provides fundamental information (Lang & Lundholm, 1993).
Transparency is not only limited to reduce the information asymmetry directly, but also indirectly via intermediaries, for the acquiring firm, target firm and investors. Most studies focus on the role of financial analyst as intermediaries (e.g. Lys & Sohn, 1990; Francis & Soffer, 1997). They use public and private information to evaluate the current performance and try to forecast future predictions. Besides the financial analyst the (financial) news can function also as intermediary. However, a recent study of McCarthy & Dolfsma (2014) showed that journalists think that they are seen as neutral, but instead the opposite is true. Journalists are in general negatively biased and select, analyses and report, intentionally or unintentionally, news in a different way, which will impact the audience view of the world. Increasing the transparency of firm’s fundamental information will reduce the information asymmetry for the journalist, whereby creating an “unbiased” picture will not be possible to make public, since all relevant information is available for everyone.
Based on above findings, this paper assumes that the level of transparency will moderate the causal relation between the media and the acquiring stock market performance. In specific, a higher level of
transparency reduces the information asymmetry for investors, but also for the journalist that publishes about an acquisition and therefore more accurate information is published to the public.
H2: Higher levels of transparency will positively moderate the relationship between the media and the acquiring firm’s stock market performance to an acquisition announcement.
METHOD
Mergers and Acquisitions: The total
sample of acquisitions was acquired from the Thomson SDC database within the period of 01/01/2012 until 31/12/2014 and only concerned deals that had a deal value higher than 10 million dollars. In addition, the focus was only on acquisitions where the ownership was 100% over another firm. A total sample size of 12,098 acquisitions were involved in the primarily dataset. Since this study analyses the moderating role of a firms’ transparency, the dataset was limited to those firms that were listed on the transparency list. The transparency list consists of the 124 largest publicly traded firms. For those firms all SEDOL numbers were collected. Based on the SEDOL numbers acquisitions were filtered from the primarily dataset. When the SEDOL number was not mentioned in the dataset we searched on the firm name. A total of 278 acquisitions were derived from the primarily dataset. For all these acquisitions newspaper articles were collected from LexisNexis.
LexisNexis: LexisNexis is the world’s largest
9 all 278 acquisitions one month before and
one month after the deal of announcement. To illustrate how the news articles were collected an example is given about one of the deals involved in the dataset, namely the acquisition between Apple Inc. and Beats Electronics on 28/5/2014. First, we searched on one month after the acquisition on the name Apple and Beats Electronics. The business entities, e.g. Ltd or Inc were deleted. To get articles of higher quality and to reduce noise in the news articles an additional search criteria was used, namely the names of the organisation must occur at least two times in the news article. The source which were used are all international English written newspapers which were available in the LexisNexis database. At first only major English newspapers were used as source, but unfortunately there was too much noise in these news articles. Furthermore, when the search criteria of at least two times occurrence was added the amount of news articles became too limited. In the case of Apple and Beats a total amount of 251 articles were published. When the amount of articles exceeded the number of 75 the word “acquisition” was added. The total amount of news articles regarding the acquisition of Apple and Beats retrieved from LexisNexis was 32. All news articles were saved as .txt file to analyse at a later stage on the amount of positive and negative words. Finally, all articles were scanned to see whether the main subject of the article was about the acquisition and no other acquisition was mentioned, this would have formed a potential bias. Duplicates were also removed when the articles were published in the same newspaper. This was not the case for Apple and Beats Electronics.
In addition, we searched on one month before the deal announcement. We only searched on the name Apple and Beats Electronics and did not use the at least two times occurrence criteria, due to the limited amount of news articles about acquisitions before the deal of announcement. When the amount of articles was above 75, the two times occurrence criteria was added. In the case of Apple and Beats Electronics the amount of news articles before the deal announcement was 34. All articles were saved as .txt file and checked on quality.
In total we found for 184 acquisitions (66,2%) articles regarding the acquisition in the newspaper. The total amount of articles one month after the announcement at first was 2975, but after analysing whether the news articles were related to the acquisition it reduced to 2275 news articles about 183 acquisitions. The total amount of articles one month before the announcement at first was 317 and after analysing on quality it reduced to 244 news articles about 43 acquisitions. The final dataset consists of 2519 articles about 184 acquisitions. A total of 322 different newspapers from 18 countries were used. A description of the amount of acquisitions per year, industry and country can be found in appendix I. In appendix II extra information is provided about how the data is collected.
MEASURES
Dependent variable Stock market
10 was used to analyse the changes in stock
prices of a firm. Stock prices are the most promising way to measure firm values after an event (Zollo & Meier, 2008). First, the event was determined, in this case it was the day of acquisition.
The 183 acquisitions in our dataset consists of a total of 69 different firms. For those firms the stock prices were obtained from Datastream for the specific event window. Furthermore, the primarily daily market indices of the firms were collected and used to calculate the dependent variable. Two events were dropped because there was no information about the stock price available, a total of 181 acquisitions remained. With the information about the stock prices and the market indices the cumulative abnormal returns (CARs) were calculated with the use of Stata. Before calculating the CARs the event window must be specified, it will show the period of which the stock prices of the firms will be examined (MacKinlay, 1997). The event window in this study was set on 20 days before and 20 days after the deal of announcement. An estimation window of 30 days was used, 60 days before the event to 30 days. The estimation window was used as a benchmark to estimate what the performance would have looked like without the event, calculated as the normal stock returns. These were compared to the returns calculated within the event, the abnormal returns. The CAR was the sum of all abnormal returns within the specified event window and was used in this study as dependent variable.
