Value Creation in M&A Transactions, Conference Calls, and Shareholder Protection

Abstract

This study investigates whether conference calls accompanying M&A announcements in Europe provide valuable information for capital market participants and hence induce an abnormal stock price revaluation on the bidder’s equity. Based on handpicked data for transactions between 2008 and 2012 we focus on the five most acquisitive country markets in Europe. Overall, our results show that bidders are more likely to conduct conference calls with increasing transaction value, for transactions with public targets and non-diversifying transactions. Further, the decision for voluntary disclosure is positively influenced by increased bidder size and the comparably weaker governance systems for German and Swiss firms. After controlling for self-selection bias and other determinants of stock returns around mergers and acquisitions (M&A) announcement, evidence is in strong support that firms with merger-related conference calls yield a higher abnormal return than firms merely publishing a press release. However, significant favourable investor reaction is only present in the UK and French subsamples and in the subsamples of industries with a focus on research and development (R&D).

1. Introduction

In recent years, conference calls emerged to a standard medium of voluntary disclosure (Bushee et al. 2003). They are mostly utilized in concordance with earnings announcements to provide detailed and additional information over the respective press release (Tasker 1998; Frankel et al. 1999; Matsumoto et al. 2011; Bassemir et al. 2013; Mayew et al. 2013; Green et al. 2014a; Cicon 2017). Confronted by the adverse selection and undervaluation problem arising from perceived information asymmetry, the management can try to enhance the investors’ understanding of the firm’s current situation and future perspectives through voluntary disclosure and thereby signal value creation (Lev and Penman 1990; Verrecchia 2001; Healy and Palepu 2001).

Conference calls, like web casts, distinguish themselves from other communication channels through the interactive nature of the disclosure process and the accessibility to a large group of market participants. The interaction between the management and the audience is thought to facilitate a customized information transfer adapted to the informational demand of capital market participants (Mayew and Venkatachalam 2012). Through the revelation of an incrementally higher informational content above the respective press release in combination with broad information dissemination through large-scale telephone conferences, conference calls are assumed to contribute to the reduction of information asymmetry (Tasker 1998; Frankel et al. 1999; Matsumoto et al. 2011).

The quality and quantity of the content disclosed in conference calls and the capital market reactions thereon have been extensively investigated for earnings calls (Tasker 1998; Frankel et al. 1999; Bushee et al. 2003; Barker et al. 2012; Chung et al. 2012; Green et al. 2014b; Cicon 2017). Consistent with the market responses to voluntary disclosure in general, most studies associate conference calls with improvement in analyst forecast accuracy and precision (Bowen et al. 2002), decreased analyst underreaction, timelier analyst response (Kimbrough 2005) and higher abnormal returns (Matsumoto et al. 2011). Consequently, conference calls are informative to the capital market. However, most research rather focuses on routinely conference calls accompanying earnings releases particularly in the United States (US), whereas conference calls and more generally voluntary disclosure in non-routine situations or outside the US or both are rarely examined and still offer rather mixed results.

Our study aims at filling this gap and obtaining knowledge about whether merger-related conference calls in the most active European M&A markets (United Kingdom (UK), France, Germany, Switzerland, and Spain) transmit additional information over the respective press release and thereby induce stock market revaluation. Furthermore, with the background of diverging governance systems within Europe we investigate whether cross-national and cross-industry differences can be observed with regard to disclosure practices and stock market responses. Consequently, we adopt the argument of another strand of disclosure research which finds that differences in legal, institutional environment, accounting practices and culture lead to country-specific disclosure patterns (Bailey et al. 2006; Ball et al. 2000). Methodologically, our study follows Kimbrough and Louis (2011) who examine merger-induced conference calls between 2002 and 2006 in the US.

The remainder of this study is organised as follows. Section 2 reviews the existing literature on voluntary disclosure and conference calls, particularly around M&A announcements in the context of institutional, regulatory and cultural determinants of country-specific disclosure environment. Section 3 describes the sample selection process and presents the descriptive statistics for the full sample and the country subsamples. Thereafter, a probit-model and the event study methodologies are explained. Results are presented in Section 4. Section 5 concludes and provides recommendations.

2. Literature Review

2.1. Conference Calls around Mergers and Acquisitions

The increasing importance of conference calls as voluntary disclosure medium has induced a growing body of disclosure research to detect the incentives and capital market consequences of conference calls (Bushee et al. 2003; Bassemir et al. 2013). In general, voluntary disclosure offers additional means of investor communication besides mandated disclosure. Mandatory disclosure typically relates to historical financial data, while voluntary disclosure often includes forward-looking details (Healy and Palepu 2001). In the setting of inefficient markets with costly information (Grossman and Stiglitz 1980), the demand for voluntary disclosure arises from information asymmetry which induces agency problems (Myers and Majluf 1984) and risk of trading against better informed investors on the part of the investors (Diamond and Verrecchia 1991; Easley and O’Hara 2004), and risk of adverse selection and undervaluation on the part of the management (Akerlof 1970; Verrecchia 2001).

Voluntary disclosure is hypothesized to reduce information asymmetry through revealing private information and thereby reducing private information search incentives (Healy and Palepu 2001; Easley and O’Hara 2004; Barker et al. 2012; Chung et al. 2012; Green et al. 2014b). Under the assumption that the management’s objective is to maximize the firm’s value, the management provides voluntary disclosure when the perceived risk of undervaluation outweighs the management’s natural aversion to disclosure attributable to proprietary costs, information dissemination costs, litigation risk and loss of reputation (Verrecchia 1983; Myers and Majluf 1984; Healy and Palepu 2001). Evidence suggests that the management’s involvement in voluntary disclosure is further encouraged by expected capital market benefits related to stock performance (Healy et al. 1999; Lang and Lundholm 2000), liquidity (Diamond and Verrecchia 1991; Healy et al. 1999), analyst coverage (Lang and Lundholm 1996; Healy et al. 1999), institutional ownership (Diamond and Verrecchia 1991) and most prominently, cost of capital (Diamond and Verrecchia 1991; Botosan 1997; Easley and O’Hara 2004; Leone et al. 2007). However, voluntary disclosure does not imply uniformly positive capital market response. Unfavourable news and low source credibility induce strong negative market reactions (Skinner 1994; Hutton et al. 2003; Kothari et al. 2009), leading in incentives for the management to optimistically skew their disclosures and reports according to Kothari et al. (2009).

Conference calls are conducted via large-scale telephone conference calls and comprise a presentation by the management and a Q&A session thereafter. In comparison to other standard voluntary disclosure media such as management forecasts, analysts’ presentations and press releases, conference calls distinguish themselves by a unique combination of interactivity, verbal transmission, lower legal restrictions and the capacity to accommodate a large number of participants (Tasker 1998). For the management, conference calls offer a platform of corporate communication which allows to flexibly adjust the disclosure level according to the demand of the audience (Frankel et al. 1999; Matsumoto et al. 2011), to support the messages with linguistic items (Camiciottoli 2011), to speak informally and candidly to the audience (Tasker 1998; Frankel et al. 1999) and to reach a vast number of market participants in a timely manner, thereby mitigating selective disclosure problems and reducing information dissemination costs (Frankel et al. 1999; Bowen et al. 2002). For investors and analysts, the Q&A session enables them to elicit additional and detailed information via follow-questions (Tasker 1998; Frankel et al. 1999) or from verbal cues (Mayew and Venkatachalam 2012). The uncovering of information may be exercised to the extent to which the disclosure of information is not entirely voluntary when the audience persists on the disclosure of information that the management is reluctant or unable to provide. The management is left in a disclosure dilemma in view of the negative impression of silence, prompting the investors to discount the firm’s value consistent with the level of disclosure (Dye 1985; Jung and Kwon 1988; Verrecchia 1990).

The informational content of conference calls and their contribution to the reduction of information asymmetry is controversially debated in view of limited accessibility for individual investors and other stakeholders (Tasker 1998; Bushee et al. 2003; Davis et al. 2012; Mayew et al. 2013), discrimination among investment professionals (Mayew 2008) and deliberate withholding of information during the call (Hollander et al. 2010; Cicon 2017). However, numerous studies support the notion that conference calls transmit additional material information to a wider audience, thereby contributing to the reduction of information asymmetry. Whereas Brown et al. (2004) use a direct proxy of information asymmetry to manifest its negative relation to conference calls, other studies establish their conclusions on heightened levels of absolute return volatility and trading volume (Frankel et al. 1999), larger abnormal returns (Matsumoto et al. 2011; Barker et al. 2012), improvement in analyst forecast accuracy and precision (Bowen et al. 2002), and reduction in the post-earnings-announcement drift as well as in the initial analyst underreaction (Kimbrough 2005).

