The Effect of Ownership Structure on Firm Performance: Evidence from the Food Industry.

(1)

1

MSc Thesis International Financial Management

University of Groningen

The Effect of Ownership Structure on Firm Performance:

Evidence from the Food Industry.

Abstract

This paper investigates the relation between ownership concentration and company performance in the food industry. Furthermore, we test if shareholder rights moderate this relationship. We use 289 companies that are located in 42 different countries. Under OLS specifications, we find that ownership concentration has a positive effect on firm performance. Further, we find some evidence that shareholder protection moderates the relationship between ownership structure and firm performance. These results enable us to argue that strong shareholder rights result in lower ownership concentration and that shareholder rights is a substitute to high ownership concentration to mitigate the agency problem. Moreover, the findings suggest that strong shareholder rights results in lower ownership concentration, however the results are only partially supported.

JEL classification: G32; G34; G38

Key words: Ownership structure, Corporate Governance, Firm Performance, Food Industry, Shareholder protection,

Name: Klaas Jan Streekstra Student number: S2968266

(2)

2

1. Introduction

The relationship between ownership structure and firm performance has been a subject of expansive inquiry in the financial literature, ever since Berle and Means (1932) have investigated the ownership changes in the United States, hereafter called the US, to develop the agency theory. Berle and Means (1932) detect an inverse correlation between the diffuseness of shareholders and company performance, since the separation of ownership and control requires effective monitoring of those who control the firm, named the management, by those who own the firm.

(4)

4 The aim of this study is to investigate the effect of ownership structure on firm performance in the food industry as one of the oldest industries in the world. Customers spent in 2010 38% of their income on food and beverages (World Bank, 2010).

Therefore, the research question is:

What is the effect of ownership structure on company performance in the food industry?

Since the problems caused by the separation of ownership and control are addressed by laws and regulations that protect shareholder rights and mitigate the agency problems, we will use shareholder protection to understand the moderating effect on the relation between ownership structure and performance.

This study is structured as follows. Section 2 provides a thorough review of the existing literature on this topic, as well as the definition of the hypotheses. Next, section 3 describes the variables, methodology, and the sample data set of this study. Section 4 presents the descriptive statistics and regression results, and the discussion of the results. Section 5 concludes the paper.

2. Literature review

2.1 Governance theory

(5)

(6)

6 field, and gives an overview of which formulas previous studies use to measure ownership concentration. In this paper, owner structure is measured by ownership concentration by the largest 1, 3, and 20 shareholders (Leech & Leahy, 1991). Morck et al., (1988) suggest that the overall firm performance (Tobin´s Q) has a positive relation between ownership and Q in the 0% to 5% board ownership range, a negative relation is found between 5% to 25% range. Beyond that, a positive relation is found. The reason that could explain the positive relationship is the convergence of interest. The negative effect from the 5% to 25% range could be explained by the agency theory. Based on the agency theory we expect that ownership concentration have a positive effect on the company performance. According to the research question and the literature review, the following hypothesis will be tested.

H1: Ownership concentration has a positive effect on the company performance.

2.2 Shareholder protection

(7)

7

Figure 1.

Ownership concentration

Shareholder protection

Firm performance

country, investors do not need to put all their investment in one company and could diversify their risk. Therefore, ownership concentration is a decent substitute for legal investor protection in poor investor protected nations. We expect that shareholder rights have a moderating effect on the relationship between ownership concentration and firm performance. Based on prior research and theories hypothesis 2 is stated:

H2: Shareholder protection has a moderating effect on the relation between ownership

concentration and firm performance.

3. Methodology

3.1 Data

(8)

8 In order to test the effect of ownership structure on the performance of a company, this study uses a regression analyses which are is in line with various previous studies (Demsetz & Lehn,1985) (Leech & Leahy, 1991), (Thomson & Pedersen, 2000), and (Demsetz & Villalonga, 2001). Demsetz & Lehn (1985) find no significant relationship between ownership concentration and accounting profit rates. To test the effect they use accounting rate of return as a dependent variable (Demsetz & Lehn, 1985). However, Demsetz & Villalonga (2001) use average Tobin´s Q as the dependent variable. Other studies like (Thomson & Pedersen, 2000) (Leech & Leahy, 1991) use the several dependent variables to measure firm performance. Thomson & Pedersen (2000) use market to book ratio, return on assets, and sales growth. Leech & Leahy (1991) use six different dependent variables to measure firm performance, which are: valuation ratio, trading profit margin, rate of growth of total sales, rate of growth of net assets, and salary of the highest paid director. However, exploring prior empirical papers, researchers use after 1985 more market measures, than accounting measures (Short, 1994). In order to test the effect of ownership, one needs a proper definition and measurement of ownership and its structure. Based on the literature review there is not one common measurement for ownership structure. Demsetz & Lehn (1985) use ownership concentration to determine the effect of ownership structure. Their work rests on the percentage of shares owned by the most significant shareholders, A5 and A20, and the calculation of the Herfindahl index AH (Demsetz & Lehn, 1985). In line with this Demsetz & Lehn, (1985) we use the same measurement of ownerships structure. Other studies like (Leech & Leahy, 1991) describe companies’ ownership structure in two ways. Through ownership concentration (Cn) and owner controlled (OCn). In line with Demsetz & Lehn

(1985) they also used the Herfindahl index and concentration ratios (Leech & Leahy, 1991). Thomson & Pedersen, (2000) also use two ways to measure ownership structure. The first one is ownership concentration, percentage of the largest shareholders. Second, they included a dummy analysis for owner identity for non-listed firms.