Independent variable Media pessimism and
media coverage: All articles obtained from LexisNexis were the input for the independent variable. The framework
11 show that positive words do not predict
anything. The media pessimism ratio per article was used to determine the average tone of the total media for every acquisition. To test whether the ratio of media pessimism is a well-developed proxy, face validity was used. Two independent students rated 15 randomly selected articles on a scale from 1 till 10. A score of 10 indicated that the article is extremely negative. The correlation test showed that both student ratings were strongly correlated with the media pessimism variable, respectively 0,7310 and 0,7525. This confirmed that media pessimism was a good proxy to use in this study. The scores of the students and the correlation table are displayed in appendix III. Additionally, another independent variable was created, media coverage. Based on the study of Fang and Peress (2009) media coverage was measured by the total number of news articles written about an acquisition within the specific event window.
Moderator Transparency: Transparency
International1 is founded in 1993 and fights as organization against corruption. Since 2012 they publish a report annually which assesses the transparency of corporate reporting of the 124 largest public listed firms with respect to their reporting on anti-corruption programs, organizational transparency and country-by-country reporting. All three categories were rated based on 3x13 questions. For the category anti-corruption programs questions like: “does the firm provide anti-corruption training for its employees?” were asked. An example question for the category organizational transparency is as follows:
1 Homepage: https://www.transparency.org/
“does the firm disclose all of its subsidiaries?”. And the for country-by-country reporting questions were asked regarding the disclosure of revenue, sales and capital expenditure in a specific country. The percentages of the categories were calculated for every firm and an overall index was derived by taking a simple un-weighted average of the results achieved from those categories. The scale was between 0 till 10, where 10 is the best. For example, Vodafone scored respectively 100%, 50% and 51%. for the categories. Their overall index score was 6,7 (((100+50+51)/3))/100).
Control variables: This study controls for
other performance indicators. The deal characteristic that was used was the relative deal value. A high deal value is often associated with a greater performance than a small deal value. Helwege, Pirinsky and Stulz (2007) assumed that analysts and investors more frequently monitor high deal value acquisitions, which reduces the risk of overvaluation. The relative deal value was calculated by dividing the deal value by the market value of the acquiring firm.
12 high q often experience a positive abnormal
return, while acquirers with a low q experience a negative abnormal return. Concerning the control variable volatility, with a high volatility it means that the stock prices can change rapidly over a short time period in a negative or positive way. A low volatility means the firm is more stable on the stock market. The last firm specific control variable was the earnings to price ratio of the acquiring firm. A high earnings/price ratio shows an expected growth (Sharpe, 2002).
Finally, some media characteristics were included. First, the total sum of words used in the articles had been used as control variable, since there were more negative words than positive words. Also the circulation of newspaper was used as quality control variable. In addition, a dummy variable was created to indicate whether the acquiring firm was the parent firm or a subsidiary. It can be assumed that a parent firm would receive more news than a subsidiary.
All statistical analysis were clustered on industry. Therefore the industry classification benchmark (ICB) was used. Several databases were used to gather the control variables. Tomson SDC database provided information regarding the deal value, country of the acquiring and target firm and whether the acquiring firms was a parent firm or a subsidiary. Datastream was
used to gather information about the market-to-book, tobinsq, volatility and earnings to price ratio. The total sum of words and circulation of the newspaper was gathered from LexisNexis.
Sample description Table 1 shows the
descriptive statistics for the whole sample of acquisitions within the period of 2012 and 2014. More specifically, it shows the deal characteristics and firms characteristics. For some variables the natural logarithm is used to reduce potential skewness.
13 Table 1. Descriptive Statistics acquisitions
Table 2. Descriptive Statistics News Articles per acquisition
Variables Observations Mean Median SD Min Max
News Length 184 335.214 314 138 45 748 Name Title 184 0.776 0.821 0.209 0 1 Byline 184 0.496 0.5 0.292 0 1 Page number 184 0.037 0 0.010 0 1 Graphic 184 0.068 0 0.099 0 1 Circulation (log) 181 11.940 12.074 1.112 7.824 14.022 Sum Words 184 5.387 2.412 8.999 0.045 76.877 Media Coverage 184 13.690 7 17.487 1 122 Negative words 184 6.786 6 5.628 0 42 Positive words 184 10.231 9.552 5.914 0 36 Ratio negative 184 0.0175 0.0159 0.0124 0 0.0869 Ratio positive 183 0.0275 0.0268 0.0221 0 0.283
Variables Observation Mean Median SD Min Max
Deal characteristics
Deal value (log) 184 6.227 6.262 1.596 2.334 9.858
Cash % 134 91.66 100 21.36 3.837 100 Stock % 28 43.75 39.40 28.66 1.292 100 International 184 0.380 0 0.487 0 1 Attitute 184 1 1 0.011 0 2 HQ / Subsidiary 184 0.417 0.4 0.331 0 1 Firm characteristics
14
RESULTS
A regression analysis will be used to test the hypothesis of this study. The dependent variable is a 41 day (-20, +20) CAR. In table 3 the CARs for the full sample and for two subsamples, high vs low media pessimism and high vs low media coverage are shown. The average CAR of the total sample is 0.00587. The difference between the CARs of acquisitions with high media coverage and low media coverage is minimal. Based on the formulated hypothesis one might have expected that the number of negative CARs for high media pessimism would be higher than for a low media pessimism, instead the opposite is true.
