• No results found

S 29525480) Abstract This paper investigates the effects of investor’s activism in the long-term performance of the targeted firms

N/A
N/A
Protected

Academic year: 2021

Share "S 29525480) Abstract This paper investigates the effects of investor’s activism in the long-term performance of the targeted firms"

Copied!
35
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

MSc Finance Thesis 2016-2017 S1 Faculty of Economics and Business

University of Groningen

The effects of shareholders’ activism on the performance of the firm on the long term. A comparison between different types of institutional investors.

by Giannakou Nikoleta ( Student Nr. S 29525480)

Abstract

This paper investigates the effects of investor’s activism in the long-term performance of the targeted firms. Moreover, it includes a comparison between the effects caused by hedge funds

activists and non-hedge funds activism. The metrics of performance that were used in this research are Tobin’s Q and ROA. In order to estimate the long-term effect on the performance

these ratios were calculated for five years after the intervention for each targeted firm.

Additionally, these ratios were calculated for every firm for two years prior to the intervention.

The outcome of this research indicates that, for the hedge funds targeted firms there is no significant difference after the intervention, both in the long term and in the short term.

Considering the firms targeted by non- hedge funds, the impact of activism found to be negative on the long-term performance.

Key Words: Shareholders’ Activism, Long-Term Firm Performance, Hedge Funds, non-Hedge Funds

Supervisor: prof.dr. C.L.M (Niels) Hermes

(2)

2 Acknowledgments

I would like to express my sincere thanks to Dr. C.L.M (Niels) Hermes for his continuous guidance, his helpful comments and the feedback he provided for the completion of this research.

I would also like to express my thanks to Giannakou Panagiotis, Electrical & Computer engineering MSc. for writing the script for the automated calculation of the industry median and

average.

(3)

3

Table of Contents

Introduction ...4

Hypothesis ...8

Proposed Model ...10

Data Collection Process Sample Construction ...13

Data Gathering and Variable Calculation ...15

Estimation outcome ...21

Firms Targeted by Hedge Funds ...25

Firms Targeted by non – Hedge Funds ...26

Limitations and Further Research ...27

Conclusion ...28

Bibliography ...29

Appendix Redundant Fixed Effects and Heteroskedasticity Tests ...31

Median/Average Calculation Script ...34

(4)

4

Introduction

This research will focus on institutional investors’ activism and the impact of this on the operating performance of the targeted firm. According to Smith, (1996), the role of institutional investors as monitors of the corporate control became very important during 1980s and 1990s, due to the rise of institutional holdings, which upgraded the role of institutional investors as shareholders, with the simultaneous decline of the corporate control market More specifically, it is reported that according to Federal Reserve Board and the Wall Street Journal, institutional ownership of stocks which belong in publicly traded firms in USA has increased rapidly, especially in 1980s and 1990s, surpassing the 50 percent level of aggregate ownership in 1992.This increase has been attributed mainly to the growth of pension funds. However, at the same time the influence of the market for corporate control that was active in 1970s and 1980s, diminished substantially. Since then, researchers examine the behavior of institutional shareholders. Particularly, literature focuses on the channels they use in order to exercise their monitoring, on what kind of industries attract active shareholders and the effect of this activism on the institutions.

In their paper Ryan, Verstegen, & Schneider., (2002) investors’ activism is defined as “the use of the power of an investor either to influence the performance or the operations of the targeted firm or to cause a large-scale change in the performance or the operations of a number of firms through the targeting of one or more firms.” Activism can be posed by independent institutional investors or in cooperation with others or they can be organized by industry groups.

The prevailing standpoint of academics is that active shareholders play a significant role in leading and monitoring the institutions they invest in. In order investors to decide, whether they will invest in a firm or not they carry out a cost-benefit analysis. In addition to the financial factors that investors take into consideration, there are other legal or social factors that affect institutional investors’ form of activism. They act in that way, because it is important that they can maintain the costs of their actions. Only in this case activist investors have the incentives and the means to help managers not to deviate from the targets of the firm. The changes that activist investors pursue on the performance and the operation of the firm, can be viewed by financial ratios such as Tobin’s Q, Return On Assets (ROA), Return On Equity (ROE), other financial measures such as operating or net income and through stock valuation. The latter is an efficient way of realizing the effect of activism because it can measure how the value of the firm is perceived by investors before and

(5)

5 after the intervention of the activist investor in the firm. However, institutional investors activism does not target only the enhancement of the economic performance of a firm but it can also affect the nonfinancial indicators, for example those represented by social performance measures.

Moreover, in many cases of investor’s activism, the aim was the reformation of the board of directors, changes in the amount and the structure of executive compensation and the repression of the anti-takeover provisions.

Apart from the existence of theories in favor of shareholders’ activism, there are researchers skeptical about this phenomenon. They claim that it undermines managerial power and forces strategic changes in institutions as the shareholders indicate. Moreover, they doubt about those shareholders’ practices, due to their lack of expertise.

Besides of the fact that the target of shareholder’s activism varies among the different cases, there are also different forms of activism presenting a range between cooperative and hostile form of intervention. Institutional investors usually start their intervention with negotiations and influence

“behind the scenes”, by having meetings with the firm’s board. This form is considered to be the cooperative form of activism. The case of activist investors going public is considered to be their last option. This form of activism can be posed with the use of, media campaigns and proxy or shareholders proposals. These forms of activism can be considered to be hostile and they signal to the market that the management of the firm is not cooperative.