3.2 Dependent variable

(9)

9 the past, while the market profit rates are based on future predictions on company data (Demsetz & Villalonga, 2001). In trying to support the effect of ownership structure on the company performance, it is preferable to look what the management has realized, and how they are going to perform in the future (Demsetz & Villalonga, 2001). The other difference between both rates is which rate does actually measure performance? The accounting rates are forced by standards that are different among countries. Tobin’s Q is determined by the acumen, pessimism, or optimism (Demsetz & Villalonga, 2001). Accounting rate methods are not influenced by the mindset of the investors and only to some extent considering future events. Tobin´s Q is affected by the mindset of the investors, that consider future events that influenced the share price positively or negatively (Demsetz & Villalonga, 2001). Each method has its benefits and drawbacks. Therefore, in this research both measurements will be included, to test if there are any differences between the results.

Approximation of Tobin’s Q=𝐸𝑞𝑢𝑖𝑡𝑦 𝑚𝑎𝑟𝑘𝑒𝑡 𝑣𝑎𝑙𝑢𝑒 _{𝐸𝑞𝑢𝑖𝑡𝑦 𝑏𝑜𝑜𝑘 𝑣𝑎𝑙𝑢𝑒} (1)

For the accounting rate method, the return on assets (ROA) is used. This ratio is used as dependent variable, because previous articles indicate that this ratio is a proper tool in order to define the firm efficiency.

Return on assets = 𝑁𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠

(2)

3.3 Independent variables

(10)

10 TOPCON 3 and TOPCON20 to check if the results are changing. The independent variables would be calculated as present in formula 3. S stands for the percentage of shares.

𝑇𝑂𝑃𝐶𝑂𝑁 = ∑ 𝑆^2, 𝑖

𝑛

𝑛=1

(3)

3.4 Shareholder protection

To test if shareholder protection affects the relation between ownership and performance, the ‘’anti-director right index’’ is used. This interaction variable is named SHRi,t in the equation. The level depends on an index combining shareholder rights that considers five different levels of shareholder rights (La Porta et al, 1997). A score of 1 indicates that the county permits shareholders to mail a proxy vote. A score of 2 indicates that shareholders are not mandatory to pay their share before the general shareholder meeting. 3 indicates that cumulative voting is accepted by the law within a country, for the 4 there is a mechanism to suppress minority shareholders. Finally, 5 indicates preemptive right of new shareholders, or when shareholders with fewer or equivalent than ten percent of the shares, are enabled to demand for an extraordinary shareholders meeting (La Porta et al, 1997). Spamann (2010) upgraded the anti-director rights index of La Porta et al, (1997). The database of Spamann (2010) will be used to determine the countries shareholder protection.

3.5 Control variables

(11)

11 3.5.1 Cash to assets ratio

The cash to assets ratio (CA) shows how much percentage cash the company holds of the total assets. A high ratio can have more reasons, like in the paper of Opler et al., (1999) is investigated. It is important for a company to hold a certain amount of cash inside the company to fulfill their short-term payments, but a too high cash ratio could indicate a loss of opportunities (Opler et al., 1999).

𝑐𝑎𝑠ℎ 𝑡𝑜 𝑎𝑠𝑠𝑒𝑡𝑠 = 𝐶𝑎𝑠ℎ 𝑎𝑛𝑑 𝑐𝑎𝑠ℎ 𝑒𝑞𝑢𝑖𝑣𝑎𝑙𝑒𝑛𝑡𝑠 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠

(4)

3.5.2 Firm size

Firm size is the control variable that measures the size of the company. The bigger the company size, the more capital investor needs to own one percentage of the shares of the firm (Demsetz & Villalonga, 2001). SIZE is the name of the variable in the equation, and is measured by the natural logarithm of total assets.

𝑆𝐼𝑍𝐸 = 𝐿𝑁 (𝑇𝑜𝑡𝑎𝑙 𝐴𝑠𝑠𝑒𝑡𝑠) (5)

3.5.3 R&D ratio

The research and development (R&D) indicates many percentages of the sales are being invested in the R&D. This ratio could be a good indicator of how the company will perform in the future. This ratio is calculated by the R&D expense divided by the total sales of the company.

𝑅&𝐷 = 𝑅&𝐷 𝑒𝑥𝑝𝑒𝑛𝑠𝑒 𝑇𝑜𝑡𝑎𝑙 𝑠𝑎𝑙𝑒𝑠

(6)

3.5.4 Leverage

(12)

12 creditors could have an influence on the management of the firm (Demsetz & Villalonga, 2001).