Correlation
The bivariate correlation of the variables that will be used in the regression can be found in table 4. The correlations show how the variables are related to one another. The most interesting results from table 4 are outlined. The variables deal value and media pessimism are significantly positively related to each other (0,137***2). This indicates that high deal value acquisitions are more likely to receive more negative news. The same explanation can be given for the positive significant relation between the deal value and media coverage (0,418***). A high deal value acquisition generates more awareness and therefore there is more media coverage. The variables media pessimism and media coverage are also highly significant (0,186***). A higher level of pessimism generates more media coverage and vice versa. In addition, the variables media pessimism and the total sum
2*** p<0.01, ** p<0.05, * p<0.1
of words about an acquisition are also highly significant (0,140***). Surprisingly, the variables CAR and media pessimism are highly positive significant (0,193***), whereas a negative relation was expected. The variables media coverage and CAR do not show any significant relation.
Regression
In table 5 the initial results are shown between the interaction of media pessimism and the media coverage on the CAR. The results show that media pessimism has a significant positive effect on the dependent variable (P-value < 0,000486), with an adjusted r-square of 3,2%. The coefficient shows a positive effect. More specifically, the increase of media pessimism by one unit causes the dependent variable, the CAR, to increase by 1,529%. The independent variable media coverage is significant as well (P-value < 0,0728) and has an adjusted r-square of 0,6%. Thus, by simply looking at the outcome of media pessimism and the amount of articles published about an acquisition on the CAR, a positive significant relation exists.
r-15 square of the model is 5,0%. In model 2 the
media pessimism variable is added as independent variable. The results show that media pessimism has a positive coefficient of 1,364 and has a significant positive impact on the CAR (P-value < 0,000105). The adjusted r-square increased with 2,2% compared to model 1 to a total of 7,2%. The control variable HQ/Subsidiary remains significant, but with a lower p-value (0,0669). In model 3 the independent variable media coverage is added and has a positive significant impact on the CAR (B = 0.00140, P-value < 0,0924), the model explains 5% of the variance in media coverage on the CAR. The control variable volatility shows a negative relation (P-value < 0,0969), relative deal value a positive (P-value 0,0752), the circulation negative (P-value < 0,0967) and the book market a positive relation (P- value < 0,0958). To summarize, after controlling for important firm and deal characteristics the empirical results do not support hypothesis 1, but support hypothesis 2. A positive significant relation between both media pessimism and media coverage on the 41 day CAR [-20, +20] has been found.
In model 4 the moderating role of transparency on media pessimism is added to test hypothesis 2. It shows that the variable media pessimism remains positive, but is not statistically significant (P-value <0,219). The moderating variable is negative, but not statistically significant (P-value <0,649). Only the control variable book market is statistically significant (P-value < 0,0746)). The adjusted r-square of the model is 6,1%. In model 5 the moderating role of transparency on media coverage is added. The results show that media coverage is positive and almost significant (P-value <
0,110), the moderating role of transparency on the media coverage is negative and insignificant (P-value < 0.425). In addition, the control variable volatility is negatively significant (P-value < 0,0712), the relative deal size positively significant (P-value < 0,0630), HQ/Subsidiary is positively significant (P-value < 0,0589) and book market is positively significant as well (P-value < 0,0720). Thus, based on these findings, hypothesis 3 cannot be supported for both media pessimism and media coverage. No moderating effect found of transparency has been found on the relation between the media and the stock market performance of the acquiring firm.
All above described regressions are tested on multicollinearity. Some variables needed to be mean-centred in order to decrease multicollinearity problems. In Stata the Variance Inflation Factor (VIF) is used to identify these problems, but no variable exceeded the limit of 10 (Kutner, Nachtsheim & Netter, 2004)
Robustness
16 and the relative deal value are the main
predictors. When adding media pessimism it still remains positively significant (P-value < 0,00228). When media coverage is added to the control variables it is still an important predictor but not significant anymore (P-value < 0,327). Also the moderating effect is in essence the same as CAR [-20, +20]. The results of CAR [-5, +5] show that book to market is the main predictor of the control variables (P-value < 0,0459). When media pessimism is added it still has a significant positive impact (p-value < 0,0198). When media coverage is added, the impact is positive, but not significant (P-value < 0,566). The moderating effect is not showing any remarkable results. The results of CAR [-2, +2] are also added in the appendix. This CAR shows some remarkable results. When adding media pessimism to the control variables a negative insignificant impact exist (P-value < 0,286). Also the media coverage is negatively insignificant (P-value < 0,616). When transparency is added to the model with media pessimism the moderator becomes positively significant (P-value < 0,0990) and the media pessimism becomes negatively significant (P-value < 0,0138). However, the adjusted r-square of this model is low (0,9%). Therefore it is difficult to draw conclusions. To summarize, the results seem robust, nevertheless when the event window becomes shorter the adjusted r-square decreases and the results become less significant. The longer the event window is, the more robust the results are.