Conference calls are primarily held in conjunction with quarterly earnings releases, which consequently have been the major object of interest (Tasker 1998; Frankel et al. 1999; Kimbrough 2005). Although the incentives of conference calls, and more generally voluntary disclosure, are assumed to be particularly pronounced for anticipated capital market transactions because of the positive relation between voluntary disclosure and cost of capital (Healy and Palepu 2001; Suijs 2007), only few studies investigate voluntary disclosure around capital market transactions such as equity offerings (Lang and Lundholm 2000; Leone et al. 2007) and M&A (Kimbrough and Louis 2011).

M&A transactions are unique events with a particularly high degree of information asymmetry to which the management may face the need to adjust their disclosure policy. In times of fundamental transitions, firms are confronted by serious investor scepticism due to high financial risk and potential compromise of the firm’s identity, further intensified by complexity and low transparency of the deal structure. Hence, the demand for material information and forward-looking details is likely to be especially high (Heldenbergh et al. 2006; Dutordoir et al. 2014).

As the investors’ perception plays a pivotal role in the deal completion through the direct impact on the bidder’s share price during the bid phase, prior research suggests an increase in the incentives of using voluntary disclosure to convince the investors of the transaction rationales (Suijs 2007; Kimbrough and Louis 2011; Fraunhoffer and Schiereck 2012). Brennan (2000) reports an elevated level of forecast disclosure around M&A resulting in industry-specific effects, whereas Amel-Zadeh et al. (2013) document the effectiveness of voluntary disclosure by shorter closing time and improved completion likelihood. Dutordoir et al. (2014) and Fraunhoffer and Schiereck (2012) document positive investor responses on synergy announcements. Kimbrough and Louis (2011) provide similar evidence for conference calls around initial M&A announcements. Consistent with the signalling hypothesis, empirical evidence is in favour of a bias towards positive forecasts related to the deal (Brennan 2000; Amel-Zadeh et al. 2013) and upward earnings management in the pre-merger period (Erickson and Wang 1999). Transactions supported by voluntary disclosure incline to be more complex, of high economic significance, and to involve stock payments (Kimbrough and Louis 2011; Dutordoir et al. 2014). These findings are consistent with results from Suijs (2007). They are suggesting that transactions, which require a greater level of disclosure, tend to be less favourable.

2.2. Firm-Specific and Country-Specific Determinants of Disclosure Environment

Disclosure decisions and the corresponding market reaction exhibit large variations across firms leading to subsequent research of determinants of disclosure patterns. With regard to firm characteristics, prior research associates firms exercising forthcoming disclosure policies with larger size (Lang and Lundholm 1993; Frankel et al. 1999), greater analysts following (Frankel et al. 1999), public status (Healy et al. 1999) and rapid growth (Frankel et al. 1999). Firms in traditional industries such as oil, gas or financials are found to be less likely to conduct conference calls contrary to R&D intensive industries such as pharmaceuticals, information technology, communication and electronics (Botosan 1997; Frankel et al. 1999; Vanstraelen et al. 2003).

With regard to the accessibility of conference calls, Bushee et al. (2003) show that closed conference calls are more likely to be held by firms with higher intangible assets, complex financial information and low financial statement informativeness as these circumstances are more prone to misinterpretation by unsophisticated investors. Open conference calls are rather conducted by firms facing greater information asymmetry, typically in high-tech industries with lower analyst following, lower institutional ownership, dispersed investor base, and greater revenue volatility. Although disclosure policy is often regarded as an “exclusively firm-specific phenomenon” (Archambault and Archambault 2003, p. 174), a growing body of international research documents national patterns of disclosure which is determined by legal and institutional factors (Ball et al. 2000; Conover et al. 2008) and cultural background (Gray 1988; Hope 2003), and remain persistent despite regulatory convergences such as the International Financial Reporting Standards (IFRS) (Kvaal and Nobes 2010).

Starting with La Porta et al. (1998) a widely investigated field is concerned with the relation between legal origin and disclosure practices based on the distinction between common-law and code-law countries. Overall, the research hypothesizes that firms in common-law countries, being imposed stringent disclosure requirements and experiencing higher pressure from the investors, exercise timelier and more forthcoming disclosure policies than in code-law countries. Higher level of disclosure in common-law countries—most exemplary, the UK and the US—is in part explainable by the prevalent shareholder-orientation, and higher levels of legal investor protection (Leuz et al. 2003), dispersion of ownership structure (Eng and Mak 2003) and takeover market activity (Santema et al. 2005). In contrast, code-law countries are typically associated with higher stakeholder-orientation, weaker legal investor protection, higher ownership concentration and lower takeover market activity with the exception of France being the second largest market after the UK in Europe. Exemplary evidences of shareholder-oriented disclosure practices in common-law countries are documented by Ball et al. (2000) for selected groups of common-law countries (Australia, Canada, UK, US) and code-law countries (France, Germany, Japan), and by Conover et al. (2008) for a larger sample of 22 countries across the world including the US, the UK, Germany, Switzerland, France and Spain, in order of descending disclosure quality.

Cultural dimensions of a disclosure decision environment are rather less researched as routinely mandatory disclosure is considered to be determined by the countries legal and institutional factors. However, cultural factors reveal an important role in explaining further variations of disclosure practices within similar legal and institutional environment (Hope 2003). A growing body of international disclosure literature sheds light on the cultural differences and their implications on disclosure practices based on the four work-related dimensions of national cultures by Hofstede (1980, 2001): power distance, uncertainty avoidance, collectivism and femininity. In particular, the disclosure level is found to be negatively affected by uncertainty avoidance (Gray and Vint 1995; Hope 2003) due to the secretiveness of the management (Gray 1988), and positively influenced by individualism due to higher competitiveness (Jaggi and Low 2000). Bridging the gap between economic, institutional, legal and cultural factors, Santema et al. (2005) offer a comprehensive survey of the UK, German and French disclosure environment. The results indicate that UK firms with high shareholder-orientation, low level of uncertainty avoidance and high level of individualism disclose most, whereas French firms with lower shareholder-orientation, high uncertainty avoidance and intermediate individualism display the lowest disclosure level.

3. Sample Selection and Descriptive Statistics

The following analysis is based on a sample of M&A transactions obtained from Thomson One Banker SDC database according to the following requirements: The bidder is a publicly quoted company-based in Europe excluding the Russian Federation and Turkey. We moreover include transactions with a deal value of more than USD 1 million (m) announced between 1 January 2008 and 30 June 2012. 6568 transactions were selected according to these criteria with an aggregate transaction value of USD 1887 billion (bn). Conference call data was also obtained from Thomson One Banker SDC database, section Company Views—News & Corporate Events. To ensure an efficient screening process and a clean data base, the regional focus has been limited to the top five acquirer nations in terms of transaction value: the UK, Switzerland, France, Germany and Spain. These countries represent 76% of the aggregate transaction value during the observation period.

From the sample of 6568 announced transactions, 3652 transactions remained after narrowing down to the top five acquirer nations. Further, entries lacking Datastream Codes, with non-retrievable stock return data and duplicate entries were excluded, whereas multiple transactions per day and company were consolidated. Self-tenders and other transactions with non-distinguishable acquirer and target, as well as entries with missing key accounting and financial data were also excluded. At last, non-discrete variables are winsorized at the 1%- and 99%- level to prevent possible outliers following the typical handling of voluntary disclosures in general (e.g., Conn et al. 2005; Matsumoto et al. 2011; Dutordoir et al. 2014). Thereafter, 2622 remaining transaction announcements were screened for supplementary conference calls. 243 merger-related conference call events were hand-collected implicitly assuming that there have been no others additional but not shown on the Thomson page. 216 conference calls were identified of being held on the same day or the next business day after the merger announcement. The rational to restrict to a relatively narrow event window follows the argument by Kimbrough and Louis (2011) that once the market perception on the merger is established, the market’s response will hardly be influenced through voluntary disclosure. The intuition is that management will rather provide material information within a short period after the announcement to yield a positive investor reaction.