At the beginning, the phenomenon of shareholders’ activism, was studied on the large public pension funds of the U.S such as CalPERS, CalSTRS, ColPERA, PSERS, SWIB, FSBA, NYSCR, TIAA-CERF. For example, Filatotchev & Dotsenko.,(2015) report that “CalPERS activism exhibit their activism following certain strategies. They aim in more board independence by requesting the firms in which they invest to compose their board predominantly of independent directors, they identify a lead director to assist the board chair and third they impose age limits on directors”. In their paper Gillan & Starks, (2002) mention that researchers were able to recognize activist behavior of institutional investors, particularly this of pension funds, from 1987 by reading their proxy proposals. These proposals were submitted by investors under the Rule 14-8, which is included in Securities and Exchange Act of 1934, issued by Securities and Exchange Commission of U.S.A. In 1992, SEC put into force new rules that granted the direct communication between

(6)

6 investors, reducing the costs of activism. This change lead to the increase of shareholders’

activism.

In their paper, McCahery, Sautner, & Starks , (2016), mention that institutional shareholders can choose between engaging with management to influence the way that a firm is being governed (also known as the “voice” or “direct intervention”) or they can leave the firm by selling their shares (also known “exit” or “voting with their feet”). According to a survey conducted by (), the most common method of influence is the engagement of the shareholders with management using the “behind the scenes” communication with management and board of directors. The method of

“exit” is the one investors prefer the least. In addition to that, as the survey indicated, various factors should exist in order this way of governance to be successful. These factors are referred in this research as follows, “investor’s equity stake size, whether other investors also exit for the same reason, managerial equity ownership and whether the other large shareholders are present.”

Furthermore, researchers claim that exit and voice are “complementary” methods of governance.

Investors which follow the method of voice tend to use voice more often than others.

Moreover, researchers have also focused their attention on the difficulties that active shareholders face. “Free Rider Problems” and legal concerns about “acting in concert” rules appear to be the most important impediments. Literature suggests that in order investors to be effective they need to be supported by the legal framework of the country that they invest in, for example they should be able to sue the directors and auditors in case of misgovernance. It is observed that, the increase in shareholders’ activism has been accompanied with shareholders’ lawsuits. Moreover, according to the existing papers there are differences between the legal environment in which the activism is exhibited and can lead to different outcomes. In general, researchers argue that in UK shareholders’

activism can be easily exhibited, due to the legal system and the small number of limitations on the action of shareholders.

However, this paper focuses on one of the main problems that can occur when investors actively participate in the governance of a firm, which is the myopic behavior which they can adopt when they are trying to enhance the performance of the targeted firm or when they want to make changes in firm’s operations.

The main contribution of this research is the comparison between the different types of institutional investors, regarding to the “myopic” behavior which they adopt. Two main categories of

(7)

7 institutional investors are formed, one of them includes the activist hedge funds and the other one includes non-hedge funds institutional investors.

More specifically, in the following sections the problem of investors’ short-termed behavior will be described and how affects the operating performance of the targeted. Moreover, there will be a categorization between the differences that can occur in the operating performance of firms targeted by different types of investors. In the section of proposed model, the model that will be used for this research will be described and the theory that supports the use of this particular model.

After that, the data collection process is described. In this section, someone can read about the method followed in order the samples of activist investors and the targeted firms to be formed.

Moreover, in the same section the method for calculating the variables for the estimation of the proposed model is described. The next section describes the method used for the estimation of the model and the outcomes. Additionally, the limitations of this research are presented and future research potentials are described. Finally, the last section of this research contains the conclusions.

(8)

8

Hypothesis

Many have argued that shareholders’ activism targets in enhancing the performance and reformatting the governance structure of a firm in the long-term. Moreover, it guarantees that the corporate governance of a firm acts in favor of the shareholders’ interests leading to the efficient governance of the firm and improvement in performance. Even though shareholders’ activism is rising and it is considered as a positive phenomenon, a large part of the papers which study the actual actions of shareholders conclude that these actions have little or no effect on the targeted firms considering both their company’s operations and efficiency. This research will focus on two main concerns that researchers have about investors’ activism. The first one is the myopic behavior that investors adopt and the second one is the existing differences in the governance of different types of investors. The first negative aspect of activism causes the decrease of the long-term firm performance whereas the second affects the efficiency of the moves made from activist investors towards the enhancement of the frim performance and the shareholders’ value.

One concern that have bothered researchers for long time is the myopic behavior that activist shareholders exhibit when they have to make decisions about the governance of the firm.

Shareholders who exhibit this kind of behavior are activists with short investment horizons, who push for actions that are profitable in the short term but are harmful for the long-term goals of the firms and their shareholders. This behavior affects certain types of changes. For example, actions driven by such a behavior may decrease the research and development expenses, the capital expenditures, market development and new business ventures, because of the fact that these kinds of expenditures pay-off on the long-term. Moreover, another problem that may rise from this kind of behavior is the reduction of the funding towards the long-term firm development because of the dividend payments, as activist investors pushes firm management to meet the dividend payment targets posed by them in their attempts to enhance the value of the shareholders on the short-term period. According to Bebchuk, Brav, & Jiang, (2015), the “myopic-activist claim” was a very important phenomenon during the discussions that dealt with shareholders’ activism and the formation of the legal rules and policies referring to activism, in order such regulations to be formed in order to prevent this kind of behavior which harms the operating performance of the firm on the long term.

(9)

9 Another main factor that can affect the outcome of shareholders’ activism according to literature is the type of the institutional investor. In Filatotchev & Dotsenko.,(2015) is stated that different types of investors may show different organizational implications, due to the fact that different owners may have different objectives and decision making horizons. The three different categories in which the institutional investors are divided, according to the business relations that grow with the firms that they invest in are the “pressure resistant”, the “pressure-sensitive” and the “pressure indeterminate”. One category of “pressure resistant” institutional investors are the hedge funds which hold an important position in the context of investor activism. However, their strategies differ from the ones exhibited by traditional institutional investors. A lot of research has been conducted in the field of hedge funds and the influence of their activism on the company in the short term and in the long term. Moreover, this outcome was compared with the mutual and pension funds activism. According to Brav, Jiang, Partnoy, & Thomas, (2008) activism of institutional investors, and especially mutual and pension fund activism, results in negative outcomes for shareholders. However, hedge funds are able to influence corporate boards and managements due to their different organizational form and incentives. Hedge funds do not work under the same regulations as mutual or pension funds and furthermore their managers suffer few conflicts of interest. This has as a result that hedge funds are better informed monitors than other institutional investors and they are able to act in a better way.