Additional regression media variables
In our dataset several media related variables are collected, for example whether the article was published on the first page of the newspaper or if a picture was included.
Therefore an additional regression analysis was carried out to see what the effects are of those variables on the CAR[-20, +20] and the control variables. These results are shown in table 10 in Appendix VI. The empirical results show that only the day has a significant positive relation (P-value < 0,0546). The day is measured as a dummy variable, where 1 indicates that the article is published on a Friday and 0 another working day. More specifically, it shows articles which are published on a Friday have a significant positive result on the CAR. This finding was also expected, because bad news is often released on Friday after the markets close.
Table 3: Average CAR [-20, +20]
Table 5: Initial results
CAR[-20,+20]
N Mean Negative CAR
Total sample 184 0.00587 92 (50%)
High media coverage 90 0.00440 45 (24%) Low media coverage 94 0.00727 47 (26%) High media pessimism 92 0.01766 41 (23%) Low media pessimism 92 -0.00596 51 (27%)
(1) (2) VARIABLES car20 car20 Media Pessimism 1.529*** (0.000486) Media Coverage 0.000602* (0.0728) Constant 0.00683 -0.00176 (0.420) (0.823) Observations 183 183 Adjusted R-squared 0.032 0.006
17 Table 4: Correlation matrix
1 2 3 4 5 5 6 7 8 10 11 12 13 CAR[-20, +20] Media Pessimis m Media Coverag e Sum
Length Size Volatility Deal Value (log) HQ or Subsidiar y EP (log) Circulatio
n (log) Book Market Tobinsq
Trans paren cy 1 CAR[-20, +20] 1.000 2 Media Pessimism 0.193*** 1.000 3 Media Coverage 0.107 0.186*** 1.000 4 Sum of words 0.105 0.140*** 0.955*** 1.000 5 Size -0.038 -0.182*** -0.005 -0.038 1.000 5 Volatility -0.096 -0.056** 0.035 0.015 -0.131*** 1.000
18 Table 6: Regression, clustered on industry, dependent variable CAR [-20, +20]
Model (1) Model (2) Model (3) Model (4) Model (5) VARIABLES
Media Pessimism 1.364*** 2.093
(0.000105) (0.219)
Media Coverage 0.00140* 0.00149
(0.0924) (0.110) Sum of Words 1.03e-06 8.32e-07 -1.48e-06 8.22e-07 -1.07e-06
(0.141) (0.213) (0.401) (0.212) (0.535) Size log (employees) -0.00352 0.00103 -0.00480 0.00175 -0.00931
(0.692) (0.904) (0.597) (0.839) (0.310) Volatility -0.00948* -0.00764 -0.00989* -0.00738 -0.00912*
(0.0888) (0.166) (0.0969) (0.112) (0.0712) Relative Deal Value 0.587* 0.509 0.563* 0.526* 0.573*
(0.0598) (0.107) (0.0752) (0.0822) (0.0630) HQ or Subsidiary 0.0537** 0.0506* 0.0527* 0.0490* 0.0506*
(0.0446) (0.0669) (0.0527) (0.0829) (0.0589) Earnings to price ratio log -0.00420 -0.00711 -0.00429 -0.00770 -0.00449 (0.737) (0.559) (0.738) (0.547) (0.726) Circulation log -0.0150 -0.0169 -0.0169* -0.0172 -0.0172 (0.136) (0.120) (0.0967) (0.116) (0.101) Book Market log 0.00955* 0.00866 0.0102* 0.00833* 0.0103* (0.0967) (0.111) (0.0958) (0.0746) (0.0720) Tobinsq 0.00981 0.0101 0.00957 0.00952 0.00717 (0.220) (0.229) (0.216) (0.195) (0.226)
Transparency -0.00193 -0.00250
(0.731) (0.670)
Media Pessimism * Transparency -0.191
(0.649)
Media Coverage * Transparency -0.176
(0.425)
Constant 0.269 0.245 0.306 0.249 0.375*
(0.157) (0.173) (0.119) (0.180) (0.0703)
Observations 173 173 173 173 173
Adjusted R-squared 0.050 0.072 0.050 0.061 0.044 Robust pval in parentheses
19
DISCUSSION AND CONCLUSIONS
The purpose of this study is to analyse the impact of media on an acquisition. Describing the media as the level of media pessimism and the amount of articles. Using the stock market performance of the acquiring firm as dependent variable, the following conclusions can be drawn: (1) a high level of media pessimism and media coverage both have a significant positive impact on the stock market performance of the acquiring firm, (2) transparency does not have a moderating effect on the relation between media and the stock market, (3) more media coverage results in more media pessimism and vice versa, especially for high deal value acquisitions.
