Keywords

1 Introduction

A two-tier (dual, Continental, Rhine) corporate governance system prevails in Poland, Germany and other European countries; supervisory and control functions are separate from decision making functions. The former belong to the supervisory board whereas the latter belong to the management board. In a two-tier system nearly all authority in the company belongs to the management board which is controlled by the supervisory board. The supervisory board is a passive body of the company; it is mainly responsible for the appointment of the management board. Therefore, it is the CEO who is responsible for the development and implementation of the company strategy, its structure and results. In most organizations the CEO is the person who identifies and implements directions of business development. Therefore, it is often emphasized in literature that leadership is an important element of corporate governance which impacts on the level of profits (Davidson et al. 2006). In public companies, CEO changes can result in different perception of the company’s development opportunities by stock investors, which, particularly in the short term, can result in a sudden change of company valuation and in the long term in limited access to capital and thus smaller potential of the company to generate profits.

In light of the more and more frequent rotation of CEO positions and since their term of office is becoming shorter and shorter, CEO succession has been studied both theoretically and empirically. However, nearly all empirical analyses relate to developed capital markets. The only study on the impact of the resignation of a management board member on market prices of stock quoted companies in the Polish and other Central European markets was conducted by Gurgul and Majdosz (2007). However, the research period was relatively short—from January 2000 to June 2005 and the study comprised only 60 successions. The study did not take into account the determinants of the market response. In this paper the research period is much longer, i.e. from 2000 to 2015, and comprises 1904 successions made in companies quoted on the Warsaw Stock Exchange. We hope that our study, which focuses on the Polish market, fills the long existing gap in research.

2 Literature Review

The beginnings of succession studies date back to the 1960s (Grusky 1960). The work of Hambrick and Mason (1984) started research into key managerial teams. In the mid 1990s Kesner and Sebora (1994) noticed that research into succession significantly increased when it began to be perceived as a key condition for the development of organizations during their life cycle. In their opinion, the following were the main contributors to that trend:

  • the work of organizations often reflects the attitudes of their management board members,

  • CEOs make final decisions,

  • CEOs most often represent companies in contacts with shareholders and other stakeholders,

  • successful companies plan succession.

Succession becomes very important in the global economy and with the growing sophistication of business structures, which is reflected in the number and diversity of academic papers which address this issue. In view of its complexity, this problem is studied by academics interested in human resources management, finance and accounting. Some theories have also been formulated.

From the point of view of organization theory, the succession of a key person in the organization can be explained by three theories—the common sense theory, the vicious circle theory and the ritual scapegoating theory. The common sense theory assumes that the results of an organization should improve whenever an ineffective CEO is replaced (Grusky 1963; Helmich 1975; Allen et al. 1979; Dalton and Kesner 1985; Kesner and Sebora 1994). In contrast, the vicious circle theory assumes that as a result of succession the situation of the enterprise will worsen, because both business relations and relations within the enterprise have been upset and relations with customers, suppliers and employees can be impaired (Grusky 1960; Beatty and Zajac 1987; Ishak and Latif 2013). According to the scapegoating theory a CEO is dismissed even when not responsible for poor results (Gamson and Scotch 1964; Boeker 1992; Khanna and Poulsen 1995). Consequently, appointment of a new CEO does not mean that the economic results of the enterprise will improve.

In finance, succession and related processes are considered to be significant market signals. The so-called signaling effect theory, coined by Spence (1973), assumes that the decision of the management body to choose a new CEO, particularly when the CEO change has been motivated by poor financial results, is a signal sent to the market about future results (Wiersema and Moliterno 2006). According to the signaling effect theory, changes in company value resulting from a CEO change can be ascribed to the informational or real effect, or a combination of the two. If the CEO change was unexpected, the signalled message can suggest that the results of the company will be worse than expected. In this situation investors will negatively respond to the informational component of the message. If the investors expect an improvement in results, a positive real effect from the message will be generated. Consequences in the form of changing share price, and consequently lower or higher company valuation, may be different and will depend on the dominant trend (Adams and Mansi 2009).

