INTRODUCTION
Problem statement
In the corporate world, the separation between owners (principals) and managers (agents) creates the agency problem, which can be mitigated through effective managerial ownership The optimal level of managerial ownership has garnered significant interest in modern corporate governance, as firms aim to maximize their value and owners' returns However, managers may prioritize their own welfare over that of the firm (Kuhnen & Zwiebel, 2006; Lambrecht & Myer, 2007) Three key approaches explain the impact of managerial ownership on firm performance: the agency approach, the managerial discretion approach, and the timing approach Empirical studies often reveal a non-linear relationship between managerial ownership and firm performance, with McConnell and Servaes (1990) identifying a reversed U-shaped relationship between Tobin’s Q and insider ownership levels, indicating that managerial ownership positively influences performance up to a threshold of 40 to 50 percent Morck, Shleifer, and Vishny (1988) further explored this relationship using sectional data.
Recent analyses of 500 Fortune firms revealed a W-shaped relationship between management ownership (MO) and Tobin’s Q However, researchers Kole (1995) and Himmelberg, Hubbard, and Palia (1999) criticized the use of cross-sectional data for inadequately capturing changes in a firm's environment and the endogeneity of MO Although various recent studies have sought to address the endogeneity issue of MO using different methodologies, limitations persist, leading to ongoing debates regarding the interpretation of the results.
Kole and Lehn (1997) explored the evolving governance structures and the driving forces behind these changes Similarly, Holderness, Kroszner, and Sheehan (1998) noted shifts in managerial ownership within US-listed firms over time Additionally, a new approach examined the relationship between changes in managerial ownership and firm performance, specifically through the lens of Tobin’s Q in American companies.
Fahlenbrach and Stulz (2008) The study attempted to examine the effect of the dynamic change in managerial ownership on the change in firm’s performance which can help eliminate endogeneity problem
Research indicates that changes in managerial ownership from the previous year significantly influence Tobin's Q in the current year, suggesting that the market fully absorbs information about managerial ownership By examining year-to-year variations in managerial ownership, we can gain insights into managers' decisions regarding stock transactions Additionally, an event study by McConnell, Servaes, and Lins (2008) revealed that managers do not buy shares solely to reach an optimal ownership level, as posited by agency theory Their findings also demonstrated that insiders fail to achieve abnormal returns, aligning with the timing approach These results echo the conclusions of Fahlenbrach and Stulz (2008), leading to the inference that managerial discretion theory partially explains changes in managerial ownership.
In Vietnam, several studies investigated the impact of the structure of ownership including the existence of state ownership, and foreign ownership on firm’s performance (Le
Research indicates an inverted U-shaped relationship between state ownership levels and Tobin's Q, with foreign ownership enhancing firm performance However, managerial ownership shows no statistically significant impact on accounting-based performance metrics like ROA and ROE Interestingly, a negative correlation exists between managerial ownership and market performance indicators, such as the P/E ratio, potentially due to the expropriation risks posed by block-holders The influence of managers extends beyond their shareholdings, as decisions may also be affected by family ownership structures prevalent in Vietnam, which have been linked to lower accounting returns Notably, managerial ownership encompasses both direct and indirect holdings, where direct ownership pertains to stocks directly owned by managers, while indirect ownership includes shares held by relatives or associated entities Previous studies often overlook the broader implications of indirect ownership, limiting their understanding of managerial power in financial decision-making.
In Vietnam, the State Securities Commission (SSC) and the Enterprise Law 2014 mandate that managerial transactions and those of related parties be disclosed This requirement aids in gathering data on changes in managerial ownership; however, the process is often time-consuming and limited to short time frames Overall, this study aims to enhance the understanding of how managerial ownership impacts firm performance and to explain the decisions managers make regarding the buying and selling of stocks.
Research objectives
This study aims to estimate the optimal level of managerial ownership and to analyze the relationship between managerial ownership and firm performance Additionally, it seeks to understand the behaviors of managers and their relatives regarding the buying and selling of company shares The research objectives are thus focused on these key areas.
(i) Estimating the optimal level for managerial ownership based on firm’s characteristics and market’s environment;
(ii) Observing the movement of actual managerial ownership level toward the optimal level;
Identifying the factors influencing stock buying and selling decisions made by the board of directors and their related parties is crucial for understanding market dynamics.
(iv) Examining the relationship between the change in managerial ownership and the change in firm’s performance (both aspects are considered as: market-based
(forward-looking) measurement and an accounting- based (backward-looking) measurement).
Research questions
To achieve the objectives of the study, the following research questions have been raised
(i) What are the determinants of optimal level of managerial ownership in Vietnamese listed firms?
(ii) Have managers adjusted their proportion of firm’s share toward the optimal level?
(iii) What factors have been considered to provide an impact on the managers’ decisions on purchasing and selling stocks?
(iv) Is there any relationship between the change in managerial ownership and the change in firm’s performance in both aspects accounting-based and market based measurement?
Contributions of the thesis
This study makes significant contributions by introducing a novel measurement of managerial ownership, encompassing both direct and indirect ownership The findings have resulted in the creation of a comprehensive database detailing managerial ownership among Vietnamese listed firms.
This study offers new empirical insights into the relationship between managerial ownership and firm performance by adopting a contemporary approach Unlike traditional studies that focus on the levels of managerial stock ownership and firm performance, this research examines how changes in managerial ownership—whether increases or decreases—impact the performance of the firm.
Research Scope
In general, the study investigates this relation by exploiting database which consisted of
285 non-financial firms listed on HOSE in the period from 2010 to 2015 Nevertheless, in several econometric regressions, some observations are excluded due to the inefficiently matched data.
Structure of the thesis
In this chapter, the gap of previous researches, the objectives of study, scope of study, the motivation, and contributions of this study are presented.
