The Effect of Sales Growth, Dividend Policy and Return on Assets on Debt Policy in Property and Real Estate Companies Listed on the Indonesia Stock Exchange in 2013 - 2017

The aim of this study was to examine and analyze the effect of Sales Growth, Dividend Policy and Return On Assets simultaneously and partially on Debt Policy in Property and Real Estate companies listed on the Indonesia Stock Exchange for the 2013–2017 period. The data analysis method used was multiple linear regression. The samples used in this study were 13 companies. The results of this study indicate that the variables Sales Growth, Dividend Policy and Return On Assets simultaneously have a significant effect on debt policy. Partially, Sales Growth and Dividend Policy have no significant effect on policy. Only Return On Assets has a significant effect on debt policy in property and real estate companies listed on the Indonesian stock exchange for the 2013-2017 period.


Background
Debt policy is a company's decision to obtain funds from third parties to invest (Hardiningsih and Oktaviani, 2012).Debt policy is a very important policy taken by managers when the company is going to expand.Most companies choose to use loans as a source of funding because loans can be tax deductible and can increase company value.Companies that use loans will be trusted by the market because they have capabilities and bright prospects and gain the trust of investors.However, the use of loans also has drawbacks, namely the higher the loan ratio makes the company more risky, because the higher the cost of borrowing.Another weakness is that if a company is going through hard times and the operating profit is not enough to cover the interest expense, the shareholders have to cover the shortfall.If the shareholders are unable to cover the shortfall, bankruptcy will occur (Indana, 2015).
Shareholders generally do not have special skills in running a company, so shareholders hire agents.Agents are management trusted by shareholders to manage the resources assigned to them.Shareholders give authority and power to agents to make decisions that can increase the prosperity of company owners.However, management often has other goals that are different from the main goals of the company, giving rise to a conflict of interest between management and shareholders.Conflicts of interest related to funding decisions are caused by shareholders only focusing on systematic risks, such as rising interest rates and inflation, while management focuses on overall company risks.Conflicts of interest can be minimized with a monitoring mechanism that will incur costs called agency costs (Keni and Dewi, 2011).
Debt policy is influenced by various factors, one of which is Sales Growth.Sales growth is the increase in sales over time.
Companies with high sales growth tend to require more investment in various elements of assets, both current assets and fixed assets, so management must seek funding sources for these investments.The higher the sales growth, the higher the loan obtained.This is because companies with high sales growth rates are able to fulfill their financial obligations (Pradhana, Taufik, and Anggaini, 2014).According to Kesuma (2009), to measure sales growth is to compare sales in year t after deducting sales in the previous period against sales in the previous period.
Dividend policy is also one of the factors that can influence debt policy.According to Sheisarvian et al. (2015) dividend policy is a policy to determine how much income will be distributed to shareholders and what will be retained by the company as retained earnings.Fixed dividend payments cause a fixed need for funds every year so that the company's need for funds will increase (Yeniantie and Nicksen, 2010).The higher the dividend, the higher the debt policy.This is because if the company increases dividend payments, the available funds for funding in the form of retained earnings will be smaller, so that to meet funding needs, management tends to use debt (Indahningrum and Handayani, 2009).Dividend policy can be measured using the dividend payout ratio (DPR).
Another factor that can affect debt policy is Return On Assets (ROA).According to Suad Husnan (in Putera, 2006), return on assets is used to measure the company's effectiveness in generating profits by utilizing the total investment made by the company.Return on assets is also a multiplication between the net income margin factor and asset turnover.Net income margin shows the ability to earn profit from every sale made by the company, while asset turnover shows how far the company is able to generate sales from its assets.If one of these factors increases (or both), then the return on assets will also increase.If the return on assets increases, it means that the company's profitability increases, so that the final impact is an increase in the profitability enjoyed by shareholders.
Property and real estate developments and developers currently have a very high investment value.So that increased performance in trade, housing, construction and other property businesses will become a business opportunity for investors to make long-term investments.Profits derived from construction and housing will increase the company's value by around 10% -20% and above from year to year so that this property company has the potential to increase enough resources so that there will be a multifold increase in the next 5-10 years.So that a company must try to increase the value of the company by providing customer satisfaction, the customer satisfaction provided by the company is being able to fulfill all the needs and desires of consumers for the development and marketing of property both large-scale and small-scale housing.
The following is average data regarding debt policy, sales growth, dividend policy, and return on assets for property and real estate companies: In accordance with table 1.1, it can be seen that the average DER from 2013 to 2017 in Property and Real Estate companies has decreased.Based on the lowest average DER value or said to be good, it was in 2017, which was 0.97%.And the highest average value of DER or said to be bad in 2013 was 1.44%.According to the trade off theory that increasing the use of debt to fund assets will reduce the value of the company.
Dealing with the average SG value from 2013 to 2017 it has decreased.The lowest average value of SG in 2016 was -11.81% and the highest in 2013 was 29.77%.If a company's sales increase or increase, it is less likely for the company to borrow external funds.Conversely, if a company has declining profits, the company will need a large source of capital for the continuity of the company, so it is likely that the company will choose external funds as a source of funding.
Based on the average score of the DPR from 2013 to 2017, it has also decreased.The lowest average value in 2017 was 12.91% and the highest in 2014 was 277.03%.If the company distributes more profits as dividend payments to shareholders, then the remaining profit is small for the continuity of its operations, then the company will choose external funding sources as additional capital.
Dealing with the average value of ROA from 2013 to 2017, it also decreased.The lowest average value in 2016 was 6.36% and the highest in 2014 was 8.46%.
By looking at the problems above, the authors are interested in conducting this research entitled: "The Effect Of Sales Growth, Dividend Policy And Return On Assets On Debt Policy In Property And Real Estate Companies Listed On The Idx In 2013-2017".