Firstly, the empirical results in this study show that media pessimism and media coverage have both a significant positive impact on the stock market performance of the acquiring firm, after controlling for important deal, firm and media characteristics. For media pessimism the exact opposite was expected, based on the findings of Tetlock (2007, 2008). Concerning the role of media coverage, the findings support the assumption made in this study (Fang & Peress, 2009; Amen, 2010). A higher level of media coverage results in a positive reaction on the stock market, since it incorporates new information and prices quicker and therefore firms obtain higher values (Ahern & Sosyura, 2014).
Thus, the more news articles are published regarding an acquisition, and the more negative the tone of those articles are, the better the stock market performance of the acquiring firm will be. The
counter-intuitive result concerning the positive role of media pessimism can be explained by the following.
20 investors often underreact to information in
the short term and overreact in the long term.
Secondly, the asymmetric response to bad news can result in a positive effect on the stock market performance. This study shows that negative words explain more than positive words. These findings are consistent with studies of Tetlock (2007, 2008) and McCarthy, Dolfsma & Huizingh (2012). Studies that made use of the financial word list of Loughran and McDonald (2011) show similar results as well (e.g. Miller, 2010; Jegadeesh & Wu, 2013). An explanation that negative words have more impact than positive could be due to the fact that there are more negative words than positive words in most wordlists. Besides, people tend to be negatively biased (Baumeister et al. 2001). We process bad news more thoroughly than good news. Other research areas, e.g. the psychology literature and political science, already found evidence of an asymmetric response to bad news and good news. The psychology literature defines this as cognitive weighting, which implies that information that is unique and novel attracts more attention, but are also perceived as more extreme (e,g, Vonk, 1996; Singh & Teoh, 2000). Impressions are based on the reference point of an individual. The reference point is different from person to person, due to differences in experience. A score of -10 is more extreme, and thus has a greater weight, than when the score is +10 at a reference point of 0. It shows that people tend to view a slightly negative tone in the news as very negative, but also more informative (Soroka, 2006). In the political science there is a lot of evidence that negative information plays a far greater role
in the voting behaviour of people (Niven, 2000; Fridkin & Kenny, 2004). In addition, Hirsch stated in his study that “the presence or absence of coverage, rather than its favourable or unfavourable interpretation, is the important variable” (1972, p. 647).
However, the question still remains, why is a negative tone positive for the stock market performance of the acquiring firm? An explanation can be that investors are more risk taking than risk avoiding. A recent study of Weber, Weber & Nosic (2012) surveyed UK online-brokerage customers for three months to analyse the determinants of changes in investor risk taking. They show that changes in risk taking are associated with changes in subjective expectations, such as news articles, in the market. It can be assumed that investors prefer to take more risk when there is more negative news about an acquisition, since they hope that the market will overreact on such an announcement. This assumption is consistent with a study of Veronesi who states that the “stock markets overreact to bad news in good times” (1999, p. 975). The study of Boyd, Jagannathan & Hu (2005) showed similar results, a high level of news of unemployment raises the stock prices during economic expansion, but lowers the stock price during a contraction. This study is conducted between 2012 and 2014, which can be considered as years of economic expansion.
21 journalist and the target and bidding firm
and therefore would have positive impact on the relation between the media and the stock market performance. However, this relation is not supported in this study. A reason why no support is found could be due to the way transparency is measured. In this study the transparency of a firm is calculated with an overall index that is obtained from a survey held by the organisation Transparency International. The study of Williams (2014) showed that transparency can be measured in a lot of different ways, like on an information level or an accountability level. Studies show that on an information the level transparency has a positive relation on the prediction of future stocks of a firm (Boyd et al. 2005). The study of Chiyachantana et al. (2013) showed that a higher level of transparency reduces the information asymmetry between informed and uninformed investors, which eventually result in a higher firm performance. They measured transparency by analysing the information that is published by a firm one hour before a trading session on the Stock Exchange of Thailand.
Another explanation can be that the firms analysed in this dataset do not differ a lot in how transparent they are (overall index ranging from 1.6 till 6.7). All the analysed firms are publicly listed and in the fortune global 500 and therefore heavily monitored by investors, analysts and other environmental factors on an daily, or even hourly, basis. They got more to lose when they do not disclose all fundamental information compared to smaller firm and therefore transparency will have less impact. Besides, publicly listed firms are subject to detailed disclosure laws.
Thirdly, the correlation matrix shows that a higher deal value attracts more media coverage and has a higher level of media pessimism. This suggest that the media has more attention for high deal value acquisitions, but is more sceptical about it as well. A reason for this could be that the risks are much higher and therefore investors, journalist and analyst will be more critical regarding whether the acquisition will succeed or not. Additionally, a higher level of media pessimism results in more media coverage and vice versa.
MANAGERIAL IMPLICATIONS
The managerial implication of this study is that the media impacts an acquisition in an important way. Especially how pessimistic the media reports about an acquisition and the amount of articles associated with it. An implication for management of especially public listed firms is to generate as much as negative attention around the acquisition. This study confirms with an analysis on a 20 days event window that it will result in a positive stock market performance for the acquiring firm.