Leverage (LEV) = 𝐷𝑒𝑏𝑡

𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠

(7)

3.5.5 Sales growth & growth of equity

The following two control variables indicate by which percentage the firm is growing each year. This indicates that the firm could increase their operations. These ratios are calculated as followed:

Sales growth (SG) = 100*(( 𝑆𝑎𝑙𝑒𝑠−𝑠𝑎𝑙𝑒𝑠 𝑙𝑎𝑠𝑡 𝑦𝑒𝑎𝑟_{𝐿𝑎𝑠𝑡 𝑦𝑒𝑎𝑟´𝑠 𝑠𝑎𝑙𝑒𝑠} )-1) (8)

Equity growth (EG) = 100*(( 𝐸𝑞𝑢𝑖𝑡𝑦 𝑛2−𝑒𝑞𝑢𝑖𝑡𝑦 𝑛1

𝐸𝑞𝑢𝑖𝑡𝑦 𝑛1 )-1 (9)

3.4 Research method

To test the effect of ownership structure on company performance an OLS regression will be used. The following equations will be used:

3.4.1 Equation

Tobin´s Q i,t = 𝛼 + 𝛽1𝑇𝑂𝑃𝐶𝑂𝑁i, t + 𝛽2𝑆𝐻𝑅𝑗, 𝑡 + 𝛽3𝑇𝑂𝑃𝐶𝑂𝑁𝑖, 𝑡 ∗ 𝑆𝐻𝑅𝑗, 𝑡 +

𝛽4𝐶𝐴𝑖, 𝑡 + 𝛽5𝑆𝐼𝑍𝐸𝑖, 𝑡 + 𝛽6𝑅&𝐷𝑖, 𝑡 + 𝛽7𝐿𝐸𝑉𝑖, 𝑡 + 𝛽8𝑆𝐺𝑖, 𝑡 + 𝛽9𝐸𝐺𝑖, 𝑡 + 𝜀i,t

(10)

ROA i,t = 𝛼 + 𝛽1𝑇𝑂𝑃𝐶𝑂𝑁i, t + 𝛽2𝑆𝐻𝑅𝑗, 𝑡 + 𝛽3𝑇𝑂𝑃𝐶𝑂𝑁𝑖, 𝑡 ∗ 𝑆𝐻𝑅𝑗, 𝑡 + 𝛽4𝐶𝐴𝑖, 𝑡 + 𝛽5𝑆𝐼𝑍𝐸𝑖, 𝑡 + 𝛽6𝑅&𝐷𝑖, 𝑡 + 𝛽7𝐿𝐸𝑉𝑖, 𝑡 + 𝛽8𝑆𝐺𝑖, 𝑡 + 𝛽9𝐸𝐺𝑖, 𝑡 + 𝜀i,t

(11)

(13)

(14)

14

4. Results

In this section, the results, and the descriptive statistics will be presented. Table 2 presents the mean and median of Tobin’s Q, return on assets, ownership concentration of the biggest shareholder, ownership concentration of the top three shareholders, shareholder protection, and cash ratio by each country. Further, there will be a discussion on the results of the regression. Finally, there will be a discussion about the result of the robustness test.

4.1 Descriptive statistics

Table 1 provides mean and median for all the variables used in this research. Table 1 offers three columns. The first one provides the descriptive statistics for the complete sample. The second column provides the descriptive statistics for companies that are located in countries with high shareholder rights, and the third column provides the descriptive statistics for firms that are located in countries with weak shareholder rights. In table 1, we define high shareholder countries with those countries with a score that is above 2.5. We define countries with low shareholder protection with a score of 2.5 or below. Further, this paper provides a statistical comparison between countries with strong shareholder protection and weak shareholder protection.

Table 1 presents the mean and median of the variables: Tobin’s Q, return on assets (ROA), ownership concentration (TOPCON1), ownership concentration with top three shareholders (TOPCON3), shareholder protection (SHR), interaction variable 1 (TOPCON*SHR), cash to assets ratio (CA), size, research and development (R&D), leverage of the firm (LEV), sales growth (SG), and equity growth (EG). With a t-test, a statistical comparisons between high shareholder right countries and low shareholder right countries. *, **, *** indicate 10%, 5%, and 1% significant level.

The mean and median of Tobin’s Q indicates that the market value is twice as much as the book value. The average q in 2010-2015 is 2.1556. This is significantly higher than the

Table 1. Descriptive Statistics

All countries high SHR Low SHR

(N=1734) (N=1668) (N=66)

Mean Median Mean Median Mean Median

(15)

(16)

16 invest more money on research and development, and on average have 15.76% more sales growth than firms in countries with low shareholder rights.