As it is mentioned above, shareholders’ activism affect the short-term and long-term operating firm performance. Thus, it is expected that the myopic behavior of the activists will decrease the performance in the long-term and. Additionally, regarding the different types of institutional investors, it is expected that the operating performance of the targeted firms will vary according to the type of activist investors. In order these differences to be observable two samples are formed;

the first one contains the firms targeted by hedge funds and the second one contains firms that are targeted by non-hedge funds institutional investors. In order to estimate the operating performance of the targeted firms, Tobin’s Q and return on assets (ROA) will be used. More specifically not only the prices of Tobin’s Q and ROA will be calculated but also the contribution of each year following the year during which the firm was targeted will be estimated, to a maximum span of five years.

(10)

10

Proposed Model

In order to investigate the impact of the myopic behavior of the activist institutional investors on the operating performance and thus the value of the frim, the model of Bebchuk, Brav, & Jiang, (2015) was followed. This particular model was chosen among others proposed in the literature because it gives the researcher the potential to distinguish the value that each year adds to the performance of the targeted firms, after the intervention of the activist institutional investors. The advantage of the model described above is achieved with the use of time dummy variables.

More specifically, the metrics used in order to measure the performance of the firm are Tobin’s Q and ROA. These two variables are adjusted for the industry performance. This adjustment allows the researcher to isolate the influences from the movements of the industry performance, which can be affected by many factors. One of the most important factors, in the examined time period is the economic recession. The economic crisis is present in the examined sample as the latter is picked in the period of 2005 to 2010. Each year’s effect on the metrics of performance will be represented by the coefficients of the time dummy variables. Apart from the five years following the intervention of the activist investor to the targeted firm, time dummy variables were built concerning the two years prior to the intervention and the year of intervention as well, in order to examine how the firm performance was affected from the years prior to the enforcement of activist shareholder’s propositions.

Apart from those variables, other control variables are used as proposed by the paper of Bhagat &

Bolton, (2008) and the paper of Bebchuk, Brav, & Jiang, (2015). These control variables are the Assets, the Market Value of the firm and the Expenses as they affect both Tobin’s Q and ROA. It is important to state that the calculation of the metrics used for the firm’s performance is based on the formula proposed by Bhagat & Bolton, (2008). Hence the dependent variables according to the adopted model will be explained by the factors presented in the following equation:

𝑸𝒊,𝒋𝒐𝒓 𝑹𝑶𝑨𝒊,𝒋= {𝐴𝑠𝑠𝑒𝑡𝑠𝑖,𝑗, 𝑀𝑎𝑟𝑘𝑒𝑡𝑉𝑎𝑙𝑢𝑒𝑖,𝑗, 𝐸𝑥𝑝𝑒𝑛𝑠𝑒𝑠𝑖,𝑗, (𝑡 − 2)𝑖,𝑗, (𝑡 − 1)𝑖,𝑗, 𝑡𝑖,𝑗, (𝑡 + 1)𝑖,𝑗, (𝑡 + 2)𝑖,𝑗, (𝑡 + 3)𝑖,𝑗, (𝑡 + 4)𝑖,𝑗, (𝑡 + 5)𝑖,𝑗}

(11)

11 Moreover, in order to be able to make the comparison between the outcomes of activism of different types of institutional investors, the regression is estimated for each of the different categories of firms targeted by different types of institutional investors. This research focuses on two different categories of firms targeted by institutional investors. One is the firms targeted by hedge funds and the other category is comprised of firms targeted by non-hedge funds institutional investors. Two regressions will be estimated for firms targeted by hedge funds; the first will have Tobin’s Q as dependent variable and the other one will have ROA as dependent variable. The same two regressions will be estimated for the firms targeted by non-hedge funds institutional investors respectively.

In order to be more specific, the model and the definition of each variable is presented below:

𝑻𝒐𝒃𝒊𝒏𝒔𝑸 𝒐𝒓 𝑹𝑶𝑨 =

𝑎0+ 𝛽1(𝑡 − 2) + 𝛽2 (𝑡 − 1) + 𝛽3𝑡 + 𝛽4(𝑡 + 1) + 𝛽5(𝑡 + 2) + 𝛽6(𝑡 + 3) + 𝛽7(𝑡 + 4) + 𝛽8(𝑡 + 5) + 𝛽9𝐴𝑠𝑠𝑒𝑡𝑠 + 𝛽10𝑀𝑎𝑟𝑘𝑒𝑡𝑉𝑎𝑙𝑢𝑒+𝛽11𝐸𝑥𝑝𝑒𝑛𝑐𝑒𝑠 + 𝜀 (1) 𝑻𝒐𝒃𝒊𝒏𝒔𝑸 ∶

[𝑩𝒐𝒐𝒌 𝑽𝒂𝒍𝒖𝒆 𝒐𝒇 𝑨𝒔𝒔𝒆𝒕𝒔+𝑴𝒂𝒓𝒌𝒆𝒕 𝑽𝒂𝒍𝒖𝒆 𝒐𝒇 𝑪𝒐𝒎𝒎𝒐𝒏 𝑬𝒒𝒖𝒊𝒕𝒚−(𝑫𝒆𝒇𝒇𝒆𝒓𝒅 𝑻𝒂𝒙𝒆𝒔+𝑩𝒐𝒐𝒌 𝑽𝒂𝒍𝒖𝒆 𝒐𝒇 𝑪𝒐𝒎𝒎𝒐𝒏 𝑬𝒒𝒖𝒊𝒕𝒚)]

𝑩𝒐𝒐𝒌 𝑽𝒂𝒍𝒖𝒆 𝒐𝒇 𝑨𝒔𝒔𝒆𝒕𝒔 (2)

In this case, we regress the industry adjusted Tobin’s Q which is the difference between each firm’s level and the average Q of the industry.