As the market response observed on the day of the conference call is presumably influenced by both the respective merger announcement and the conference call, the effect of the latter will be isolated by comparing the abnormal return of the conference calls sample to a control sample which encompasses all other merger announcements not supplemented by a conference call. The filtering process is summarized in

Table 1. Sample Selection Procedure.

Characteristic	Number of Observations
Grand Sample	6568
Acquirer nation not included in the list of top 5 acquirer nations	(2916)
Top 5 acquirer nations (UK, Switzerland, France, Germany, Spain)	3652
Acquirer Datastream Code not available	(120)
Stock data not retrievable	(21)
Duplicates (Identical acquirer, target and announcement date)	(80)
Consolidation (Identical acquirer and announcement date, several targets)	(107)
Self-tender (Identical target and acquirer name and CUSIP)	(295)
Missing key accounting data (Acquirer market cap, total assets)	(407)
Screening Sample	2622
Insufficient stock data	(104)
Full Sample	2518
- Event Sample: Conference calls held on the day of the transaction announcement or the next business day	216
- Control Sample: Other transaction announcements	2302

.

In

Table 2. Descriptive Statistics.

Variable	Full Sample			Event Sample (CC = 1)			Control Sample (CC = 0)			Difference to Event Sample
	Obs.	Mean	Median	Obs.	Mean	Median	Obs.	Mean	Median	Mean	Median
Acquirer Nation
UK	1605	0.637	1.000	93	0.431	0.000	1512	0.657	1.000	−0.226 ***	−1.000 ***
FR	344	0.137	0.000	38	0.176	0.000	306	0.133	0.000	0.043 *	0.000 *
DE	260	0.103	0.000	37	0.171	0.000	223	0.097	0.000	0.074 ***	0.000 ***
CH	156	0.062	0.000	34	0.157	0.000	122	0.053	0.000	0.104 ***	0.000 ***
ES	153	0.061	0.000	14	0.065	0.000	139	0.060	0.000	0.004	0.000
Acquirer Industry
CPS	278	0.110	0.000	14	0.065	0.000	264	0.115	0.000	−0.050 **	0.000 **
ENERGY	198	0.079	0.000	27	0.125	0.000	171	0.074	0.000	0.051 ***	0.000 ***
FINANCE	346	0.137	0.000	25	0.116	0.000	321	0.139	0.000	−0.024	0.000
HEALTH	187	0.074	0.000	33	0.153	0.000	154	0.067	0.000	0.086 ***	0.000 ***
HT	310	0.123	0.000	27	0.125	0.000	283	0.123	0.000	0.002	0.000
IND	350	0.139	0.000	26	0.120	0.000	324	0.141	0.000	−0.020	0.000
MATERLS	199	0.079	0.000	21	0.097	0.000	178	0.077	0.000	0.020	0.000
MEDIA	200	0.079	0.000	10	0.046	0.000	190	0.083	0.000	−0.036 *	0.000 *
REALEST	154	0.061	0.000	10	0.046	0.000	144	0.063	0.000	−0.016	0.000
RETAIL	103	0.041	0.000	6	0.028	0.000	97	0.042	0.000	−0.014	0.000
STAPLES	103	0.041	0.000	10	0.046	0.000	93	0.040	0.000	0.006	0.000
TELECOM	90	0.036	0.000	7	0.032	0.000	83	0.036	0.000	−0.004	0.000
Announcement Year
2008	686	0.272	0.000	40	0.185	0.000	646	0.281	0.000	−0.095 ***	0.000 ***
2009	452	0.180	0.000	29	0.134	0.000	423	0.184	0.000	−0.049 *	0.000 *
2010	608	0.241	0.000	63	0.292	0.000	545	0.237	0.000	0.055 *	0.000 *
2011	554	0.220	0.000	57	0.264	0.000	497	0.216	0.000	0.048	0.000
Deal Characteristics
CC	2518	0.086	0.000	216	1.000	1.000	2302	0.000	0.000	1.000	1.000 ***
TV	2518	0.403	0.026	216	2.330	0.368	2302	0.222	0.022	2.107 ***	0.346 ***
DEALRATIO	2518	0.268	0.041	216	0.247	0.092	2302	0.270	0.036	−0.023	0.056 ***
PUBLIC	2518	0.168	0.000	216	0.361	0.000	2302	0.150	0.000	0.211 ***	0.000 ***
CASH	2518	0.362	0.000	216	0.444	0.000	2302	0.354	0.000	0.090 ***	0.000 ***
FOREIGN	2518	0.519	1.000	216	0.671	1.000	2302	0.505	1.000	0.167 ***	0.000 ***
INDR	2518	0.641	1.000	216	0.801	1.000	2302	0.626	1.000	0.175 ***	0.000 ***
AMKTCAP	2518	11.788	0.910	216	19.680	3.747	2302	11.048	0.749	8.633 ***	2.998 ***
ABM	2518	1.476	0.571	216	0.566	0.446	2302	1.561	0.591	−0.995 ***	−0.145 ***
AEqMultiplier	2518	10.415	1.352	216	3.532	1.153	2302	11.061	1.365	−7.529	−0.212
Observations	2518			216			2302			2518

* p < 0.10, ** p < 0.05, *** p < 0.01. Mean differences and median differences are based on t-test and Wilcoxon rank sum test, respectively. p-values are two-tailed. Acquirer Nation: Variable name according to the country code by ISO3166 (UK: United Kingdom; FR: France; DE: Germany; CH: Switzerland; ES: Spain). VARIABLE = 1 if acquirer nation is respective nation, and 0 otherwise; Acquirer Industry: Variable name according to the Thomson One Banker Industry Coding (CPS: Consumer Products and Services; ENERGY: Energy and Power; FINANCE: Financials; HEALTH: Healthcare; HT: High Technology; IND: Industrials; MATERLS: Materials; MEDIA: Media and Entertainment; REALEST: Real Estate; RETAIL: Retail; STAPLES: Consumer Staples; TELECOM: Telecommunications). VARIABLE = 1 if acquirer industry is respective industry, and 0 otherwise; Announcement Year: VARIABLE = 1 if transaction was announced in the respective year, and 0 otherwise; Transaction Characteristics: CC = 1 when merger announcement is accompanied by a conference call, and 0 otherwise, TV = Transaction value (expressed in USD bn), DEALRATIO = Ratio of the total transaction value to the acquirer’s pre-announcement market value as of four weeks prior to the transaction announcement, PRIV = 1 when the target firm is a private company, and 0 otherwise, CASH = 1 for all-cash consideration, and 0 otherwise, FOREIGN = 1 when the target firm is a foreign company, and 0 otherwise, INDR = 1 when acquirer and target firm have the same industry specification, AMKTCAP = Acquirer’s market value of equity four weeks prior to the transaction announcement (expressed in USD bn), ABM = Acquirer’s book-to-market ratio computed as book value last 12 months over acquirer’s market value of equity four weeks prior to the transaction announcement, AEqMultiplier = Leverage ratio of the acquirer’s total asset to the acquirer’s pre-announcement market value as of four weeks prior to the transaction announcement.

, descriptive statistics regarding acquirer origin, industry, announcement year and deal characteristics are presented for the full sample, the conference call subsample (Event Sample) and the non-conference call subsample (Control Sample). Overall, supplementing conference calls are held for 8.6% of the full sample.1

In absolute values (Obs.), the UK displays the highest disclosure frequency of 93 conference calls in strong contrast to Spain, where only 14 conference calls have occurred. France, Germany, and Switzerland take an intermediate stance with about 35 conference calls each. Industry subsamples exhibit a more evenly distribution of conference calls with six industries associated with more than 20 conference calls and other six industries with less than 20 conference calls. The most active industry, being the healthcare sector, accounts for 33 conference calls. The analysis of relative conference call activities offers a more in-depth view. While acquirers from the UK are substantially less represented in the conference call sample (5.8%), acquirers from France (11.0%), Germany (14.2%), and Switzerland (21.8%) have a higher portion in the conference call sample than non-conference call sample. Furthermore, acquirers from energy (13.6%) and healthcare sectors (17.6%) are more heavily represented in the event sample. On the contrary, the sectors consumer products and services (5.0%), and media and entertainment (5.0%), appear to conduct conference calls less frequently.