Literature Review
Nina Purwaningsih and Tumpal Manik (2016)  possessed by the population, or a small part of the population members taken according to certain procedures so that they can represent the population, the researchers used 13 companies from 48 populations as samples in this study.With 5 years of observation, from 2013-2017, 5 years of observation x 13 samples = 65 observations.

Data Types and Sources
The type of data used in this research was quantitative data.Data sources were secondary data, namely data obtained from sources related to research, such as financial report data taken from the official website www.idx.co.id, journals and previous research results.

Data collection technique
The data collection technique in this study was a documentation technique.Documentation, namely the collection of data and information was done by examining the documents obtained without re-processing the data.Data obtained from print and electronic media.

Data analysis technique
The data that has been collected will be analyzed by performing descriptive statistical analysis and classical assumption tests.Descriptive statistical analysis was carried out to determine the dispersion and distribution of data.While the classical assumption test was conducted to test the research.Data analysis in this study was carried out using computer assistance, namely Microsoft Excel and the SPSS (Statistical Package For Social Science) computer program version 20.0.

A. Descriptive Statistical Analysis
This study used descriptive statistics to describe the standard deviation, minimum and maximum mean and the variables studied.Descriptive statistics describe data into information that is clearer and easier to understand.Descriptive statistics are used to develop company profiles that are samples of descriptive statistics related to data collection and improvement, as well as presentation of the results of these improvements.(Ghozali 2006).
B. Classic Assumption Test This test was carried out in order to get reliable and unbiased regression results or called BLUES (Best Linear Unbiased Estimator).From this test the assumptions that must be fulfilled are that there is no strong correlation between the independent variables (multicollinearity), there are no residual periods t with t-1 (autocorrelation) and there is no unequal variance from the residuals of one observation to another (heteroscedasticity). , the resulting data is normally distributed.
a. Normality Test The normality test aims to test whether in a Regression Model, Dependent Variables, Independent Variables both have normal data distribution or not.A good regression model is a method with normal or close to normal data distribution.P-P plot graphs can be used to detect normality which compares the cumulative distribution of the actual data with the cumulative distribution of the normal distribution.The normal distribution will form a straight diagonal and plotting the data will be compared with the diagonal line.If the data spreads around the diagonal line and follows the direction of the diagonal line, the regression model meets the normality assumption.If the data spreads (dots) away from the diagonal line and does not follow the direction of the diagonal line, the regression model does not meet the normality assumption (Ghoz ali, 2009). b