𝑹𝑶𝑨: 𝑶𝒑𝒆𝒓𝒕𝒊𝒏𝒈 𝑰𝒏𝒄𝒐𝒎𝒆 𝒃𝒆𝒇𝒐𝒓𝒆 𝒅𝒆𝒐𝒑𝒓𝒆𝒄𝒊𝒂𝒕𝒊𝒐𝒏 𝑻𝒐𝒕𝒂𝒍 𝑨𝒔𝒔𝒕𝒆𝒔 𝒕 𝒕𝒉𝒆 𝒆𝒏𝒅 𝒐𝒇 𝒕𝒉𝒆 𝒚𝒆𝒂𝒓 (3) Additionally, we regress the industry adjusted ROA which is the difference between each firm’s level and the average ROA of the industry.

𝒕: dummy variable that is 1 for the year during which the firm was first targeted by shareholders and zero otherwise.

t+1 to t+5: dummy variables which are 1 for targeted firms of interest for the years 1 to 5 after the year during which firms were targeted, otherwise they are zero.

(12)

12 (𝒕 − 𝟏), (𝒕 − 𝟐): dummy variables which are one for 1,2 and 3 years prior to the year the firm was targeted. The coefficients of these variables will help the researcher to understand how Tobin’s Q was during the years before the intervention of the activist shareholder.

𝑴𝒂𝒓𝒌𝒆𝒕 𝑽𝒂𝒍𝒖𝒆: The market value of the company each year.

𝑨𝒔𝒔𝒆𝒕𝒔: The value of the assets that the firm has each year.

𝑬𝒙𝒑𝒆𝒏𝒄𝒆𝒔: According to literature this variable is calculated from the following formula 𝑹&𝑫 + 𝑨𝒅𝒗𝒆𝒓𝒕𝒊𝒔𝒊𝒏𝒈 𝑬𝒙𝒑𝒆𝒏𝒔𝒆𝒔

𝑻𝒐𝒕𝒂𝒍 𝑨𝒔𝒔𝒆𝒕𝒔

(4)

For all three control variables, the log values were used in the regression.

(13)

13

Data Collection Process

Sample construction

The main purpose of this research is to find in what ways institutional investors’ activism affects the long-term firm performance and to compare the effects between activism posed by hedge funds and activism posed by non- hedge funds institutional investors. In order to construct the appropriate sample of activist institutional investors, the method of Brav, Jiang, Partnoy, &

Thomas, (2008) is followed. This method was chosen, because it gives the potential to recognize the cases of investors activism by reading the incentives of the investors and their actions.

Moreover, the sample’s source of the activist institutional investors are firms which operate in United States of America or Canada. The previous sample locations were chosen as the required information -which can indicate that an investor is activist according to the method that is going to be followed, are exclusively available through the public database of the Security and Exchange Commission of the U.S.A.

During this process of data collection, Schedule 13D filings were accessed in order to distinguish the activist institutional investors among the other investors. These filings were introduced by the Securities and Exchange Act of 1934 in U.S.A. According to the regulations a person or a group have to submit such a file when they acquire the beneficial ownership of more than 5% of a voting class of a company’s equity securities registered under the Section 12 of the Securities and Exchange Act of 1934.

In the case of this research, such filings were accessed from SEC’s database (EDGAR) during the period 2005-2010. More specifically, in order to construct the sample of institutional investors the behavior of which can be characterized as activist, emphasis was given on Items 2 and 4. Item 2 contains information about the investor who submits the filing and Item 4 provides information about the interests of the investors. In this item, someone can be informed about contacts that investors had with board of the targeted firm prior to the investment, the goals that the investor has set for the particular firm, the reason of the intervention and finally strategies that investors want to implement in order to reach their goals. Thus, based on these two items it is possible to identify whether the person or the firm submitting the filing is an institutional investor or a representative of an institutional investor, as well as the type of the institutional investor. Furthermore, these

(14)

14 items are helpful to distinguish the activist investors among the other investors who have filed a schedule 13d. From the filings reviewed during the sampling process, the ones filed by institutional investors were included in this research. The cases that were rejected are those of hostile acquisitions, of funding lending programs, as it is considered not to be representative cases for shareholders’ activism. However, there was a further filtering among the remaining Schedule 13 D filings, as cases of investors who had made only one filing during years 2005 to 2010 were excluded, because it was not possible to decide whether the investing firm was indeed a regular investor or there was just one case of activism. The filings in which the targeted firms were closed- end funds or other non-regular firms were excluded as well, as closed-end funds are organized in a different way and their performance should be analyzed in a way different than the one followed in this research. After this procedure, a research about the investors was conducted to find more elaborate about the activist investor to match the proper subsamples (hedge-funds and non-hedge fund institutional investors). The sources of the search were the websites of the firms or websites that have information useful for personal investors containing details about public firms.