(17)

17

Table 2. Descriptive Statistics

Tobin's Q ROA TOPCON1 TOPCON3 SHR CA

N Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Code law countries

Argentina 18 2.7394 1.8152 3.26% 3.93% 0.4611 0.4729 0.4729 0.6030 2 2 11.50% 11.35% Austria 12 1.6775 1.6605 3.59% 3.52% 0.1069 0.0576 0.1576 0.1490 2.5 2.5 4.01% 3.99% Belgium 24 1.0319 0.9922 3.32% 3.53% 0.2004 0.1631 0.2711 0.2791 3 3 6.74% 6.14% Brazil 24 1.8481 1.8855 5.54% 3.05% 0.2500 0.0400 0.2515 0.0709 5 5 11.29% 11.26% Switzerland 48 1.9692 1.6953 6.57% 5.73% 0.2341 0.2511 0.2443 0.2606 3 3 6.91% 5.98% Chili 12 0.6104 0.3954 4.68% 5.07% 0.1026 0.1018 0.1241 0.1226 4 4 6.61% 6.51% Germany 30 1.6462 1.4027 3.14% 3.36% 0.3362 0.1523 0.3908 0.2495 3.5 3.5 5.27% 5.94% Egypt 36 1.2743 1.2796 8.92% 9.92% 0.1977 0.2500 0.2237 0.2669 3 3 23.09% 24.57% Spain 24 1.6962 1.4108 4.20% 4.65% 0.0104 0.0680 0.0289 0.0805 5 5 7.62% 5.27% Finland 30 1.0220 0.7475 2.61% 2.68% 0.1178 0.0124 0.1399 0.0196 3.5 3.5 7.11% 4.92% France 66 2.0930 2.0689 3.33% 3.22% 0.3043 0.2014 0.3494 0.2476 3.5 3.5 10.04% 9.03% Greece 18 0.5031 0.4013 -5.23% -2.10% 0.2426 0.0151 0.3175 0.0316 5 5 2.20% 1.78% Indonesia 78 3.0233 2.2171 6.27% 5.71% 0.2889 0.2510 0.3089 0.2515 4 4 8.63% 6.53% Italy 12 0.7399 0.5729 1.92% 1.13% 0.1607 0.1403 0.1845 0.1515 2 2 6.95% 6.21% Jordan 6 0.7961 0.8102 3.69% 4.79% 0.9729 0.9729 0.9728 0.9728 1 1 0.66% 0.65% Japan 360 1.0633 0.9156 2.69% 2.76% 0.0053 0.1666 0.0104 0.1835 4.5 4.5 12.69% 11.79% South Korea 78 1.4021 1.0397 2.41% 2.07% 0.1656 0.1563 0.1820 0.1660 4.5 4.5 7.67% 6.33% Mexico 36 2.0650 1.5754 7.29% 7.28% 0.2365 0.2500 0.2591 0.2500 3 3 7.92% 3.63% Netherlands 18 1.6792 1.4681 2.44% 1.75% 0.0675 0.0656 0.1021 0.0688 2.5 2.5 5.23% 3.14% Norway 12 1.3600 1.3977 5.64% 5.59% 0.3008 0.3273 0.3261 0.3754 3.5 3.5 7.08% 6.44% Peru 12 1.6400 1.6492 7.49% 5.90% 0.1324 0.2145 0.1428 0.2246 4.5 4.5 3.13% 1.94% Philippines 18 9.7776 0.7124 7.08% 5.93% 0.1424 0.0473 0.1870 0.1198 4 4 16.50% 13.82% Sweden 12 2.2063 2.0486 4.38% 5.00% 0.1778 0.1619 0.1999 0.1838 3.5 3.5 6.36% 3.29% Thailand 66 2.7909 2.3148 10.21% 9.86% 0.0490 0.0234 0.0691 0.0409 4 4 12.59% 4.49% Turkey 42 1.4009 1.3898 3.97% 5.63% 0.3007 0.2906 0.3255 0.3098 3 3 6.77% 1.74% Taiwan 72 1.8221 1.3617 5.40% 3.58% 0.0403 0.0087 0.0516 0.0144 3 3 10.52% 10.01% Total 1164 1.9184 1.3549 4.41% 4.37% 0.2156 0.1870 0.2421 0.2190 3.5 3.5 8.27% 6.80%

Common law countries

Australia 24 1.1194 1.1205 4.08% 4.44% 0.0333 0.0254 0.0593 0.0463 4 4 4.40% 2.73% Canada 6 4.0663 4.0538 9.69% 9.11% 0.1173 0.1156 0.1182 0.1164 4 4 1.83% 1.62% United Kingdom 48 2.0930 2.0689 5.98% 6.05% 0.0171 0.3473 0.0281 0.4026 5 5 5.67% 4.00% Hong Kong 18 2.4189 1.1577 5.37% 5.08% 0.2171 0.2266 0.2840 0.2325 5 5 15.42% 13.46% Ireland 12 3.3308 3.2012 6.37% 6.93% 0.1380 0.0193 0.1464 0.0321 5 5 8.76% 5.94% Israel 30 1.6782 1.3192 3.41% 3.53% 0.2771 0.2809 0.3054 0.4142 4 4 5.98% 5.80% India 168 2.7143 0.9752 2.83% 1.33% 0.0853 0.0221 0.0977 0.0396 5 5 4.84% 2.11% Sri Lanka 18 8.2991 2.3778 15.41% 11.08% 0.4770 0.4201 0.4850 0.4319 4 4 5.06% 5.31% Malaysia 48 4.8466 1.0782 10.52% 6.32% 0.2494 0.2508 0.2659 0.2546 5 5 14.56% 6.61% Nigeria 12 10.4087 9.9575 17.11% 19.01% 0.2631 0.3540 0.2639 0.3551 4 4 9.32% 8.57% New Zealand 6 6.3706 7.3173 -1.69% 1.25% 0.0860 0.0697 0.1498 0.1439 4 4 25.78% 21.83% Pakistan 12 5.7791 1.4282 16.23% 17.85% 0.4212 0.4212 0.4252 0.4261 4 4 4.83% 1.44% Singapore 66 1.5200 1.0767 4.66% 4.31% 0.1195 0.0480 0.1513 0.1078 5 5 14.99% 15.40% United States 60 3.5598 2.5458 7.13% 7.21% 0.0805 0.0193 0.0878 0.0298 5 5 5.56% 3.44% South Africa 36 2.6794 2.7483 7.42% 6.76% 0.0464 0.0202 0.0788 0.0346 5 5 6.14% 5.70% Zimbabwe 6 0.9966 0.0228 9.40% 10.82% 0.0151 0.0065 0.0316 0.0070 4 4 7.45% 7.47% Total 570 3.8676 2.6531 7.74% 7.57% 0.1652 0.1654 0.1861 0.1922 4.5 4.5 8.79% 6.96%

