The Effect of Antitakeover Provisions on Firm Value:
Evidence from the Netherlands
Doeko Kwast S2366495 MSc Finance University of Groningen Faculty of Economics and Business Supervisor: prof. dr. C.L.M. Hermes
Date: 07-06-2018
Abstract
This study examines the relation between the presence of antitakeover provisions (ATPs) and firm value in the Netherlands. Six types of ATPs have been observed. The effect of each type will be examined to identify potential different effects among the various types of ATPs. Possible entrenchment effects will be examined. Furthermore, mediating effects and moderating effects of respectively long-term investment intensity and bargaining power will be researched. The research will be conducted on a sample of 66 listed firms that have been observed over a period from 2008 to 2016. Five hypotheses will be tested in separate models. The effects of the ATPs are examined in comprehensive models testing the effects of all ATPs simultaneously, and in isolated models testing the effects of the various types, independently of the other ATPs. All models are period-fixed effects models. The results indicate a negative relation between the presence of two types of ATPs and firm value. Moreover, long-term investment intensity positively mediates the relation between the presence of one type of ATPs and firm value, and negatively mediates the relation between the presence of another ATP and firm value. The study provides no supportive evidence for a moderating effect of bargaining power.
Keywords: Antitakeover provisions (ATPs); Firm value; Tobin’s Q; Entrenchment; Managerial myopia; Bargaining power; Corporate governance
1. Introduction
In the past year, several large Dutch firms, such as Unilever and AkzoNobel, were target firms in a hostile takeover attempt. As a response, former Minister of Economic Affairs Henk Kamp proposed to strengthen Dutch antitakeover laws to protect firms against hostile takeovers (Kamerbrief, May 20, 2017). One of the proposed measures is a one-year period for consideration after a hostile takeover bid has been submitted. During this period, the shareholders cannot replace the Board of Directors. Large asset managers have written a letter to Mr. Kamp, persuading him to reconsider the proposed laws, saying it ‘would impact the efficiency of the financial markets, entrench ineffective company managers and disenfranchise institutional investors’ (Financial Times, June 15, 2017).
Apart from the proposed measures, the Civil Code of the Netherlands already facilitates several antitakeover provisions (ATPs) that firms can adopt, among which issuance of depositary receipts of shares, preference shares, and priority shares. Buijn and Storm (2013) define ATPs as ‘measures that – temporarily or permanently – affect the balance of power’ between shareholders, the Board of Directors, and the Supervisory Board. Over the past decades, both academics and investors have debated about the effects of ATPs on firm value. Most of the large listed firms in the Netherlands have adopted ATPs. This has led to the terminology ‘Dutch discount’, which implies that most Dutch listed firms are worth less than they potentially could be, due to the restriction of shareholder power.
According to the Dutch Corporate Governance Code (2016), a firm’s management should pursue long-term value creation while taking all stakeholders’ interests into consideration. Advocates of ATPs argue that ATPs are important to enable the management in pursuing this goal (De Brauw, 2017). It can be argued that the management has more information than the shareholders, making the management more suitable to assess the interests of all stakeholders. ATPs lead to less fear of replacement, enabling the management to act in the interests of all stakeholders.
The academic literature identifies three main views on the effects of ATPs on firm value: the entrenchment hypothesis, the managerial myopia hypothesis, and the bargaining power hypothesis (Stráska and Waller, 2014; Chemmanur and Tian, 2018). The entrenchment hypothesis states that the presence of ATPs entrenches inefficient managers, which negatively influences firm value. The managerial myopia hypothesis states that managers who are protected by ATPs behave less myopic, leading to higher firm value (Stráska and Waller, 2014). A third possible effect of ATPs is predicted by the bargaining power hypothesis, which states that ATPs provide more bargaining power to the boards of target firms, which can lead to higher takeover premiums and therefore, higher valuation of these firms (Stráska and Waller, 2010).
research on the influence of ATPs on firm value in the Dutch context is scarce (De Jong, DeJong, Mertens, and Wasley, 2005; Kabir, Kantrijn, and Jeunink; 1997). The hostile takeover attempts in 2017 have stimulated the debate about the protection of Dutch firms and the role of ATPs (Het Financieele Dagblad, May 23, 2017). The aim of this study is to examine the influence of ATPs on firm value in the Netherlands. The effect of each type of ATP will be examined individually, to identify possible different effects among the various types of ATPs. The relation between ATPs and firm value might be contingent on certain firm characteristics (Stráska and Waller, 2010; Sundaramurthy, 2000). In examining the influence of ATPs on firm value, the mediating effect of long-term investment intensity will be researched, using capital expenditure as a proxy. Moreover, the moderating effect of bargaining power will be investigated, using a modified measure of shareholder concentration as a proxy. To examine the above-mentioned relations, data on 66 Dutch firms over a sample period of 2008 to 2016 are collected and analyzed. All observed firms are or have been listed on the AEX-index, the AMX-index, or on the AScX-index, between 2008 and 2016.
The results of the study indicate negative effects of the presence of protective preference shares, and of the presence of depository receipts of shares that give voting rights under all circumstances, on firm value. The results partly support the entrenchment hypothesis. The results provide supportive evidence for positive effects of the presence of certified preference finance shares, and for negative effects of the presence of a supermajority provision, on term investment intensity. The results indicate a positive effect of long-term investment intensity on firm value. Furthermore, the study provides evidence for a positive mediating effect of long-term investment intensity on the relation between certified preference finance shares and firm value, and for a negative mediating effect of long-term investment intensity on the relation between a supermajority provision and firm value. The results partly support the managerial myopia hypothesis. The study does not provide evidence for a moderating effect of bargaining power.
The paper proceeds in the following structure: section 2 provides an overview of the prior research and theory regarding the influence of ATPs on firm value, a brief overview of the Dutch corporate governance model, and the hypotheses are developed and explained. Section 3 presents the methodology. The results are discussed in section 4, followed by the conclusion and limitations in section 5.
2. Literature review
2.1 Entrenchment hypothesis
firm’s shareholders (Jensen and Meckling, 1976). If a manager manages the firm inefficiently, the share price will drop, making the firm more attractive for a takeover and replacement of its management. This phenomenon is called the market for corporate control (Manne, 1965). Jensen and Ruback (1983) define corporate control as ‘the rights to determine the management of corporate resources.’ Adoption of ATPs hinders the takeover of a firm and reduces the threat of disciplinary replacement of inefficient managers. This lack of fear of external discipline could lead to value-destroying behavior by the management (Humphery-Jenner, 2014; Stráska and Waller, 2014). The inefficient behavior of entrenched management increases the agency costs, which will, according to the entrenchment hypothesis, result in lower firm value.