Because of the fact that the amount of filings in each year, of the examined time period is large, however they should have been reviewed within a limited period of time, thus, the selection of filings was made randomly. Finally, the groups of hedge funds and non-hedge funds institutional investors formed after the described sampling process contain the same number of activist institutional investors, 75 in each group. The non-hedge fund institutional investors are comprised of investment banks, insurance companies and mutual funds. Following the original method of Bebchuk, Brav, & Jiang, (2015), investors that mange pooled investment vehicles and was not easy to decide whether they belong in hedge funds category or not, due to the luck of information, they are included in the sample of hedge funds. The number of the firms targeted by hedge funds active investors were 222 and the firms targeted by non-hedge funds institutional investors were 209.

According to Item 4 of the filings selected there are five main reasons for which the activist institutional investors, filed a Schedule 13D. The reasons of filings examined in the sample formed for this research’s purposes, found to be the same as appearing in the paper of Brav, Jiang, Partnoy,

& Thomas, (2008). The first and most common reason noticed in the filings selected for this sample is the investor’s belief that the value of the targeted firm is undervalued, thus, as an activist

(15)

15 investor, aims to opt for the maximization of the firm value and the value of the shareholders. The second reason is that the investor wants to make changes in the payout policy and the capital structure of the targeted firm. The third reason is the fact that activist investors want to make changes in the business strategy of the targeted firm, such as the involvement in a merger or an acquisition and proposals to pursue it’s a more optimal growth strategy. The fourth reason noted in some cases, is the fact that activist investors impel the sale of the targeted firm. Finally, the fifth reason for which investors decide to intervene actively in a firm is, in order to make changes in the governance structure of the firm.

Data Gathering and Variable Calculation

The data demanded for the regression of the model were retrieved from Compustat and the metrics of the performance were calculated using the formulas presented earlier. The total amount of observations for the firms targeted by hedge funds is 2885 and for the non-hedge funds institutional investor is 2717. Moreover, a limitation that exists in this sample is the fact that there are prices missing for the variables of the firms, as there are certain firms which stop to be public from a specific year and after, thus there are no available data in Compustat. However, it safe to claim that the sample used for this research is large enough to provide statistically significant results.

According to literature, the adjustment of Tobin’s Q and ROA for the performance of the industry, is achieved by subtracting the median Q and ROA of the industry from each Q and ROA of the corresponding firm which belongs to the same industry. However, after the completion of the data gathering and the calculation of the variables, the metrics of performance found to be skewed. In the case of a positively skewed distribution, the mean is greater than the median and in the case of negatively distributed data, mean is lower than the median. Thus, in this research, in most of the cases the data are positively skewed as it can be seen in Tables 1,2,3, and 4. For this reason, the industry average was used for the adjustment, in order the calculations to be more robust.

(16)

16 Table 1: Tobin’s Q and ROA before the adjustments for Hedge fund targeted firms

Panel A: Q t-2 t-1 t t+1 t+2 t+3 t+4 t+5

Mean 1.591106 1.677739 1.272079 1.14242 0.894331 0.841288 0.732189 0.63475 Standard

Error 0.099496 0.152589 0.085357 0.0914 0.071274 0.084717 0.073713 0.065302 Median 1.278336 1.202659 1.045571 0.970064 0.904182 0.474914 0 0 Standard

Deviation 1.508939 2.319157 1.294501 1.389156 1.083265 1.287581 1.117916 0.990348 Sample

Variance 2.276897 5.378489 1.675733 1.929755 1.173464 1.657864 1.249736 0.980789

Panel B: ROA t-2 t-1 t t+1 t+2 t+3 t+4 t+5

Mean -0.00711 -0.00302 0.019298 0.02195 0.027547 0.025771 0.041026 0.045493 Standard

Error 0.02796 0.026321 0.014329 0.014256 0.012323 0.011673 0.010591 0.01734

Median 0.024396 0.021275 0.017023 0 0 0 0 0

Standard

Deviation 0.42587 0.400914 0.218248 0.217133 0.187692 0.177799 0.161311 0.264115 Sample

Variance 0.181365 0.160732 0.047632 0.047147 0.035228 0.031612 0.026021 0.069757

Table 2: Adjusted Tobin’s Q and ROA for Hedge fund targeted firms

Panel A:Q t-2 t-1 t t+1 t+2 t+3 t+4 t+5

Mean -7.3011 -15.06 -3.5202 -3.89307 -4.05563 -16.8899 -92.4586 -8.30164 Standard

Error 2.738286 6.053852 1.034219 1.132451 1.255174 5.930118 77.4775 2.008506

Median -0.57591 -0.46818 -0.15071 0 0 0 0 0

Standard

Deviation 41.70834 92.20953 15.75275 17.24898 19.11825 90.32486 1180.102 30.59266 Sample

Variance 1739.585 8502.597 248.1492 297.5272 365.5074 8158.581 1392641 935.9108

Panel B:ROA t-2 t-1 t t+1 t+2 t+3 t+4 t+5

Mean 0.930522 1.204939 0.525511 0.4853 0.2811 0.362435 0.596235 0.408376 Standard

Error 0.306473 0.387172 0.124474 0.11775 0.048696 0.070822 0.156555 0.086166

Median 0.080413 0.044324 0.034736 0 0 0 0 0

Standard

Deviation 4.657982 5.884502 1.89184 1.789652 0.74012 1.076401 2.379428 1.309606 Sample

Variance 21.69679 34.62736 3.579059 3.202855 0.547778 1.15864 5.661679 1.715067

In order to calculate the industry average, the variables needed for the calculation of Q and ROA were retrieved from Compustat for all the firms available for each industry.

(17)

17 After that, the metrics of the performance were calculated and then the averages for each industry, for each year. This adjustment in the prices of Tobin’s Q and ROA compensates for the effect that the economic crisis had on the performance of the firms. More specifically, the effect of the recession on the particular industry that each firm belongs is subtracted from the metrics of its performance. For this reason, it is safe to claim that the differences that exist on the performance before and after the intervention will be due to the implementation of activist investors’ strategies.

