THE EFFECT OF COMPANY GROWTH, EARNINGS PER SHARE, AND CAPITAL STRUCTURE ON DIVIDEND POLICY WITH COMPANY AGE AS A CONTROL VARIABLE

ABSTRACT


Introduction
In the current era of globalization, economic development in Indonesia in building a company requires capital that can be obtained in several ways, both from personal capital, bank loans, and investor search.One of the tools to measure economic development in Indonesia is to determine the level of investment development in the capital market (Bulkia, Kurniaty, Rahmah, & Abdan, 2022).In investing, investors first see the condition of the company that continues to grow and progress.Therefore, a company that goes public must inform all forms of corporate obligations that concern the interests of shareholders, including announcing the distribution of dividends to be paid to investors.Conversely, having a low DPR means having a small amount of cash when the incident states that the company is in the process of expanding the company's territory, so it requires retained earnings to help finance the process that is being carried out (Sulistiyowati, Anggraini, & Utaminingtyas, 2010).
There is currently a 44% drop in profits at Saudi Aramco, but the company continues to pay dividends.In 2020, the company showed reports that it had experienced a 44.3% decline in profits.The phenomenon of issuers choosing not to share net profit in the form of dividends has received reactions from the IDX.The action taken by IDX is to seek information from issuers that have not distributed dividends in the last three years.Dividend policy is still an interesting study related to dividend payment obligations for issuers.
The company's growth ratio describes the company's ability to maintain its economic position amid economic growth and the company's sector (Kasmir, 2012).If the company has good growth, it will attract more attention from investors because a company that is growing rapidly portrays dividend income that will be better than a company that does not portray growth.Company growth can affect dividend policy in companies with good growth rates, which will allocate the funds obtained by the company to invest to reduce dividend distribution to shareholders (Pertami et al., 2018).Research conducted by (Maulany, Budiana, & Azizah, 2023) states that company growth affects dividend policy, this allows companies to hold their profits for expansion and not pay them as dividends.The results of research conducted by Wahyuliza and Fahyani (2019) that the company's growth does not affect the dividend policy, this means that even if the company experiences company growth or not, the company still distributes dividends according to the results of the general meeting of shareholders (GMS).The results of another study conducted by (Ramandini & Yuyetta, 2019) that company growth has a significant negative influence on dividend policy, this is under life cycle theory where dividend distribution tends to follow the company's life cycle pattern so that dividend distribution decisions are strongly influenced by the company's needs in distributing cash flows.
Earnings per share can also affect dividend policy.According to Fahmi (2012), earnings per share is a form of providing benefits given to shareholders from each share owned.Earnings per share describes the profitability of the company depicted on each share, the higher the value of earnings per share reflects that the company has managed its finances so that it can distribute profits in the form of dividends and increase the interest of shareholders to invest (Wiagustini, 2010).Accurate stock valuation can minimize risk in decision-making, therefore to evaluate the company's financial condition, investors can do so by calculating the company's financial ratio, namely earnings per share.Empirical evidence from the research of Pamungkas et al., (2017) states that earnings per share do not have a significant effect on dividend policy, this means that the size of earnings per share significantly does not affect dividend policy.Unlike the case of (Hardianto, 2021) states that earnings per share have a positive and significant effect on dividend policy, this shows that the higher value of earnings per share will cause greater profits and the possibility of increasing the amount of dividends received by shareholders.
Capital structure is the result or result of a financing decision that essentially chooses to use debt or external sources to fund company operations (Syamsuddin, 2019).The company's capital structure can be proxied by the debt-to-equity ratio which is a comparison of the company's total debt capital with the company's total own capital.When a company has a high debt ratio, it will affect liquidity directly.The use of large debt will reduce the company's cash in the form of interest and principal payments which ultimately has an impact on reducing dividends distributed.According to (Wahyudin, 2023), capital structure affects dividend policy.The results of the analysis show that a higher capital structure will cause the company to reduce or not increase its dividend payout ratio.Conversely, in the research of (Firdaus & Sediaz, 2017) that the capital structure (DER) does not influence dividend policy (DPR) where the results show that the higher the DER level, it means that the composition of debt will also be higher, resulting in the lower the company's ability to pay dividends.This research is in line with the research of (Nurfadillah & Al Farauqi, 2022), (Hartawan, 2021), and (Lubis, 2020) which does not affect dividend policy.
The study also used the company's age as a control variable.The longer the life of the company will be able to increase the ability to pay dividends when the profits generated are high, the company will tend to distribute dividends.Long-established companies will be trusted by investors compared to newly established companies because long-established companies are assumed to be able to generate higher profits than newly established companies (Novyanny and Turagan, 2019).The addition of this control variable is motivated by the fact that the company's age has a significant positive effect on dividend policy (Idawati and Sudiarta, 2014).
Based on the description above, the results of the study still vary due to the possibility of differences between the company's sample and the variables tested.In addition, there are still many companies that have never distributed dividends since being listed on the Indonesian stock exchange.Thus, it can make investors speculate on health companies, health sector stocks are attractive in the eyes of investors because there is an increase in sales of medical devices, medicines, demand for medical check-ups, and so on that are needed by the community during the Covid 19 pandemic so that companies in this sector experience an increase in profits and can distribute dividends to shareholders, even during the pandemic, many companies suffered losses (Fissholah & Hascaryani, 2022).
Capital structure is added as an independent variable because the higher the capital structure, it will cause the company to reduce or not increase its dividend payout ratio and the addition of the company's age control variable in this study is based on the fact that long-established companies will be trusted by investors compared to newly established companies because long-established companies can generate high profits and company capital management can run effectively and efficiently than newly established companies (Novyanny and Turagan, 2019).This study aims to examine the effect of company growth, earnings per share, and capital structure on dividend policy.The benefits of this research include contributions to updated research around dividend policy, references for academics, assistance for companies in decision-making, and consideration for investors in investing.The results of the study are expected to increase the company's awareness of dividend distribution and inform the interests of shareholders.