LITERATURE REVIEW
The theoretical background of managerial ownership and firm’s performance
Agency theory, as articulated by Jensen and Meckling (1976), posits that ownership structure significantly influences agency costs, highlighting the conflict between managers (agents) and shareholders (principals) in firm management Managers, while accountable for their decisions, may prioritize personal benefits over maximizing shareholder value, leading to detrimental investment choices One proposed solution to mitigate this conflict is increasing managerial ownership, which can lower monitoring costs and align managers' interests with those of shareholders Zwiebel (1995) noted that owner monitoring cannot completely eliminate the risks posed by managerial decisions Beyer, Czarnitzki, and Kraft (2012) further suggested that higher levels of managerial ownership encourage managers to act more like owners, thereby reducing principal-agent divergence up to a certain threshold However, they also warned that when boards of directors gain excessive power, managers may prioritize their own interests over the firm's value maximization.
(2007) also advocated the combination of effects between innovation and entrenchment during the increasing managerial ownership level
In conclusion, an increase in managerial ownership leads to both alignment of interests and entrenchment effects, creating two opposing impacts on firm performance These effects arise from the separation between principals (owners) and agents (managers), as agents make decisions aimed at maximizing the welfare of the principals.
Jensen and Meckling (1976) posited that when managers own stocks, they are motivated to implement investment strategies that enhance the company's cash flow and minimize external payments This alignment of interests leads to a positive correlation between higher managerial ownership and improved firm performance.
Leland and Pyle (1977) suggested that managerial ownership serves as a signal of company quality, as insiders invest in shares to maximize their welfare while being risk-averse and cautious with their investment choices By increasing their shareholding, managers indicate a more valuable firm, thereby attracting outside investors Additionally, Stulz (1988) noted that higher managerial ownership reduces the likelihood of hostile takeovers.
A significant negative correlation exists between managerial ownership and a firm's profitability or value, particularly when managerial ownership is at elevated levels As managerial ownership increases, it becomes increasingly challenging for external shareholders to exert control over management, leading to decisions that may favor managers' personal interests over the overall well-being of the firm.
Morck, Shleifer, and Vishny (1988) along with Stulz (1988) posited that increased voting rights can lead to an entrenchment effect, where higher managerial ownership negatively impacts a firm's performance This occurs because it becomes challenging for external and minority shareholders to effectively monitor and control the company.
Ineffective management can lead to the inefficient acquisition of valuable information in the takeover market, as noted by Hirshleifer and Thakor (1994) Additionally, Fahlenbrach and Stulz (2009) argue that the costs associated with holding more shares rise due to decreased diversification in managers' portfolios, making them more inclined to retain additional stock when their compensation is proportional or higher.
Agency theory suggests a trade-off between the advantages and disadvantages of higher managerial ownership, indicating a nonlinear relationship with firm performance and the existence of an optimal ownership level McConnell and Servaes (1990) identified an inverted U-shaped relationship, where lower managerial ownership enhances incentive effects, while higher ownership leads to negative entrenchment effects Additionally, Larcker, Randall, and Itner (2003) noted that factors such as a firm's lifecycle stage, R&D expenditures, asset structure, and growth opportunities influence the optimal level of managerial ownership.
The effect of managerial ownership on firm’s performance could be demonstrated as figure1.1
Figure 1.1 The relationship between insider ownership and firm’s performance
INOWNS Convergences of Entrenchment Convergences of interests interests effect
Source: Iturralde , Maseda , and Arosa (2011)
Managerial discretion theory, introduced by Hambrick and Finkelstein (1987), defines the latitude that managers have in making strategic decisions, influenced by three key factors Firstly, the variance in the business environment affects managerial freedom; for instance, firms with higher R&D and advertising expenditures often signal greater managerial discretion Secondly, organizational characteristics, such as available resources and inertial forces, shape managerial actions, where limited financial resources restrict strategic choices and strong corporate cultures impose constraints, especially in larger organizations Lastly, the characteristics of managers themselves play a crucial role, as Finkelstein and Boyd (1998) found a correlation between managerial compensation and discretion, indicating that higher firm performance is associated with increased managerial pay Further developments by Stulz (1990) and Zwiebel (1996) suggest that managers prioritize their utility over firm value, leading to endogenous managerial ownership This behavior can result in a negative relationship between managerial ownership and firm performance, as managers may exploit their discretion for personal gain, as noted by Fama (1980), Jensen (1986), and Brush et al (2000).
According to Fahlenbrach and Stulz (2008), three key motivations are explored when managers held stock under managerial discretion approach
In times of constrained financial resources, particularly for start-ups, managerial ownership can provide a more cost-effective source of capital than external financing options Firms facing higher information asymmetries often struggle to secure external resources, such as bank loans or equity from outside investors As a company matures, the costs associated with issuing shares or obtaining bank loans decrease, diminishing the financing motivation for managers Consequently, it is anticipated that the percentage of stock held by managers will decline over time.
Managerial ownership can align the interests of managers and minority shareholders, provided that their stock holdings do not exceed a certain threshold The bonding motivation is particularly significant when managers have lower reputations or possess greater managerial discretion However, this motivation diminishes in firms with a higher ratio of intangible assets, fewer growth opportunities, or when managers are more well-known This trend is observed in organizations that are stable and relatively mature.
When faced with threats to their control, boards of directors often increase their share ownership This trend is particularly noticeable during periods of poor business performance and when managerial capabilities are not fully acknowledged By acquiring more shares, managers aim to reassure owners of their commitment to improving the company's prospects and mitigating the risk of hostile takeovers.