With regard to the deal characteristics, some statistically significant differences are present between the event sample and the control sample. Overall, transactions accompanied by conference calls display substantially higher transaction values similar to Kimbrough and Louis (2011), suggesting that bidders engaging in conference calls seek for more economically important deals. Also consistent with Kimbrough and Louis (2011), the conference call sample contains more public or foreign targets. However, transactions with conference calls appear to involve more all-cash deals and consolidating transactions which is in contrast to the observations by Kimbrough and Louis (2011). Bidder companies engaging in conference calls exhibit a substantially higher market value and lower book-to-market ratio in conjunction with prior research (Lang and Lundholm 1993; Frankel et al. 1999; Kimbrough and Louis 2011).

Methodology

We apply a probit model similar to Kimbrough and Louis (2011) to explain which factors determine the management’s decision to conduct conference calls. Therefore, the full sample of 2518 observations is employed. The probit model includes deal characteristics, acquirer financial status, selected industry membership and country fixed effects. Additionally, the model allows controlling for the influence of economic crisis by clustering to crisis and non-crisis periods. The likelihood of conference calls is estimated for the full sample as a whole and subsample specific by bidder countries using the following equation:

$$\begin{array}{ll}C{C}_i={\beta }_0+{\beta }_{1}T{V}_i& +{\beta }_{2}DEALRATI{O}_i+{\beta }_{3}PUBLI{C}_i+{\beta }_{4}CAS{H}_i+{\beta }_{5}FOREIG{N}_i\\ & +{\beta }_{6}IND{R}_i+{\beta }_{7}LNAMKTCA{P}_i+{\beta }_{8}AB{M}_i+{\beta }_{9}AcqMultiplie{r}_i\\ & +{\beta }_{10}CRISI{S}_i+{\beta }_{11}FINANC{E}_i+{\beta }_{12}HEALT{H}_i\\ & +{\beta }_{13}H{T}_i+countryfixedeffects+{\epsilon }_i\end{array}$$

(1)

with:

• CC = 1 when merger announcement is accompanied by a conference call on the day of the announcement or the day after, and 0 otherwise;
• TV = Transaction value (expressed in USD bn);
• DEALRATIO = Ratio of the total transaction value to the acquirer’s pre-announcement market value as of four weeks prior to the transaction announcement;
• PUBLIC = 1 when the target firm is a public company, and 0 otherwise;
• CASH = 1 for all-cash consideration, and 0 otherwise;
• FOREIGN = 1 when the target firm is a foreign company, and 0 otherwise;
• INDR = 1 when acquirer and target firm have the same industry specification;
• LNAMKTCAP = Log of the acquirer’s market value of equity four weeks prior to the transaction announcement (expressed in USD bn);
• ABM = Acquirer’s book-to-market ratio computed as book value last 12 months divided by acquirer’s market value of equity four weeks prior to the transaction announcement;
• AEqMultiplier = Leverage ratio of the acquirer’s total asset to the acquirer’s pre-announcement market value as of four weeks prior to the transaction announcement;
• CRISIS = 1 when the announcement year is 2008, 2009 or 2010, and 0 otherwise.
• FINANCE = 1 when the acquirer is a financial firm, and 0 otherwise;
• HEALTH = 1 when the acquirer is a healthcare firm, and 0 otherwise;
• HT = 1 when the acquirer is a high technology firm, and 0 otherwise.

Following Kimbrough and Louis (2011), the model controls for deal characteristics which may influence the intensity of the demand for additional disclosure. While transaction value (TV) and deal ratio (DEALRATIO) indicate the economic importance, other characteristics such as the public status (PUBLIC), target nationality (FOREIGN), and industry relatedness (INDR) are included as determinants for the level of information asymmetry inherent in the deal environment. Public companies and non-diversifying targets supposedly exhibit a lower degree of information opaqueness opposite to foreign targets. Further, the method of payment (CASH) is considered based on the notion of Kimbrough and Louis (2011) that cash considerations induce lower informational needs than stock considerations due to the simpler transaction structure. Additionally, several metrics of the acquirer’s financial status such as market capitalization (LNAMKTCAP) and book-to-market-ratio (ABM) are included to control for general disclosure patterns according to previous research which documents that larger companies and companies with lower book-to-market-ratios are more likely to engage in voluntary disclosure, in particular, conference calls (Lang and Lundholm 1993; Frankel et al. 1999).

A measure for the firm’s leverage (AEqMultiplier) is taken into account following the evidences of Zarzeski (1996) and Hope (2003) that financial leverage is negatively related to disclosure as debtors are more likely to have direct access to information and do not require a high level of public disclosure. Leverage is measured by total assets over market capitalization of the acquirer. To capture the potential impact of the global financial crisis and the Eurozone crisis at their climax, the variable CRISIS clusters 2008, 2009 and 2010 to a period of economic crisis. With regard to the acquirer industry, the sectors healthcare (HEALTH), financials (FINANCE) and high technology (HT) are chosen due to their highly contrasting profiles similar to Kimbrough and Louis (2011). While the financial sector is representative for traditional sectors with a typically higher degree of information transparency, healthcare companies and high technology companies are exemplary for sectors with higher cost of proprietary costs and thus, lower disclosure incentives. The healthcare and high technology sector insofar contrast each other as the former appears to exercise a more forthcoming disclosure policy according to Botosan (1997). Further, dummy-variables are used for countries to capture the country-specific likelihood of conference calls. The UK serves as base country, hence a positive coefficient of a country indicates that the acquirer firm-based in that respective country is more likely to conduct conference calls than UK firms.

We apply an event study framework to assess the impact of merger-related conference calls on the firm’s value. The impact is proxied by abnormal returns which are computed as the deviation between the actual returns and the market’s prior expectations according to MacKinlay (1997) and Kothari and Warner (2008). Our event study takes an estimation window of 200 days before the event ranging from day −230 to day −31 and an event window from day −10 to day +10 relative to the event date (t = 0). The event date is defined as the date of conference call for the event sample and the date of M&A announcement for the control sample. Following MacKinlay (1997), a market model is applied to estimate the expected returns based on Standard and Poor’s (S&P) Europe 350 as the benchmark index which includes all five stock markets under consideration. Further, country-specific indices FTSE100, CAC40, DAX30, SMI20 and IBEX35 are employed for sensitivity checks while using heteroscedasticity-consistent standard errors (Huber 1967; White 1980). These abnormal returns are accumulated over different time periods (i.e., cumulated abnormal return (CAR)) and then averaged over all firms (i.e., cumulated average abnormal return (CAAR)).

Both univariate tests and multivariate tests are used to isolate stock market reaction attributable to conference calls. Similar to Kimbrough and Louis (2011) only the time periods subsequent to the conference calls are utilized as for the multivariate tests, because conference calls are hypothesized to induce unexpected stock price reactions through the quantity and quality of information transmitted (Tasker 1998; Frankel et al. 1999). Unlike M&A announcements, these factors cannot be anticipated, hence, no abnormal returns prior to the conference calls are expected. For the multivariate regression on country- and industry-level we decided to impose a minimum threshold of at least 20 conference calls for the country and industry subsamples following Jaggi and Low (2000) in order to avoid a skewed representation of countries and industries. Consequently, the Spanish subsample and the industry subsamples of consumer products and services, media and entertainment, real estate, retail, consumer staples, and telecommunications are dropped in the country-level and industry-level analyses. For all other analyses, these subsamples remain included in the full sample.

4. Empirical Results

4.1. Determinants of Conference Calls

Table 3. Probit Estimation—Determinants of Merger-Related Conference Calls.

