The Influence of Profitability, Composition of the Independent Board of Commissioners and Managerial Ownership on Profit Management in Food and Beverage Companies on the Indonesian Stock Exchange Period 2017-2021

The purpose of this study is to identify and analyze the effect of profitability, independent board of commissioners composition and managerial ownership on earnings management. This study uses data analysis method using multiple linear regression analysis. Partial research results show that profitability has no effect and is not significant on earnings management. The composition of the independent board of commissioners also has no influence and is not significant on earnings management. Managerial ownership has no effect and is not significant on profitability.


INTRODUCTION
Business and the world economy continue to experience developments and changes, followed by increasing business competition which encourages companies to provide their best performance.Companies have various types of resources, both tangible, such as cash, inventory and machinery, and intangible, such as patents and human resources.
These resources are generally obtained at the expense of part of the company's funds which are generally obtained from stakeholders, namely investors and creditors.Before investing their funds in a company, investors and creditors need information to assess the company's suitability for obtaining funds so that they can calculate future profits and returns from the investments or loans they provide.
The main medium for obtaining this information is from the financial reports presented by the company (Subramanyam, 2014: 74) The report that the company wants to submit is generally a report that shows the company's best performance, but there are times when the company cannot achieve this best condition, such as the profit achieved is not as good as the target.This condition can encourage management to make various efforts so that the financial report becomes better.attractive in the eyes of investors and creditors.

Background of the Problem
Indonesia also has several earnings management scandals that have occurred.One example of an earnings management scandal in Indonesia is the earnings management carried out by PT Indofood Tbk in its 2018 financial report which recorded a net profit of 809.84 thousand US Dollars from what should have been a loss of 244.95 million US Dollars.The company and all directors and commissioners who signed the financial report were subject to sanctions from the Financial Services Authority (OJK) and the Indonesian Stock Exchange (BEI).Apart from that, the Public Accounting Firm (KAP) had its auditor's license frozen for 1 year and received a written warning and was also obliged to improve its quality control system (Kumparan, 2019).Apart from PT Indofood Tbk, Indonesian companies such as Tiga Pilar Sejahtera Food Tbk, GarudaFood Putra Putri Jaya Tbk, and others have also been proven to have overstated their reported profits.
From the research results, it was concluded that Indonesia was in 8th place for accrual earnings management and 23rd for real earnings management out of the 31 countries in the research sample, so that Indonesia was still at a high level for accrual earnings management in 1991-2010.If the company records discretionary accruals to reduce profits then the discretionary accruals will have a negative sign and if the company records discretionary accruals to increase profits then the discretionary accruals will have a positive sign Scott (2015:453).Therefore, to measure the average discretionary accruals, it is more appropriate to use the absolute value of discretionary accruals.
There are several factors that can motivate management to carry out earnings management, some of which include profitability, the composition of independent commissioners, and managerial ownership.Factors that can influence earnings management are the independent board of commissioners, namely members of the board of commissioners who act independently and do not have relationships such as management, financial and family relationships with other members in a company (PBI No. 8/4/PBI/2006).

Formulation of the problem
Based on the background of the problem described above, the formulation of the problem in this research is as follows: 1. Does profitability affect earnings management? 2. Does the composition of the independent board of commissioners affect earnings management? 3. Does managerial ownership have an effect on earnings management? 4. Do profitability, the composition of the independent board of commissioners, and managerial ownership simultaneously influence earnings management?

Research Purposes
Based on the problem formulation described above, the aim of this research is to: 1. To determine the effect of profitability on earnings management.2. To determine the influence of the composition of the independent board of commissioners on earnings management.3. To determine the influence of managerial influence on earnings management.4. To determine the influence of profitability, the composition of the independent board of commissioners, and managerial ownership.

LITERATURE REVIEW 1. Agency Theory
An agency relationship is a contract with a person or several people called a principal who employs another person called an agent to perform several services on their behalf (Jensen & Meckling,1976).