Two influential studies regarding entrenchment by ATPs and its influence on firm value were conducted by Gompers, Ishii, and Metrick (2003) and Bebchuk, Cohen, and Ferrell (2009). Gompers et al. (2003) argue that firms are republics, with shareholders as voters and elected representatives as directors. The directors appoint managers, who are regarded as bureaucrats by the model. The rules of governance determine the balance of power between shareholders, directors and managers. The authors construct a spectrum with two extremes: democracy, with high shareholder power, and dictatorship, with low shareholder power. In democracy firms, the shareholders can easily replace the directors, and the management has limited power. Gompers et al. (2003) describe these firms as having the ‘strongest shareholder rights’ and the ‘lowest management power.’ In contrast, shareholders of dictatorship firms cannot easily replace the directors, and the management has ‘extensive power.’ Dictatorship firms are defined as having the ‘weakest shareholder rights’ and the ‘highest management power.’ The main aim of the research is to identify the optimal balance of power. In order to identify the balance of power of each firm, the authors construct the so-called Governance Index (G index), in which they add a point for every ATP adopted by a firm. For two provisions that increase shareholder rights, a point is added when firms have not adopted the provision. The observed ATPs are divided in five categories. Firms with five or less ATPs are regarded as democracy firms. Firms with 14 ATPs or more are regarded as dictatorship firms. The G index is based on the database of the Investor Responsibility Research Center (IRRC), which contains information on governance provisions of US firms. To examine the relation between the G index and firm value, the authors use Tobin’s Q as a measure of firm value. Tobin’s Q is calculated by dividing the market value of a firm’s assets by the book value of a firm’s assets. Gompers et al. (2003) observe that dictatorship firms underperformed democracy firms during the sample period, that the G index is negatively correlated with firm value, and that ATPs cause higher agency costs. However, the authors conclude that no strong conclusions can be made about causality.
for mergers, and supermajority requirements for charter amendments’, poison pills, and golden parachute provisions. Based on these six ATPs, Bebchuk et al. (2009) construct the Entrenchment Index (E index), in which they add one point for each of the six ATPs that a firm has adopted. Similar to the research of Gompers et al. (2003), the observed ATPs are obtained from the IRRC database. The authors observe a negative correlation between the E index and firm value. Firm value is measured by Tobin’s Q. Bebchuk et al. (2009) conclude that although the observed relation is consistent with the entrenchment hypothesis, the data do not allow the authors to make conclusions about causality.
De Jong et al. (2005) have researched the influence of various types of ATPs on the value of Dutch listed firms between 1992 and 1999. The authors find negative effects on firm value for priority shares, preference shares and depository receipts of shares. De Jong et al. (2005) mention the lack of exposure to the market for corporate control as a possible explanation of the observed results. The lack of exposure to the market for corporate control reduces shareholder power and therefore, results in a reduced threat of disciplinary replacement. The lack of this discipline-enforcing mechanism can result in value-destroying behavior, effecting in higher agency costs and hence, lower firm value.
2.2 Managerial myopia hypothesis
In contrast to the entrenchment hypothesis, the managerial myopia hypothesis predicts that the presence of ATPs positively influences firm value. The managerial myopia hypothesis predicts that the presence of ATPs reduces the threat of a hostile takeover, which leads to a more long-term focus of a firm’s management, resulting in higher firm value. In the absence of ATPs, the fear of a takeover would result in myopic behavior of the management (Stein, 1988). Stein defines managerial myopia as the focus on short-term results instead of long-term goals. There are two assumptions underlying the managerial myopia hypothesis. The first assumption is the market value maximization hypothesis, which states that managers aim to maximize the market value of the firm in making investment decisions (McConnell and Muscarella, 1985). McConnell and Muscarella find that capital expenditure is positively related to the valuation of the firm, measured by the reaction of stock prices to investment accouchements. The opposite of the market value maximization hypothesis is the size maximization hypothesis, which predicts that managers aim to maximize the size of the firm when making investment decisions, leading to overinvestment and destruction of value.
short-term investments. Impatient, uninformed shareholders are not able to assess the true value of long-term investments.
Due to this inability, a drop in current earnings due to long-term investments leads to undervaluation of the stock, making the firm a potential takeover target (Shleifer and Vishny, 1990). The threat of a takeover leads to an incentive for managers to boost current earnings instead of investing in value-enhancing long-term projects, resulting in destruction of value (Stráska and Waller, 2014). According to the managerial myopia hypothesis, ATPs reduce the threat of a takeover and enable the management to make long-term investments, leading to value creation for shareholders and thus higher firm value.
Gompers et al. (2003) find that capital expenditure is higher for ‘high-G firms’ than for ‘low-G firms’. Sundaramurthy (2000) argues that ATPs result in a ‘long-term contract’ between the shareholders and the management, which enables the management to invest in firm-specific human capital, resulting in better long-term firm performance. Chemmanur and Tian (2018) find that ATPs positively influence firm innovation. Firm innovation requires making long-term investments. The observed effect is stronger for firms where the information asymmetry between the management and the shareholders is larger, and for firms that operate in more competitive product markets. Moreover, the authors find that ATPs are enhancing for firms that engage in innovative activities, whereas ATPs are value-destroying to other firms. The findings support both the entrenchment hypothesis and the managerial myopia hypothesis.
2.3 Bargaining power hypothesis
increased bargaining power enables managers to bargain more effectively, resulting in a higher takeover premium, compared to the takeover premium paid if all shareholders act individually (Grossman and Hart, 1980; Stráska and Waller, 2014). The possibility of receiving a higher takeover premium leads to a higher valuation of the firm.
Comment and Schwert (1995) find that ATPs increase the management’s bargaining power, resulting in a higher takeover premium, while ATPs do not deter potential takeovers. Therefore, ATPs would be beneficial to the shareholders. Stráska and Waller (2010) conclude that adoption of ATPs can be advantageous for firms with low levels of bargaining power,
while they observe a negative relation between ATPs and firm value in the whole sample. The authors use a combined measure of shareholder concentration, managerial ownership, and price-earnings ratio as a proxy for bargaining power. The observed negative relation supports the entrenchment hypothesis, since the presence of ATPs increases a firm’s agency costs. The effect of ATPs on firm value is contingent on the level of bargaining power. For firms with low levels of bargaining power, the benefit of the increased bargaining power due to the presence of ATPs outweighs the increased agency costs. However, for firms with high levels of bargaining power, the benefits of increased bargaining power are limited, and adoption of ATPs will be value-destroying due to managerial entrenchment.
This study uses a modified measure of shareholder concentration as a proxy for bargaining power. If a firm has high shareholder concentration, a cartelized response is more likely, since the shareholders are able to calibrate their response to a takeover offer more easily than when the firm is owned by a large number of shareholders who each own a small fraction of the shares. Dispersed shareholders are more likely to respond individually, rather than to respond in a cartelized manner. Kabir et al. (1997) find that firms with dispersed shareholders adopt a larger number of ATPs than firms with high levels of shareholder concentration do. Concentrated shareholders are more actively involved in managerial decision-making (Kabir et al., 1997). Concentrated shareholders can use their influence to communicate their interests to the negotiating management. Hereby, shareholders respond in a cartelized manner, and the management acts in their interest. Therefore, shareholder concentration is used as a proxy for bargaining power. The specific measurement of bargaining power will be discussed in subsection 3.2.3.