$\mathit{C}{\mathit{C}}_{\mathit{i}}={\mathit{\beta }}_0+{\mathit{\beta }}_1\mathit{T}{\mathit{V}}_{\mathit{i}}+{\mathit{\beta }}_2\mathit{D}\mathit{E}\mathit{A}\mathit{L}\mathit{R}\mathit{A}\mathit{T}\mathit{I}{\mathit{O}}_{\mathit{i}}+{\mathit{\beta }}_3\mathit{P}\mathit{U}\mathit{B}\mathit{L}\mathit{I}{\mathit{C}}_{\mathit{i}}+{\mathit{\beta }}_4\mathit{C}\mathit{A}\mathit{S}{\mathit{H}}_{\mathit{i}}+{\mathit{\beta }}_5\mathit{F}\mathit{O}\mathit{R}\mathit{E}\mathit{I}\mathit{G}{\mathit{N}}_{\mathit{i}}+{\mathit{\beta }}_6\mathit{I}\mathit{N}\mathit{D}{\mathit{R}}_{\mathit{i}}+{\mathit{\beta }}_7\mathit{L}\mathit{N}\mathit{A}\mathit{M}\mathit{K}\mathit{T}\mathit{C}\mathit{A}{\mathit{P}}_{\mathit{i}}+{\mathit{\beta }}_8\mathit{A}\mathit{B}{\mathit{M}}_{\mathit{i}}+{\mathit{\beta }}_9\mathit{A}\mathit{c}\mathit{q}\mathit{M}\mathit{u}\mathit{l}\mathit{t}\mathit{i}\mathit{p}\mathit{l}\mathit{i}\mathit{e}{\mathit{r}}_{\mathit{i}}+{\mathit{\beta }}_{10}\mathit{C}\mathit{R}\mathit{I}\mathit{S}\mathit{I}{\mathit{S}}_{\mathit{i}}+{\mathit{\beta }}_{11}\mathit{F}\mathit{I}\mathit{N}\mathit{A}\mathit{N}\mathit{C}{\mathit{E}}_{\mathit{i}}+{\mathit{\beta }}_{12}\mathit{H}\mathit{E}\mathit{A}\mathit{L}\mathit{T}{\mathit{H}}_{\mathit{i}}+{\mathit{\beta }}_{13}\mathit{H}{\mathit{T}}_{\mathit{i}}+\mathit{n}\mathit{a}\mathit{t}\mathit{i}\mathit{o}\mathit{n}\mathit{a}\mathit{l}\mathit{i}\mathit{t}\mathit{y}+{\mathit{\epsilon }}_{\mathit{i}}$
Dependant Variable: CC	Full Sample		UK		France		Germany		Switzerland		Spain
	Coeff. Estimate	Marginal Effect	Coeff. Estimate	Marginal Effect	Coeff. Estimate	Marginal Effect	Coeff. Estimate	Marginal Effect	Coeff. Estimate	Marginal Effect	Coeff. Estimate	Marginal Effect
Deal Characteristics
TV	0.109	0.015 ***	0.061	0.006 **	0.230	0.033 ***	0.170	0.030 **	0.652	0.126 **	0.242	0.029 **
DEALRATIO	0.032	0.004	0.024	0.002	0.191	0.028	0.291	0.051	2.015	0.391 **	−0.193	−0.023
PUBLIC	0.371	0.049 ***	0.385	0.040 ***	0.128	0.018	0.103	0.018	0.994	0.193 ***	0.446	0.053
CASH	0.127	0.017	0.094	0.010	0.276	0.040	0.099	0.017	0.467	0.090	0.185	0.022
FOREIGN	0.115	0.015	0.249	0.026 **	0.697	0.100 **	−0.838	−0.146 ***	0.036	0.007	0.453	0.054
INDR	0.300	0.040 ***	0.233	0.024 **	0.333	0.048	0.565	0.098 **	0.367	0.071	0.451	0.054
Other Characteristics
LNAMKTCAP	0.052	0.007 **	0.089	0.009 ***	0.080	0.012 *	0.126	0.022 **	−0.115	−0.022	−0.057	−0.007
ABM	−0.138	−0.018	−0.145	−0.015	−0.252	−0.036	0.072	0.013	0.901	0.175	0.091	0.011
AEqMultiplier	−0.003	−0.000	−0.001	−0.000	−0.094	−0.014 ***	−0.010	−0.002	−0.679	−0.132 **	−0.012	−0.001
CRISIS	−0.286	−0.038 ***	−0.220	−0.023 **	−0.473	−0.068 **	0.057	0.010	−0.421	−0.082	−0.818	−0.097 **
FINANCE	−0.018	−0.002	0.064	0.007	1.362	0.196 **	−0.385	−0.067	Omitted +	Omitted +	0.993	0.118 *
HEALTH	0.218	0.029 *	−0.199	−0.020	0.499	0.072 *	0.849	0.147 ***	−0.010	−0.002	Omitted+	Omitted+
HT	0.182	0.024	0.225	0.023	0.650	0.094 **	0.225	0.039	−1.000	−0.194	0.566	0.067
FR	0.117	0.016	Not applicable		Not applicable		Not applicable		Not applicable		Not applicable
DE	0.378	0.050 ***	Not applicable		Not applicable		Not applicable		Not applicable		Not applicable
CH	0.485	0.064 ***	Not applicable		Not applicable		Not applicable		Not applicable		Not applicable
ES	0.071	0.009	Not applicable		Not applicable		Not applicable		Not applicable		Not applicable
Constant	−2.103 ***		−2.382 ***		−2.601 ***		−2.429 ***		−0.519		−1.612 *
Observations	2518		1605		344		260		133		148
Pseudo R2	0.153		0.117		0.241		0.219		0.379		0.300

* p < 0.1, ** p < 0.05, *** p < 0.01. Estimates are derived in a probit estimation. Significance levels are based on two-tailed tests. ⁺ Variable omitted due to data collinearity. See Table 2 for the definition of variables.

summarizes the results of the probit model which largely correspond to the implications of the descriptive statistics on Table 2.

Overall, conference calls are more likely for transactions with higher transaction value and for transactions involving public targets (both statistically significant at the 1% level). However, contradicting results are obtained for industry consolidations, which in this study positively influence the likelihood of conference calls. A possible explanation is that the management uses conference calls more often to emphasize the synergies being one of the most valued merits in an intra-industry consolidation. The results further confirm the findings by Lang and Lundholm (1993) and Frankel et al. (1999) that the likelihood of conference calls increases with company size measured by market capitalization of the acquirer, and decreases with book-to-market-ratio. The evidences are strongly in support that economic crisis lessens the likelihood of conference calls. Furthermore, acquirers in the healthcare sector appear to be keener to conference calls consistent with Botosan (1997), and firms from Germany or Switzerland are more likely to conduct conference calls than UK firms as the coefficients are significantly positive for these two countries. In view of the national patterns of disclosure, the results are unexpected as common-law countries, represented by the UK in this study, are evidenced to provide a higher level of disclosure (Ball et al. 2000; Conover et al. 2008). Given the lower takeover market activity (La Porta et al. 1998; Santema et al. 2005) and an intrinsically higher degree of information asymmetry of code-law countries (La Porta et al. 1998), the result appears less surprising based on the assumption that the investors’ demand for disclosure may be particularly high around such strategic moves when the level of informedness is generally lower in these markets. Conversely, the investors’ demand for disclosure may be lower in the UK and France due to the higher takeover market activity, and higher disclosure requirements on the part of the UK. The intuition here is that the intrinsic information asymmetry in the UK and France is lower because the amount of private information revealed to the market rises with the number of transactions (Diamond and Verrecchia 1991; Santema et al. 2005). On the part of the management, the incentives for providing additional information may be even higher in code-law countries, like Germany, in the light of stricter judicial systems as observed by La Porta et al. (1998).

Country-level estimates exhibit varying patterns of conference call determinants although the overall inferences are mostly confirmed. However, the inferences need to be interpreted with caution as the sample size of the country subsamples differs substantially and neither mean nor median differences are tested between the subsamples. In all countries, the results are supportive of the economic importance to drive the likelihood of conference calls. Transaction value appears to be a key determinant of the disclosure decision, whereas the influence of deal ratio is particularly pronounced for Swiss bidders. The involvement of public targets has a significantly positive influence on the conference call likelihood, particularly in the UK and Switzerland. In contrast to the inferences from the overall sample, foreign status of the target has a significant, albeit mixed effect on the disclosure decision. Cross-border acquisitions positively influence the UK and the French acquirers’ decision to hold conference calls contrary to German acquirers, for whom foreign targets have a strongly negative effect on the likelihood of conference calls. Industry consolidation inclines to a positive effect on the disclosure choice across all countries, mostly pronounced in the UK and Germany. As for the financial status of the acquirer, a positive relation between the market capitalization and the conference call likelihood is affirmed for the UK, France and Germany, whereas the negative influence of the leverage is valid for France and Switzerland. The impact of economic crisis appears to be strongest in the UK, France and Spain. The membership to a certain industry sector does not induce a uniform effect on the likelihood of conference calls across countries. While industry effects are not at all present in the UK and Switzerland, French firms appear to be strongly influenced by industry membership. In conclusion, the most important implications of the country-wise probit estimation is that the economic importance of the transaction and the acquirer’s size have a uniformly positive influence on the likelihood of conference calls sustained across all countries.