The managerial discretion approach suggests that increased managerial ownership is often observed in younger or financially constrained firms, as well as in poorly performing companies where the board's skills are not publicly acknowledged This approach indicates that managers are likely to sell their shares when the firm performs well or when the market is more liquid By focusing on changes in managerial ownership and the firm's performance, rather than the absolute levels of these metrics, researchers can gain a clearer understanding of managerial actions related to buying and selling shares.
The concentrated timing approach posits that insiders possess greater operational knowledge of their enterprises than external investors, enabling them to achieve abnormal returns This suggests that managers may buy company stock during periods of strong performance, potentially indicating overvaluation, and vice versa (Jenter, 2005) According to market timing theory, managers can effectively outperform the market to secure exceptional returns While this theory aligns with the managerial discretion approach, they stem from different foundations Research by McConnell, Servaes, and Lin (2008) examined the relationship between changes in insider ownership and abnormal returns, finding that variations in insider ownership significantly influence a firm's performance.
Endogeneity of managerial ownership
Roberts and Whited (2013) identified that endogeneity arises from the correlation between the error term and an explanatory variable This issue can stem from omitted variables that correlate with both the error term and the independent variable, as well as from measurement errors in capturing proxies Furthermore, simultaneity, where two variables are mutually determined, also contributes to the problem of endogeneity.
Demsetz and Lehn (1985) posited that in a contracting environment, managerial ownership is an endogenous variable influenced by managers' decisions aimed at maximizing firm value Similarly, Jensen and Meckling (1976) argued that ownership structure, particularly managerial ownership, significantly impacts corporate performance, thereby raising questions about causality.
The endogeneity problem arises from the co-determinants of managerial ownership and firm performance, with monitoring technology playing a crucial role in addressing unobservable factors Specifically, firms with superior monitoring capabilities can achieve optimal managerial ownership levels that align the interests of managers and owners, leading to enhanced market value This is because potential investors recognize that less company resources are needed for oversight, thereby reducing costs If the quality of monitoring technology is not adequately accounted for, the observed negative correlation between managerial ownership and market-based performance may be misleading.
A key example of firm heterogeneity is the proportion of intangible assets, as companies with higher levels of intangible fixed assets necessitate greater managerial ownership to curb managerial discretion This relationship is evident when assessing market performance through Tobin's Q, where the market value of assets often exceeds their book value, particularly for intangible assets Consequently, the unmeasured ratio of intangible assets contributes to a misleading positive correlation between managerial ownership and Tobin's Q.
Himmelberg, Hubbard, and Palia (1999) developed an econometric model to address the endogeneity of managerial ownership, highlighting the diverse environments of firms In this model, xit represents observable characteristics, while uit signifies unobservable traits for firm i at year t, with the assumption that these unobservable characteristics remain constant over time.
mit is the level of managerial ownership;
According to optimal contract, the manager’s effort can be represented by the following equation:
The firm’s performance of firm i at year t (PERit) is the function of managers’ effort, observable and unobservable firm characteristics:
So, we can combine (2) and (3):
𝑃𝐸𝑅 𝑖,𝑡 = 𝜕𝜃 𝑚 𝑖,𝑡 + (𝜕𝛽 2 + 𝛽 3 ) 𝑥 𝑖,𝑡 + (𝜕𝛾 2 + 𝛾 3 )𝑢 𝑖 + 𝜕𝜖 𝑖,𝑡 + 𝜗 𝑖,𝑡 (3a) The short version of equation (3a)
To examine the link between managerial ownership and firm performance using equation (3b), it is essential that the error term 𝜏it remains uncorrelated with both managerial ownership (independent variable) and firm performance (dependent variable) However, since managerial ownership is influenced by unobservable characteristics, 𝜏it is likely to be correlated with managerial ownership (m), which may affect the consistency of the parameters in the analysis.
In term of econometrics, the result could be:
As a result, we cannot estimate equation (3b) by OLS because the coefficients are inconsistent and biased So, the research focused on the change instead of level of managerial ownership
2.3 The empirical evidences of relationship between managerial ownership and firm’s performance and limitations
2.3.1 The research in worldwide and the limitations 2.3.1.1 The exogenous managerial ownership
Numerous empirical studies have explored the nonlinear relationship between managerial ownership and firm performance using cross-sectional data While findings vary across different research, a general consensus indicates that the connection between managerial ownership and firm performance is not straightforward.
Morck et al (1988) investigated the link between managerial ownership and Tobin's Q using cross-sectional data from 500 stable Fortune firms Their findings indicated that a managerial ownership threshold ranging from 0% to 5% positively impacts Tobin's Q, a key market-based measure of firm performance.
Q would reduce if managerial ownership increases up to 25 percent and the repeatedly positive impact again in the case managers owned 25 percent excessively The nonlinear relationship is illustrated similarly Figure 2.1
A study by McConnell and Servaes (1990) analyzing 1,173 enterprises listed on NYSE and AMEX in 1976 and 1986 revealed a reversed U-shaped relationship between Tobin’s Q and insider ownership levels Their findings indicated that managerial ownership positively impacts firm performance, with an optimal threshold between 40% and 50% The research examined Tobin’s Q as the dependent variable alongside two control variables: the fraction of insider ownership and its square The significant positive coefficient of insider ownership, coupled with the negative coefficient of its square, confirmed the non-monotonic relationship observed.
Short and Keasey (1999) bolstered their argument with empirical evidence from the United Kingdom, using two performance proxies: the market-to-book ratio of total assets (VAL) and return on equity (ROE), both regressed against a cubic function of managerial ownership levels Their findings revealed that VAL's market-based measurements fluctuated in relation to managerial ownership, echoing Morck et al (1988) They identified distinct thresholds for managerial ownership, with a positive threshold at 12.99 percent, followed by a negative correlation until reaching 41.99 percent, after which the positive impact resumed Similarly, ROE exhibited a correlated graph with VAL, showing different turning points at 15.58 percent and 41.84 percent, respectively.