Based on a detailed analysis of 227 academic articles (of which 181 were empirical and 46 were conceptual), Cragun et al. (2016) created a typology of research on succession. They started their literature study with an analysis of Finkelstein et al. (2009), who conducted their study in the following dimensions: (1) Will there be a succession?, (2) How will the succession be conducted (as a result of which process)?, (3) Who will become CEO?, and (4) What will be the consequences of succession?

Cragun et al. (2016) expanded the four dimensions and added second- and third-level problems, thus arriving at a multi-layer structure with 32 types of studies. Each of the empirical articles they analyzed was classified into at least one of the third-level components, and a few articles were classified into more than five types (multi-aspect publications). The articles authored by Berry et al. (2006) and Hillier et al. (2005) were considered to be most comprehensive.

When we look at the initial division into four dimensions, we notice that the authors had most often studied the consequences of succession—156 articles were classified in this group. On the other hand, only 46 articles addressed the features and qualifications of the CEO successor. It should also be pointed out that succession from the point of view of the effective functioning of the entire company and circumstances in which succession takes place, were most frequently studied (Table 1).

Table 1 Statistical summary of the typology proposed by Cragun et al. (2016)
Full size table

This article is the first in a series on CEO successions in the Polish capital market. We focused on identification of the greatest number of CEO appointments in 2000–2015. In addition to the evaluation of market responses, we also identified factors which affected the response to actual succession, so as to be able to try to answer the question whether the CEO change affected the long term financial results of the enterprise. The entire publication series, in accordance with the typology proposed by Finkelstein et al. (2009) and Cragun et al. (2016), will be part of the studies which analyze succession from the point of view of the consequences of the CEO change and which also take into account such aspects as the circumstances of the change, CEO profile and economic situation of the company before succession.

3 Sample Selection

A database of CEO successions comprising companies quoted on the Warsaw Stock Exchange was developed by the authors from scratch. This decision was made after an analysis of available lists of changes in the governance structure of Polish public companies offered by commercial providers. However, all of them proved to be incomplete, both with respect to the event itself (many CEO appointments were not included) and with respect to its description (no specific characteristics of succession). Consequently, it was considered that an adoption of such lists as a basic research material would impact negatively upon the entire inference process. Thus, it became important to determine the right search, identification and selection procedures to obtain a high quality database of CEO successions. The process was formulated on the basis of our experience described in Byrka-Kita et al. (2017).

Our study only comprises companies whose shares in 2000–2015 were traded on the regulated public capital market. This was done on purpose—we wanted to differentiate between companies with different information requirements. Polish public companies, which operate under the Polish legal regime, must comply with the provisions of Article 56(1) of the Act on Public Offering (Dz.U. 184/2005, item 1539 as amended), according to which the issuer […] is obliged […] to simultaneously provide the following information to the Financial Supervision Authority, to the company operating the regulated market and to the public:

  1. 1.

    confidential information within the meaning of Art. 154 of the Act on Trading in Financial Instruments, hereinafter referred to as “confidential information” Footnote 1 ;

  2. 2.

    current and periodic information.

A precise catalogue of events, which, if they take place requires generation of current information by the company, is given in the implementing acts of the minister responsible for supervision over the capital market. The legal grounds for the mandatory publication of information about changes in management board structures, including in the CEO position, require companies to immediately disclose this situation. Investors, on the other hand, must verify the valuation of financial assets in their portfolios. These regulations apply mainly to companies traded on the unregulated market which makes them less transparent, and was the main factor on the basis of which they were excluded from the studyFootnote 2.

CEO successions were identified on the basis of newswires published by the companies. Preparation of the CEO succession database was divided into four stages.

First a newswire search engine available on the website of GPWInfoStrefa was used.Footnote 3 The newswires were a secondary informational materials compared to the information provided by the companies in the form of current and periodic reports. We searched the database using such key words as “powołano” (was appointed), “powołanie” (appointment) or “prezes” (CEO). At the study stage no equally efficient tool to search “raw” company reports had been found. We identified over 10,000 newswires,Footnote 4 and verified them to see if they could be used in our study. Some of the hits were rejected on the basis of the title because they clearly referred to an event other than CEO appointment. We also rejected some results when the newswire informed about an appointment in a company operating in the unregulated market. In this way we obtained a so-called base list of 1324 appointments in the period January 2005 to June 2015.