Research Methods
This study aims to examine the effect of company growth, earnings per share, and capital structure on dividend policy.The data used in this study is quantitative data obtained from the financial statements of trading and investment service sector companies listed on the Indonesia Stock Exchange in the 2018-2022 period.The analysis method used is a multiple linear regression method using the Statistical Product and Service Solution (SPSS) version 25 application.The results of this regression analysis will be used to test the hypothesis proposed in this study.
In this study, several variables are the main focus, namely dependent variables, namely dividend policy, and independent variables, namely company growth, earnings per share, and capital structure.In addition, there is also a control variable, namely the age of the company.Hypothesis testing includes a coefficient of determination test, a simultaneous test (statistical test F), and a partial test (statistical test t) to test the effect of these variables on dividend policy.In addition, before conducting a hypothesis test, classical assumption tests are also carried out to ensure that the regression model used is free from statistical assumption disorders such as normality, multicollinearity, autocorrelation, and heteroscedasticity.The population taken for use in this study is trading and investment service sector companies listed on the Indonesia Stock Exchange for the period 2018 -2022.Of all the populations taken, not all of them will be used as research objects, so that they are taken under the criteria for research needs.Sampling using purposive sampling techniques, where the samples taken are samples under specified criteria.

Normality Test Results
The normality test is carried out to find out whether in the regression model, the two variables, namely the independent variable and the dependent variable, have a normal or near-normal data distribution (Ghozali, 2018: 154).The tools used in the normality test in this study used normal p-plot graphs and One-Sample Kolmogorov-Smirnov Test.Decision-making regarding the normality of the One-Sample Kolmogorov-Smirnov Test is as follows: 1.If the sig (2-tailed) value > 0.05; then the data distribution is normal.2. If the sig (2-tailed) value < 0.05; then the data distribution is abnormal.
The results of the normality test using non-parametric Kolmogorov-Smirnov (K-S) statistics can be seen in Table 1: As a result of the Kolmogorov-Smirnov test above, the significant value obtained is 0.000 which means that the significance level is less than 0.05 (0.000 < 0.05).So it can be concluded that the data processed in the equation is not normally distributed.This means that the data used in this study must use outliers to normalize the data.The number of samples obtained previously amounted to 170, in determining outliers obtained several in SPSS which are indicated as extreme data.The following extrim data are identified From the data above then in outliers gradually.First, the 75th data is company data PT.Surya Citra Media Tbk in 2020 was removed from the sample.Then the data amounted to 169 and was reprocessed.The results show that there is no normality in the normal probability plot figure and the significance value is still below 0.05.Here are the results of data processing with 169 samples: From the results of data processing with 169 samples, it turned out that there were still no results of normally distributed data, then carried out the 97th data sample outlier, namely PT.Dyandra Media Internasional Tbk in 2020.After that, it was reprocessed and tested for normality of sample data totaling 168.Here are the results of data processing with 168 data: Based on the results of data processing with 168 samples of normal images, the probability plot has shown data points that follow the diagonal line, but the significance number is still below 0.05, which is 0.048, so the data results still cannot be said to be normal.The next step is to continue the data outlier, namely the 168th data sample, namely PT.Pudjiadi Prestige Tbk in 2022.Then the number of data samples becomes 167 data.After the outlier, the number of samples became 167.Ghozali (2016: 41) stated, there are four causes of data outliers, namely: Annamali Lolo Ramma, Juniati Gunawan 1. Error in data entry 2. Failure to specify a missing value in a computer program 3. Outliers are not members of the population we sampled 1) Outliers come from the population we sample, but the distribution of variables in that population has extreme values and is not normally distributed.
Judging from the sample data used in this study, the cause of outliers in this study is number 4. Outliers come from the population we sample, but in that population, there are extreme values that are not normally distributed.The results of data normality testing after outliers are shown as follows:

Multicollinearity Test Results
Multicollinearity testing aims to test whether the regression model correlates with independent variables.This test is performed using a VIF with criteria.If the VIF of a free variable < 10, it can be concluded that the free variable does not occur in multicollinearity.This test is also carried out using Tolerance with criteria, if the Tolerance of a free variable > 0.10 then it can be concluded that the independent variable is not multicollinearity.Based on the results of the analysis that has been done, the tolerance value and VIF value are shown in the table, as follows: Thus it can be concluded that there is no symptom of Multicollinearity between independent variables in the regression model.

Heteroscedasticity Test Results
The heteroscedasticity test is performed to determine whether, in the regression model, there is an inequality of variance from the residuals of one observation to another.If the variance from residual one observation to another is fixed, then it is called homoscedasticity and if different it is called heteroscedasticity.The results of the heteroscedasticity test using the Glacier test are as follows:  7, the results of the heteroscedasticity test with the Glacier method above show the significance value of each independent variable and the control variable is greater than 0.05 so it can be concluded that the model does not occur heteroscedasticity.

Autocorrelation Test
The autocorrelation test is intended to test whether there is a correlation between errors in period t and errors in the previous period t-1, this test uses Durbin Watson Test criteria.

Table 8. Autocorrelation Decision Value d Information 0 < d < dL
There is autocorrelation dL < d < dU

No conclusions 4-dL < < 4
There is autocorrelation 4-dU < d < 4-dL No conclusions dU < dw < 4-dU No autocorrelation The results of the autocorrelation test above show that the equation studied has some observations of 167, with the number of variables totaling 5 variables.Therefore, the lower limit value dL is 1.6857 with the upper limit du of 1.8089, the value of 4-dl is 2.3134, and 4-du is 2.1911.Durbin Watson test results were obtained of 1.969 in the area of no autocorrelation (du < DW < 4-du) or 1.8089 < 1.969 < 2.1911.So it can be concluded that autocorrelation does not occur.

Multiple Regression Analysis
This study uses multiple linear regression analysis because there is more than one independent variable.The independent variables in this study are company growth, earnings per share, and capital structure to dividend policy with company age as the dividend policy control variable.The regression equation model to test the hypothesis that has been formulated in this study is as follows: DPR = α + β1 GROW + β2 EPS + β3 DER + β4 AGE + e  10, the regression equation is as follows: DPR = 1,224 -0,102 GROW + 0,097 EPS + 0,271 DER + 0,770 AGE + e

Test Results of Coefficient of Determination (R2)
The coefficient of determination test is used to measure how far the model can explain the variation of the independent variable against the dependent variable.The value of the coefficient of determination indicated by the R² value of the regression model is used to determine the magnitude of variability of the dependent variable that can be explained by the independent variables.
Based on the table shows that the R² value is 0.414 which means that the variability of the variables Company Growth, Earnings Per Share, and Capital Structure to Dividend Policy with Company Age as the Control Variable of Dividend Policy in this study is 41.4% whereas, the remaining 58.6% is explained by variables outside the research model.

Simultaneous Test Results (Test F)
The simultaneous significance test (statistical F test) serves to find the presence or absence of influence together (simultaneous) independent variables on the dependent variable (Ghozali, 2018).11 above shows that there is a simultaneous influence with a calculated F value of 7.891 where the value is greater than the table f-number of 2.27 (7.891 > 2.27) and the significance level is smaller than 0.05 which is 0.000.So it can be concluded that the equation affects simultaneously or together the variables of company growth, earnings per share, and capital structure on dividend policy with the age of the company as a dividend policy control variable.