Kole (1995) studied 352 firms, contrasting with the 500 Fortune firms examined by Morck et al (1988), and identified an N-shaped relationship between managerial ownership and Tobin’s Q, with notable differences in the turning points The research revealed that the positive impact of managerial ownership on Tobin’s Q is more pronounced in smaller firms compared to larger ones.
Hermalin and Weisbach (1991) conducted a study on triennial data from 142 NYSE-listed firms across the years 1971, 1974, 1977, 1980, and 1983, revealing a complex W-shaped relationship between managerial ownership and Tobin’s Q Initially, a positive correlation exists with managerial ownership up to 1 percent, which then reverses as ownership increases from 1 percent to 5 percent Beyond 5 percent, the positive effect of managerial ownership becomes dominant, but this is ultimately overshadowed by a negative impact once ownership exceeds the 20 percent threshold.
In summary, research indicates that the link between managerial ownership or insider ownership and firm performance is non-monotonic These findings are based on the assumption that managerial ownership is exogenous and rely solely on cross-sectional data A key limitation of this research is the inability to adequately account for unobservable firm heterogeneity, which reflects changes in the firm's environment.
Demsetz and Lehn (1985) emphasized the significance of unobserved heterogeneity in a contracting environment In a study by Himmelberg, Hubbard, and Palia (1999), the relationship between managerial ownership and firm performance was analyzed using panel data from 600 randomly selected US firms between 1982 and 1992, addressing endogeneity issues Utilizing a fixed effect model, the study found no econometrically significant impact of insider ownership on firm value, suggesting that previous regression results may reflect a spurious relationship.
Numerous researchers have validated the concept of endogenous managerial ownership, leading to various econometric solutions to address this issue However, the interpretation of the estimated results, particularly regarding how changes in managerial ownership can influence firm value, has sparked considerable debate.
The corporate governance of Vietnamese listed firms
Emerging markets have prioritized the enhancement of corporate governance to safeguard investors and promote market transparency Listed companies operate under a dual structure of governance, comprising the General Meeting of Shareholders (GMS) and the Board of Management (BOM) The Law on Enterprise 2014, effective from 2015, introduced significant changes compared to the 2005 version, aimed at improving information transparency and management quality.
The Law of Enterprise 2014 in Vietnam has effectively reduced the gap between local regulations and international standards by lowering the minimum voting rights and quorum requirements It mandates that at least 20% of the board of directors must be independent members, enhancing oversight of enterprise operations Additionally, the new legislation emphasizes transparency and information disclosure, requiring CEOs, chairpersons, and other management personnel to disclose their ownership stakes and those of related parties in any firms where their total shareholding exceeds 10%.
Figure 1.2 The management structure of Shareholding Company
Le and Walker (2008) highlighted that Vietnam's capital markets are still in the early stages of development, emphasizing that listed companies must comply with the Law of Securities 2006, which has led to issues with flexibility and accountability The Law of Enterprise 2014 introduced reforms aimed at enhancing administrative procedures and facilitating mergers and acquisitions (M&A), such as permitting multiple legal representatives and removing restrictions on the types of firms that can merge.
Figure 1.3 The internal governance structure of a listed company
This diagram illustrates the relationships within a listed company, highlighting the powers and responsibilities of various parties involved Solid lines indicate the rights related to appointment and dismissal, while dashed lines represent the monitoring functions that are essential for effective governance.
According to the mechanism, to separate the control and supervisory of firm operation, requirement of one third of members of Board of Management must be non-executive independent member
Secretary Sub-Committees General Director
RESEARCH METHODOLOGY AND DATA
Data sources
In Vietnam, firms are listed on two stock exchanges: HNX (Hanoi Stock Exchange) and HOSE (Ho Chi Minh Stock Exchange) This study focuses exclusively on firms listed on HOSE to ensure a homogeneous sample, excluding securities and financial companies due to their different capitalization rules and regulations Data was collected from various sources, including annual reports and firm prospectuses, with particular attention to the ownership rates of board members and related parties The research faced challenges due to incomplete financial information, resulting in unbalanced panel data By the end of 2015, there were 341 companies listed on HOSE; however, after removing 28 financial firms and 28 companies that violated security regulations, the final sample consisted of 285 companies from 2010 to 2015.
This study examines managerial ownership (MO), which encompasses both direct and indirect ownership as indicated by the percentage of shares held by board members and related parties, in accordance with Article 28 of Circular 52/2012-BTC and its amendment by Circular 155/2015-BTC Indirect ownership refers to shares held by board members representing the organization (Neely, Gregory & Platts, 1995) Koufopoulous, Zoumbos, and Argyropoulous (2008) argue that performance in management can be assessed through quantification and accounting, emphasizing the need for firms to effectively manage their operational processes to meet objectives Numerous researchers (Demirbag & Zaim, 2006; Gedennes & Sharma, 2002) agree on the importance of measuring business performance to evaluate the success of resource management over time, leading to the development of performance measurement concepts that facilitate comparisons of a firm's achievements across different periods.
Effective corporate governance significantly influences a firm's performance, as it establishes a framework that enhances operational effectiveness (Ehikioya, 2009) By measuring firm performance, valuable insights are communicated to external stakeholders, reflecting the organization's operational efficiency These performance metrics facilitate the quantification of complex concepts, making evaluation more straightforward and accessible (Lebas, 1995).