. Data Multicollinearity Test
The multicollinearity test is used to determine whether or not there is a deviation from the classic multicollinearity assumption, namely the existence of a linear relationship between the independent variables in the regression model (Priyatno, 2013).There are several testing methods that can be used, one of which is to look at the magnitude of the Variane Inflation Factor (VIF) and Tolerance.A Regression Model that is free of Multicollinearity has VIF values around numbers and has a tolerance number close to 1.According to Priyatno, in general, if VIF > 5, the variable has a multicollinearity problem with other independent variables.c.Heteroscedasticity Test To detect the presence or absence of heteroscedasticity by looking at the scatterplot graph between the predicted value of the dependent variable and its residual value.Heteroscedasticity occurs when it forms a certain regular pattern (wavy then narrows), whereas if there is no clear pattern, and the points spread above and below the number 0 on the Y axis, then there is no heteroscedasticity (Ghozali, 2005) in Ayu 2014.

d. Autocorrelation Test
The aim is to test whether in the linear regression model there is a correlation between the confounding errors in period t and the confounding errors in period t-1.Autocorrelation arises because of other observations (Ghozali, 2005)  The coefficient of determination (2) is to find out how much influence the variable X has on the variable Y.The coefficient of determination from the results of multiple regression shows how much the dependent variable is and can be explained by the independent variables.The value of the coefficient of determination is between zero and one.The value of 2 is equal to 0, so there is not the slightest influence exerted by the independent variable on the dependent.Conversely, the value of 2 is equal to 1, so the influence exerted by the independent variable on the dependent variable is perfect.
D It is used to test the first, second and third hypotheses to test the independent variables individually on the dependent variable using T-test analysis, using a 95% confidence level, this test is carried out by comparing the value of count with the value e so that it can be seen or determined whether the hypothesis is significant or not significant.Ho is rejected if count >ble , and Ha is accepted meaning that the independent variable concerned affects the value of the dependent variable.Ho is accepted if count < ble , then Ha is rejected, which means that the independent variable has no effect on the dependent variable.

B. Simultaneous Test (F Test)
The F statistical test shows whether all independent variables have a joint or simultaneous influence on the dependent variable.This test uses the F test by comparing the calculated F test with F table, with the following conditions: If count <e then H0 is accepted and Ha is rejected, meaning that together the independent variables have no significant effect on the dependent variable.If count >ble, then H0 is rejected and Ha is accepted, meaning that together the independent variables affect the dependent variable.Y = α + β1X1 + β2X2 + β3X3 + ε

Variable Descriptive Analysis
Research on the effect of sales growth, dividend policy and return on assets on debt policy in property and real estate companies listed on the Indonesia Stock Exchange consists of one dependent variable and three independent variables.The dependent variable is debt policy (Y) and the independent variables are sales growth (X1), dividend policy (X2), and return on assets (X3).The objects in this study are property and real estate companies listed on the IDX in 2013-2017.In this study, 13 samples were taken from 48 property and real estate companies listed on the IDX in 2013-2017 based on predetermined criteria.The data taken is the financial statement data of property and real estate companies obtained from the Capital Market Information Center (PIPM) BEI Pekanbaru.From this data, it will be processed again to produce data for research variables, namely the dependent variable (debt policy) and independent variables (sales growth, dividend policy and return on assets).
Based on the results of descriptive data processing in table 3 above, it is known that the number of N (data) is 65, with the highest average value being DER -0.0093 with a standard deviation of 1.00000.

Analysis of Dependent Variables and Independent Variables
The variables used in this study include the dependent and independent variables.The dependent variable is Debt Policy and the independent variables include Sales Growth, Dividend Policy and Return On Assets.
1. DER Analysis of Property and Real Estate Companies Listed on the Indonesia Stock Exchange.Debt policy is a very important decision in a company where debt policy is one part of the company's funding policy.In addition, the company's debt policy also functions as a monitoring mechanism for the actions of managers in managing the company (Pithaloka, 2009).
Descriptive 3. DPR Analysis of Property and Real Estate Companies Listed on the Indonesia Stock Exchange Dividend policy is a policy to determine the amount of company profits that must be distributed to shareholders in the form of dividends and which must be retained in the form of retained earnings.However, retaining a larger amount of current profits within the company also means less money that will be available for dividend payments.So the company will need more funds for further operational activities.