Despite the fact that public companies are obliged to publish such information in the form of current reports, the Polish Securities and Exchange Commission (presently the Financial Supervision Authority) launched their Electronic Information Transfer System (EITS) as late as the end of 2004. Its resources were made available on the GPWInfoStrefa website. Previously, communications from public companies were provided to the EMITENT system, but they were not arranged in the way they are arranged in the EITS system. Furthermore, they could be identified only by browsing the publications of company reports on the websites of Bankier.pl and Money.pl services. At the same time current reports for 2000–2004 prepared by companies, which were no longer quoted in 2016, were unavailable in the databases of GPWInfoStrefy, Bankier.pl and Money.pl, Polish Press Agency and the Rzeczpospolita daily. The missing reports were obtained directly from the Financial Supervision Authority. In this way the number of newswires was increased to 1511 and the analysis comprised the period starting 1 January 2000.

At the next stage of the selection process the completeness of the base list was verified. The risk that a significant part of newswires were omitted resulted from the preliminary estimation of the potential number of successions. If we assumed that each year a CEO was appointed in half of the companies operating in the regulated market, the complete set was estimated to include approx. 2700 newswires. In this situation, the base list represented 55% of all the possible appointments, which was found to be insufficient to reliably generalize the results. In order to identify the omitted events, we used the resources of Notoria Serwis On-Line, to generate a list of newswires qualified to the “change in governing bodies” group. This list was compared with the base list, which was then supplemented with the missing events. In total, at the second stage of the selection procedure, the number of observations was increased by 393 appointments, which produced a database of 1904 events. This number suggests that on average CEO appointments occurred in more than every third but in less than half of the companies traded on the Polish capital market (Fig. 1).

Fig. 1
figure 1

Total number of CEO appointments on the Polish capital market vs. an estimated number of appointments if changes were made in every fourth or every third (etc.) company

Full size image

Subsequently, we made a detailed selection of the newswires and supplemented individual observations with additional characteristics of the companies and their environment, succession process and CEO appointments (Table 2). At the fourth stage of the procedure we included the period in which companies were quoted on the stock exchange. The companies whose shares were traded for a relatively short time and therefore there was insufficient information necessary to calculate additional rates of return and volumes, were rejected. The final number of companies is not fixed; it depends on the research aspect. In subsequent publications we redefine the details of the final stage of selection as it depends on the research hypotheses.

Table 2 Variable definitions
Full size table

The results described in the subsequent part of the article relate to the sample generated in the third stage of the procedure.

4 Results

The database comprising 1904 CEO appointments in companies traded on the Polish capital market is not a sample representing the actual successions. Detailed verification revealed that 2.89% of the events (n = 55) should have been rejected at the earlier selection stages or that there were no detailed data about them. 112 (5.88%) of the newswires informed about appointment of acting CEOs, which in fact means that the appointment was made temporarily and the company continued its CEO selection process and searched for the right candidate or waited for the approval of relevant governing bodies or determination of the management board composition. On the basis of the data it is also possible to say that the appointment of an acting CEO is not synonymous with the appointment of the same person for the CEO position for the next term of office. Therefore, there are valid doubts to consider such an event as a CEO succession. In principle, such a situation is temporary and requires the successor to exercise his/her statutory duties with caution and prudence. Such events can be classified as quasi successions (Fig. 2).

Fig. 2
figure 2

Structure of CEO appointment database after the third selection of newswires

Full size image

CEO re-appointments (850—44.64%), i.e. appointments of the same person for a new term of office, are a significant part of the sample. Unlike in the case of the appointment of an acting CEO, re-appointment for the next term of office should not be connected with succession because there is no change in the key position in the company and the necessary condition for succession is not met.