Information Asymmetry Theory
Information asymmetry is a situation that explains that one party has more information and access than the other party (Scott,2015:22).Scott (2015:22-23) divides information asymmetry into two types, namely Adverse selection and Moral hazard.

Signal Theory
A signal is an action taken by company management that gives investors a clue about how management views the company's prospects (Brigham & Houston, 2019: 500).A signal is an action that a good manager can take that a bad manager cannot do Scott (2015:503).

METHODOLOGY
The type of research used in this research is causal research which is carried out to test whether one variable causes another variable to change, namely whether X causes or influences variable Y (Sekaran & Bougie,2016:44).This research is also included in quantitative research, namely research where the data is in the form of numbers (Sekaran & Bougie,2016:2).This research approach was carried out by studying company records and documents as required, namely secondary data from financial reports, annual reports and Monthly Reports on the Registration of Securities Holders of manufacturing companies on the Indonesian Stock Exchange via the internet.Secondary data in this research comes from annual reports, financial reports and Monthly Securities Holder Registration Reports of manufacturing companies listed on the Indonesia Stock Exchange for the 2017-2021 period which were collected via the official website of the Indonesia Stock Exchange, namely https://www.idx.co.id/, articles, journals, previous research and books related to research.The analysis technique used is a quantitative data analysis technique using data testing tools, namely the SPSS program consisting of descriptive statistical tests, classical assumption tests, simple linear regression analysis, multiple linear regression analysis, and significance testing

RESEARCH RESULT Classic assumption test
cThe steps taken before linear regression, namely the classical assumption test, need to be carried out.Ghozali (2016:167) states "a regression model using the Ordinary Least Square (OLS) estimation method will provide Best Linear Unbiased Estimator (BLUE) results if it meets all classical assumptions."The data must meet classical assumptions so that the regression analysis model can be accounted for.The classical assumption test consists of four tests, namely: normality test, multicollinearity test, heteroscedasticity test, and autocorrelation test.

Normality test
The normality test results in this research regression model used three tests, namely P-P Plot Picture 1. P-P Plot Normality Test Results Based on Figure 1 which shows the normal P-Plot graph before the outliers, it can be seen that the points have spread around the diagonal line, but the spread tends to move away from the diagonal line.This indicates that the residual data is not normally distributed.Based on graphic analysis, it can be concluded that the regression model violates the normality assumption.Likewise with the results of the histogram graph.

Picture 2. P-P Plot Normality Test Results
Based on Figure 1.2 which shows the histogram graph before the outliers, it can be seen that the histogram graph gives a distribution pattern that is skewed to the left and is not normal.This is also confirmed by the results of the Kolmogorov-Smirnov test which has a significance of 0.000 <(0.005), which means the residuals are not normally distributed.2021).In this study, the data was discarded (outliers) using a square root transformation.The square root transformation is used if the data does not meet the assumption of homogeneity of variance, in other words the root transformation functions to make the variance homogeneous and approach a normal distribution.The results of the normality test using the P-P Plot show that the data moves close to a diagonal line, which means that the data distribution is normally distributed.

Picture 4. Normality Test Results Histogram
From Figure 4 above, it can be seen that the bars in the histogram are generally below the normal curve, so it can be concluded that the research variable data is normally distributed.2. above shows that the results of the normality test using the Kolmogorov-Smirnov test obtained a significant value (Asymp.Sig.(2-tailed)) of 0.200 which is greater than 0.05, from these results it is concluded that the data in this study is normally distributed.

Heteroscedasticity Test
The heteroscedasticity test in this study was carried out using a statistical approach through the Glejser test using a significance level of 0.05.The assessment criteria in heteroscedasticity testing with Glejser are as follows: Table 3 shows that the sig value.profitability is 0.209, and the composition of the independent board of commissioners is 0.912 and managerial ownership is 0.452, where both independent variables have a significant value greater than 0.05 so that the results of this calculation meet the test requirements for heteroscedasticity symptoms.Thus, it can be concluded that this research data does not contain symptoms of heteroscedasticity.