In order to calculate the average and the median of this sample a script was constructed for the calculation of the average and the median of each different category of SIC code. This script has as inputs the years of the examined period, the SIC codes and Tobin’s Q and ROA for each firm belonging in the same industry for each year. The script is described in the Appendix section.

Considering the mean values of Tobin’s Q and ROA for firms targeted by hedge funds, it can be securely inferred that there is not a clear pattern of how the above metrics evolve through time. As it can be clearly observed in Tables 1 and 2 and in Graph1, Tobin’s Q both before and after the adjustment after the year of intervention becomes almost stable for the following 3 years. After that that period, in the fourth year after the intervention, there is an abrupt decline followed by an increase once more. However, the pattern of evolution of ROA over the five years following the intervention of the hedge fund in the targeted firm, is rather different. In this case ROA is continuously declining after the intervention of the activist investor and starts to increase after year (t+2) until year (t+4). Then a small decline is observed in year (t+5).

In the case of the firms targeted by non-hedge funds activist investors, the patterns of the performance metrics’ evolution over the five years following the intervention of the investor is completely different from the pattern observed from Q and ROA of the hedge fund targeted firm.

More specifically, as it is obvious in Tables 3,4 and Graph 2 Tobin’s Q continues to decline after the year of intervention for year (t+1) and then it follows an increasing trajectory until year (t+4) followed by a slight decrease. The ROA value presents an increase after year t of intervention until year (t+2). Then it declines for the following 2 years and then there is a slight increase for year (t+5).

In both hedge-fund and non-hedge fund targeted firms, the mean prices of Q and ROA after the industry adjustment are significantly different from the unadjusted ones. In the case of Tobin’s Q, the adjusted values are lower than the unadjusted prices. This implies that the industry mean is

(18)

18 positive and larger than the median Q of each targeted firm as well. This can indicate that the effect of the industry on Tobin’s Q is stronger than the one generated by the operations of the targeted firm. However, it is observed that in the case of ROA the adjusted median prices are higher than the unadjusted ones. Contrary to Tobin’s Q observations, this fact implies that the industry median prices are negative and smaller than median ROA prices of each targeted firm. This indicates that the effect of industry on ROA is weaker than this generated by the operations of the targeted firms.

Graph1 : Adjusted Q and ROA of hedge fund targeted firm

0 0.2 0.4 0.6 0.8 1 1.2 1.4

t-2 t-1 t t+1 t+2 t+3 t+4 t+5

ROA

-100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0

t-2 t-1 t t+1 t+2 t+3 t+4 t+5

Q

(19)

19 Table 3: Tobin’s Q and ROA before the adjustments for Non-Hedge fund targeted firms

Table 4: Adjusted Tobin’s Q and ROA for Non-Hedge fund targeted firms

Panel A: Q t-2 t-1 t t+1 t+2 t+3 t+4 t+5

Mean -3.65092 -1.65005 -7.49099 -21.7817 -6.81799 -4.00537 -3.68304 -4.62648 Standard Error 1.797924 1.467628 5.461084 14.07906 3.81945 1.750788 1.654705 1.67762

Median -0.04554 -0.06798 0 0 0 0 0 0

Standard

Deviation 26.30136 21.46955 79.88877 205.9589 55.87374 25.61182 24.20624 24.54147 Sample

Variance 691.7618 460.9416 6382.215 42419.05 3121.875 655.9652 585.9422 602.2836

Panel B: ROA t-2 t-1 t t+1 t+2 t+3 t+4 t+5

Mean 0.634096 0.623825 0.462075 0.539079 0.660354 0.602134 0.274247 0.357122 Standard Error 0.215815 0.247437 0.173078 0.265773 0.322532 0.304884 0.086086 0.147811

Median 0.022795 0.014566 0 0 0 0 0 0

Standard

Deviation 3.142308 12.97972 2.525993 3.869719 4.718233 19.79922 1.259329 2.162288 Sample

Variance 9.874096 145.8029 6.380641 14.97473 22.26172 138.1841 1.58591 4.675491

Panel A: Q t-2 t-1 t t+1 t+2 t+3 t+4 t+5

Mean 1.297832 2.004776 1.397624 -0.23566 -0.01077 0.224116 0.173107 0.029373 Standard Error 0.807909 0.268082 0.147843 0.960251 0.691837 0.641013 0.668821 0.670589

Median 1.424065 1.386678 1.111534 1.048872 0.994396 0 0 0

Standard

Deviation 11.76333 3.903328 2.147539 13.91536 9.953794 9.267015 9.645868 9.648099 Sample

Variance 138.376 15.23597 4.611925 193.6373 99.07802 85.87756 93.04276 93.08582

Panel B: ROA t-2 t-1 t t+1 t+2 t+3 t+4 t+5

Mean -0.02357 -0.01225 -0.01689 0.009644 0.005081 0.005229 0.014486 0.003675 Standard Error 0.050697 0.047131 0.035599 0.015258 0.018005 0.021707 0.015827 0.013895

Median 0.065188 0.063423 0.013379 0 0 0 0 0

Standard

Deviation 0.738166 0.684613 0.517111 0.220047 0.260917 0.313069 0.227161 0.199913 Sample

Variance 0.544889 0.468694 0.267404 0.048421 0.068078 0.098012 0.051602 0.039965

(20)

20 Graph 2: Adjusted Q and ROA of non-hedge fund targeted firm

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

t-2 t-1 t t+1 t+2 t+3 t+4 t+5

ROA

-25 -20 -15 -10 -5 0

t-2 t-1 t t+1 t+2 t+3 t+4 t+5

Q

(21)

21

Estimation Outcome

As it is stated in the section where the model is described for the hedge fund targeted firms, two types of models were estimated, one having Tobin’s Q as dependent variable and one having ROA.

