The Effect of Profitability, Financial Leverage, and Winner/Loser Stock to Profit Altering

: This study aims to find out: (1) the effect of profitability on the occurrence of income smoothing practices, (2) the effect of financial leverage on the occurrence of income smoothing practices, (3) the effect of winner/loser stock on the occurrence of income smoothing practices, and (4) the effect of profitability , financial leverage, winner/loser stock on the occurrence of income smoothing practices simultaneously. This research was conducted using Statistical Analysis method. The statistical analysis method used here is logistic regression analysis. Where in the statistical regression analysis test will be tested in it: (1) testing the feasibility of the regression model, (2) assessing the entire model, (3) Coefficient of Determination, (4) Multicollinearity Testing, (5) Classification Matrix, and (6) Model Formed Regression. The results obtained from this study: (1) profitability has no effect on the occurrence of income smoothing practices, (2) financial leverage does not affect the occurrence of income smoothing practices, (3) winner/loser stock does not affect the occurrence of income smoothing practices, and (4 ) profitability, financial leverage, winner/loser stock have no effect on the occurrence of income smoothing practices simultaneously.


INTRODUCTION
In Indonesia, there have been many phenomena of large companies performing income smoothing. For example, in 2015 PT Timah (Persero) Tbk (TINS) provided information on the company's financial condition that was different from what actually happened, where since 2013 the directors of PT Timah (Persero) Tbk (TINS) according to the Timah Employees Association (IKT) who came from The provinces of Bangka Belitung and Riau Islands, have made many mistakes and omissions during their three years in office since 2013, namely by providing different information to the public regarding the achievement of the company's financial condition so that they judge that the directors have committed many public lies through the media. An example is the press release of the financial statements for the first half of 2015 which states that efficiency and strategies have resulted in positive performance. In fact, in the first semester of 2015, the operating profit and loss amounted to Rp. 59 billion. This is done so that the company's performance is considered good by the public so that it can attract investors' interest in the company. For information, in addition to experiencing a decrease in profit, PT Timah also recorded a debt increase of almost 100 percent compared to 2013. In 2013, the company's debt only reached Rp. 263 billion. However, this amount of debt increased to Rp. 2.3 trillion in 2015.
Income smoothing carried out by the management of a company is influenced by several factors. One of the factors that influence income smoothing is profitability. For investors who will invest their capital, profitability is a reference to assess the company's performance because it is directly related to income smoothing. Thus, each company will try to achieve the level of profitability according to the planned target, to ensure the continuity of its business. Therefore, management performs income smoothing to avoid the possibility of periods that are less profitable for the company.
In addition to the company's profits, investors also pay attention to the company's leverage. Income smoothing is also related to the level of company leverage. A high leverage ratio puts pressure on the company so that management will take steps to shift future period profits to current period profits to reduce the high ratio (Sitorus, et al, 2014: 91).
The existence of a go public company is also reflected by the performance of the shares issued on the IDX, namely whether the company's shares are classified as winner stock or loser stock. Winner stock companies are companies whose shares have returns above or higher than market returns, while loser stock companies are companies that have returns equal to or less than market returns. Iskandar & Suardana (2016: 818) states that the winner stock company performs income smoothing to maintain its position and maintain company variability, while the loser stock company performs income smoothing to be in the winner stock position so as to increase the value of the company and investors see that the company's performance good.
Based on this explanation, the authors are interested in conducting research on the Effect of Profitability, Financial Leverage, and Winner/Loser Stock on Income Smoothing. The specific purpose of this study is for investors in making decisions to invest in a company, which in this study will show whether or not there is an influence of changes in profitability, financial leverage and winner/loser stock on the occurrence of income smoothing practices in a company.
The targets for the findings are: 1. Does profitability affect income smoothing? 2. Does financial leverage affect income smoothing? 3. Does winner/loser stock affect income smoothing? 4. Do profitability, financial leverage, and winner/loser stock have a simultaneous effect on income smoothing?

RESEARCH METHODS
This research was conducted using Statistical Analysis method. The statistical analysis method used here is logistic regression analysis, with the object of research being a manufacturing company which was reduced by using sample criteria in accordance with the required research data. And the year studied in this study was 4 years starting from 2014-2017.
Because there are variables that are dichotomous (performing income smoothing and not income smoothing, and winner stock or loser stock), the testing of the target findings is carried out using logistic regression tests. The test was carried out with a significant level (α) of 5 percent.

RESEARCH RESULTS AND DISCUSSION Logistics Regression Test
The stages in testing using logistic regression can be explained as follows (Ghozali, 2013): a. Assessing the feasibility of the regression model The feasibility of the regression model was assessed using the Hosmer and Lemeshow test. The test results are shown in table 1.