2.4 Corporate governance in the Netherlands
reliant on the applicable circumstances and should be assessed case by case (De Brauw, 2017).
The fundamental goal of the Dutch corporate governance model is the long-term value creation of the firm (Corporate Governance Code, 2016). In pursuing this goal, the board of a firm should always pay attention to the interests of all stakeholders. In comparison: in the United Kingdom, a firm’s management is obliged to take into account long-term stakeholders’ interests, but only if these stakeholders’ interests are relevant for shareholders’ interests (De Brauw, 2017). US-states with a so-called shareholder jurisdiction have a similar approach to corporate governance.
Kabir et al. (1997) observe a difference in discipline-enforcing mechanisms between the Anglo-American model and the Rhineland model. The authors argue that the market for corporate control is a relatively more important mechanism in the US for the enforcement of managerial discipline. In the Rhineland model, supervisory boards and concentrated ownership are relatively more important in enforcing managerial discipline. Moreover, the authors state that there is a difference between the two models with respect to the pursued goals of the management when adopting ATPs. According to the authors, managers of US firms adopt ATPs mainly to protect shareholders’ interests during takeover attempts, while managers in the Rhineland model adopt ATPs mainly to limit shareholder power. For this study, nine types of ATPs adopted by large Dutch listed firms have been observed. An overview and a description of the observed ATPs are presented in Appendix 1.
2.5 Hypotheses
This study tests three hypotheses: the entrenchment hypothesis, the managerial myopia hypothesis, and the bargaining power hypothesis. The main aim of the study is to examine if there is a relation between the presence of ATPs and firm value and, if present, whether this relation is negative or positive. A negative relation supports the entrenchment hypothesis. A positive relation might indicate supportive evidence for the managerial myopia hypothesis, or for the bargaining power hypothesis.
𝐇𝟏𝐀: The presence of ATPs negatively influences firm value
𝐇𝟏𝐁: The presence of ATPs positively influences firm value
A positive relation between the presence of ATPs and firm value might be resulting from less myopic behavior by the management, leading to higher long-term investment intensity. To examine the relation between the presence of ATPs and long-term investment intensity, the following hypothesis will be tested:
The relation between the presence of ATPs and firm value might be mediated by long-term investment intensity. In that case, ATPs result in higher long-long-term investment intensity, leading to higher firm value. To test the mediating effect of long-term investment intensity, the following hypothesis will be tested:
𝐇𝟑: The relation between the presence of ATPs and firm value is positively mediated by the intensity of long-term investments
An alternative explanation for a possible positive relation between the presence of ATPs and firm value is provided by the bargaining power hypothesis. The bargaining power hypothesis predicts that the presence of ATPs increases the level of bargaining power. The presence of ATPs is advantageous if the benefits of increased bargaining power outweigh the increased agency costs. Hence, ATPs are expected to be value-enhancing for firms with low levels of bargaining power. Therefore, the level of bargaining power is expected to have a negative moderating effect on the relation between the presence of ATPs and firm value.
𝐇𝟒: The relation between the presence of ATPs and firm value is negatively moderated by the level of bargaining power
3. Methodology and data
This section describes the methodology and data of this study. Firstly, the dependent variable, the independent variables, and the control variables will be defined, and the measures that will be used to quantify these variables will be explained. Secondly, the regression models that will be used to test the hypotheses developed in subsection 2.5 will be developed and explained. A description of all variables used in the equations that specify the regression models is provided in Appendix 2. Finally, the sample selection, data transformations, and summary statistics will be discussed.
3.1 Dependent variable
Similar to the research of Gompers et al. (2003) and Bebchuk et al. (2009), the dependent variable is firm value. Tobin’s Q is used as a measure of firm value. Tobin’s Q is calculated by the market value of a firm’s assets divided by the book value of the firm’s assets. The market value of the assets is the combined market value of the equity and the liabilities. The book value of the assets represents the replacement value of the assets.
𝑉!" = !"#$%& !"#$% !" !""#$"!"
Tobin’s Q can be an indicator of management’s performance (Hasbrouck, 1985). If a firm has a high Tobin’s Q, the market value is high relative to the book value, which indicates good performance. The data to measure Tobin’s Q are obtained from the financial database Orbis. Orbis contains market data and accounting data on all firms of the sample. Orbis obtains the accounting data from the firms’ financial reports. Both market values and book values are measured at the end of the year.
In the model that tests the relation between the presence of ATPs and the intensity of long-term investments, long-term investment intensity is the dependent variable. Long-term investment intensity is defined and explained in subsection 3.2.2.
3.2 Independent variables 3.2.1 ATPs
The number and types of ATPs adopted by each firm are obtained from a data file made available by Eumedion, a Dutch foundation that aims to ‘develop good corporate governance and sustainability performance’ (Eumedion, 2018). Eumedion has obtained the data on ATPs from the firms’ bylaws. An overview and a description of the observed types of ATPs are presented in Appendix 1. There are nine observed types of ATPs. However, the ATPs pyramid structure, dual class shares, and absolute binding nomination are excluded from the sample, since each of these ATPs are observed at only one firm. Possible observed effects of these ATPs are most likely resulting from idiosyncratic characteristics, rather than resulting from the presence of these ATPs.
3.2.2 Long-term investment intensity
A firm’s capital expenditure divided by its total assets, is used as a measure of long-term investment intensity. Capital expenditure is widely used as a proxy for long-long-term investment intensity (McConnell and Muscarella, 1985; Sundaramurthy, 2000). Total assets are measured in book value at the end of the year. Capital expenditure is the net amount of funds invested in fixed assets, other long-term assets, and other investments during a certain year, including R&D expenditure. Therefore, capital expenditure relative to the total value of the assets is used as a proxy for long-term investment intensity.
𝐶𝐴𝑃𝐸𝑋!" =!"#$%"& !"#$%&'()*$!"
!"#$% !""#$"!" (2)
The firms’ cash flow statements and balance sheets are used to obtain the firms’ capital expenditure and total assets. Both financial statements are obtained from the financial database Orbis.
3.2.3 Bargaining power
A modified measure of shareholder concentration, the Herfindahl-Hirschman Index (HHI), will be used as a proxy for bargaining power. Concentrated shareholders are more likely to respond in a cartelized manner, providing the management with more bargaining power. In the Netherlands, investors must report so-called substantial holdings in a public limited company to the Dutch Authority of Financial Markets (AFM). Each time a holding in a public limited company ‘reaches, exceeds or falls below’ a certain threshold, the investor is obliged to report this to the AFM. The threshold values are: 3%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50%, 60%, 75%, and 95%. The observed holdings are measured at the end of the year, similar to the measurement moment of firm value. The substantial holdings represent the percentage of common shares outstanding held as a substantial holding. However, for firms that have issued depository receipts of shares, the substantial holdings represent the percentage of depository receipts held as a substantial holding. The investors that own the depository receipts of shares are the beneficial owners of a firm, rather than the trust office that owns the common shares. Therefore, for these firms, shareholder concentration is measured by the percentage of depository receipts held as a substantial holding.
for bargaining power. The HHI is a widely used measure for concentration (Rhoades, 1993). The HHI in this study is calculated by squaring the holdings of all shareholders that hold substantial holdings, and subsequently summing the squared holdings.