22
LIMITATIONS AND FUTURE RESEARCH
Despite the interesting results this study shows, there are still some limitations. Firstly, the relative small sample size is a limitation of this study. To keep this research feasible within the timeframe only a limited amount of 184 acquisitions were analysed between 2012 and 2014. Besides, only firms listed on the fortune 500 were used and acquisitions with a deal value above 10 million. For future research a larger sample size should be analysed including lower deal values as well, to examine whether media still has a significant impact. Secondly, the way of measuring the tone in the media was rather simplistic. The tone of the media per article was analysed by the frequency of words listed on the negative wordlist in the Harvard Dictionary. For example when an article mentioned: There is bad news or there is no bad news, the program would analyse both “bad” as negative. It is possible that this formed a bias in this research. Nowadays more sophisticated linguistic software programs can measure the tone of an article more accurate by analysing the full sentence and not only separated words. Thirdly, this study only examined the short-term performance of the acquisitions, however it could be interesting to analyse the impact of news on the long-term performance of the acquiring firm as well (Zollo & Meier, 2008). This should give a more comprehensive overview of the impact of news. Fourthly, this study made use of the cumulative abnormal return as dependent variable, it would be more accurate and interesting to use the abnormal return as dependent variable. Beside the fact that the abnormal return is more accurate, it could give an answer on the question why the study of
23
REFERENCES
Ahern, K. R., & Sosyura, D. (2014). Who writes the news? Corporate press releases during merger negotiations. The Journal of Finance, 69(1), 241-291.
Andrade, G., Mitchell, M. L., & Stafford, E. (2001). New evidence and perspectives on mergers. Journal of Economic Perspectives, 15(2), 103-120.
Bai, C. E., Liu, Q., Lu, J., Song, F. M., & Zhang, J. (2004). Corporate governance and market valuation in China. Journal of Comparative Economics, 32(4), 599-616.
Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D. (2001). Bad is stronger than good. Review of general psychology, 5(4), 323. Bollen, J., Mao, H., & Zeng, X. (2011). Twitter mood predicts the stock market. Journal of Computational Science, 2(1), 1-8.
Boyd, J., Hu, J., Jagannathan, R., (2005). The stock market’s reaction to unemployment news: why bad news is usually good for stocks. The Journal of Finance, 60(2), 649– 672.
Brown, L. D., & Caylor, M. L. (2004). Corporate governance and firm performance. Working paper.
Buehlmaier, M. M. (2013). The role of the media in takeovers: Theory and evidence. Working paper.
Cartwright, S., & Schoenberg, R. (2006). Thirty years of mergers and acquisitions research: Recent advances and future opportunities. British journal of management, 17(S1), S1-S5.
Chi, L. C. (2009). Do transparency and disclosure predict firm performance? Evidence from the Taiwan market. Expert Systems with Applications, 36(8), 11198-11203.
Chiyachantana, C. N., Nuengwang, N., Taechapiroontong, N., & Thanarung, P. (2013). The effect of information disclosure on information asymmetry. Investment Management and Financial Innovations, (1), 225-234.
Cornell, B. (2013). What moves stock prices: Another look. The Journal of Portfolio Management, 39(3), 32-38.
Cutler, D. M., Poterba, J. M., & Summers, L. H. (1989). What moves stock prices? The Journal of Portfolio Management, 15(3), 4-12. Daniel, K., Hirshleifer & D., Subrahmanyam, A., (1998). Investor psychology and security market under- and overreactions. Journal of Finance, 53(6), 1839 – 1885.
De Bondt, W. F., & Thaler, R. (1985). Does the stock market overreact?. The Journal of finance, 40(3), 793-805.
Engelberg, J. E., & Parsons, C. A. (2011). The causal impact of media in financial markets. The Journal of Finance, 66(1), 67-97.
Fama, E. F. (1970). Efficient capital markets: A review of theory and empirical work. The journal of Finance, 25(2), 383-417.
24 Florini, A., Birdsall, N., Flynn, S., Haufler, V.,
Lipton, D., Morrow, D., & Sharma, S. (2000). Does the invisible hand need a transparent glove? The politics of transparency. World Banks Annual Conference on Development Economics, 163-184.
Francis, J. & Soffer, L. (1997). The relative informativeness of analysts' stock recommendations and earnings forecast revisions. Journal of Accounting Research, 35(2) 193-211.
Fridkin, K. L & Kenney, P. J. (2004). ‘Do Negative Messages Work? The Impact of Negativity on Citizens' Evaluations of Candidates’, American Politics Research, 32(5), 570-606.
Garcia, D. (2013). Sentiment during recessions. The Journal of Finance, 68(3), 1267-1300.
Gelb, D. S., & Zarowin, P. (2002). Corporate disclosure policy and the informativeness of stock prices. Review of Accounting Studies, 7(1), 33-52.
Gupta, P. K. (2013). Mergers and acquisitions (m&a): the strategic concepts for the nuptials of corporate sector. Innovative Journal of Business and Management, 1(4), 60-80. Gutierrez, R. C., & Kelley, E. K. (2008). The Long‐Lasting Momentum in Weekly Returns. The Journal of Finance, 63(1), 415-447.