Table 1. Hosmer and Lemeshow test Hosmer and Lemeshow Test
Step Chi-square df Sig. Hosmer and Lemeshow test results show the Chi-square value of 9.224 with a significance of 0.324. Based on these results the significance value is greater than 0.05, it can be concluded that the logistic regression model is appropriate and feasible to use because it can predict the value of the observations. b. Assess the overall model (overall model fit) This test aims to determine whether the model used as a whole is good. The test is done by comparing the value between -2log Likelihood (-2LL) at the beginning (block number = 0) with a value of -2log Likelihood (-2LL) at the end (block number = 1). If there is a decrease in the value of -2 log likelihood, it can be said that the logistic regression model is good to use. The test results are shown in table 2: Based on the results of the overall model test above, there was a decrease in the value of -2 log likelihood of 5,699, from 175,911 to 170,212. So, it can be concluded that the overall regression model used is a good model.

Coefficient of Determination (Nagelkerke R Square)
The value of the coefficient of determination in the logistic regression model is indicated by the value of Nagelkerke R Square. The test results are shown in table 3: The value of Nagelkerke R Square is 0.058, which means that the variability of the dependent variable that can be explained by the independent variable is 5.8%, while the remaining 94.2% is explained by other variables outside this research model.

Multicollinearity Test
The multicollinearity test aims to determine the correlation or relationship between one independent variable and other independent variables. A good regression model is a regression with no symptoms of a strong correlation between the independent variables. This test uses a correlation matrix between independent variables to see the magnitude of the correlation between the independent variables. The test results are shown in table 4: The test results show that there is no correlation coefficient value between variables whose value is greater than 0.9, so it can be concluded that there are no serious multicollinearity symptoms between independent variables.

Classification Matrix
classification matrix will show the predictive power of the regression model to predict the possibility of receiving a going concern audit opinion on the auditee. The test results are shown in table 5:  Table 5 shows that according to predictions, there are 57 companies that do not perform income smoothing, while from the observations there are only 12 companies that do not perform income smoothing with a classification accuracy of 21.1% (12/57). While the companies that perform income smoothing are 71 companies, but from the observations there are only 61 companies that perform income smoothing with an accuracy of classification of 85.9% (61/71) or overall classification accuracy is 57%.

Formed Regression Model
The logistic regression model that is formed produces regression coefficient values and significance. The regression coefficient of each of the tested variables shows the form of the relationship between the variables. The results of the regression formed can be seen in  Table 6 showing results testing with regression logistics on level error 5 percent. Results testing logistics produce model as following :

Discussion
Test hypothesis target findings conducted with method compare Among level significance (sig) with error rate (α) = 5%. If sig < 0.05 so target findings could said take effect, whereas If sig > 0.05 then the target of the finding can be said no take effect. Based on Table 4.6 can interpreted results as following in:

Does profitability affect income smoothing?
The target of the first finding is to find out whether there is an effect of profitability on income smoothing. From table 4.6 above, it is obtained that profitability (ROA) shows a negative regression coefficient of 4.144 with a significance level of 0.062 greater than (5%), so it can be concluded that profitability has no significant effect on income smoothing.

Does financial leverage affect income smoothing?
The target of the second finding is to find out whether there is an effect of financial leverage on the income smoothing audit opinion. From table 4.17 above, financial leverage (DER) shows a negative regression coefficient of 0.146 with a significance level of 0.485 greater than (5%), it can be concluded that financial leverage has no significant effect on income smoothing.

Is it a winner/loser stock take effect to alignment profit?
The target of the third finding is to find out whether there is an effect of winner/loser stock on income smoothing. From table 4.6 above, the winner/loser stock (wls) shows a positive regression coefficient of 0.030 with a significance level of 0.940 greater than (5%), it can be concluded that the winner/loser stock has no significant effect on income smoothing. Do profitability, financial leverage and winner/loser stock have a simultaneous effect on income smoothing?
The target of the fourth finding is to find out whether there is an effect of income smoothing, financial leverage, winner/loser stock on income smoothing. From table 4.7 above, the significance of 0.175 is greater than (5%), it can be concluded that the profitability of financial leverage, winner/loser stock has no simultaneous effect on income smoothing.

CONCLUSIONS AND RECOMMENDATIONS
From the results of research that has been carried out up to 80%, it can be concluded that profitability does not partially affect income smoothing, financial leverage does not partially affect income smoothing, winner/loser stock does not partially affect income smoothing, and profitability, financial leverage and Winner/Loser Stock have no simultaneous effect on Income Smoothing. from these results it can be seen that there are many other factors that can be a factor in the occurrence of income smoothing practices in a company. And this can be seen from the results of the determination where the variables studied at this time only have an influence on the occurrence of income smoothing practices by 5.8%, and 94.2% again there are other variables that affect the occurrence of income smoothing.