𝐻𝐻𝐼 = !!!!(𝑆𝐻!)! (3)
Subsequently, the values of the HHI are transformed into a percentile rank to improve the comparability of the observed values, similar to the method used in the research of Stráska and Waller (2010). The level of bargaining power is specified by the following equation:
𝐵𝑃!" = 𝐻𝐻𝐼!"∗ (4) 3.3 Control variables
Next to the independent variables of interest, six control variables are included, namely size, industry, revenue, age, solvency ratio, and return on equity (ROE). The data on the control variables are obtained from the financial database Orbis. Market capitalization is used to measure size. Market capitalization is the market value of a firm’s equity, which is calculated by multiplying the current share price by the number of common shares outstanding. Analogous to firm value and bargaining power, size is measured at the end of each year. The second control variable is industry. To control for industry specific effects, the sample is divided into nine industry categories, which are defined in Table 4 of subsection 3.5. Revenue is measured on a yearly basis. Age is measured in years since the incorporation of the firm. Solvency ratio is a measure of the financial structure, and is calculated by dividing a firm’s shareholder funds by a firm’s total assets. ROE measures the firm’s profitability relative to its equity. ROE is calculated by the ratio between the profit before tax and shareholders funds.
3.4 Research method
The following subsections present and explain the regression models that test the hypotheses developed in subsection 2.5. All hypotheses will be tested using two approaches. The first approach uses a comprehensive model, which tests the effects of the various ATPs simultaneously. The second approach uses six isolated models for the various types of ATPs, to measure the effect of a specific ATP, independently of the other ATPs. The ATP variables specified in the regression models are binary, that take on the value 1 if the ATP is adopted by firm i at time t, and zero otherwise.
variables with constant values over time for a specific entity, such as industry variables or certain ATPs, cancel out when an entity fixed effects model is used. Therefore, the redundant fixed effects test can only be conducted for period fixed effects, and it is inappropriate to use pooled OLS, since it is impossible to draw conclusions about the homogeneity among the observed firms.
A random effects model is appropriate only if the error terms are uncorrelated with the explanatory variables. The correlation of the error terms with the explanatory variables will be examined by conducting a Hausman test. Under the null hypothesis, the error terms are uncorrelated with the explanatory variables. If the error terms are correlated with one or more explanatory variables, a period fixed effects model will be used. To avoid implications regarding heteroskedasticity of the error terms, White’s period heteroskedasticity consistent standard error estimates will be used. The mediating effect of long-term investment intensity will be tested using a two stage least squares (2SLS) model. A 2SLS model is appropriate when one of the explanatory variables is endogenous. The fist stage results in an estimate of the mediator, long-term investment intensity. In the second stage, the estimate is used as the value of the mediator (Wooldridge, 2015).
3.4.1 Entrenchment hypothesis
The main aim of this study is to examine the influence of the presence of ATPs on firm value. The comprehensive regression model designed to test the hypotheses H1a and H1b is specified by equation 5.
𝑉!" = 𝛼 + β!𝑃𝑃𝑆!" + β!𝑃𝑅𝐼𝑂!"+ β!𝐶𝑃𝐹𝑆!" + β!𝐷𝑅𝑆!"+
β!𝐷𝑅𝑆𝑅!"+ β!𝑆𝑃!"+ 𝜆!𝑆𝐼𝑍𝐸!"+ 𝜆!𝐼𝑁𝐷𝑈𝑆𝑇𝑅𝑌! + 𝜆!𝑅𝐸𝑉𝐸𝑁𝑈𝐸!"+ (5) 𝜆!𝐴𝐺𝐸!" + 𝜆!𝑆𝑂𝐿𝑉𝐸𝑁𝐶𝑌!" + 𝜆!𝑅𝑂𝐸!" + 𝜀!"
The regression coefficients β! to β! measure the effect of the presence of a certain ATP on firm value. The entrenchment hypothesis predicts that the presence of ATPs negatively influences firm value. The managerial myopia hypothesis and the bargaining power hypothesis predict that the presence of ATPs positively influences firm value. According to the prediction of hypothesis H1a, the values of the regression coefficients β! to β! are expected to be negative, whereas hypothesis H1b predicts that the values of the regression coefficients will be positive.
To identify the influence of a certain ATP on firm value, independently of the other ATPs, six regressions will be conducted, specified by equation 6.
The regression coefficient β! measures the effect of the presence of a certain ATP on firm
value. According to hypothesis H1a, the value of the regression coefficient β! is expected to be negative, whereas hypothesis H1b predicts that the value of the regression coefficient will be positive.
3.4.2 Managerial myopia hypothesis
The managerial myopia hypothesis predicts that the presence of ATPs encourages the management to make long-term investments. Hypothesis H2 tests the relation between the presence of ATPs and long-term investment intensity. The comprehensive regression model designed to test hypothesis H2 is specified by equation 7.
𝐶𝐴𝑃𝐸𝑋!" = 𝛼 + β!𝑃𝑃𝑆!"+ β!𝑃𝑅𝐼𝑂!" + β!𝐶𝑃𝐹𝑆!"+ β!𝐷𝑅𝑆!"+ β!𝐷𝑅𝑆𝑅!"
+β!𝑆𝑃!"+ 𝜆!𝑆𝐼𝑍𝐸!"+ 𝜆!𝐼𝑁𝐷𝑈𝑆𝑇𝑅𝑌!+ 𝜆!𝑅𝐸𝑉𝐸𝑁𝑈𝐸!" + 𝜆!𝐴𝐺𝐸!"+ (7) 𝜆!𝑆𝑂𝐿𝑉𝐸𝑁𝐶𝑌!"+ 𝜆!𝑅𝑂𝐸!"+ 𝜀!"
The regression coefficients β! to β! measure the effect of the presence of a certain ATP on long-term investment intensity. The managerial myopia hypothesis predicts that the presence of ATPs positively influences long-term investment intensity. According to the prediction of hypothesis H2, the values of the regression coefficients β! to β! are expected to be positive.
To identify the effect of a certain ATP on long-term investment intensity, independently of the other ATPs, six regressions will be conducted, specified by equation 8.
𝐶𝐴𝑃𝐸𝑋!" = 𝛼 + β!𝐴𝑇𝑃!"+ 𝜆!𝑆𝐼𝑍𝐸!"+ 𝜆!𝐼𝑁𝐷𝑈𝑆𝑇𝑅𝑌! + 𝜆!𝑅𝐸𝑉𝐸𝑁𝑈𝐸!" (8) + 𝜆!𝐴𝐺𝐸!"+ 𝜆!𝑆𝑂𝐿𝑉𝐸𝑁𝐶𝑌!"+ 𝜆!𝑅𝑂𝐸!"+ 𝜀!"
The regression coefficient β! measures the effect of the presence of a certain ATP on long-term investment intensity, independently of the other ATPs. According to the prediction of hypothesis H2, the value of the regression coefficient β! is expected to be positive.