Measurement variables
3.2.1 Definition and measurements of firm’s performance 3.2.1.1 Accounting – based measurements
Accounting-based measurements primarily focus on profitability, serving as benchmarks for competitor comparisons and risk assessment However, these indicators can be affected by future expenditure estimates, including depreciation and provisions Additionally, their effectiveness is constrained by accounting conventions and the methods used to record asset values (Kapopoulos & Lazaretou, 2009).
Return on Assets (ROA) is a key accounting metric that measures a company's profitability by comparing its earnings after tax to the book value of total assets, reflecting how effectively a firm generates profit from its assets This ratio is widely used to assess firm performance (Hu & Zhou, 2008; Mehran, 1995; Demsetz & Lehn, 1985; Vo & Nguyen, 2014) According to Ibrahim and AbdulSamad (2011), ROA indicates how efficiently a company utilizes its assets to provide economic benefits to shareholders, irrespective of its capital structure Notably, a statistical analysis by Macrothink Institute reveals that ROA is the most frequently employed measure of firm performance among scholars.
Market-based measurements capture investors' expectations regarding a firm's future profit-generating capabilities, emphasizing long-term profitability In contrast, accounting-based indicators primarily assess short-term profits, highlighting the distinction between immediate financial performance and future growth potential (Bozec, Dia, & Bozec, 2010).
Tobin's Q is a key metric used to assess a firm's market performance, defined as the ratio of the market value of a firm to the replacement value of its assets, as introduced by Tobin and Brainard in 1969 This ratio is calculated by dividing the total market value of common and preferred stock, along with total liabilities, by the book value of total assets, representing the replacement cost of production capacity However, due to the inefficiencies in the Vietnamese debt market, the market value of liabilities is often unattainable, which complicates the calculation of Tobin's Q.
𝑇𝑜𝑏𝑖𝑛 ′ 𝑠 𝑄 = Market value of common and preferred stock+book value of liabilities
Book value of total assets
Tobin’s Q is a widely recognized metric for assessing firm performance, utilized by numerous researchers including McConnell, Servaes, and Lins (2008), Kole (1997), Fahlenbrach and Stulz (2009), Coles, Lemmon, and Meschke (2012), Firdaus and Kusumastuti (2013), and Hoang, Nguyen, and Hu (2016).
3.2.2 Definition and measurement of managerial ownership
Managerial ownership refers to the proportion of stock held by block holders and insiders, including managers and firm officers (Holderness, 2008) It can also be defined as the stock held by board members, excluding stock options (Cho, 1998) Empirical studies, such as those by Agrawal and Knoeber (1996), have used alternative proxies, like the fraction of stock owned by the CEO Additionally, another measurement includes the total percentage of stock held by board directors and their families (Short & Keasey, 1996) However, these measurements primarily capture direct ownership, necessitating further clarification of managerial ownership patterns as outlined by the Securities and Exchange Act (SEC).
1934 required the public firms register the percentage of stock held by board of directors Moreover, SEC’s definition of managerial ownership included indirect and direct ownership
Direct ownership refers to managers having the title, voting rights, and financial benefits from stocks, such as dividends or capital gains In contrast, indirect ownership allows managers to influence the firm through voting rights without actually holding the stock title or receiving financial benefits This type of ownership often includes shares held by family members and their representative organizations.
This study followed the definition developed by Holderness, Kroszner and Sheehan
(1998) which measured total indirect and direct ownership of members of board of directors excluding chief accountant as managerial ownership level.
Research methodology
This article outlines a comprehensive analysis of managerial ownership through three distinct parts Firstly, it examines the determinants of the optimal level of managerial ownership and assesses whether the actual levels are aligning with this optimal threshold Secondly, it employs Probit regression to explore the factors influencing stock transactions among related parties and managerial decisions Lastly, it investigates the correlation between changes in managerial ownership and firm performance using POLS, FE, and RE methodologies, with data analysis conducted using Stata 12 software The study begins with descriptive statistics and correlation analysis, including VIF calculations, to summarize data across years and industries To further elucidate the specific impacts of these factors, multivariate regressions are applied, forming the foundation of the econometric model presented.
Yit is the dependent variable;
POLS regression assumes that the error term and independent variables are uncorrelated, leading to unbiased and consistent estimators However, if unobserved individual effects are present, Random Effects (RE) or Fixed Effects (FE) models may be more suitable than POLS To determine the most efficient model, the F-test is used to compare POLS and FE, while the Breusch-Pagan-Lagrange Multiplier (LM) test evaluates whether RE or POLS is preferable, with the null hypothesis indicating that POLS is superior Additionally, the Hausman test is employed to assess the efficiency of FE versus RE An overview of these tests is summarized in the accompanying table.
Table 3.1 Tests are utilized to find the appropriate model
Breusch –Pagan test (RE vs POLS)
Hausman test (RE vs FE)
H0: POLS is not rejected H0: POLS is not rejected POLS
H0: POLS is not rejected H0: POLS is rejected RE model
H0: POLS is rejected H0: POLS is not rejected FE model
H0: POLS is rejected H0: POLS is rejected H0: RE is reject FE model
H0: POLS is rejected H0: POLS is rejected H0: RE is not reject RE model
To obtain the best linear unbiased estimator (BLUE), various diagnostics are performed, including the Wald test for group-wise heteroskedasticity and the Wooldridge test for autocorrelation By utilizing robust standard errors, we achieve more efficient estimators While the magnitude of the estimators remains constant, the standard error values are reduced, enhancing the reliability of the results.
RESULTS AND DISCUSSIONS
Data description
Our study analyzes 1,554 observations from 285 listed firms on the HOSE stock exchange between 2010 and 2015 These firms are categorized into 19 distinct industries, including real estate, rubber, information technology, oil and gas, tourism, building materials, construction, health and chemistry, education, mining and quarrying, energy and electricity, plastic packaging, manufacturing, seafood, transportation and warehousing, steel, food and agriculture, commerce, and various other industries.