(18)

18

4.1.1 Correlation matrix

The correlation matrix of all variables is presented in Appendix 1. This table indicates that there is no large correlation between the control variables. High correlation exists between TOPCON1 and TOPCON 3. These variables measure the largest shareholder and the three largest shareholders. Moreover, there is no large correlation between the independent variable and the control variables, or among the control variables.

4.2 Regression

The following section indicates the results of the regressions. We present two regressions in this thesis. The first regression analyzes the relationship between Tobin’s q and ownership concentration. The second regression analyzes presents the relationship between return on assets and ownership concentration. Both analyses use panel data. There are some significant different results between both regressions. However, these will be elaborated in the following subsections.

4.2.1 Regression ownership concentration

(19)

(20)

20 Figure 2. Regression Pooled OLS Pooled OLS

Quadratic

term Pooled OLS Fixed effects Random effects

TOBIN'S Q (1) (2) (2a) (3) (4) (5)

Variable Coefficient Coefficient Coefficient Coefficient Coefficient Coefficient

C -0.03311 -0.9847 -0.692 -1.2159 -1.9301 0.7325 (1.9378) (1.7329) (1.0165) (1.7167) (13.7763) (2.1062) CA 3.4598 3.6236* 3.8461*** 3.3619 1.8217 1.0136 (2.1618) (2.1286) (1.2303) (2.0610) (1.1456) (1.3879) SIZE 0.1420 0.1847 0.1727** 0.1762 0.4192*** 0.1289 (0.1374) (0.1281) (0.0688) (0.1295) (0.1580) (0.1266) R&D -13.7226 -7.9781 -10.1022 -8.1053 -0.0340 -8.5471 (9.4008) (8.2226) (8.7414) (8.2924) (8.9951) (8.5420) LEV -0.01337 -0.1912 -0.0671 -0.4684 0.2418 0.4780 (1.0052) (1.0038) (0.5793) (1.0067) (0.8510) (0.8183) SG 0.0018 0.0043 0.0038 0.0039 -0.0158 0.0251 (0.0072) (0.0066) (0.0253) (0.0062) (0.0193) (0.0174) EG 0.4156 0.3746 0.3812** 0.3589 0.2188 0.1773* (0.3347) (0.3171) (0.1585) (0.3096) (0.1194)* (0.1075) TOPCON1 2.8782** 3.5915*** -3.7685 0.0504 -6.5753** (1.3291) (0.7164) (2.9321) (5.7054) (3.1553) SHR 0.0954 -0.4793 -0.0803 (0.1481) (3.2033) (0.2986) TOPCON*SHR 1.9001* 0.2368 2.0222** (0.9741) (1.4239) (0.7981) Adjusted R-squared 0.0082 0.0251 0.0219 0.0391 0.5416 0.0045 Observations 1734 1734 1734 1734 1734 1734

The dependent variable is the Tobin’s Q formula. C is standing for the constant, the other variable are: ownership concentration (TOPCON), shareholder protection (SHR), interaction variable (TOPCON*SHR), cash to assets ratio (CA), size, research and development (R&D), leverage of the firm (LEV), sales growth (SG), and equity growth (EG). The meanings of ***, **, * indicate the significant level at 1*%, 5%, and 10%.

(21)

21 level. Research and development still have a negative effect on the dependent variable, but the coefficient is small compared to the other models. The sign of leverage is also changing from negative to positive, this indicates that the more debt the companies have, the higher the Tobin’s Q ratio is. This is in line with the Modigliani and Miller proposition 2, that argues that 100% leverage companies have the highest firm value (Hillier et al., 2016). Further, the sign of the ownership concentration variable is changing from negative (-3.7685) to positive (0.0504), also the coefficient of shareholder protection is changing from positive (0.0954) to negative (-0.4793). The sign of the interaction variable remains the same, but the coefficient is decreasing. All findings are insignificant. In column 5, we included random effects to the model. The random effects model can estimate the influence of variables, which do not vary over time. The fixed effects model throws away information (Brooks, 2014).The results of column 5 are similar to the results of column 3, but the sign of the constant is positive(0.1177), and the control variable equity growth is significant at the 10% level. Further, the ownership concentration variable has a strong significant negative effect (-6.5753) on the Tobin’s Q. Shareholder rights has a small negative effect on the dependent variable. The interaction variable is significant at the 5% level. To test which model is preferred we have done a redundant fixed effect test and a Hausman test. Based on the results of these tests the fixed effect model is preferred.