Partial Test Results (Test t )
A partial test (t-test) is used to determine whether the independent variable has a significant effect on the dependent variable.The condition of the t-test decision is that if the sig t value < 0.05, then the independent variable partially affects the dependent variable (Ho is rejected) and vice versa.
Based on the results of regression analysis in the table, the regression equation can be obtained, as follows: The partial test results are as follows: H1: Company Growth has a significant negative effect on Dividend Policy Based on the results of the t-test (partial) in the regression model, a calculated t value of -2.686 > 1.65403 was obtained with the significance of the Company Growth variable of 0.024 < 0.05 (significance level of 5%), it can be concluded that partially the Company's Growth variable affects the Dividend Policy variable with an unstandardized coefficient beta of -0.102.The results of this study show that H1 is accepted, which means "Company Growth has a significant negative effect on Dividend Policy".H2: Earnings Per Share has a significant positive effect on Dividend Policy Based on the results of the t-test (partial) in the regression model, a calculated t value of 2.719 > 1.65403 was obtained with the significance of the Earning Per Share variable of 0.013 < 0.05 (significance level of 5%), it can be concluded that partially the Earning Per Share variable affects the Dividend Policy variable with an unstandardized coefficient beta of 0.097.The results of this study show that H2 is accepted, which means "Earnings Per Share has a significant positive effect on Dividend Policy".H3: Capital Structure has no effect on Dividend Policy Based on the results of the t-test (partial) in the regression model, a calculated t value of 1.577 < 1.65403 was obtained with the significance of the Capital Structure variable of 0.058 < 0.05 (significance level of 5%), it can be concluded that partially the Capital Structure variable has no effect on the Dividend Policy variable with an unstandardized coefficient beta of 0.271.The results of this study show that H3 is rejected, which means "Capital Structure does not affect Dividend Policy".

Control Variable: Company Age
Based on the results of the t-test (partial) in the regression model, a calculated t value of 2.175 > 1.65403 was obtained with the significance of the Company Age variable of 0.015 < 0.05 (significance level of 5%), it was partially concluded that the Company's Age variable affected the Dividend Policy variable with an unstandardized coefficient beta of 0.770.The results of this study show "Company Age has a significant positive effect on Dividend Policy" and can be a control variable of Dividend Policy.

The Company's growth has a significant negative effect on the Dividend Policy
Based on the results of the above study, it can be concluded that partially the Company's Growth variable is 0.024 < 0.05, It can be concluded that partially, the Company's Growth variable affects the Dividend Policy variable with an unstandardized coefficient beta of -0.102.The results of this study show that H1 is accepted, which means that the Company's Growth has a significant negative effect on the Dividend Policy.From the representative data, it is said that the increase in assets occurs due to an increase in debt so the company's growth becomes a variable that negatively affects dividend policy.
In this case, the company's management must carefully consider reinvesting profits to expand the business and distribute profits to shareholders.Sarimelati Kencana Tbk at the company's growth value in 2021 was -0.01 and then in 2022 it increased by 0.13 but the dividend policy figure decreased from 2021 1.09 to -2.56 in 2022.This means that these results are in line with existing data.
The results of this study are in line with the results of research by Devi and Mispiuyanti (2020) stating that the company's growth hurts dividend policy because there is separate financing related to the addition of company assets so that the company's high and low growth rate does not affect the dividend payout ratio Or whether or not the company's growth will continue to distribute dividends with the same results because the growth that occurs further increases the search for investment that will be channeled for expansion and improvement of company operations.

Earnings Per Share has a significant positive effect on Dividend Policy
Based on the results of the study above, it is concluded that partially the variable Earning Per Share is 0.013 < 0.05, It can be concluded that partially the variable Earning Per Share affects the variable Dividend Policy with an unstandardized coefficient beta of 0.097.The results of this study show that H2 is accepted, which means that Earnings Per Share has a significant positive effect on Dividend Policy.With a significant influence, it shows that Earnings Per Share is one of the important aspects of stock returns (Badjra, 2016).This is under the sample data used where the average or mean value in EPS shows a high number of 233.3259 or 233% of the total sample of 167.Back to the previous explanation when the company decides to distribute dividends, it becomes a positive value from investors, shareholders, and stakeholders who want to invest in the company because the company is considered to have strong financial performance so that it can pay consistent dividends.
The results of this study are in line with the results of research conducted by Sari and Isroah (2017) which states that earnings per share have a significant positive effect on dividend policy because a higher value of earnings per share will cause greater profits in the form of dividends received by shareholders.This result is also in line with other studies that state that earnings per share can positively affect the dividend policy of Badjra (2016), Olivia and Ida (2016), Stefan andSiti (2012), andYusuf andRahmawati (2016).