4.1). Therefore, the control variable revenue is excluded from the model. The second stage model is specified by equation 9.
𝑉!" = 𝛼 + β!𝑃𝑃𝑆!"+ β!𝑃𝑅𝐼𝑂!"+ β!𝐶𝑃𝐹𝑆!" + β!𝐷𝑅𝑆!"+ β!𝐷𝑅𝑆𝑅!"
+β!𝑆𝑃!"+ β!𝐶𝐴𝑃𝐸𝑋!"+ 𝜆!𝑆𝐼𝑍𝐸!" + 𝜆!𝐼𝑁𝐷𝑈𝑆𝑇𝑅𝑌! + 𝜆!𝐴𝐺𝐸!" (9) +𝜆!𝑆𝑂𝐿𝑉𝐸𝑁𝐶𝑌!" + 𝜆!𝑅𝑂𝐸!" + 𝜀!"
The regression coefficients β! to β! measure the effect of the presence of a certain ATP on firm value, when controlling for the effects of long-term investment intensity on firm value. The regression coefficient β! measures the effect of the level of long-term investment intensity on firm value. Long-term investment intensity is expected to positively influence firm value, and therefore, the value of β! is expected to be positive. Under the managerial myopia hypothesis, the relation between the presence of ATPs and firm value is expected to be positive. Long-term investment intensity is expected to positively mediate this relation. The values of the regression coefficients β! to β! are expected to be smaller than the values of the regression coefficients β! to β! of equation 5, due to the mediating effect of long-term investment intensity.
To identify the mediating effect of long-term investment intensity on the relation between the presence of a certain ATP and firm value, independently of the other ATPs, six 2SLS models are tested. The first stage obtains an estimate of the value of long-term investment intensity. The models to obtain the estimate are similar to the models specified by equation 8. In the second stage, the value of long-term investment intensity is the value estimated in stage 1. Similar to the comprehensive model, the control variable revenue is excluded from the second stage models. The second stage models are specified by equation 10.
𝑉!" = 𝛼 + β!𝐴𝑇𝑃!"+ β!𝐶𝐴𝑃𝐸𝑋!"+ 𝜆!𝑆𝐼𝑍𝐸!"+ 𝜆!𝐼𝑁𝐷𝑈𝑆𝑇𝑅𝑌! + 𝜆!𝑅𝐸𝑉𝐸𝑁𝑈𝐸!" + 𝜆!𝐴𝐺𝐸!" + 𝜆!𝑆𝑂𝐿𝑉𝐸𝑁𝐶𝑌!" + 𝜆!𝑅𝑂𝐸!" + 𝜀!" (10)
The regression coefficient β! measures the effect of the presence of a certain ATP on firm value, independently of the other ATPs, when controlling for the effects of long-term investment intensity on firm value. The regression coefficient β! measures the effect of long-term investment intensity on firm value. The value of β! is expected to be smaller than the value of β! of equation 6, due to the mediating effect of long-term investment intensity. According to the prediction of hypothesis H3, the value of β! is expected to be positive.
3.4.3 Bargaining power hypothesis
moderating effect of bargaining power. The comprehensive regression model designed to test hypothesis H4 is specified by equation 11.
𝑉!" = 𝛼 + β!𝑃𝑃𝑆!"+ β!𝑃𝑅𝐼𝑂!"+ β!𝐶𝑃𝐹𝑆!"+ β!𝐷𝑅𝑆!" + β!𝐷𝑅𝑆𝑅!"+ β!𝑆𝑃!" + β!𝑃𝑃𝑆!" ∗ 𝐵𝑃!" + β!𝑃𝑅𝐼𝑂!"∗ 𝐵𝑃!"+ β!𝐶𝑃𝐹𝑆!"∗ 𝐵𝑃!" + β!"𝐷𝑅𝑆!"∗ 𝐵𝑃!"+ (11)
β!!𝐷𝑅𝑆𝑅!" ∗ 𝐵𝑃!"+ β!"𝑆𝑃!"∗ 𝐵𝑃!"+ 𝛾𝐵𝑃!"+ 𝜆!𝑆𝐼𝑍𝐸!"+ 𝜆!𝐼𝑁𝐷𝑈𝑆𝑇𝑅𝑌! + 𝜆! 𝑅𝐸𝑉𝐸𝑁𝑈𝐸!" + 𝜆!𝐴𝐺𝐸!" + 𝜆!𝑆𝑂𝐿𝑉𝐸𝑁𝐶𝑌!" + 𝜆!𝑅𝑂𝐸!" + 𝜀!"
The regression coefficients β! to β! measure the influence of the presence of a certain ATP on firm value. The regression coefficients β! to β!" measure the difference in the relation between the presence of a certain ATP and firm value when the level of bargaining power changes (Stráska and Waller, 2010). The bargaining power hypothesis predicts that the presence of ATPs is value-enhancing for firms with low levels of bargaining power. Therefore, the values of β! to β! are expected to be positive. Bargaining power is expected to have a negative moderating effect on the relation between the presence of ATPs and firm value. The relation is expected to be weaker for firms with higher levels of bargaining power, as the benefits of increased bargaining power due to the adoption of ATPs do not outweigh the increased agency costs. Therefore, the values of the regression coefficients β! to β!" are expected to be negative, as higher levels of bargaining power reduce the advantage of the adoption of ATPs (Stráska and Waller, 2010). The regression coefficient γ measures the difference in the intercept as a result of including bargaining power. High bargaining power results in higher possible takeover premiums, since the management can negotiate more effectively. Therefore, the value of the regression coefficient γ is expected to be positive.
To identify the moderating effect of bargaining power on the relation between a certain ATP and firm value, independently of the other ATPs, six regressions will be conducted, specified by equation 12.
𝑉!" = 𝛼 + β!𝐴𝑇𝑃!"+ β!𝐴𝑇𝑃!"∗ 𝐵𝑃!"+ 𝛾𝐵𝑃!!+ 𝜆!𝑆𝐼𝑍𝐸!"+ 𝜆!𝐼𝑁𝐷𝑈𝑆𝑇𝑅𝑌! (12) +𝜆!𝑅𝐸𝑉𝐸𝑁𝑈𝐸!"+ 𝜆!𝐴𝐺𝐸!"+ 𝜆!𝑆𝑂𝐿𝑉𝐸𝑁𝐶𝑌!" + 𝜆!𝑅𝑂𝐸!" + 𝜀!"
3.5 Sample selection and summary statistics
The sample consists of 66 firms, which are or have been listed on the AEX-index, the AMX-index, or on the AScX-index, between 2008 and 2016. The 66 firms and the sample period are selected on the availability of data on the two main variables of interest: the presence of ATPs and firm value. As explained in subsection 3.2.1, three types of ATPs are excluded from the sample. The three firms that have adopted one of the excluded ATPs are excluded from the sample. The data used for this study are panel data, which contain both time series data and cross-sectional data. The data are obtained from an unbalanced panel, mainly because not all firms were listed over the full sample period. There are 377 observed values of firm value and ATPs. Similar to the research of Stráska and Waller (2010), observations for which there is no available information about the presence of ATPs or on firm value are excluded, since the aim of this study is to examine the relation between the presence of ATPs and firm value.