4.2. Event Study

4.2.1. Full Sample Analysis

Table 4. Univariate Analysis of Event-Period Returns.

Event Window	CAAR Event Sample (CC = 1)		CAAR Control Sample (CC = 0)		Differences
	Mean	Median	Mean	Median	Mean	Median
[−10; 0]	0.0105	0.0054	0.0116 ***	0.0029 ***	−0.0012	0.0025
[−5; 0]	0.0060	0.0046	0.0071 ***	0.0030 ***	−0.0012	0.0015
[−4; 0]	0.0074	0.0054	0.0070 ***	0.0021 ***	0.0004	0.0033
[−3; 0]	0.0076	0.0062	0.0077 ***	0.0036 ***	−0.0001	0.0026
[−2; 0]	0.0099 **	0.0078 **	0.0081 ***	0.0031 ***	0.0018	0.0047
[−1; 0]	0.0108 ***	0.0084 **	0.0080 ***	0.0037 ***	0.0028	0.0048
[0; 0]	0.0090 **	0.0060 ***	0.0071 ***	0.0031 ***	0.0019	0.0029
[0; 1]	0.0135 ***	0.0097 ***	0.0114 ***	0.0039 ***	0.0021	0.0058 *
[0; 2]	0.0171 ***	0.0116 ***	0.0103 ***	0.0041 ***	0.0068	0.0076 ***
[0; 3]	0.0162 ***	0.0104 ***	0.0101 ***	0.0041 ***	0.0062	0.0063 **
[0; 4]	0.0160 ***	0.0097 ***	0.0111 ***	0.0045 ***	0.0049	0.0051 **
[0; 5]	0.0154 ***	0.0116 ***	0.0099 ***	0.0038 ***	0.0055	0.0078 **
[0; 6]	0.0135 **	0.0135 ***	0.0092 ***	0.0041 ***	0.0043	0.0094 **
[0; 7]	0.0112 **	0.0120 ***	0.0072 ***	0.0027 **	0.0040	0.0093 *
[0; 8]	0.0099 *	0.0107 **	0.0079 ***	0.0027 ***	0.0020	0.0079
[0; 9]	0.0098	0.0109 **	0.0083 ***	0.0030 ***	0.0015	0.0079
[0; 10]	0.0112 *	0.0120 **	0.0073 ***	0.0032 **	0.0039	0.0089 *
[−1; 1]	0.0152 ***	0.0139 ***	0.0123 ***	0.0044 ***	0.0029	0.0095
[−2; 2]	0.0179 ***	0.0144 ***	0.0112 ***	0.0037 ***	0.0067	0.0106 ***
[−3; 3]	0.0148 **	0.0128 ***	0.0106 ***	0.0042 ***	0.0042	0.0086 **
[−4; 4]	0.0144 **	0.0124 ***	0.0110 ***	0.0047 ***	0.0034	0.0077 *
[−5; 5]	0.0123 **	0.0158 ***	0.0098 ***	0.0052 ***	0.0024	0.0106
[−10; 10]	0.0126	0.0208 *	0.0118 ***	0.0020 *	0.0008	0.0188
Observations	216		2302		2518	217

* p < 0.1, ** p < 0.05, *** p < 0.01. Significance levels are based on two-tailed tests. CAAR = Cumulated average abnormal returns over the respective event period.

reports the results of the comparison of the CAAR based on S&P Europe 350. The event sample and the control sample both exhibit significant positive CAAR around M&A announcements both in terms of mean and median. The findings indicate that the abnormal returns for the event sample are rather concentrated within a short period around the event and thereafter. Whereas the control subsample demonstrates strong significant pre-announcement abnormal returns over all event windows, such evidences are not observable for the event sample until −2 days relative to the event.

Contrary to Kimbrough and Louis (2011) who find a significantly lower average CAR for conference calls compared to non-conference call announcements in terms of mean and median values, the average CAR of the present event sample inclines to be higher than the control sample, although no statistically significant differences are present. The test on median differences, however, establishes a strongly favourable investor reaction to the conference calls sample. Most notably, significant differences in the CAR are present over the event windows up to +7 days and ±4 days relative to the event date. The inconsistency in the outcomes are not surprising given the substantial differences in the sample size and the generalizing nature of univariate tests leading to limited interpretability.

Next, multivariate tests are performed for several event periods control for other determinants of the market reaction using the following model:

$$\begin{array}{ll}CA{R}_i={\beta }_0+{\beta }_{1}C{C}_i& +{\beta }_{2}T{V}_i+{\beta }_{3}DEALRATI{O}_i+{\beta }_{4}PUBLI{C}_i+{\beta }_{5}CAS{H}_i\\ & +{\beta }_{6}FOREIG{N}_i+{\beta }_{7}IND{R}_i+{\beta }_{8}LNAMKTCA{P}_i+{\beta }_{9}INVMILL{S}_i\\ & +{\epsilon }_i\end{array}$$

(2)

The variable CC serves to reflect the impact of conference calls on the investor reaction. Transaction value (TV) and the ratio of the transaction value to the acquirer’s market capitalization (DEALRATIO) are included in expectation that investors will rather react in an unfavourable manner to transactions involving larger targets due to increased post-merger operational and financial risk (Scanlon et al. 1989; Fuller et al. 2002). Target’s public status (PUBLIC) is included based on prior evidence that acquisition of public targets is associated with negative bidder’s announcement return because of the higher share liquidity and hence, increased bid premium offered for public targets (Fuller et al. 2002; Conn et al. 2005). The model contains a variable to control for the method of payment (CASH) because of prior outcome indicating on bidders’ underperformance for stock-based considerations and reversed tendency for cash considerations which is attributable to the investors’ general perception of stock-to-stock merger to be motivated by overvaluation of the bidder’s share (Travlos 1987; Chang 1998). Target’s origin (FOREIGN) is included based on the observations that cross-border acquisitions underperform domestic transactions due to higher informational opaqueness and operational riskiness of such international expansion (Conn et al. 2005; Moeller and Schlingemann 2005). Industry relatedness of bidder and target (INDR) serves to control for the notion that diversifying acquisitions are associated with lower return to the bidder’s shareholders (Scanlon et al. 1989). Furthermore, a measure for the firm size (LNMKTCAP) is included because Moeller et al. (2004) find that larger bidder companies experience smaller announcement return irrespective of the form of financing and private status of the target. At last, the inverse Mills-ratio (INVMILLS) is employed to correct the self-selection bias associated with the determinants of management’s decision for merger-related conference calls. Therefore, the inverse Mills-ratio is computed according to Heckman (1979) based on the probit-regression in Table 3. Excluded are yearly, country or industry fixed effects following the argumentation of Kimbrough and Louis (2011) that these factors rather act as determinants of the management’s decision to conduct conference calls but do not directly affect the announcement returns.

Table 5. Multivariate Analysis of Event Period Returns—Cross-Country Analysis.