The average managerial ownership (MO) across all observations is approximately 20%, slightly below the 22.4% average of US firms, but significantly higher than British firms at 16.7% and China's civilian-run firms at 9.31% The MO level varies widely, from 0% to nearly 99% in family-owned companies Notably, there was a decline in MO from 25% in 2010 to around 12% by 2015, with industry variations evident—41% in the seafood sector versus only 8% in education Firm performance is assessed through accounting-based metrics like ROA, averaging around 6% with fluctuations from -74% to 77% during the period Additionally, Tobin’s Q, a market-based measure, typically exceeds 1, indicating market values are higher than book values, particularly in the real estate, rubber, and mining sectors where board ownership significantly surpasses average levels.
Table 4.1 Summary statistics: the firm’s characteristics of 285 firms listed on
Variable Obs Mean Std Dev Min Max
1 Some observations are excluded of sample due to inefficiently matching data
Fiscal year Number of firms MO level ROA Tobin Q Positive change Negative change DIV
Mean Std Mean Std Mean Std Mean Std Mean Std Mean Std
Industry Number of firms MO level ROA Tobin Q Positive change
Mean Std Mean Std Mean Std Mean Std Mean Std Mean Std
Information technology 9 0.271 0.344 0.058 0.068 1.102 0.674 0.030 0.036 -0.059 0.072 1081.633 1117.034 Oil and gas 9 0.103 0.085 0.042 0.106 1.357 1.065 0.009 0.023 -0.027 0.040 1096.509 1294.765 Tourism 4 0.327 0.342 0.110 0.184 1.801 1.083 0.025 0.036 -0.166 0.188 2085.762 2968.341 Building and materials 9 0.263 0.406 0.061 0.087 1.853 2.673 0.134 0.237 -0.074 0.141 1812.500 1053.090 Education 1 0.077 0.034 0.070 0.031 0.938 0.168 0.018 0.015 -0.018 0.013 350.000 364.005 Mining and quarrying of mineral 9 0.089 0.140 0.075 0.085 4.237 20.241 0.011 0.017 -0.080 0.115 1595.556 1539.943 Energy and electricity 12 0.191 0.244 0.099 0.099 1.224 0.505 0.089 0.138 -0.058 0.124 1671.647 1616.282 Plastic packing 9 0.384 0.371 0.098 0.088 1.254 0.843 0.074 0.148 -0.080 0.061 1247.892 1005.053 Manufacturing business 21 0.183 0.233 0.069 0.063 1.167 0.611 0.062 0.117 -0.093 0.140 1225.899 1028.398
Food and nurture 14 0.101 0.153 0.080 0.100 1.781 1.627 0.106 0.147 -0.061 0.080 1692.857 2210.057 Commerce 13 0.158 0.208 0.056 0.054 0.972 0.373 0.032 0.040 -0.061 0.113 1401.389 1439.376 Sea-food 16 0.413 0.337 0.055 0.120 0.985 0.385 0.144 0.218 -0.171 0.159 1446.970 1661.033 Transportation and warehousing 22 0.118 0.159 0.066 0.078 1.074 0.507 0.031 0.054 -0.046 0.092 1135.231 1066.320 Building and materials 13 0.141 0.156 0.047 0.074 1.028 0.471 0.060 0.157 -0.047 0.083 1032.418 1441.485 Construction 22 0.192 0.329 0.037 0.097 0.976 0.416 0.032 0.048 -0.054 0.088 917.959 876.612 Other industry 43 0.153 0.237 0.060 0.076 1.167 0.910 0.092 0.161 -0.091 0.156 1207.653 1282.901
This table illustrates the relationship between managerial ownership and various firm attributes, where "m" signifies the total percentage of stock owned directly and indirectly by all managers Key ratios include KTA, which measures the ratio of tangible assets to total assets; SIGMA, representing the idiosyncratic stock price risk; and YS, indicating the ratio of operating income to sales as a proxy for market power Additionally, RDTA reflects the ratio of R&D spending to total assets, while RDUM serves as a dummy variable indicating whether firms report R&D expenditures Finally, CAPEXTA denotes the ratio of capital expenditure to total assets.
Table 4.3 outlines the characteristics of enterprises across various industries, revealing that the average change in managerial ownership levels—both positive and negative—averages around 8 percent in absolute value This significant shift contrasts with Zhou's (2001) findings and may be attributed to adjustments in the percentage of shares held by managers and their related parties.
The pairwise correlation matrix reveals the relationships among variables in regression (11) aimed at estimating the optimal managerial ownership level Variance inflation factors (VIF) are calculated to assess multicollinearity, with all VIF values remaining below 10, except for the correlation between Ln(S) and its square, as well as (KTA) and (KTA)² Most correlation coefficients are relatively low, with the highest at approximately 0.35 This matrix suggests that idiosyncratic stock price risk and the ratio of R&D expenditure negatively impact managerial ownership levels, while the market power proxy, represented by the ratio of operating income to sales, shows a positive relationship with managerial ownership levels.
The determinants and movement of managerial ownership
The analysis starts with the premise that managerial ownership is regarded as an exogenous variable Utilizing OLS regression, the study examines the nonlinear relationship between the level of managerial ownership and the performance of firms, with Return on Assets (ROA) and Tobin’s Q serving as key performance indicators.
The findings do not establish a clear relationship between managerial ownership and firm performance, as measured by both market-based and accounting-based metrics using OLS This outcome contradicts previous research results.
Vo and Nguyen (2013) question the rationality of their model based on OLS regression results Furthermore, Le (2013) suggested that managerial ownership and capital structure may be endogenous factors influencing a firm's performance.