Undoubtedly, re-appointment is an important event from the point of view of company stakeholders since it confirms confidence in supervisory bodies and owners in the work performed by the CEO. Furthermore, on the basis of the data for 2013–2015 we can say that the importance of re-appointments is growing, which definitely stabilizes corporate governance in companies quoted on the Warsaw Stock Exchange. With some simplification we can say that the Polish capital market is moving away from the trends of the so-called one-time CEOs and if somebody is appointed CEO in a public company, it is very likely that that person will be approved by the supervisory board for the next term of office. However, this conclusion should not be treated as fully binding since it must be verified empirically (which requires a separate study); preliminary analysis does suggest existence of this regularity.

However, from the point of view of the research problem, re-appointments are not actual events, which is the case with 46.59% of the observations (887) of a new CEO appointment. Therefore, the subsequent analysis will pertain to the group of observations that are actual successions (Table 3).

Table 3 Descriptive statistics of the analyzed variables (only new CEO appointments)
Full size table

Most CEO successions were identified in 2008 (81) and 2012 (85), whereas in the entire period from 2000 to the end of 2004 there were only 130 successions in the sample. This lack of regularity in the distribution in the entire period under analysis is understandable if two objective conditions are taken into account. On the one hand, we must take into account the number of companies whose shares were quoted on the regulated market of the Warsaw Stock Exchange in individual years, and on the other we must take into account the greater availability of current reports of public companies after 2005, i.e. when the ESPI system was implemented (Fig. 3).

Fig. 3
figure 3

New CEO appointments and re-appointments in the Polish capital market in 2000–2015

Full size image

A more difficult access to current reports can be observed if the number of appointments is divided by the number of public companies on the regulated market at the end of each year. From 2005 onwards, the result of the division has been around 20% whereas in previous years it was 12% on average. It is natural that 25 appointments were selected for 2001 and 2004 since at that time only nearly 200 companies were quoted on the Polish regulated capital market. It should be expected that in each of those years about 20 appointments were omitted (0.2 × 230 − 25). In total, about 90 appointments could have been omitted, i.e. 40% of the likely number of all events in the period. In the opinion of the present authors, this result is satisfactory and the research sample can be considered credible and well prepared.

During the analysis of the basic statistics, which characterize the research sample, it is important to notice that a typical CEO is younger than 45 years at the time of appointment. Most often it is a man (92.66%) with general education (77.71%), obtained in Polish educational institutions (73.51%). Furthermore, very often it is a person with experience in a given sector (90.52%) who had acted as a CEO or management board member before (85.21%). Special attention should be paid to the structure of appointments with respect to the attachment of the successor to the company. It was determined that in just under half of the events (44.58%) an insider was appointed, i.e. a person who was attached to the company in the earlier period. In just one year, i.e. in 2006, the number of appointed outsiders was lower than the number of appointed insiders. Thus, despite the fact that the difference between these groups in the entire sample is only 10% points, the preference for outsiders is a regularity stable over time, also in recent years in which re-appointments are more frequent. Appointment of insiders in turn does not mean that the succession is planned (Fig. 4).

Fig. 4
figure 4

Number of insider and outsider CEO successions in 2000–2015 on the Polish capital market

Full size image

The analysis of CEO succession processes was made on the basis of current reports of companies and other categories of generally available information. This means that the predictability of the change in the CEO position was based without any access to internal documents, decisions or notes made by company authorities. As a result of the adopted criteria, only every tenth (11.41%) succession was considered to be planned. In such cases it was the supervisory boards that announced the person who will become the new CEO, or the CEO was changed because the existing CEO reached retirement age. The officially communicated reasons for the change included resignation (52.49%) and removal (35.37%).

The official number of management board members at the time of succession is another issue worthy of analysis. In half of the cases (57.01%) the management board was composed of not more than two members, and in one case the newly appointed CEO was the only management board member. This is probably due to the fact that although top management teams with many members do not guarantee higher effectiveness of the company, they do allow for greater differentiation of the responsibilities between management board members. Furthermore, if we assume that the management board is a group of persons with key importance for the organization development (Kesner and Sebora 1994), the smaller it is the more difficult it is to plan succession as a continuous process since there is a small set of CEO candidates. Theoretically, management boards with just a few members are cheaper (fewer salaries have to be paid), provided that management support services are not outsourced to external advisory and consulting companies. The biggest management board consisted of 12 persons, which is similar to the solution adopted by English boards of directors (Fig. 5).