$\mathit{C}\mathit{A}{\mathit{R}}_{\mathit{i}}={\mathit{\beta }}_0+{\mathit{\beta }}_1\mathit{C}{\mathit{C}}_{\mathit{i}}+{\mathit{\beta }}_2\mathit{T}{\mathit{V}}_{\mathit{i}}+{\mathit{\beta }}_3\mathit{D}\mathit{E}\mathit{A}\mathit{L}\mathit{R}\mathit{A}\mathit{T}\mathit{I}{\mathit{O}}_{\mathit{i}}+{\mathit{\beta }}_4\mathit{P}\mathit{U}\mathit{B}\mathit{L}\mathit{I}{\mathit{C}}_{\mathit{i}}+{\mathit{\beta }}_5\mathit{C}\mathit{A}\mathit{S}{\mathit{H}}_{\mathit{i}}+{\mathit{\beta }}_6\mathit{F}\mathit{O}\mathit{R}\mathit{E}\mathit{I}\mathit{G}{\mathit{N}}_{\mathit{i}}+{\mathit{\beta }}_7\mathit{I}\mathit{N}\mathit{D}{\mathit{R}}_{\mathit{i}}+{\mathit{\beta }}_8\mathit{L}\mathit{N}\mathit{A}\mathit{M}\mathit{K}\mathit{T}\mathit{C}\mathit{A}{\mathit{P}}_{\mathit{i}}+{\mathit{\beta }}_9\mathit{I}\mathit{N}\mathit{V}\mathit{M}\mathit{I}\mathit{L}\mathit{L}{\mathit{S}}_{\mathit{i}}+{\mathit{\epsilon }}_{\mathit{i}}$
Variable	[0; 0]	[0; 1]	[0; 2]	Variance Influenced Factors
	Coefficient Estimate	Coefficient Estimate	Coefficient Estimate
CC	0.008 **	0.012 **	0.017 ***	1.11
	(2.08)	(2.44)	(3.48)
TV	−0.001 ***	−0.002 **	−0.002 **	1.14
	(−3.24)	(−2.28)	(−2.30)
DEALRATIO	−0.000	0.001	0.000	1.05
	(−0.37)	(0.50)	(0.24)
PUBLIC	−0.006 **	−0.001	−0.005	1.09
	(−2.07)	(−0.22)	(−0.76)
CASH	0.001	−0.001	0.001	1.03
	(0.46)	(−0.51)	(0.22)
FOREIGN	0.004 *	0.010 ***	0.011 **	1.13
	(1.65)	(2.75)	(2.53)
INDR	0.004 *	−0.000	−0.003	1.05
	(1.75)	(−0.04)	(−0.62)
LNAMKTCAP	−0.003 ***	−0.005 ***	−0.006 ***	1.43
	(−3.89)	(−4.58)	(−4.00)
INVMILLS	−0.002	−0.002	−0.003	1.33
	(−0.93)	(−0.97)	(−1.45)
Constant	0.025 ***	0.048 ***	0.052 ***
	(3.04)	(3.55)	(3.39)
Observations	2518	2518	2518
Adjusted R2	0.018	0.027	0.024

* p < 0.1, ** p < 0.05, *** p < 0.01. p-values are two-tailed. CAR = Cumulated abnormal return over the event period; LNAMKTCAP = Log of the acquirer’s market capitalization four weeks prior to the transaction announcement; INVMILLS = Inverse Mills-ratio according to Heckman (1979) based on the probit estimation of Equation (1). See Table 2 for the definition of other variables.

contains the outcome of the multivariate regression of the CAR on the event date and for the periods from the event date up to +1 and +2 days relative to the event date. Most importantly, conference calls account for a strongly significant positive abnormal return across all event windows indicating that conference calls transmit material information and prompt investors to trade on the released information. Bidders providing conference calls concurrent with merger announcement yield up to 1.7% higher abnormal returns than other firms relying on press releases only. Compared to Kimbrough and Louis (2011) who detect 6.5% additional abnormal return over a three-day period for merger-related conference calls in the US, the investor response in the European market appears somewhat muted. A possible explanation may be that disclosure generally induces larger stock market responses in the US market attributable to the enhanced liquidity in the US market compared to the European market (Frost and Pownall 1994). Alternatively, the large stock market response may be also credited to the investors’ confidence in firms listed on the US stock market and thereby complying with higher disclosure requirements (Frost and Pownall 1994; Roosenboom and Dijk 2009). However, as a direct comparison of the magnitude of abnormal stock performance between the results of Kimbrough and Louis (2011) and the present results lacks reliability due to differences in the estimation model and sampling period, the main inference to be drawn here is that conference calls accompanying European merger announcement demonstrate a significant positive market response comparable to prior research.

The coefficients and the signs of other determinants are largely consistent with expectations. The evidences are in support of a negative relation between bidder return and transaction value, public targets and size of the bidder firm. Additionally, cash offer and non-diversifying acquisitions with higher industry relatedness tend to exhibit favourable investor response, although not significantly different from zero for all event windows.

One major exception is the behaviour of stock response to foreign targets. While the prevalent notion in current M&A literature is that foreign targets prompt a negative abnormal return due to the naturally higher information asymmetry inherent in cross-border acquisitions, the present results indicate on a significantly higher abnormal return. However, the findings are insofar not surprising as several other studies confirm positive bidder return for acquisition of foreign targets attributable to potential synergies arising from the expansion of the bidding firm’s multinational network (Doukas and Travlos 1988; Morck and Yeung 1991).

Another notable result in relation to the control variables is that the inverse Mills-ratio does not exhibit any significant influence implying that the results are unlikely to be driven by self-selection bias. Overall, the behaviour of the control variables is in line with prior literature and lends validity to the present model.

4.2.2. Country-Level Analysis

Table 6. Multivariate Analysis of Event Period Returns—Country Subsamples.

Variable	Overall	UK	France	Germany	Switzerland
	Coefficient Estimate	Coefficient Estimate	Coefficient Estimate	Coefficient Estimate	Coefficient Estimate
CC	0.017 ***	0.022 ***	0.029 ***	0.014	0.002
	(3.48)	(2.76)	(2.81)	(1.32)	(0.17)
TV	−0.002 **	−0.002	−0.002 *	−0.005	−0.001
	(−2.30)	(−1.59)	(−1.84)	(−1.55)	(−0.77)
DEALRATIO	0.000	0.000	0.022 ***	0.004	0.011
	(0.24)	(0.03)	(3.13)	(0.46)	(0.30)
PUBLIC	−0.005	0.004	−0.012	−0.015 *	−0.021
	(−0.76)	(0.33)	(−1.57)	(−1.92)	(−1.64)
CASH	0.001	0.004	0.000	−0.004	−0.001
	(0.22)	(1.09)	(0.07)	(−0.59)	(−0.06)
FOREIGN	0.011 **	0.013 *	0.002	0.012	−0.004
	(2.53)	(1.95)	(0.41)	(1.51)	(−0.34)
INDR	−0.003	−0.006	0.003	0.001	0.000
	(−0.62)	(−0.82)	(0.50)	(0.08)	(0.00)
LNAMKTCAP	−0.006 ***	−0.008 ***	−0.003 *	−0.003 *	−0.005
	(−4.00)	(−3.35)	(−1.67)	(−1.82)	(−1.47)
INVMILLS	−0.003	−0.004	−0.006	−0.002 **	−0.003
	(−1.45)	(−1.37)	(−0.39)	(−2.43)	(−0.11)
Constant	0.052 ***	0.063 ***	0.037	0.029 **	0.066
	(3.39)	(2.70)	(0.83)	(2.24)	(1.00)
Observations	2518	1605	344	260	156
Adjusted R2	0.022	0.020	0.078	0.034	0.041

* p < 0.1, ** p < 0.05, *** p < 0.01. p-values are two-tailed. CAR = Cumulated abnormal return over a three-day period spanning the event date to day +2; LNAMKTCAP = Log of the acquirer’s market capitalization four weeks prior to the transaction announcement; INVMILLS = Inverse Mills-ratio according to Heckman (1979) based on the probit estimation of Equation (1). See Table 2 for the definition of other variables.

provides the results of the multivariate regression for country-subsamples over a three-day event window beginning on to the event day. Most strikingly, abnormal returns induced by conference calls are significant positive only in the UK and France whereas such abnormal stock performance is not evident in all other countries. In the context of the prior probit estimation, the results are even more puzzling since Germany and Switzerland do not earn any significantly positive excess return credited to conference calls although these countries seem to have a higher commitment in providing conference calls.

However, there exist diverse possible sources of varying investor responses between the countries. The simplest interpretation following the assumptions of Kimbrough and Louis (2011) and Frankel et al. (1999) would be that merger-related conference calls in the UK and France tend to transmit a higher informational content than the respective press release, inducing investors and analyst to trade on the newly released information. Conversely, firms in Germany and Switzerland may rather recapitulate public information, thus participating investors and analyst do not obtain superior information to gain excess returns. However, this hypothesis needs further supporting evidence which we unfortunately are unable to provide given our data restrictions. Following the observations by Skinner (1994) and Hutton et al. (2003) that stock return is majorly dependent on whether the disclosure contains bad or good news, another possible implication is that conference calls in the UK and France rather involve beneficial news whereas firms in Germany and Switzerland disclose a mix of good and bad news.