Capital Structure does not affect Dividend Policy
Capital structure and dividend policy are two different things in a company's financial management.Capital structure includes how a company finances its assets, including financing with debt or equity, while dividend policy covers how the company decides to distribute profits to shareholders.This is illustrated by one of the companies that has the lowest DER value, namely PT.When viewed at the company's DER value for 5 periods, namely 2018-2022, it has a variable DER value.Agency theory in this study shows that there is an agency relationship between shareholders as company owners and company management as agents responsible for company management.In this study, company management tends to make decisions that benefit itself rather than decisions that benefit shareholders.Signal theory in this study focuses on the messages conveyed by companies through financial decisions such as capital structure and dividend policy.Then the company can provide signals or messages about the company's financial condition to the market through these financial decisions.
In this case, decisions regarding capital structure can provide signals about the company's ability to manage risk and capital availability, while dividend policy can provide signals about the stability of the company's profits.Although the capital structure does not directly affect dividend policy, the company's management may choose to make financial decisions that can give positive signals to shareholders.The results of this study are in line with the results of (Hartawan, 2021), and (Firdaus & Sediaz, 2017) research that capital structure has no influence on dividend policy, which shows that the higher the DER level, it means that the composition of debt will also be higher, resulting in the company's lower ability to pay dividends.

Dividend Policy Control Variables Company Age
Based on the results of the study above, it is concluded that partially the Company's Age variable is 0.015 < 0.05, It can be concluded that partially the Company's Age variable affects the Dividend Policy variable with an unstandardized coefficient beta of 0.770.Long-established companies are assumed to have more competitors, thus causing a continuous decline in the company's financial performance within a certain time which can lead to opportunities to get a going concern view (Kuntari, 2017).The age of a company can affect dividend policy in many different ways.Several studies show that a company's lifespan has a positive relationship with dividend policy.This can be due to several reasons, such as Financial Stability: where a company that has been operating for a long time may have more resources and sufficient financial stability to pay dividends consistently.

Conclusion
This study examines the effect of company growth, earnings per share, and capital structure on dividend policy with company age as a control variable in trading and investment service sector companies on the Indonesia Stock Exchange for the 2018-2022 period.The results of the analysis using the multiple linear regression method show that the company's growth has a significant negative effect on dividend policy, while earnings per share have a significant positive effect on dividend policy.However, there is no significant influence between capital structure and dividend policy.That is, companies that experience high growth will still distribute dividends with the same results regardless of growth rate.In addition, companies with high earnings per share tend to pay consistent dividends to increase investor confidence, while the company's capital structure does not affect dividend policy.
This study has several limitations that need to be considered by future researchers, such as limited samples because some companies use foreign currencies and do not present complete financial statements.In addition, the disposal of outlier data also reduces the number of samples tested.Although the results of the analysis showed that the independent variable could only explain about 41.4% of the variability of the dependent variable, this study can provide useful implications.For companies, the results of this study remind them to increase the value of EPS and consider the right capital structure to manage risk.For investors, this research provides more detailed information about the company's financial statements that can help in making investment decisions.Suggestions for future research include extending the study period to avoid data outliers, considering other industry sectors, and adding independent and moderating variables to strengthen the relationship between variables and test other guesses regarding dividend policy.

Table 5 . Kolmogorov-Smirnov One-Sample Normality Test Results
Based on the normality test table using the One-Sample Kolmogorov-Smirnov Test shown in the Table shows that the Kolmogorov-Smirnov value obtained sig value = 0.200 > 0.05.The Kolmogorov-Smirnov value is greater than the Kolmogorov-Smirnov table value of 0.05.This means it can be concluded that the regression model is normally distributed.

Table 6 .
Based on the table, it is shown that there are no variables that have a tolerance value of less than 0.10 and no independent variables that have a VIF value of more than 10.The Effect Of Company Growth, Earnings Per Share, And Capital Structure On Dividend Policy With Company Age As A Control Variable

Table 10 .
Based on the table of regression test results in Table