To improve the robustness of the data, a log transformation is used for the values of the variables firm value, size, and revenue. A log transformation results in less skewed distributions. The values of the variables long-term investment intensity, solvency, and ROE cannot be transformed into decimal logarithms, since there are respectively forty-five, seven and seventy negative values observed for these variables. Likewise, the values of the variable age cannot be transformed into decimal logarithms, as there are two observations with the value zero.1 As explained in subsection 3.2.3, the values of bargaining power are transformed into a percentile rank. The effect of outliers is moderated by winsorizing the variables for the 2.5% tails of the distributions. The minimum values and maximum values are the values of the 2.5th percentile and the 97.5th percentile, respectively. The ATP variables and the industry variables are not winsorized, since these variables are measured as dummy variables. Summary statistics are reported in Table 1. Table 15 in Appendix 3 presents the summary statistics for untransformed and unwinsorized values.
Table 2 presents the number of firms per year that have adopted zero to three ATPs. On average, 80.56% of the observed firms has adopted at least one ATP. Most of the observed firms have adopted one ATP. The maximum observed number of ATPs adopted by a firm is three. The increase in the number of observed firms between 2011 and 2012 is resulting from an expansion of the observed firms by Eumedion. Table 3 provides an overview of the adopted ATPs. The most-adopted ATP is protective preference shares with 219 observations over the sample period. The least-adopted ATP is depository receipts of shares with restricted voting rights during a hostile takeover attempt, with 17 observations over the sample period.
1 To perform robustness checks, the values of the variabels long-term investment intensity, solvency, ROE, and age are
Table 1: Summary statistics
Descriptive statistics per variable. Firm value is the decimal logarithm of Tobin’s Q, which is calculated as the market value of assets divided by the book value of assets. Investment intensity is capital expenditure to total assets. Bargaining power is the percentile rank of the HHI score. Size is the decimal logarithm of market capitalization in million US dollars. Revenue is the decimal logarithm of revenue in million US dollars. Age is the number in years since the incorporation of the firm. Solvency is the ratio between shareholder funds and total assets. ROE is the return on equity, which is measured as the ratio between profit before tax and shareholder funds. Only firms with observations on both firm value and ATPs are included in the sample. All variables are winsorized for the 2.5% tails of the distributions. N*T represents the number of observations.
N*T Mean Median St.dev. Max Min Skewness Kurtosis
Firm value 377 -0.2794 -0.1688 0.4738 0.4939 -1.5229 -1.0225 0.7706 Investment intensity 325 0.0500 0.0428 0.0732 0.2458 -0.1659 -0.0654 2.0630 Bargaining power 377 0.4987 0.4970 0.2888 0.9748 0.0210 -0.0025 -1.2152 Size 375 3.1569 3.1357 0.7363 4.5193 1.6876 -0.0550 -0.7907 Revenue 369 6.2894 6.3478 0.7283 7.6315 4.9828 0.0212 -0.8526 Age 377 67.4987 45.0000 58.2354 247.6000 5.0000 1.3326 1.4860 Solvency 375 0.3890 0.4084 0.1937 0.8766 0.0449 0.0979 -0.2587 ROE 368 0.1098 0.1101 0.1807 0.5201 -0.3776 -0.2976 0.9545
Table 2: Number of ATPs adopted
Number of firms per year that have adopted 0, 1, 2, or 3 ATPs. ATP represents the number of ATPs adopted by a firm. N represents the number of observations in a certain year. Only firms with observations on both firm value and ATPs are included in the sample. The total number of observations is 377.
ATP 2008 2009 2010 2011 2012 2013 2014 2015 2016 Total 0 2 3 2 2 12 14 15 17 18 85 1 14 13 15 16 19 18 21 23 23 162 2 8 9 8 8 12 13 14 13 15 100 3 3 3 3 3 4 4 3 3 4 30 N 28 29 29 30 48 49 53 56 60
Table 3: Observed ATPs over the period 2008-2016
Types of ATPs observed per year. NO is the number of firms that have not adopted ATPs. PPS represents the number of firms that have the possibility to issue protective preference shares. PRIO represents the number of firms that have issued priority shares. CPFS represents the number of firms that have issued certified preference finance shares. DRS represents the number of firms that have issued depository receipts of shares that give voting rights under all circumstances. DRSR represents the number of firms that have issued depository receipts of shares with restricted voting rights during a hostile takeover attempt. SP represents the number of firms that have adopted supermajority provisions related to the nomination or discharge of board members. N represents the number of firms observed in a certain year. Only firms with observations on both firm value and ATPs are included in the sample. The total number of observations is 377.
ATP 2008 2009 2010 2011 2012 2013 2014 2015 2016 Tota l NO 2 3 2 2 12 14 15 17 18 85 PPS 18 18 19 19 27 28 28 29 33 219 PRIO 6 6 6 7 7 7 6 6 7 58 CPFS 6 6 5 5 6 6 6 5 7 52 DRS 3 3 3 4 6 6 6 6 5 42 DRSR 1 1 1 1 2 2 3 3 3 17 SP 5 5 5 5 7 7 9 9 9 61 N 28 29 29 30 48 49 53 56 60
observations, 25.99% of the total observations. Media is the industry with the lowest number observations, with 17 observations, 4.51% of the total observations.
Table 4: Firms per industry
Number of firms per industry per year. Only firms with observations on both firm value and ATPs are included in the sample. The total number of observations is 377.
Industry 2008 2009 2010 2011 2012 2013 2014 2015 2016 Total
Agricultural and biochemical 2 2 2 2 4 5 7 7 7 38
Communication and postal services 2 2 2 2 2 2 2 2 2 18 Construction 4 4 4 4 6 6 5 5 5 43 Financial industry 1 1 1 3 5 5 5 6 10 37 Industrial 8 8 8 8 12 12 13 14 15 98 Media 1 2 2 2 2 2 2 2 2 17 Real estate 3 3 3 3 4 4 5 5 4 34 Retail 3 3 3 1 5 5 5 6 7 35 Services 4 4 4 4 7 8 9 9 8 57 N 28 29 29 30 48 49 53 56 60
4. Results
In this section, the results of this study are reported. Firstly, the results of the multivariate analyses are reported. Subsequently, the robustness of the results will be discussed. The effects of the control variables on firm value are not subject of the research of this study. Therefore, there will be no elaboration on the results of the control variables. 4.1 Multivariate analyses
Before testing the models, the correlations between the variables are observed to detect possible multicollinearity issues. The pairwise correlations are reported in Table 5. In case of perfect multicollinearity, the correlation between two variables is 1.0000. Near multicollinearity results in high standard errors of the regression coefficients. The highest correlation between two variables is between size and revenue, with a value of 0.7284. Therefore, multicollinearity issues are absent in the models.