Multicollinearity Test
The multicollinearity test aims to test whether there is a correlation between independent variables in the regression model.To detect multicollinearity problems in this research, the Tolerance and Variance Inflation Factor (VIF) values were used.The following table presents the results of the multicollinearity test using Tolerance and VIF values, namely: Table 4

. Multicollinearity Test Results
The multicollinearity test produced a tolerance for profitability of 0.614 and the composition of the independent board of commissioners of 0.585 and for managerial ownership of 0.891 and a VIF for profitability of 1.628 for managerial ownership of 1.122.the composition of the independent board of commissioners is 1,710.These results show that Tolerance is above 0.10 and VIF is below 10, so it can be concluded that there is no multicollinearity in the regression model of this research.

Autocorrelation test
The autocorrelation test aims to test whether there is a correlation between confounding errors between one period and the previous period in the regression model Ghozali (2016:107).Autocorrelation test results on the Summary Model using the Runs Test.The assessment criteria in heteroscedasticity testing with Glejser are as follows: 1.If the value of Asymp.sig < 0.05, it can be concluded that the regression model used has symptoms of autocorrelation.2. If the value of Asymp.sig> 0.05, it can be concluded that the regression model used does not contain symptoms of autocorrelation.

Table 5. Autocorrelation Test Results
Asymp value.The sig (2-tailed) obtained is 0.270, which is greater than 0.05, so the regression model does not experience symptoms of autocorrelation.

Multiple Linear Regression Analysis Results
The results of the regression analysis are in the form of regression coefficients for each independent variable.Based on the equation above, it can be interpreted as follows: 1.The constant (a) is 0.039, which means that if Profit Management is not influenced by other variables or factors it will still be worth 0.039 units.2. The Profitability regression coefficient (b1) has a positive sign of 0.085, which means that a one unit increase in profitability will increase profit management by 0.085 units.3. The regression coefficient for the Composition of the Independent Board of Commissioners (b2) has a positive sign of (-0.023), which means that a one unit increase in the composition of the independent board of commissioners will reduce Profit Management by (-0.023) units.4. The regression coefficient for Managerial Ownership (b4) has a positive sign of 0.033, which means that an increase in one unit of Managerial Ownership will increase Profit Management by 0.033 units.

Simultaneous Hypothesis Test Results (F-Test)
In determining the Ftable size obtained at alpha (0.05) using the formula: df1=k-1; df1=3-1; df2=2, and df2=n-k; df3=15-3; df3=12 so Ftable¬ is 3.885.The Ftable¬ value will be compared with the Fcount¬ value below, which aims to determine whether or not there is a simultaneous influence between the independent variables on the dependent variable.Based on the results of the F test, the Fcount value is (0.108) > Ftable (3.88) and the significance value is (0.955) > alpha (0.05).Based on these results, it can be concluded that Profitability and the Composition of the Independent Board of Commissioners simultaneously have no effect on Profit Management.

Partial Hypothesis Test Results (t-Test)
In determining the ttable size obtained at alpha (0.05) using the formula: df=n-k; df=15-3; df=12, so ttable¬ is 2.160.The ttable value will be compared with the tcount value below, which aims to determine whether there is a partial influence between the independent variable and the dependent variable.

Coefficients a
The results of the partial hypothesis test can be interpreted as follows: 1. Based on the table above, the calculated t value is 1.288 < t table 2.160 and the significant value is 0.209 < 0.05, so it can be concluded that profitability has no effect on earnings management.2. Based on the table above, the calculated t value is 0.763 < t table 2.160 and the significance value is 0.452 < 0.05, so it can be concluded that Managerial Ownership has no significant effect on Profit Management. 3. Based on the table above, the calculated t value is -0.112 < t table 2.160 and the significance value is 0.912> 0.05, so it can be concluded that the composition of the independent board of commissioners does not have a significant effect on Profit Management.