Healy, P. M., Palepu, K. G., & Ruback, R. S. (1992). Does corporate performance improve after mergers?. Journal of financial economics, 31(2), 135-175.
Helwege, J., Pirinsky, C., & Stulz, R. M. (2007). Why do firms become widely held? An analysis of the dynamics of corporate ownership. The Journal of Finance, 62(3), 995-1028.
Hirsch, P. M. (1972). Processing fads and fashions: An organization-set analysis of cultural industry systems. American journal of sociology, 639-659.
Hviid, M., & Prendergast, C. (1993). Merger failure and merger profitability. Journal Of Industrial Economics, 41(4), 371-386.
Ikenberry, D. L., & Ramnath, S. (2002). Underreaction to Self‐Selected News Events: The Case of Stock Splits. Review of Financial Studies, 15(2), 489-526.
Jegadeesh, N., & Wu, D. (2013). Word power: A new approach for content analysis. Journal of Financial Economics, 110(3), 712-729. Kopits, G., & Craig, J. D. (1998). Transparency in government operations (No. 158). Washington, DC: International monetary fund.
Kothari, S. P. (2000). The role of financial reporting in reducing financial risks in the market. Conference series-federal reserve bank of Boston, 44, 89-102.
Kummer, C., & Steger, U. (2008). Why merger and acquisition (M&A) waves reoccur: the vicious circle from pressure to failure. Strategic Management Review, 2(1), 44-63.
Kutner, M. H., Nachtsheim, C. J. & Neter, J. (2004). Applied Linear Regression Models (4th ed.). McGraw-Hill Irwin.
25 Lang, L. H., Stulz, R., & Walkling, R. A. (1991).
A test of the free cash flow hypothesis: The case of bidder returns. Journal of Financial Economics, 29(2), 315-335.
Lang, M., & Lundholm, R. (1993). Cross-sectional determinants of analyst ratings of corporate disclosures. Journal of accounting research, 31(2), 246-271.
Li, F. (2010). The information content of forward‐looking statements in corporate filings—A naïve Bayesian machine learning approach. Journal of Accounting Research, 48(5), 1049-1102.
Loughran, T., & McDonald, B. (2011). When is a liability not a liability? Textual analysis, dictionaries, and 10‐Ks. The Journal of Finance, 66(1), 35-65.
Lys, T., & Sohn, S. (1990). The association between revisions of financial analysts' earnings forecasts and security-price changes. Journal of Accounting and Economics, 13(4), 341-363.
MacKinlay, A. C. (1997). Event studies in economics and finance. Journal of economic literature, 35(1), 13-39.
Malkiel, B. G. (2003). The efficient market hypothesis and its critics. Journal of economic perspectives, 17(1), 59-82.
Mallikarjunappa, T. & P. Nayak, (2007) “Why do Mergers and Acquisitions quite often fail?”, AIMS International, 1(1), 53-69. McCarthy, K. J., Dolfsma, W. A., & Huizingh, K. R. E. (2012). Media Sentiment and the Propensity to Patent. Working Paper, Presented at the DRUID.
McCarthy, K. J., & Dolfsma, W. (2014). Neutral Media? Evidence of Media Bias and its Economic Impact. Review of Social Economy, 72(1), 42-54.
Mian, G. M., & Sankaraguruswamy, S. (2012). Investor Sentiment and Stock Market Response to Earnings News. Accounting Review, 87(4), 1357-1384.
Miller, B. P. (2010). The effects of reporting complexity on small and large investor trading. The Accounting Review, 85(6), 2107-2143.
Niven, D. (2000). “The Other Side of Optimism: High Expectations and the Rejection of Status Quo Politics.” Political Behavior, 22(1), 71–88.
Odean, T., (1998). Volume, volatility, price and profit when all traders are above average. Journal of Finance, 53(6), 1887 – 1934.
Organisation for Economic Co-operation and Development. (1999). OECD principles of corporate governance. OECD.
Pritamani, M., & Singal, V. (2001). Return predictability following large price changes and information releases. Journal of Banking & Finance, 25(4), 631-656.
Rosen, R. J. (2006). Merger Momentum and Investor Sentiment: The Stock Market Reaction to Merger Announcements. The Journal of Business, 79(2), 987-1017.
26 Schoenberg, R. (2006). Measuring the
Performance of Corporate Acquisitions: An Empirical Comparison of Alternative Metrics. British Journal of Management, 17(4), 361-370.
Schumaker, R.P & H. Chen (2009), Textual analysis of stock market prediction using breaking financial news: the AZFin text system, ACM Transactions on Information Systems, 27(2), 1-12.
Sharpe, S. A. (2002). Re-examining stock valuation and inflation: the implications of analysts' earnings forecasts. Review of Economics and Statistics, 84(4), 632-648. Shleifer, A. (2000). Inefficient markets: An introduction to behavioural finance. Oxford university press.
Singh, R., & J.B.P. Teoh. (2000). “Impression Formation from Intellectual and Social Traits: Evidence for Behavioural Adaptation and Cognitive Processing.” British Journal of Social Psychology 39(4), 537–54.