Alternative interpretations are offered by the international disclosure literature in view of the institutional, regulatory and cultural environment. Beginning with the differentiation of common-law and code-law countries, common-law countries—typically represented by the UK or the US—are characterized by higher investor protection, rigorous enforcement of these corporate laws, more liquid stock market, dispersed ownership structure, and higher requirement of disclosure (La Porta et al. 1998; Ball et al. 2000; Bailey et al. 2006; Conover et al. 2008). Consequently, the larger stock market response in the UK may be attributed to the higher liquidity of its stock market and investors’ overall confidence in firms complying with higher standard of disclosure and exercising shareholder-oriented disclosure practices in routine.

However, a mere classification of countries in common-law and code-law countries requires further elaboration regarding the strongly significant positive market response in France compared to other code-law countries. France code-law countries are typically characterized by poorer legal protection, low public enforcement and block building of large shareholders according to La Porta et al. (1998). French companies evidently exercise less timely disclosure and provide a lower amount of strategic information (Santema et al. 2005; Conover et al. 2008). Furthermore, France displays a higher level of uncertainty according to Hofstede (2001, p. 500), which should further information asymmetry following the argument of Gray (1988) that societies with highly uncertainty avoiding cultures tend to be more secretive. Under the assumption that the intrinsic information asymmetry is particularly high in France as neither the legal system nor the ownership structure require higher level of public disclosure and cultural sentiment even fosters non-disclosure, the French market response is anything but unexpected in the light of investors’ enthusiasm towards additional information in an environment of high information asymmetry. Under the assumption that French investors generally tend to higher uncertainty avoidance according to Hofstede (2001, p. 500) than any other country in this study, favourable response may be even larger. The response is then to be regarded as a “reward for conducting conference calls” following Kimbrough and Louis (2011, p. 649). Conversely, the greater tolerance for ambiguity displayed by German and Swiss investors may contribute to a lower degree of enthusiasm and hence lead to an insignificant reaction as observed in the utilized sample. However, as an alternative explanation for our findings it is important to recognize that the relation between acquirer size and target size is significant only in France indicating size dependent information asymmetry divergences.

In conclusion, stock returns associated with conference calls appear to be highly influenced by the country-specific investor reaction to disclosure as strong divergences of stock returns can be observed between countries. The present study suggests various approaches to explain this phenomenon by adopting different strands of economic theories about information asymmetry, voluntary disclosure and international disclosure patterns. The validation of these approaches remains a task for future research.

4.2.3. Industry-Level Analysis

At last,

Table 7. Multivariate Analysis of Event Period Returns—Industry Subsamples.

Variables	Overall	ENERGY	FINANCE	HEALTH	HT	IND	MATERLS
	Coefficient Estimate	Coefficient Estimate	Coefficient Estimate	Coefficient Estimate	Coefficient Estimate	Coefficient Estimate	Coefficient Estimate
CC	0.017 ***	0.003	0.016	−0.006	0.036 **	0.023 *	0.041 *
	(3.48)	(0.22)	(1.27)	(−0.47)	(2.19)	(1.92)	(1.97)
TV	−0.002 **	−0.001	−0.004 ***	−0.002 **	−0.117 *	−0.010	0.001
	(−2.30)	(−0.49)	(−2.61)	(−2.43)	(−1.67)	(−0.96)	(0.78)
DEALRATIO	0.000	−0.010	−0.005	−0.025	0.272 **	−0.000	0.001
	(0.24)	(−0.83)	(−1.32)	(−1.14)	(2.08)	(−0.42)	(0.54)
PUBLIC	−0.005	0.002	−0.015 *	−0.027 *	−0.018	−0.004	0.003
	(−0.76)	(0.14)	(−1.96)	(−1.83)	(−0.94)	(−0.35)	(0.15)
CASH	0.001	0.003	−0.005	0.013	0.019 *	−0.004	−0.009
	(0.22)	(0.31)	(−0.81)	(1.28)	(1.94)	(−0.66)	(−0.39)
FOREIGN	0.011 **	0.014	0.010	0.034 **	0.003	−0.006	0.045
	(2.53)	(1.38)	(1.10)	(2.42)	(0.21)	(−0.96)	(0.91)
INDR	−0.003	−0.005	0.012	−0.000	0.007	0.002	−0.059
	(−0.62)	(−0.49)	(1.53)	(−0.02)	(0.52)	(0.23)	(−1.06)
LNAMKTCAP	−0.006 ***	−0.006 **	−0.006 ***	−0.007 ***	0.012	−0.008 *	−0.018
	(−4.00)	(−2.08)	(−2.68)	(−2.86)	(1.52)	(−1.92)	(−1.16)
INVMILLS	−0.003	−0.007	−0.006 **	−0.050 **	0.033 ***	−0.022	0.002
	(−1.45)	(−0.40)	(−2.00)	(−2.02)	(4.05)	(−1.53)	(0.10)
Constant	0.052 ***	0.069 *	0.054 ***	0.124 *	−0.159 ***	0.111 **	0.149
	(3.39)	(1.68)	(2.66)	(1.82)	(−2.97)	(2.02)	(0.90)
Observations	2518	198	346	187	310	350	199
Adjusted R2	0.022	0.015	0.085	0.058	0.365	0.042	0.035

* p < 0.1 ** p < 0.05 *** p < 0.01. p-values are two-tailed. CAR = Cumulated abnormal return over a three-day period spanning the event date to day +2; LNAMKTCAP = Log of the acquirer’s market capitalization four weeks prior to the transaction announcement; INVMILLS = Inverse Mills-ratio according to Heckman (1979) based on the probit estimation of Equation (1). See Table 2 for the definition of other variables.

presents the results of multivariate regression for the industry subsamples. Similar to the country-by-country analysis, the industry subsamples also show a varying degree of investor response to merger-related conference calls, although the significance is not as high as observed in the country-samples. Most strikingly, significant abnormal returns are present only in industries with a high R&D-intensity: High technology, industrials, and materials. These industries are typically characterized by an inherently higher information asymmetry most saliently due to the uniqueness of their business and intangibility of their R&D activities (Aboody and Lev 2000). Hence, the results are consistent with expectations that investors reward firms for their commitment in voluntary disclosure, in particular when the information asymmetry is high (Botosan 1997; Leone et al. 2007). Alternatively, these companies usually have higher growth potentials compared to traditional industries and thus, increased synergy potentials, which may also contribute to higher abnormal returns upon the synergy disclosure consistent with Fraunhoffer and Schiereck (2012).

The argument of R&D-intensive firms, however, does not hold for the healthcare industry in view of the under-reaction despite the presumably high R&D intensity. A possible explanation may be that healthcare companies are perceived to exercise a forthcoming disclosure policy as suggested by Botosan (1997) and evidenced in the prior probit model of present study. Their policy may have an attenuating influence on the perceived information asymmetry, thereby mitigating an urgent demand for disclosure and lowering the reward for disclosure.2

5. Conclusions

The present study investigates bidders’ conference call practices and investors’ response around M&A announcements. This period is typically characterized by an inherently higher information asymmetry and increased demand of investor for information due to the operational and financial risk associated with such substantial strategic transition. The geographical focus is set on five Western European countries, i.e., the UK, France, Germany, Switzerland and Spain. The analysis is performed both from cross-national and country-specific perspectives to offer a comprehensive understanding of institutional, legal and cultural circumstances as potential sources of varying disclosure practices and investor response across countries. Investor reaction has been further examined for industry differences.

Overall, bidders are more likely to conduct conference calls with increasing transaction value, for transaction with public targets and non-diversifying transactions. Further, the decision for voluntary disclosure is positively influenced by higher bidder size and bidder’s industry membership to the healthcare industry, but appears to be discouraged in times of economic crises. Additionally, German and Swiss firms are more likely to hold conference calls. The country-by-country analysis presents a slightly varying weight of disclosure factors partly attributable to the country-specific disclosure environment. However, the role of transaction value and bidder’s size as key determinant remains valid across all countries.

After controlling the self-selection bias and other determinants of stock returns around M&A announcements, there is support that firms with merger-related conference calls yield a higher abnormal return than firms merely publishing a press release. Favourable investor reaction is, however, only present in the UK and French subsamples and in the subsamples of industries with high R&D-intensity, with national divergence being more pronounced than industry differences.

The results open an avenue for future research. As many recent studies of conference calls, e.g., have found that the tone of the call is informative, this could be an additional explanatory factor of capital market reactions. The inclusion of tonality could go in line with an extension of the sample period to include more recent transactions and to increase the sample size.