As explained in subsection 3.5, the values of the variables firm value, long-term investment intensity, bargaining power, size, revenue, age, solvency, and ROE are winsorized for the 2.5% tails of the distributions for all regression models. For all models, a Hausman test is conducted. The results of these tests indicate that the error terms are correlated with one or more explanatory variables for all models. Therefore, a random effects model is inappropriate, and a period fixed effects model is used for all models.
4.1.1 Entrenchment hypothesis
The results of the comprehensive regression model and of the isolated regression models are reported in Table 6 and Table 7, respectively. The results of the comprehensive model indicate statistically significant negative effects of the presence of protective preference shares on firm value. The results of both the comprehensive model and the isolated models indicate statistically significant negative effects of the presence of depository receipts of shares that give voting rights under all circumstances on firm value. The indicated effect of the presence of protective preference shares is statistically significant at the 10% level. The difference between the effects indicated by the comprehensive model and the effects indicated by the isolated models testing the hypotheses H1a and H1b will be discussed in subsection 4.2. The indicated effect of the presence of depository receipts of shares that give voting rights under all circumstances is statistically significant at the 5% level, for both the comprehensive model and the isolated model. The results do not provide evidence for a statistically significant relation between the presence of four ATPs and firm value. The adjusted R-squared value of the comprehensive model is 0.8989, indicating that 89.89% of the variance of firm value is explained by the explanatory variables. The adjusted R-squared values of the isolated regression models are within a range of 0.8905 and 0.8965, indicating that the isolated regression models explain 89.05% to 89.65% of the variance of firm value.
Table 6: Results comprehensive regression model entrenchment hypothesis
The dependent variable is firm value. Firm value is the decimal logarithm of Tobin’s Q. PPS represents protective preference shares. PRIO represents priority shares. CPFS represents certified preference finance shares. DRS represents depository receipts of shares that give voting rights under all circumstances. DRSR represents depository receipts of shares with restricted voting rights during a hostile takeover attempt. SP represents supermajority provisions related to the nomination or discharge of board members. Standard errors are reported in parentheses. The results are controlled for size, industry specific effects, revenue, age, solvency, and return on equity (ROE). The variables firm value, size, revenue, age, solvency, and ROE are winsorized for the 2.5% tails of the distributions. R² is the adjusted R-squared, and reflects the proportion of the variance explained by the explanatory variables. N*T represents the number of observations. P-values smaller than 0.01, 0.05 and 0.10 are indicated by ***, ** and *, respectively.
Variable Coefficient Variable Coefficient
The indicated statistically significant effects are consistent with the prediction of H1a and therefore, provide supportive evidence for the entrenchment hypothesis, as the results indicate that there is a negative relation between the presence of these ATPs and firm value. The results are inconsistent with the prediction of hypothesis H1b, and therefore, do not provide supportive evidence for the managerial myopia hypothesis or the bargaining power hypothesis, as these hypotheses predict value-enhancing effects of the presence of ATPs. According to the entrenchment hypothesis, the indicated effects result from the entrenchment of the management of the firms that have adopted protective preference shares, depository receipts of shares that give voting rights under all circumstances, or both. The presence of these ATPs leads to the management being less fearful of disciplinary replacement, resulting in inefficient behavior. The inefficient behavior leads to increased agency costs, which are ultimately borne by the firms’ shareholders, resulting in lower valuations of these firms.
However, the results do not provide evidence for a statistically significant relation between the presence of four types of ATPs and firm value. The absence of a relation might indicate that the behavior of a firm’s management is unaffected by the presence of these ATPs. Not all ATPs provide the same level of protection. For example, the protective effects of priority shares and of supermajority provisions can be circumvented if the acquirer acquires a stake providing two-thirds of the voting rights, representing more than half of the firm’s capital (see Appendix 1). Therefore, these ATPs might be less entrenching. Unaffected behavior does not result in higher agency costs, explaining the absence of an observed effect on firm value. However, this explanation does not fully explain the observed effects. There are two observed types of depository receipts of shares. The results only indicate an effect of the presence of depository receipts that give voting rights under all circumstances on firm value, although this is theoretically the least entrenching type of the two.
A more plausible explanation for the absence of an observed effect of the presences of the four ATPs on firm value is that the presences of these ATPs entrench the firm’s management, but at the same time lead to higher long-term investment intensity or more bargaining power. In that case, the negative effects are offset by the positive effects, resulting in no statistically significant relation between the presence of these ATPs and firm value. The results of the models that test the managerial myopia hypothesis and the bargaining power hypothesis are reported in the subsections 4.1.2 and 4.1.3, respectively.
4.1.2 Managerial myopia hypothesis
This subsection reports the results of the regression models that test the hypotheses H2 and H3. Hypothesis H2 predicts that the presence of ATPs positively influences long-term investment intensity. A positive relation between the presence of ATPs and long-term investment intensity provides supportive evidence for the managerial myopia hypothesis.
comprehensive model and the isolated models indicate a statistically significant positive effect of the presence of certified preference finance shares on long-term investment intensity, and a statistically significant negative effect of the presence of a supermajority provision related to the nomination or discharge of board members on long-term investment intensity. The indicated effect of the presence of certified preference finance shares is statistically significant at the 1% level, for both the comprehensive model and the isolated models. The effects of the presence of a supermajority provision, indicated by the results of the comprehensive model and the isolated model, are statistically significant at the 10% level and at the 5% level, respectively. The results do not provide evidence for a statistically significant relation between the presence of four types of ATPs and long-term investment intensity. The adjusted R-squared value of the comprehensive model is 0.2170, indicating that 21.70% of the variance of long-term investment intensity is explained by the explanatory variables. The adjusted R-squared values of the isolated regression models are within a range of 0.1992 and 0.2198, indicating that the isolated regression models explain 19.92% to 21.98% of the variance of long-term investment intensity.
The indicated statistically significant positive effect of the presence of certified preference finance shares on long-term investment intensity is consistent with the prediction of hypothesis H2 and therefore, provides supportive evidence for the managerial myopia
Table 8: Results comprehensive regression model managerial myopia hypothesis
The dependent variable is long-term investment intensity. Long-term investment intensity is capital expenditure to total assets. PPS represents protective preference shares. PRIO represents priority shares. CPFS represents certified preference finance shares. DRS represents depository receipts of shares that give voting rights under all circumstances. DRSR represents depository receipts of shares with restricted voting rights during a hostile takeover attempt. SP represents supermajority provisions related to the nomination or discharge of board members. Standard errors are reported in parentheses. The results are controlled for size, industry specific effects, revenue, age, solvency, and return on equity (ROE). The variables long-term investment intensity, size, revenue, age, solvency, and ROE are winsorized for the 2.5% tails of the distributions. R² is the adjusted R-squared, and reflects the proportion of the variance explained by the explanatory variables. N*T represents the number of observations. P-values smaller than 0.01, 0.05 and 0.10 are indicated by ***, ** and *, respectively.