(22)

22 the 1% level. The results of the independent variable, shareholder protection, are similar, but only the sign of the coefficient of the interaction variable changes to negative (-0.0008).

Figure 3. Regression Pooled OLS Pooled OLS

Quadratic

term Pooled OLS Fixed effects Random effects

ROA (1) (2) (2a) (3) (4) (5)

Variable Coefficient Coefficient Coefficient Coefficient Coefficient Coefficient

C 0.0976*** 0.0831*** 0.0892*** 0.0875*** 0.2986 0.1177*** (0.0253) (0.0230) (0.0143) (0.0258) (0.1969) (0.0259) CA 0.0771* 0.0796* 0.0821*** 0.0771* 0.0327 0.0480*** (0.0430) (0.0427) (0.0173) (0.0420) (0.0208) (0.0184) SIZE 0.0003 0.0009 0.0006 0.0009 0.0004 0.0001 (0.0018) (0.0016) (0.0009) (0.0017) (0.0023) (0.0014) R&D -1.0900*** -1.0024*** -1.0433*** -0.0996*** -0.9716*** -0.9933*** (0.1586) (0.1713) (0.1229) (0.1719) (0.1285) (0.1184) LEV -0.1196*** -0.1223*** -0.1203*** -0.1246*** -0.1808*** -01563*** (0.0143) (0.0142) (0.0081) (0.0137) (0.0122) (0.0099) SG -0.0001 -0.0001 -0.0001 -0.0001 0.0001 0.0001 (0.0002) (0.0002) (0.0004) (0.0002) (0.0003) (0.0003) EG 0.0100 0.0094 0.0096*** 0.0092 0.0081*** 0.0087*** (0.0099) (0.0095) (0.0022) (0.0094) (0.0017) (0.0017) TOPCON 0.0439** 0.0462*** -0.0329 0.0099 -0.0271 (0.0186) (0.0100) (0.0422) (0.0815) (0.0443) SHR -0.0008 -0.0411 -0.0006 (0.0033) (0.0458) (0.0040) TOPCON*SHR 0.0214 -0.0008 0.0146 (0.0140) -0.0203 (0.0112) Adjusted R-squared 0.1805 0.1971 0.1899 0.2019 0.6076 0.1865 Observations 1734 1734 1734 1734 1734 1734

Return on assets (ROA) is the dependent variable. C is standing for the constant, the other variables are: ownership concentration (TOPCON), cash to assets ratio (CA), size, research and development (R&D), leverage of the firm (DA), sales growth (SG), equity growth (EG), ), shareholder protection (SHR), and interaction variable (TOPCON*SHR),. The meanings of ***, **, * indicate the significant level at 1%, 5%, and 10%.

(23)

23 that the equity companies have a higher return on assets than debt financed companies. The sign of the independent variable, shareholder protection, and the interaction variable, are similar to figure 2, but the results are insignificant. To test which model is preferred we have done a redundant fixed effect test and a Hausman test. Based on the results of these tests the fixed effect model is preferred.

Under OLS specifications, ownership concentration has a positive effect on company performance, but including an interaction variable in the model the coefficient is changing from positive to negative. The results under the pooled OLS and random effects show that shareholder rights have a positive effect on the dependent variable. When investors purchase shares of a company, they obtain some certain power or rights. Normally, these powers or rights are protected through the enforcement of law and regulation. These rights include voting for a director, to participate in shareholder meetings, to fire the director, and a call for extraordinary shareholder meetings (La Porta et al., 2000). This result suggests that shareholder rights are a substitute to ownership concentration to mitigate the agency problem. Therefore, investors could diversify their risk because, their rights are protected by law. These results also show that companies located in countries with strong shareholder rights have a higher company performance than those which are located in countries with low shareholder rights. This suggests when shareholders have more power and rights the managers’ work more on behalf of the company than for their own benefit. These findings are in line with the agency theory of Jensen & Meckling, (1976). The results of the preferred model (Fixed effects) show results that are in opposite. These results suggest that shareholder rights have a negative effect on the company performance, but the relationship with the dependent variable increases when the ownership concentration increases. This suggests that shareholders power inside the company have more influence on the company performance than shareholder rights do. Moreover, when the shareholder rights increases the relationship decreases, however, these results are only partially supported.

(24)

24 from the food industry. Including the interaction variable in the equation, this finding gets insignificant. For the second hypothesis, we find only weak empirical evidence that shareholder protection moderates the relationship between ownership concentration and company performance. Only the random effects model of figure 2 gives significant results at the 5% level. These results present that when using all information (random effect) in the model the results change significantly. The results are in line with the agency theory (Jensen & Meckling, 1976). Using return on assets as the dependent variable, the coefficient decreases and the results are insignificant. Therefore, we can only partially support hypotheses 2. Because there is some evidence, that shareholder rights moderates the relationship between ownership concentration and company performance. Moreover, the findings suggest that strong shareholder rights result in lower ownership concentration and that shareholder rights is a substitute to high ownership concentration to mitigate the agency problem. For both figures, we obtain similar findings when we run the regression with the top three shareholders and top twenty shareholders as independent variable.