The endogenous test conducted using the ivreg2 command in Stata 12 indicates that managerial ownership is endogenous, as detailed in the appendix The tangible asset to total asset ratio serves as the instrument, likely correlated with managerial ownership Validity tests, including the underidentification test for instrument relevance and the Sargan-Hansen test for overidentifying restrictions, were performed using the ivreg2 command with the endog option The underidentification test confirms that managerial ownership is significantly endogenous at a 5 percent level The Sargan statistic, at approximately 7 percent, suggests that we cannot reject the null hypothesis of instrument validity at a 5 percent significance level However, at a 10 percent significance level, the rejection of the null hypothesis indicates potential instrument invalidity Ultimately, the endog option reveals that the instrument, KTA, can be treated as an exogenous variable.
Table 4.5 The relationship between level of managerial ownership and firm’s performance
The study examines the impact of various factors on firm performance, measured through Return on Assets (ROA) and Tobin’s Q Key independent variables include the level of Market Orientation (MO) and its square, alongside control variables such as firm size (represented by the natural logarithm of total assets and its square), leverage (long-term debt to total assets ratio), R&D expenditure (as a ratio of total assets), capital expenditure (also as a ratio of total assets), and risk (assessed through the variation of residuals from the CAPM model) The results of the pooled ordinary least squares (POLS) regression analysis will be presented, with standard errors indicated in parentheses, and significance levels denoted by ***, **, and * for 1%, 5%, and 10%, respectively.
4.2.1 The determinants of managerial ownership
Using the model developed by Himmelberg, Hubbard, and Palia (1999), we estimated the optimal level of managerial ownership for each firm, applying a transformed log of managerial ownership to enhance data accuracy Our analysis employed POLS, RE, and FE methods to assess the determinants influencing managerial ownership levels, with various tests conducted to identify the most suitable model Ultimately, the FE model was recommended, and the regression details are provided in Table 4.6.
The regression results reveal a slight divergence from the initial correlation matrix analysis, indicating a nonlinear inverse-U shaped relationship between firm size and optimal managerial ownership levels This suggests that larger firms face increased operational complexities and organizational structures, leading to higher monitoring and agency costs However, these larger firms can also benefit from economies of scale regarding these costs This finding aligns with the arguments presented by Himmelberg, Hubbard, and Palia (1999) and Do and Wu.
In 2014, a study revealed a nonlinear relationship between the ratio of tangible assets to total assets (KTA) and the level of managerial ownership, as determined by POLS regression However, the significance of these coefficients increased to 10 percent when a fixed effects model was utilized, following the application of robust standard errors.
The FE model reveals a significantly negative relationship between the ratio of tangible assets to total assets, indicating that firms with a higher proportion of tangible assets tend to have lower levels of managerial ownership Conversely, the analysis of R&D expenditure, measured by the ratio of R&D expenditure to total assets (RDTA) and a dummy variable (RDUM), shows a significantly positive correlation, suggesting that firms that invest more in R&D are likely to experience higher managerial ownership levels Gertler and Hubbard (1988) further emphasize that a higher R&D expenditure ratio highlights the critical importance of "soft capital - technology," which may increase managerial discretion and vulnerability.
The RDUM variable is incorporated into the regression analysis to ensure all observations are included, even from firms that do not separately report their R&D expenditures, thereby enhancing the sample size and reducing bias RDUM is assigned a value of 1 for firms that disclose R&D spending and 0 for those that do not A negative coefficient for RDUM suggests that transparent reporting of R&D expenditures reduces agency problems and managerial discretion, which correlates with a decrease in managerial ownership levels.
Table 4.6 The determinants of managerial ownership level
This table presents the estimation results for the optimal level of managerial ownership, using ln(1-m) as the dependent variable, where m represents the proportion of managerial ownership The regression analysis incorporates industry and year fixed effects, detailed in the appendix To ensure efficient estimators, robust standard errors were utilized, with standard errors indicated in parentheses Significance levels are denoted by ***, **, and * for 1%, 5%, and 10%, respectively.
VARIABLE POLS RE FE FE ROBUST
4.2.2 The movement of actual managerial ownership
This study examines the determinants of managerial ownership (MO) to estimate the optimal level of MO for each case It calculates the gap between actual managerial ownership and the optimal level, investigating how managerial ownership levels shift toward this optimal point, as indicated by the results in Table 4.6 To address this question, a simple regression analysis is employed, grounded in agency theory, which supports the establishment of the optimal managerial ownership level and the dynamics of movement toward it.
Table 4.7 The movement actual managerial ownership level toward to estimated optimal level
The analysis calculates the gaps (surplus/deficit) by comparing actual levels to optimal Managerial Ownership (MO) levels, derived from three estimation methods: (1) fixed effects with robust standard errors, (2) POLS controlling for year dummies, and (3) POLS controlling for industry dummies The actual change reflects the percentage variation in shares held by all managers within the year Additional results for optimal MO levels are available in the appendix Fixed effects and robust standard errors were utilized across three regressions, with tests for heteroscedasticity and the Hausman test conducted to ensure appropriate estimations Standard errors are provided in parentheses, with significance levels indicated by ***, **, and * for 1%, 5%, and 10%, respectively.
Research by McConnell, Servaes, and Lins (2008) suggests that actual managerial ownership is influenced by changes in a firm's characteristics, leading to the implementation of regressions comparing actual changes to ownership gaps The findings reveal that all coefficients in three separate regressions are negative and significant at the 1 percent level, indicating a rejection of the hypothesis that managerial ownership adjusts to an optimal level This implies that actual ownership changes often deviate from the optimal ownership levels.