Fig. 5
figure 5

Sample selection procedure

Full size image

Women, although they are rarely appointed as CEOs on the Polish capital market, are elected as members of management boards. In total in one fourth of the cases (24.97%) where a CEO was appointed, the management board consisted of at least one woman. The following conclusions were drawn on the basis of additional calculations:

  • no relation between the presence of women on management boards and the number of management board members was confirmed (Spearman’s linear correlation matrix was applied); statistically, these two phenomena are independent;

  • in the majority of cases (76.93%) the appointment of a woman as CEO meant that all the other persons were men;

  • there is no statistical difference between the average (median) age at which a woman or a man is appointed CEO. In both groups it is usually 44–45 years (t-test and U Mann-Whitney two-sample test);

  • at a significance level α = 0.1 the hypothesis about no relation between the sex of a new CEO and former attachment to the company must be rejected (Table 4). Based on relative quantities (percentage of subgroups) we can notice that women are more often insiders while men are more often outsiders (the differences between the groups were verified by the chi-square and Tau Kendall tests).

Table 4 The structure of new CEO appointments according to gender and attachment of the successor to the company
Full size table

The characteristics of successors, circumstances of changes or distribution of CEO appointments over time are not the only aspects worthy of consideration in studies of public companies. The problem of dates appeared at the very beginning of the data collection and selection process—it was necessary to determine the actual date of succession. Three dates were possible—(1) the day on which information about CEO appointment was published, (2) the day on which the resolution on CEO appointment was adopted by the supervisory board and (3) the day on which the decision on CEO appointment became effective. If all these three events take place on the same day, there are no methodological doubts. However, any movement in time between these events is a reason for discussion about their hierarchy (classification).

In most cases information about succession was published on the same day on which the supervisory board made its decision (62.82%) or when the decision became effective (48.19%). However, very frequently information was made available in the ESPI system (previously EMITENT system) on the day following the day on which a relevant resolution was adopted (32.66%) or on the day before it became effective (23.76%). There were 99.43% of cases when information about the appointment was announced either 1 week before or 1 week after the adoption of a relevant resolution and 85.29% of cases when information about the appointment was announced either 1 week before or 1 week after the decision became effective. Polish law does not directly stipulate the time at which the management board has to provide relevant current information to the market, but only says that it should be done immediately. The statistics presented above indicate significant nonchalance exhibited by companies in this respect.

Further research, connected with the complete publication cycle, was conducted using each date category. However, the present authors are of the opinion that in the context of the capital market the day on which the information is published is of key importance. When a decision is made about a specific date, all the events which interfere with the impact of succession on the financial result and company goodwill should be taken into account. Therefore, for each appointment we analyzed the period of 3 months before and after the event, trying to find significant economic events, e.g. decision about dividend, distribution of shares, information about the conclusion of a significant commercial agreement, etc. Eventually, occurrence of events which could impact on the investor response in the short run was confirmed for 224 observations.

5 Conclusions

The Polish capital market, although being the largest in Central and East European countries, is still in a phase of development and quality transformation. This is proven by the observation of the evolution of the structure of CEO appointments in public companies in 2000–2015. Up until the years of the world financial crisis, supervisory boards, representing the interests of shareholders, more often decided to appoint a new person rather than extend the term of office of the existing CEO. However, the statistics for recent years reveal a change in this trend. It is possible that the period of increased uncertainty in the financial markets, also in the political context, prompted owners to stabilize leadership in companies. If the political environment becomes less predictable, it is irrational to increase business risk with more frequent rotations in the top and key position in the company.

In conclusion it is worth pointing out that generally a CEO succession in the Polish capital market is not as frequent as in markets which were analyzed empirically. Ultimately, we identified 887 cases which can be considered to be actual succession. Furthermore, there is very little differentiation in this sample with respect to sex, education and experience of the appointees, and in the context of the predictability of the entire process. Further research will answer the question whether investors perceive differences between successors or circumstances in which the change was made.