Variable Coefficient Variable Coefficient
hypothesis. The indicated statistically significant negative effect of the presence of a supermajority provision on long-term investment intensity is inconsistent with the prediction of hypothesis H2. According to the managerial myopia hypothesis, the indicated effect of the presence of certified preference finance shares is explained by less myopic behavior of the management of the firms that have adopted this ATP. Due to information asymmetry between the shareholders and the management, the management has an incentive to boost current earnings instead of investing in value-enhancing long-term projects, to avoid becoming a potential takeover target due to undervaluation (Stráska and Waller, 2014). According to the managerial myopia hypothesis, the presence of certified preference finance shares provides protection to the management, reducing the threat of a hostile takeover and of replacement, resulting in higher long-term investment intensity. In contrast to the prediction of hypothesis H2, the results indicate that the presence of a supermajority provision is negatively related to long-term investment intensity. The theory of the managerial myopia hypothesis does not provide an explanation for this indicated effect. The indicated effect is most likely resulting from unobserved characteristics of the presence of a supermajority provision, that cause a negative relation between the presence of this ATP and long-term investment intensity.
Hypothesis H3 predicts that there is a positive relation between the presence of ATPs and firm value, and that this relation is positively mediated by the long-term investment intensity. The results reported in Table 6 and Table 7 do not provide evidence for a positive relation between the presence of ATPs and firm value. Moreover, evidence for a statistically significant positive effect between the presence of ATPs and long-term investment intensity is only found for certified preference finance shares, as presented in Table 8 and Table 9. The mediating effect of long-term investment intensity on the relation between the presence of ATPs and firm value is tested by using a 2SLS model. The first stage computes an estimate for the value of long-term investment intensity. In the second stage, the value of the variable long-term investment intensity is the value estimated in stage 1.
within a range of 0.3044 and 0.4618, indicating that the isolated regression models explain 30.44% to 46.18% of the variance of firm value.
The results reported in Table 6 to Table 11 provide supportive evidence for a mediating effect of long-term investment intensity on the relation between the presence of the two ATPs and firm value. Without inclusion of long-term investment intensity, the results do not indicate a relation between the presence of certified preference finance shares and firm value, or between the presence of a supermajority provision and firm value. The results indicate a statistically significant positive effect of the presence of certified preference finance shares on long-term investment intensity, and a statistically significant negative effect of the presence of a supermajority provision on long-term investment intensity. Furthermore, the results provide supportive evidence for a positive relation between long-term investment intensity and firm value. Moreover, the results indicate that the effect of the presence of certified preference finance shares on firm value is negative, and the effect of the presence of a supermajority provision on firm value is positive, when controlling for the effect of long-term investment intensity.
Table 10: Results comprehensive 2SLS regression model managerial myopia hypothesis
The dependent variable is firm value. Firm value is the decimal logarithm of Tobin’s Q. Long-term investment intensity is denoted by CAPEX. Long-term investment intensity is capital expenditure to total assets. PPS represents protective preference shares. PRIO represents priority shares. CPFS represents certified preference finance shares. DRS represents depository receipts of shares that give voting rights under all circumstances. DRSR represents depository receipts of shares with restricted voting rights during a hostile takeover attempt. SP represents supermajority provisions related to the nomination or discharge of board members. Standard errors are reported in parentheses. The results are controlled for size, industry specific effects, age, solvency, and return on equity (ROE). The variables firm value, long-term investment intensity, size, age, solvency, and ROE are winsorized for the 2.5% tails of the distributions. R² is the adjusted R-squared, and reflects the proportion of the variance explained by the explanatory variables. N*T represents the number of observations. P-values smaller than 0.01, 0.05 and 0.10 are indicated by ***, ** and *, respectively.
Variable Coefficient Variable Coefficient
The presence of certified preference finance shares leads to higher long-term investment intensity, which subsequently positively influences firm value. The positive effects on firm value are most likely offset by the negative effects on firm value, resulting from entrenchment effects, explaining the absence of an indicated relation by the results reported in Table 6 and Table 7. The results indicate an opposite effect of the presence of a supermajority provision of firm value. The theory of the managerial myopia hypothesis does not provide an explanation for the observed effects. One might argue that managers who are protected by a supermajority provision spend the firm’s resources on private benefits, instead of making long-term investments. However, the underlying market value maximization assumption states that the firm’s management aims to maximize the value of the firm in making investment decisions. Moreover, this explanation does not clarify the absence of a negative relation between the presence of a supermajority provision and firm value, reported in Table 6 and Table 7, as shareholders would take this money-squandering behavior into account when they value a firm. Therefore, the effects are most likely caused by unobserved characteristics of a supermajority provision, that have positive effects on firm value, and negative effects on long-term investment intensity, which subsequently negatively influences firm value. The positive effects on firm value are most likely offset by the negative effects on firm value, explaining the absence of an indicated relation by the results reported in Table 6 and Table 7.
4.1.3 Bargaining power hypothesis
This subsection presents the results of the regression models that test hypothesis H4. Hypothesis H4 predicts that the relation between the presence of ATPs and firm value is negatively moderated by the level of bargaining power.
The results of the comprehensive regression model and of the isolated regression models are reported in Table 12 and Table 13, respectively. The results of both the comprehensive model and of the isolated models do not provide evidence for statistically significant relations between the presence of the various types of ATPs and firm value, or for a moderating effect of bargaining power on these relations. The negative effects of two types of ATPs on firm value, indicated by the results reported in Table 6 and Table 7, are not indicated by the results reported in Table 12 and Table 13, due to the inclusion of bargaining power. Moreover, the results do not indicate a statistically significant effect of bargaining power on the intercept of the models. The adjusted R-squared value of the comprehensive model is 0.8994, indicating that 89.94% of the variance of firm value is explained by the explanatory variables. The adjusted R-squared values of the isolated regression models are within a range of 0.8914 and 0.8973, indicating that the isolated regression models explain 89.14% to 89.73% of the variance of firm value.
of bargaining power on firm value, with exclusion of the ATPs. The results reported in Table 14 do not provide evidence for a statistically significant relation between bargaining power and firm value, explaining the results reported in Table 12 and Table 13.
Table 12: Results comprehensive regression model bargaining power hypothesis
The dependent variable is firm value. Firm value is the decimal logarithm of Tobin’s Q. Bargaining power is denoted by BP. Bargaining power is the percentile rank of the HHI score. The moderating effect of bargaining power is measured by the multiplication of the ATP and bargaining power. PPS represents protective preference shares. PRIO represents priority shares. CPFS represents certified preference finance shares. DRS represents depository receipts of shares that give voting rights under all circumstances. DRSR represents depository receipts of shares with restricted voting rights during a hostile takeover attempt. SP represents supermajority provisions related to the nomination or discharge of board members. Standard errors are reported in parentheses. The results are controlled for size, industry specific effects, revenue, age, solvency, and return on equity (ROE). The variables firm value, bargaining power, size, revenue, age, solvency, and ROE are winsorized for the 2.5% tails of the distributions. R² is the adjusted R-squared, and reflects the proportion of the variance explained by the explanatory variables. N*T represents the number of observations. P-values smaller than 0.01, 0.05 and 0.10 are indicated by ***, ** and *, respectively.
Variable Coefficient ATP*BP Variable Coefficient