The same types of models were estimated for the firms targeted by non-hedge funds intuitional investors, as well.

Table 5: Evolution of Tobin’s Q and ROA of firms targeted by Hedge Funds

Q Q ROA ROA

t-2 -0.917828

(0.993389) 0.223810

(0.931961) 0.009272

(0.081700) 0.013944

(0.070696)

t-1 -1.872249*

(1.014574) -1.238759

(1.029835) 0.186107**

(0.082932) 0.184127***

(0.058075)

t -1.130534

(1.017905)

-0.066888 (0.482051)

-0.030942 (0.077923)

-0.014427 (0.072852)

t+1 -0.436834

(1.027263) -0.485520

(0.813003) 0.026427

(0.077403) 0.035919

(0.031081)

t+2 -0.313667

(1.041115)

0.087326 (0.415947)

-0.079128 (0.079678)

-0.060339 (0.059921)

t+3 -0.605014

(1.081486) -0.564213

(1.291995) 0.009915

(0.079096) 0.011983

(0.044374)

t+4 -1.147372

(1.288184)

-2.155957***

(0.829914)

0.105250 (0.072661)

0.120546 (0.071741)

t+5 -1.095950

(1.393410) -0.378178

(1.073191) 0.026085

(0.071528) 0.063560

(0.060456)

Assets 0.701715

(0.148476) -0.240562

(0.221247) -0.062054***

(0.012571) -0.006779 (0.014014) Market Value -0.577590**

(0.152026)

-0.100153 (0.189410)

0.059944***

(0.012744)

0.020696 (0.013080) Expenses -0.049257**

(0.112623) 0.068203

(0.110597) 0.000730

(0.009209) 0.002473

(0.009283)

Year Fixed Effects Y Y Y Y

Firm Fixed Effects - Y - Y

R-Squared 0.037296 0.451080 0.032060 0.423548

F-statistic 4.819000 8.891175 4.120084 7.949744

Prob(F-statistic) 0.000000 0.000000 0.000000 0.000000

*, **,*** indicates significance at 10, 5 and 1 percent levels respectively

(22)

22 Before calculating the estimation of any of those regressions, the correlation between the variables was calculated in order to check whether the variables are highly correlated and thus, they are not appropriate to be included in the regression.

From the inspection of the calculated correlation matrix, was made clear that the variables were not highly correlated. Moreover, outliers were detected in both subsamples. Thus, the data were Winsorised in the 5% and 95% level.

For the estimation of this model, panel data were constructed and the method of Ordinary Least Squares were used for the estimation of the coefficients. According to literature, in the estimated models time, fixed effects and firm fixed effects were used in one version of the estimated regressions while industry fixed effects were added for another version. However, in Eviews the use of industry fixed effects is not supported. Thus, a small modification in the models proposed from the original paper had to be done. In this research, one version of the estimated model contains only time fixed effects and the other version contains firm and period fixed effects.

Redundant Fixed Effects tests were used in the models as it is important to check whether the use of time and firm fixed effects is appropriate for the data that are included in this research.

According to the outcomes of the tests in both cases of firms targeted by hedge funds and non- hedge funds, firm fixed effects were appropriate to be used in the estimation of regression. The tests considering the use of firm fixed effects resulted in the fact that, time fixed effects were proper to be used, in the case of the hedge funds targeted firms. However, in the case of firms targeted by non-hedge funds, time fixed effects were not proper technique to use for the estimation of the model.

(23)

23 Table 6: Hypothesis Testing for coefficients of firms targeted by Hedge Funds

Year Fixed Effects

Y Y Y Y

Firm Fixed Effects - Y - Y

*, **,*** indicates significance at 10, 5 and 1 percent levels respectively

Time fixed effects were implemented to account for time trends. In the regressions, including firm fixed effects, the coefficients of time dummy variables can be interpreted as the excess performance of the targeted firm over its own all-time averages and at the same time because of the existence of time fixed effects they are adjusted for conditions which apply to the whole market.

Another set of tests was employed, to check for possible existence of heteroskedasticity and autocorrelation. The outcomes of these tests indicate that the problems of heteroskedasticity and autocorrelation are present. In order to deal with the problem of heteroskedasticity White cross- section standard errors and covariance were used, as this method was the one giving more statistically significant coefficients with lower Standard Errors. However, despite the attempts made in order to estimate statistically significant coefficients, it is a fact that the coefficients of time dummy variables are statistically insignificant in most of the cases for both versions of models

Wald Test Q Q ROA ROA

Relative to (t) t+2

P-Value 0.816867

0.3608

0.154215 0.8045

-0.048186 0.5823

-0.045912 0.6733 t+3

P-Value 0.525520

0.5832

-0.497325 0.6834

0.040856 0.6498

0.026410 0.7757 t+4

P-Value -0.016838

0.9896 -2.089069**

0.0344 0.136192

0.1265 0.134973

0.1947 t+5

P-Value 0.034584

0.9815

-0.311290 0.7425

0.057027 0.5400

0.077988 0.4219 Relative (t-1)

t+2

P-Value 1.558582*

0.0982 1.326086

0.1544 -0.265235***

0.0054 -0.244467**

0.0119 t+3

P-Value

1.267236 0.2074

0.674547 0.6593

-0.176193*

0.0721

-0.172144**

0.0180 t+4

P-Value 0.724878

0.5870

-0.917198 0.4862

-0.080857 0.4055

-0.063581 0.4976 t+5

P-Value 0.776299

0.6105 0.860581

0.5120 -0.160022

0.1112 -0.120567

(24)

24 either for hedge fund targeted firms and for non-hedge fund targeted firms respectively. Although the slope of the estimated models is statistically significant.