Cheung and Wei (2006) investigated the difference of optimal and observed level managerial
R-squared 0.128 0.157 0.149 ownership and their explanation was the survivals of ownership adjustment cost Additionally, the authors identified that the existing theories inefficiently explain this issue So, this phenomenon induces the question what are the determinants for the large change (increase / decrease) in managerial ownership.
The explanation of the large change (decrease or increase)
The sample of 1409 observations consists of 464 large decrease, 201 large increase and
744 no large change In this study, 1 percent designated as threshold for large change because according to Article 15 of regulation of disclosure information in HOSE issued with Decision
According to Decision No 7/2013 issued by the Ho Chi Minh City Department of Education and Training, insiders are required to disclose their anticipated transaction volumes and outcomes Furthermore, Article 13 stipulates that large shareholders must announce trading information when their transactions result in a change of at least 1 percent, or when their ownership reaches nearly 5 percent.
The Mann-Whitney-Wilcoxon rank-sum test, conducted using the ranksum command in Stata 12, assesses the equality of distribution between independent samples This analysis provides insights into the characteristics of firms and market conditions influencing changes in managerial ownership levels Evidence indicates notable differences in firm attributes among various groups, such as those with large drops versus no changes, large increases versus no changes, and large drops versus large increases Notably, firms experiencing significant changes exhibited higher concurrent and lagged managerial ownership levels compared to the unchanged group Initially, the large increase group's managerial ownership level was lower than that of the large drop group; however, after adjustments, the concurrent managerial ownership level surpassed that of the large drop group.
The pay-out policy reveals no significant differences in dividends across three groups Gertler and Hubbard (1988) suggested that firms with higher fixed investments typically exhibit lower levels of managerial ownership due to easier monitoring compared to intangible assets Conversely, firms with a higher ratio of R&D expenditure and lower tangible assets tend to see a significant increase in managerial ownership When considering firm size, no notable differences were observed between groups with no change and those with a large increase; however, a significant disparity exists between large increases and large decreases in firm size Larger enterprises benefit from external monitoring, such as by rating agencies, leading to reduced managerial ownership levels Overall, the impact of firm size on managerial ownership remains ambiguous.
The study examines the impact of market conditions, including concurrent and lagged returns, as well as stock market turnover A significant decline in stock prices may correlate with improved market liquidity, indicated by a higher turnover of the Vn-Index Conversely, contemporaneous stock returns show minimal significance across three pairwise groups Notably, managers are inclined to purchase stocks following poor performance and sell them when firms and the overall market are thriving This behavior suggests a stability in managerial ownership levels during periods of steady market and firm operations.
Table 4.8 Summary statistics of data by data source
The table displays the average values of key variables analyzed in the regressions, categorizing all observations into three groups: large increase, no change, and large decrease Additionally, the last three columns present the p-values from the Mann-Whitney-Wilcoxon rank-sum test, which assesses the equality of distribution among these groups.
Large increase No change Mann - Whitney - Wilcoxon rank-sum test of equality of distribution (p- value )
4.3.2 The likelihood regression of large change (increase or decrease) against the change in firms’ characteristics and market condition
Table 4.9 indicates that firms with higher managerial ownership levels experienced significant declines As enterprises scale up, the board of directors' shareholding percentage decreases, suggesting a diversification of managers' portfolios to leverage external monitoring advantages It becomes challenging for individuals to hold substantial shares in large corporations, as managers' welfare cannot justify larger ownership stakes This finding contrasts with the research of Fahlenbrach and Stulz (2009) The impact of changes in R&D budgets remains unclear, leading to notable fluctuations in both large increase and large drop groups Increased total investment expenditure and book leverage are associated with declines in managerial ownership, while companies can benefit from bank supervision Additionally, higher capital expenditure may lead to reduced managerial ownership and diminished managerial rights over business operations Dividend factors, including initiation and termination, show no significant relationship with changes in managerial ownership.
Table 4.9 illustrates the marginal effects from Probit regression, with columns (1) and (2) focusing on large increases (at least 1 percent) in the dependent variable, while columns (3) and (4) examine large decreases The significantly positive coefficients for lagged managerial ownership in columns (3) and (4) suggest that firms with higher managerial ownership are likely to experience a reduction Additionally, larger firms tend to face a decline in managerial ownership due to limitations on managers' properties The positive coefficients of the fixed assets ratio in these columns indicate that firms with more substantial hard investments have a higher probability of experiencing a drop in managerial ownership levels Furthermore, as businesses increase capital spending, the likelihood of a reduction in managerial ownership rates diminishes.
Managerial risk aversion is influenced by the level of stock ownership, as a higher percentage of specific stock held by managers indicates a less diversified portfolio This creates a trade-off between the advantages of diversification and the compensation linked to stock performance Consequently, an increase in idiosyncratic risk (SIGMA) within a firm is likely to result in a decrease in managerial ownership levels, and vice versa This relationship is supported by the findings of Demsetz and Lehn (1985), which reveal significant coefficients in both the large drop and large increase groups at a 5 percent level in both RE and PA regressions.
Market conditions, including liquidity and current rates of return, were found to be insignificant at the 10 percent level However, the lagged performance of the overall market likely influences managers' decisions to buy or sell stocks Specifically, managers are inclined to purchase stocks following a period of poor market performance and tend to sell when the market is performing better.
Notes: the changes are classified into large increase and large decrease with threshold of change being
The analysis reveals that in columns (1) and (2), a significant increase in the dependent variable is observed when the percentage of MO exceeds 1%, while it is marked as 0 otherwise Conversely, in columns (3) and (4), a substantial decrease in the dependent variable is indicated when MO drops more than 1%, with a value of 0 in other cases Standard errors are provided in parentheses, and significance levels are denoted by ***, **, and * for 1%, 5%, and 10%, respectively.
VARIABLE RE PA RE PA