Soroka, S. N. (2006). Good news and bad news: Asymmetric responses to economic information. Journal of Politics, 68(2), 372-385.
Stiglitz, J. E. (2000). The contributions of the economics of information to twentieth century economics. Quarterly Journal of economics, (115)4, 1441-1478.
Tetlock, P. C. (2007). Giving content to investor sentiment: The role of media in the stock market. The Journal of Finance, 62(3), 1139-1168.
Tetlock, P.C., M. Saar-Tsechansky & S. Macskassy (2008), More Than Words: Quantifying Language to Measure Firms' Fundamentals, The Journal of Finance, 63(3), 1437–1468.
Tetlock, P. C. (2010). Does public financial news resolve asymmetric information?. Review of Financial Studies, 23(9), 3520-3557.
Tian, G. Y., & Twite, G. (2011). Corporate governance, external market discipline and firm productivity. Journal of Corporate Finance, 17(3), 403-417.
Van de Kauter, M., Breesch, D., & Hoste, V. (2015). Fine-grained analysis of explicit and implicit sentiment in financial news articles. Expert Systems with Applications, 42(11), 4999-5010.
Veronesi, P. (1999). Stock market overreactions to bad news in good times: a rational expectations equilibrium model. Review of Financial Studies, 12(5), 975-1007.
Vonk, R. (1996). “Negativity and Potency Effects in Impression Formation.” European Journal of Social Psychology, 26(6), 851–65. Weber, R. A., & Camerer, C. F. (2003). Cultural conflict and merger failure: An experimental approach. Management Science, 49(4), 400-415.
27 Yu, Y., Duan, W., & Cao, Q. (2013). The impact
of social and conventional media on firm equity value: A sentiment analysis approach. Decision Support Systems, 55(4), 919-926.
Zollo, M., & Meier, D. (2008). What is M&A performance?. The Academy of Management Perspectives, 22(3), 55-77.
28
APPENDIX I
Acquisitions per year:
Acquisitions per industry:
Acquisitions per country:
29
APPENDIX II
Steps Criteria Acquisitions Articles
Step 1: Collect acquisitions from SDC Thomson
database. (1) between 01/01/2012 and 31/12/2014, (2) deal value > 10 million dollars, (3) 100%
ownership 12,098
Step 2: Filter on firms that are listed on Transparency
list. (1) SEDOL number or organization name 278
Step 3: Collect articles one month after deal of announcement
(1) Delete business entities, (2) at least two time occurrence of acquiring and target firm, (3) > 75 articles, add "acquisition”
183 2975
Step 4: Check quality of articles one month after deal
of announcement (1) duplicates, (2) no other acquisitions mentioned 183 2275 Step 5: Collect articles one month before deal of
announcement
(1) Delete business entities, (2) > 75 articles, add "acquisition”
43 317
Step 6: Check quality of articles one month after deal
of announcement (1) duplicates, (2) no other acquisition mentioned 43 244
30
APPENDIX III
15 randomly selected articles are rated by two students independently on a scale of 0 till 10, whereas 0 means not negative at all and 10 that the article is highly negative. In the other table the correlation is shown.
Acquisition ID Article Number of words Media Pessimism Rating Student 1 Rating Student 2
1 4 253 0,023715 1 1 31 3 229 0,017467 2 2 62 7 236 0,059322 10 8 69 35 114 0,052632 4 4 89 9 334 0,128743 9 9 101 12 524 0,03626 1 2 125 19 261 0,015326 1 3 132 8 536 0,020522 4 7 150 1 281 0,010676 1 5 157 7 437 0,02746 4 7 164 4 672 0,028274 6 5 171 22 228 0,035088 4 4 178 39 210 0,07619 6 4 180 45 471 0,016985 4 7 184 15 262 0,034351 3 3
Media Pessimism Student 1 Student 2 Media Pessimism 1.000
Student1 0.7310 1.000
31
APPENDIX IV
Variable Description Database
Cumulative Abnormal
Return The sum of the difference between the normal and expected return. Datastream, calculated with Stata Media Pessimism Number of negative words divided by the total number of words in an article (Tetlock, 2007)
LexisNexis Media Coverage Total amount of articles per acquisition (Fang & Peress, 2009)
LexisNexis Sum of Words Total sum of words used in all the articles per acquisition.
LexisNexis
Size (employees) The size of the firm in terms of employees Datastream
WC07011
Volatility The change in security values per firm Datastream
WC080234 Relative Deal Value Calculated by dividing the deal value of the acquisition by the market value of the acquiring firm Deal value –Thompson SDC
Market Value – Datastream MV
HQ/Subsidiary If the acquiring firm is a parent firm or subsidiary, 1 = HQ, 0 = subsidiary
Thompson SDC Earnings to price ratio The expected earnings of the acquiring firm in a coming period. Datastream
WC01250 / MV Circulation The average of the circulation of the newspapers involved per acquisition
LexisNexis Book to Market Ratio of the book value of the acquiring firm to its market value Datastream
1/MTBV TobinsQ The market value of a firm divided by the replacement value of the firm’s assets Datastream Transparency The availability of all firm specific information, based on three categories: anti-corruption programs, organizational