Discussion of Research Results
The Influence of Profitability on Earnings Management Based on the results of the t-test carried out, it can be concluded that profitability has no significant effect on earnings management.Based on these results, hypothesis 1 is rejected.Profitability as measured by the ROA (return on assets) indicator explains the company's ability to generate profits by making its assets more efficient Perdana (2018).When a company is able to manage ROA well, it will increase the company's net profit.However, when ROA is not managed efficiently it will affect the company's net profit.One of the goals of earnings management is to get a good response from investors towards the company.

The influence of the composition of the independent board of commissioners on earnings management
Based on the results of the t-test, it can be concluded that the composition of the independent board of commissioners does not have a significant effect on earnings management.Based on these results, hypothesis 2 is rejected.The results of this research indicate that the existence of independent boards of commissioners in food and beverage companies in Indonesia has not been able to become a mechanism for reducing earnings management practices.The results of this research are not in line with the assumptions of agency theory where the agent (management) has different interests from the principal (independent board of commissioners).Both management and the independent board of commissioners will try to fulfill their personal interests, less effective financial performance can influence investors' interest in investing so that investors will choose companies with good financial performance, therefore companies that have poor financial performance will tend to carry out earnings management.compared to companies with good financial performance.

The influence of managerial ownership on earnings management
Based on the results of the t-test, it can be concluded that Managerial Ownership has no significant effect on earnings management.Based on these results, hypothesis 3 is rejected.The research results show that partial managerial ownership has a significant negative effect on earnings management, so it can be concluded that the higher the managerial ownership of a company, the lower the earnings management will be and vice versa.Increasing managerial ownership means that the number of shares owned by management also increases, so that their interests as shareholders will also increase.The results of this research are not in line with the assumptions of agency theory which states that agents and principals who each try to maximize their different interests can give rise to conflicts of interest.
Thus, managerial and investor ownership have the same goal, namely profit for each party, so that companies with less effective financial performance can influence investors' interest in investing so that investors will choose companies with good financial performance, therefore companies that have good financial performance.not good.Management will tend to carry out earnings management.

CONCLUSIONS AND RECOMMENDATIONS
The influence of profitability, the composition of the independent board of commissioners, and managerial ownership on earnings management in food and beverage companies on the Indonesia Stock Exchange for the 2017-2021 period is that partially profitability has no effect on company earnings management, partially the composition of the independent board of commissioners has no effect on earnings management in food and beverage companies on the Indonesia Stock Exchange for the 2017-2021 period, partially managerial ownership has no effect on earnings management in food and beverage companies on the Indonesia Stock Exchange for the 2017-2021 period, simultaneously profitability, the composition of the independent board of commissioners, and managerial ownership have no effect on earnings management in food and beverage companies on the Indonesia Stock Exchange for the 2017-2021 period.With the results and conclusions of the research, suggestions that can be useful for this research are for future researchers to add other factors that can influence earnings management, because considering that there are still 109.8% of other factors or variables that can influence earnings management, next can add company sectors to increase the number of research objects and can increase the research period which will increase the amount of research data so that it covers trends each year.

Table . 1
Normality Test Results Using the Kolmogorov-Smirnov Test

Table 1 .
Residual values that are not normally distributed can be corrected by transforming the entire data or removing outlier data Luhur (

One-Sample Kolmogorov-Smirnov Test
c. Lilliefors Significance Correction.d.This is a lower bound of the true significance.

Table 2 .
Normality Test Results Using the Kolmogorov-Smirnov Test

One-Sample Kolmogorov-Smirnov Test
. This is a lower bound of the true significance.1.If the significant value is <0.05, it can be concluded that the regression model used is not feasible.2. If the significant value is > 0.05, it can be concluded that the regression model used is feasible. d

Table 3 .
Heteroscedasticity Test Results

Table 6 .
Multiple Linear Regression Results