Classic assumption test
This test is carried out in order to obtain reliable regression results and obtain results that are not biased or called BLUES (Best Linear Unbiased Estimator).From this test, the assumptions that must be met are that there is no strong correlation between the independent variables (multicollinearity), there are no residual periods t with t-1 (autocorrelation) and there is no variance inequality from the residual of one observation to another (heteroscedasticity), the resulting data is normally distributed.

A. Data Normality Test Results
The purpose of the Normality Test is to find out whether in the regression model the independent variables and the dependent variable both have a normal or abnormal distribution of the data in the sample.The normality test in this study uses a normal probability plot graph.Normality can be detected by looking at the distribution of data or points on the diagonal axis of the graph (Ghozali 2005).
B. Multicollinearity Test Results Multicollinearity is defined as a perfect linear relationship between several independent variables.The aim is to test whether the regression model found a correlation between the independent variables.A good regression model should have no correlation between the independent variables.Multicollinearity can be detected by looking at the VIF (variant inflation factor) and tolerance values.If the VIF value (variant of inflation factor) is below 10 or (VIF <10) and the tolerance value is more than 0.1 or close to 1, then multicollinearity does not occur.Multicollinearity test results can be seen in the following table: To detect the presence or absence of heteroscedasticity by looking at the scatterplot graph between the predicted value of the dependent variable and its residual value.Heteroscedasticity occurs when it forms a certain regular pattern (wavy then narrows), whereas if there is no clear pattern, and the points spread above and below the number 0 on the Y axis, then there is no heteroscedasticity (Ghozali, 2005) in Ayu 2014.

B. Autocorrelation Test Results
The aim is to test whether in the linear regression model there is a correlation between the confounding errors in period t and the confounding errors in period t-1.Autocorrelation arises because of other observations (Ghozali, 2005)

Determination Coefficient Test
The coefficient of determination (2) is to find out how much influence the X variable has on the Y variable.The coefficient of determination from the results of multiple regression shows how much the dependent variable is and can be explained by the independent variables.

A. Simultaneous Test (f-statistical Test)
The f-statistical test shows whether all independent variables have a joint or simultaneous influence on the dependent variable, namely SG, DPR and ROA on Debt Policy.To find out the effect can be seen in the ANOVA table below: The t test is used to determine the effect of each independent variable on the dependent variable, namely SG, DPR and ROA on Debt Policy (DER) in the Property and Real Estate sector companies listed on the IDX with the t test or partial test.Thus, it is concluded regarding partial hypothesis testing that has been made previously as follows: a.The effect of f SG on Debt Policy The results of the SG regression analysis on Debt Policy show a significance value of > α or 0.109 > 0.05, so according to the framework outlined in the research model, it can be concluded that SG has no partial effect on Debt Policy.

b. The Effect of the House of Representatives on Debt Policy
The results of the DPR's analysis of Debt Policy show a significance value of > α or 0.146 > 0.05, so according to the framework outlined in the research model, it can be concluded that the DPR has no partial effect on Debt Policy.

c. The Effect of ROA on Debt Policy
The results of the analysis of ROA on Debt Policy show a significance value of <α or 0.000 <0.005, so it is in accordance with the research model presented, so it can be concluded that ROA has a partial effect on Debt Policy.

Conclusion
This study aims to examine the effect of Sales Growth, Dividend Policy and Return On Assets on Debt Policy in Property and Real Estate companies listed on the Indonesia Stock Exchange in 2013-2017, 13 Property and Real Estate sector companies were used as samples in this study.Based on the results of research and discussion, it can be concluded as follows: 1. Simultaneously the independent variables, namely sales growth, dividend policy and return on assets have a significant effect on debt policy.
2. The results of multiple linear regression analysis show that partially the sales growth variable (SG) has a positive and insignificant effect on debt policy (DER), the dividend policy variable (DPR) shows negative and insignificant results on debt policy (DER) and the variable Return On Assets (ROA) has a negative and significant effect on debt policy (DER).