4.3 Robustness check

(25)

25 figure 5 confirm the main results of the regression. Further, we have run the regression again for companies that are located in countries with weak shareholder rights (≤2.5) and companies that are located in countries with strong shareholder rights (≥4). First, we compare the results of the first model which each other (appendix 3). In this model, we use Tobin’s Q as dependent variable. The results indicate that companies located in countries with strong shareholder rights, ownership concentration and the interaction variable have a positive effect on the dependent variable. Using return on assets as the dependent variable, the sign of the interaction variable changes to negative. Next, we analyze the results of the companies that are located in countries with weak shareholder rights. The results indicate that ownership concentration has a positive effect on the Tobin’s Q, shareholder protection has a negative sign and the interaction variable has a positive effect on the dependent variable. All findings are significant. Using return on assets as the dependent variable (appendix 4), the results are similar but insignificant.

The results of appendix 3 indicate that ownership concentration has a strong positive effect (13.7908) on the company performance when they are located in countries with weak shareholder rights. These results are significant at the 1% level. The coefficient of ownership concentration decreases when shareholder rights increases. This result confirms our findings that shareholder protection has a positive effect on the relationship between ownership concentration and firm performance. Both robustness tests are in line with the main findings of this paper.

5. Conclusion and limitations

(26)

(27)

(28)

28

6. Reverences

Adams, R. B., Hermalin, B. E., & Weisbach, M. S. (2010). The role of boards of directors in corporate governace: A conceptual framework and survey. Journal of Economic Literature, 48(1), 58-107.

Ang, J. S., Cole, R. A., & Lin, J. W. (2000). Agency costs and ownership structure. Journal of Finance,

55(1), 81-106.

Basu, N., Paeglis, I., & Rahnamaei, M. (2016). Multiple blockholders, power, and firm value. Journal of

Banking & Finance(66), 66-78.

Berle, A., & Means, G. (1932). The modern Corporation and Private Property. New York: Macmillan Publishing Company.

Brooks, C. (2014). Introductory Econometrics for Finance. Cambridge: Cambridge University Press. Demsetz, H. (1983). The Stucture of Ownership and the Theory of the Firm. The Journal of Law and

Economics, XXVI, 375-390.

Demsetz, H., & Lehn, K. (1985). The structure of Corporate Ownership: Causes and Consequences.

Journal of Political Economy(93), 1155-1177.

Demsetz, H., & Villalonga, B. (2001). Ownership structure and corporate perfrmance. Journal of

Corporate Finance(7), 209-233.

Dittmar, A., Mahrt-Smith, J., & Servaes, H. (2003). International Corporate Governance and Corporate Cash Holdings. Journal of Financial and Quantitatve Analysis, 38(1), 111-133. Fama, E. F., & Jensen, M. C. (1983). Separation of Ownership and Control. The Journal of Law and

Economics, 26(2), 301-325.

Gozzi, J., Levine, R., & Schmukler, S. (2008). Internationalization and the evolution of corporate valuation. Journal of Financial Economics(88), 607-632.

Graff, M. (2008). Law and Finance: Common Law and Civil Law Countries Compared an Empirical Critigue. Economica, 75(297), 60-83.

Harford, J., Mansi, S., & Maxwell, W. (2008). Corporate governance and firms cash holdings in the US.

Journal of Financial Economics, 87, 535-555.

Hillier, H., Ross, S., Westerfield, R., Jaffe, J., & Jordan, B. (2016). Corporate Finance (3e ed.). Berkshire: McGraw-Hill Education.

Jensen, M. C., & Meckling, W. H. (1976). Theory of the Firm: Managerial Behavior, Angency Costs and Ownership Structure. Journal of Financial Economics(3), pp. 305-360.

Keynes, J. (1936). The General Theory of Employment, Interest, and Money. Harcourt Brace London. La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (1997). Legal Determinants of External Finance.

(29)

29 La Porta, R., Lopez-De-Silanes, F., & Shleifer, A. (1999). Corporate Owenership Around the World. The

Jorunal Of Finance, LIV(2), 471-517.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (2000). Investor protection and corporate governance. Journal of Financial Economics(58), 3-27.

Leech, D., & Leahy, J. (1991). Ownership Structure, Control Type Classification and the Performance of Large British Companies. The Economic Journal(101), 1418-1437.

Lindenberg, E. B., & Ross, S. A. (1981). Tobin's q ratio and industrial organization. Journal of Business, 1-32.

Millet-Reyes, B., & Zhoa, R. (2010). A Comparison Between One-Tier and Two-Ties Board Stuctures in France. Journal of International Financial Management and Accounting, 21(3), 279-310. Morbey, G. K. (1988). R&D: Its relationship to company performance. Journal of product and

innovation management, 5(3), 191-200.

Morck, R., Shleifer, A., & Vishny, R. (1988). Management Ownership and Market Valuation: an Empirical Analysis. Journal of Financial Economics(20), 293-315.

Myers, S., & Majluf, N. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13, 187-200. Opler, T., Pinkowitz, L., Stulz, R., & Williamson, R. (1999). The determinants and implications of

corporate cash holdings. Journal of Financial Economics(52), 3-46.

Ozkan, A., & Ozkan, N. (2004). Corporate cash holdings: An empirical investigation of UK companies.

Journal of Banking and Finance, 28(9), 2103-2134.

Pratt, J. W., & Zeckhauser, R. J. (1985). Principals and agents: The structure of business. Boston:

Harvard Business School Press(37).

Shleifer, A., & Vishney, R. W. (1997). A Survey of Corporate Governace. The Journal Of Finanace,

LII(2), 737-783.

Short, H. (1994). Ownership, Control, Financial Structure and the Performance of Firms. Journal of

Economic Surveys(8).

Spamann, H. (2010). The ´´Antidirector Rights Index`` Revisted. The review of Financial Studies, 23(2), 467-486.

Thomson, S., & Pedersen, T. (2000). Ownership Structure and Economic Performance in the Largest European Companies. Strategic management Journal(21), 689-705.

Van Veen, K., & Elbertsen, J. (2008). Governance Regimes and Nationality Diversity in Corporate Boards: A Comparative Study of Germany, the Netherlands and the United Kingdom. Journal

Compilation, 16(5), 386-399.

(30)

30 Williamson, O. E. (1964). The Economics of Discretionary Behaviour: Managerial Objectives in a

Theory of the Firm. Prentice-Hall.

Zeckhauser, R. J., & Pound, J. (1990). Are Large Shareholders Effective Monitors? An Investigation of

(31)

31

Figure.4

Tobin's Q ROA TOCON1 TOPCON3 TOPCON20SHR TOPCON1*SHR TOPCON3*SHR CA SIZE R&D DA SG EG

Tobin's Q 1 ROA 0.4035 1 TOCON1 0.1280 0.1247 1 TOPCON3 0.1221 0.1197 0.9822 1 TOPCON20 0.1211 0.1196 0.9788 0.9990 1 SHR 0.0445 -0.0387 -0.2704 -0.2796 -0.2834 1 TOPCON1*SHR 0.1618 0.1355 0.9427 0.9275 0.9236 -0.0662 1 TOPCON3*SHR 0.1550 0.1287 0.9185 0.9415 0.9399 -0.0612 0.9785 1 CA 0.0649 0.2072 -0.0553 -0.0475 -0.0516 0.0010 -0.0490 -0.0389 1 SIZE 0.0369 -0.0195 -0.1251 -0.1411 -0.1415 0.0976 -0.1147 -0.1307 -0.0791 1 R&D -0.0259 -0.1338 -0.1466 -0.1368 -0.1405 0.0904 -0.1425 -0.1274 0.1410 0.1125 1 DA -0.0253 -0.3574 0.0884 0.0863 0.0855 0.0604 0.1087 0.1062 -0.3731 -0.0250 -0.1546 1 SG 0.0081 0.0178 -0.0072 0.0050 0.0046 -0.0034 -0.0057 0.0071 0.0322 -0.0562 -0.0154 -0.0245 1 EG 0.0642 0.1219 0.0533 0.0583 0.0581 -0.0342 0.0570 0.0629 0.0177 -0.0445 -0.0555 -0.0340 0.1003 1

Appendix

Appendix 1. Correlation matrix

(32)

32

Appendix 2 logistic regression

Figure 5. Logistic regression

(1) Variable Coefficient C 3.5263*** (1.2867) CA 4.1075** (1.6879) SIZE -0.0401 (0.0772) R&D 10.2374 (19.5696) LEV 0.1786 (0.5485) SG 1.4966** (0.7192) EG 0.0965 (0.0961) TOPCON -2.9308*** (0.5728) Observations 0 66 Observations 1 1260 Observations 1326

(33)

33

Appendix 3 Regression (Tobin’s Q)

Figure 6. Regression

Shareholder rights ≤2.5 ≥4

TOBIN'S Q (1) (2)

Variable Coefficient Coefficient

C -10.9084*** -0.5043 (3.2669) (3.0930) CA 16.5889*** 5.0588 (4.3132) (2.7540) SIZE -0.1602 0.2536 (0.1548) (0.1762) R&D -10.4631 -8.6638 (22.2618) (8.2736) LEV 6.8874*** 0.6437 (1.4863) (1.1039) SG 0.2851 0.0016 (0.8170) (0.0077) EG 0.7987 0.4576 (0.6290) (0.3946) TOPCON 13.7908*** 4.5316 (3.3763) (25.4633) SHR 4.3816*** -0.4350 (1.0211) (0.9117) TOPCON*SHR -5.9686*** 0.3031 (1.7143) (5.6620) Adjusted R-squared 0.4136 0.0536 Observations 66 1260

(34)

34

Appendix 4 Regression (Return on assets)

Figure 7. Regression