Table 7: Evolution of Tobin’s Q and ROA of firms targeted by Non-Hedge Funds

*, **,*** indicates significance at 10, 5 and 1 percent levels respectively

The main purpose of this research is not only to find the coefficients of time dummy variables but also to examine whether there are significant differences in the coefficients of time dummy variables between the years prior to the intervention and after the intervention. The estimation of the time dummy variables aims to the measurement of the contribution of activism during the years following the intervention of an activist investor on the performance of a firm. Essentially, the

Q ROA

t-2 1.550265***

(0.572419)

0.032915 (0.051017)

t-1 2.333372***

(0.825666)

3.65E-05 (0.062193)

t 1.538800***

(0.420624) -0.038285 (0.033602)

t+1 0.660859

(0.799571)

-0.041633 (0.044387)

t+2 0.333847

(0.331999)

-0.034928 (0.028653)

t+3 0.825615

(0.404618)

-0.019572 (0.031217)

t+4 0.384642

(0.277738)

-0.028158 (0.025746)

t+5 -0.608688

(0.748277)

-0.025907 (0.040987) Assets -0.800940***

(0.112790) 0.031798***

(0.005166) Market Value 0.452075***

(0.138219)

0.008570 (0.007634)

Expenses 0.298688**

(0.122688) -0.013317*

(0.007026)

Year Fixed Effects - -

Firm Fixed Effects Y Y

R-Squared 0.323808 0.402554

F-statistic 5.459984 7.682461

Prob(F-statistic) 0.000000 0.000000

(25)

25 statistically significant differences between the coefficients indicate the change in the level of contribution in the performance of the targeted firm. The statistical significance of the differences between the coefficients was tested using the Wald test, which is available in EViews .

These differences are calculated between the coefficients of variables t+1 to t+5 and of variable t.

In addition to these tests, also tests for the differences between the coefficients of time dummy variables(t+2) to (t+5) and the coefficient of dummy (t-1) were employed. These additional differences are important to be estimated, because of the fact that the Tobin’s Q and ROA were measured at the end of the fiscal year and it is critical to have the difference between the year before the intervention and the year exactly after the intervention. It is important to focus on the differences from (t+3) and further, because the aim of this research is to find the effect on the long term operating performance of the targeted firm.

From the estimation of the model used by Bebchuk, Brav, & Jiang, (2015) the coefficients of the time dummy variables overall found to be statistically insignificant. This outcome indicates that during the years which followed the intervention of the activist shareholder the operating performance of the targeted firms was not affected significantly, thus from this research is not obvious whether the strategies implemented by activists affected the performance of the targeted firm.

Firms Targeted by Hedge Funds

The performance of the firms targeted by hedge funds does not exhibit a clear pattern of how each year following the intervention affects the operating performance of the targeted firm. This is obvious both from the regression output and from the Wald tests. Especially, as it can be seen in Table 6, the differences between the coefficients are not statistically significant in general.

However, there are some exceptions which show that the differences between the coefficients in the long-term for specific cases are negative and statistically significant. Although this outcome is the opposite of the outcome found in the original paper of (Bebchuk, Brav, & Jiang, 2015), it is expected for the firms included in this subsample. Someone can claim this, because of the unstable time trends exhibited by the median Qs and ROAs, as it can be seen in graph 1.

(26)

26 Firms Targeted by non Hedge Funds

In the case of the firms targeted by non-hedge funds institutional investors we find statistically significant differences between the coefficients of time dummy variables when Tobin’s Q is the dependent variable of the regression. These differences, apart from the fact that they are statistically significant they are negative as well. This, indicates that the operating performance of the firm is affected negatively in the long term by the intervention of the activist investors.

However, this is not the case for the differences estimated from the coefficients of the regression which has as dependent variable ROA. In that particular case, the differences were again statistically insignificant.

Table 8: Hypothesis Testing for coefficients of firms targeted by Non-Hedge Funds

Year Fixed Effects

- -

Firm Fixed Effects Y Y

*, **,*** indicates significance at 10, 5 and 1 percent levels respectively

Wald Test Q ROA

Relative to (t) t+2

P-Value -1.204952**

0.0143 0.003357

0.9247 t+3

P-Value -0.713184

0.1557 0.018714

0.4850 t+4

P-Value -1.154157***

0.0007

0.010127 0.6873 t+5

P-Value -2.147488**

0.0103 0.012378

0.7876 Relative (t-1)

t+2

P-Value -1.999525**

0.0242 -0.034965

0.5409 t+3

P-Value

-1.507757*

0.0943

-0.019608 0.7593 t+4

P-Value -2.942061*

0.0155 -0.028195

0.6310 t+5

P-Value -2.942061**

0.0155 -0.025944

0.7069

Referenties

GERELATEERDE DOCUMENTEN

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication:.. • A submitted manuscript is

ROE is the return on equity, calculated as Earnings per Share for the most recent fiscal year divided by the previous year’s book value per share.. Net profit is the net

These results indicate that the board size of the parent firm does not have an incremental effect on the change in market capitalization of t he spun-off firm in the second and

Literature research facilitated the identification of three key concepts: namely, the characteristics of a family firm, individual long-term orientation and management control

Beside the benefits for green innovating firms, this study also showed that having inbound board interlocks with a top 100 green innovating firm or firms from the KIBS-sector

The negative tone of the media is expected to have a negative influence on the stock market performance, while the volume of media coverage is expected to have a

Overall, hedge fund activism in Germany is accompanied by positive short-term shareholder value creation around the initial disclosure date, consistent with

Figure 7 shows the efficient frontiers for these portfolios. From table 3 and Figure 7 it becomes clear that hedge funds do certainly add value in a portfolio. Also, it is clear