2 Suggestion
Regarding to the results of the research, there are several suggestions that should be made for the development of this system to be better, including the following: 1.For future researchers, it is hoped that they will conduct another study regarding Debt Policy with (DER) using a different independent variable and a larger number of samples in order to obtain a more definite conclusion which independent variable is more influential.2. For companies to be able to manage their company's debt policies to be able to stabilize the company and advance the company.So that shareholders, keep investing their capital and the rate of return of money can run smoothly.
3. Collect more references in order to support theory and cover existing limitations.

Table 1 .
The Average of DER, SG, DPR and ROA in Property and Real Estate Companies in 2013 -2017

Table 2 .
Cash Flow, Dividends and Performance on Debt Policy.Free Cash Flow has a positive effect on debt policy.Meanwhile, Managerial Ownership, Institutional Ownership, Dividend Policy, and sales growth have a negative effect on debt policy.Research conducted by Nurul Fadillah, Asri Eka Ratih, and Tumpal Manik (2016) resulted in Current Ratio, Return On Assets having a positive influence on debt policy.Meanwhile Firm Size and Firm Growth have no negative effect on debt policy.Operational Variable Property and Real Estate sector registered on the Indonesia Stock Exchange (IDX) during the observation period, 2013-2017, namely 48 Property and Real Estate sector companies listed on the Indonesia Stock Exchange B. SampleThe sample is part of the number and characteristics in Sri Tegia Imala 2016.A good model is a model that is free from autocorrelation.Autocorrelation testing uses the Durbin-Watson model.C. Coefficient of Determination (R²) . Multiple Linear Regression Analysis This study uses statistical data analysis techniques using the SPSS (Statistic Social Science) program, namely multiple regression to see the effect of Sales Growth, Dividend Policy and Return On Assets on Debt Policy in Property and Real Estate Companies listed on the Indonesia Stock Exchange in 2013 -2017 will use Multiple Regression as follows:

Table 4 .
DER data of Property and Real Estate companies listed on the IDX 2013 -2017 is known as follows: Based on table 4.2 it can be seen that the DER in Property and Real Estate companies during the 2013-2017 period has decreased.2. SG Analysis of Property and Real Estate Companies Listed on the Indonesia Stock Exchange Descriptive SG data of Property and Real Estate companies listed on the IDX 2013 -2017 are known as follows: Development of DER in Property and Real Estate Companies Listed on the IDX 2013 -2017

Table 5 .
Development of SG in Property and Real Estate Sector Companies Listed on IDX 2013 -2017

Table 7 .
The Result of Multicollinearity Test Coefficients aBased on table 4.5 the results of calculating the tolerance values of the three variables show more than 0.10 and the VIF value is less than 10.So it can be concluded that the regression and independent variables in this study are free from the influence of multicollinearity.A. Heteroscedasticity Test Results in Sri Tegia Imala 2016.A good model is a model that is free from autocorrelation.Autocorrelation testing uses the Durbin-Watson model.

Table 8 .
The Result of Autocorrelation Test Model Summary b Based on the results of the autocorrelation test, it can be seen that the Durbin-Watson value is 1.687.This value is compared with the value of the Durbin-Watson table with a significance level of 5%, the number of samples (n) = 65, and the number of independent variables (k) = 3.From the Durbin-Watson table, the dl value is 1.535, the du value is 1.662 and 4-du of 2.338.The results obtained are DL < DU < DW or 1.535 < 1.662 < 1.687, so there is no autocorrelation.4.Multiple Linear Regression Analysis This analysis aims to determine whether Sales Growth, Dividend Policy and Return On Assets affect Debt Policy.In this research, the technique used is multiple linear analysis technique, because this study uses more than one independent variable.Following are the results of multiple linear regression analysis data calculations using the SPSS version 20 program, namely: