The Influence of Earning Management, Dividend Policy, and Free Float Ratio on Stock Returns

: Analysis of company performance is one of the considerations for investors in making investment decisions. Several issues related to earnings management arise due to the flexibility in accounting standards application becomes an investor’s special concern in analyzing financial statements. On the other side, investors’ attention to dividend distribution has various effects on several studies. Another issue that also attracts the investors’ attention is the new rules of minimum free float ratio for companies listed on the IDX to be 7.5% in 2019. Based on this phenomenon, this study aims to examine the effect of earnings management, dividend policy


INTRODUCTION
To provide convenience and realize fairness in the preparation of financial reports, management has the flexibility to apply accounting.However, Adiwibowo (2018) says that there is a risk of flexibility in the application of accounting in the form of opportunities for management to manipulate earnings better known as earnings management.This opportunity can become greater if the specified profit target cannot be met.This condition increases monitoring costs due to differences in interests between owners and management (Berle & Means, 1932) as research results by (Weinstein, 2012) stated that agreements between shareholders and external parties such as policymakers and public opinion will control and direct management behavior.
The property and real estate sector is one of the vulnerable sectors due to earnings management (Abbas et al., 2018).This is because the property and real estate sector was rapidly changing, persistent, and had intense competition.The same thing was conveyed by (Nurbaiti & Suatkab, 2019) that the property and real estate sector had several fraudulent financial statement cases (financial statement fraud) which is the largest from year to year.
Several frauds in the property and real estate sector were reported by Dwidinda et al. (2017) which stated that the Chief Executive Officer (CEO) of Cabot Investment Properties (CIP) in the United States committed fraud against elderly investors and misused US$17 million of these investors' funds for his personal needs.Apart from the CIP case, securities law firm (Edward & Kantas, 2019) wrote that American Realty Capital Property (ARCP) in 2014 admitted to accounting errors amounting to US$23 million at the end of 2014 and made investors sell their shares.
The shares received by investors can be influenced by the company's earnings management (Istiqomah & Adhariani, 2017).This phenomenon happened when the share price of PT Hanson International Tbk declined after news emerged that the company's main director was involved in the profit manipulation case of PT Asuransi Jiwasraya.In this case, the Financial Audit Agency (BPK) projected that potential losses due to the PT Jiwasraya Insurance cases reached IDR 16.8 trillion (Putri, 2020).Apart from the involvement of the main director, PT.Hanson International Tbk has a potential overstatement revenue of IDR 613 billion over the Sale and Purchase Agreement (PPJB) for ready-to-build plots which have not been disclosed in the financial statements.In this period, the share price of PT.Hanson International Tbk (MYRX) continued to fall and touched the lowest share price permitted by the Indonesian Stock Exchange (BEI) at IDR 50 in 2019.Figure 1 shows the drastic decline in MYRX share prices since 2018.The influence of earnings management on stock returns has been analyzed in several studies.The research of (Istiqomah & Adhariani, 2017) and (Salim & Rusman, 2019) shows that earnings management has a significant negative influence on stock return.However, different results were shown by other studies.(Adiwibowo, 2018) in his research concluded that earnings management does not have a significant influence on stock return, while the research of (Alami, 2021) shows that earnings management has a significant influence with a positive value on stock return.Various results on stock return research allowed for carrying out similar research for several industrial sectors.
Apart from earnings management, in several studies, another factor that can influence stock return is a dividend policy.In his hypothesis, (Gordon, 1959) states that the connection between profit strategy and stock costs can be affected by how much data is contained in profit circulation declarations.This is in line with (Ouattara, 2017) that profit declarations on normal outcomes in a decrease in share costs during the occasion time frame, which is characterized as the exchanging a very long time when the declaration.Investors were dissatisfied with the dividend yield, which led to a decline in share prices following the announcement of the dividend.If there is a lot of information contained in the dividend distribution announcement and the information is interesting to investors, the share price will increase.One of the information contained in dividend distribution is the amount of dividends distributed/Dividend Payout Ratio (DPR).
The dividend policy had a significant positive effect on stock return (Astarina et al., 2019).However, these results have different conclusions from (Sari, 2016) which dividend policy has a significant negative effect on stock return.On the other hand, the dividend policy of each company was different even in the same industry.The dividend policy for the Property and Real estate sector is shown in Table 1.  1 indicates that a large DPR value does not mean will produce a higher result on stock return.Table 1 shows that Alam Sutera Realty (ASRI) stock return increased by 1.68% even though they did not distribute dividends during 2020.Meanwhile, Metropolitan Land (MTLA), which distributed 17.71% of its profit, experienced a decline and resulted in a decrease of -25.35% in stock returns.These relationships open up research opportunities for dividend policy effect on stock return in the property and real estate sector.
Apart from earnings management and dividend policy, another variable thought to influence stock return is a free float ratio.(El-Nader, 2018) in his research stated the free float ratio can attract investors referring to the concept that the larger free float ratio can minimize stock price games carried out by investors with large funds.However, research on free float ratios to stock return is https://www.ilomata.org/index.php/ijtcstill limited.Several studies, such as (Bostanci & Kilic, 2010;Wang & Xu, 2004), concluded that the free float ratio has a significant positive effect on stock return.However different results were presented by (Çalişkan & Keresteciȯğlu, 2013) that show the free float ratio does not have a significant effect on stock return.
The changes in the minimum free float ratio limit for IDX are shown in the Decree of the Board of Directors of PT Bursa Efek Indonesia Number 59/BEI/07-2019.Regulation Number I-V concerning Special Provisions for the Listing of Shares and Equity Securities states that the number of listed company shares owned by the public who are not controlling at least 7.5% of the number of shares that constitute the paid-up capital of the company.
The relationship of dividend policy effects on stock return, the changes in free float ratio, and many cases of earnings management, especially in the property and real estate sectors, allowing conducting this research.This research seeks to examine the relationship between earnings management, dividend policy and free float ratios to stock returns in the property and real estate sector listed on the Indonesian Stock Exchange (IDX).Hopefully, this research can provide additional literature for investors and companies in the property and real estate sector.To maintain the level of accuracy and control over the influence of variables not used in the research, control variables were used in this research.The control variables used in this research are the profitability ratio and firm value.
The different interests of the company owner (ownership) and management as supervisors of company performance resulted in costs (Berle & Means, 1932).In the following period, Berle and Means' research was refined by Jensen and Meckling (1976) born the agency theory.In agency theory, the company management (agent) will be given authority on behalf of the shareholders (principal) as a decision maker in running a company with compensation in the form of incentives received in the form of salaries, bonuses, and other incentives (Jensen & Meckling, 1976).
The incentives given to agent performance based on financial reports will encourage agents to manipulate profits in financial reports (Healy, 1985).The emergence of an agent authority in agency theory can cause limitations on the principal information add-on.In this situation, The principal doesn't know what the agents doing related to company performance.This is a form of information asymmetry known as moral hazard (Scott, 2015).
The indication of earnings management carried out by the company is a signal for investors to take future steps.The signals are an effort to provide information to other parties regarding an accurate picture of the company's performance and any problems they have so that other parties are willing to invest even under uncertainty (Spence, 1973).A good signal from the company will be a consideration for investors when investing in a company.In brief, this will influence on stock return to be received.Scott (2015) defines earnings management as the selection of accounting policies by management for certain purposes.Furthermore, earnings management practices themselves are divided into two (2), namely efficient and opportunistic contract earnings management practices.Earnings management practices that are efficient contracts can improve the quality of financial information presented, while opportunistic nature can cause losses to readers because they receive information that does not match reality (Istiqomah & Adhariani, 2017).
Information obtained from the presentation of financial reports will be a consideration for investors' decisions.Opportunistic earnings management practices can lead to wrong decisions.In this case, investors are at risk of experiencing losses due to this decision.Information about opportunistic earnings management practices can alienate investors from the company.Thus, this research would hypothesize that earnings management has a significant negative influence on stock return.This hypothesis is in line with (Adhariani, 2021;Salim & Rusman, 2019).The first hypothesis of this research is Earnings management has a significant negative effect on stock returns (H1).
Apart from earnings management, (Bhattacharya, 1979) stated dividend policy is also a signal for investors.Many investors use certain preferences in assessing a company's dividend policy.(Modigliani & Miller, 1961) revealed the use of preferences in assessing dividend policy by investors, referred to as client effect theory.Some of the investor preferences are the tax rates on dividend policy proposed by (Litzenberger & Ramaswamy, 1979), the different conditions for each investor proposed by (Pettit, 1977), and the company sector proposed by (Baker & Powell, 1999).
According to Gup & Agrrawal, (1996), the amount of dividends distributed from the company's profits will depend on the company's life cycle.The company life cycle period is divided into 4, namely pioneering, expansion, stabilization, and decline.In the pioneering stage, companies generally do not distribute dividends.Then, in the expansion stage, the company will start distributing dividends in small amounts.And, lastly, getting closer to the cycle decline then the dividends distributed will be even greater.
According to Wijaya & Tel (2017), the dividend policy is a company policy to use the company's net profits from both the previous year and the current year.The dividend is a return given by the company to shareholders so that dividends and their policies are taken into consideration by shareholders.This is approved by (Astarina et al., 2019;Retnaningrum & Haryanto, 2018), who concluded that the dividend policy variable had a significant positive effect on stock returns.Based on this, the second hypothesis in this research is Dividend policy has a significant positive effect on stock returns (H2).
Apart from dividends, shareholders can also get returns on share ownership from the capital gain of share price differences.In practice, regulation of share prices can also occur.(Biswal, 2003) reveals that companies with large market capitalization and do not comply with free float ratio regulation can lead to manipulation of stock prices by groups of investors with large funds.The free float is the number of shares traded on the stock market minus shares held by the parent company which is used to exercise control over subsidiaries (Çalişkan & Keresteciȯğlu, 2013).
In other words, (Bostanci & Kilic, 2010) stated that the free float ratio is the number of shares released by the company for buying and selling on the stock market.While the free float ratio is felt it can attract the attention of investors because the greater the free float ratio of a company, the public ownership of the company becomes greater (El-Nader, 2018).According to (Fitriani et al., 2020), the higher the free float ratio it has the potential to increase the company's market capitalization.This happens if the number of shares circulating in the public increases, the company will receive a greater amount of funds.Thus the third hypothesis of this research is: The free float ratio has a significant positive effect on stock returns (H3).
To obtain more solid results, this research uses profitability and company value as control variables.Arifin & Agustami, (2017) revealed that profitability ratios provide an overview of the company's ability to generate profits in a certain period based on the assets or capital owned by the company.The results of (Dana & Artini, 2021) show that the profitability ratio proxied by ROE shows a significant influence with a positive value on stock return.
Apart from company performance, company value is also a concern for investors in making decisions.Utami (2017) revealed that the public's assessment of companies that have good performance can be shown by the amount of company value.This can be seen when the public assesses the market price of a company's shares above its book value.In line with this, Tobin's Q is often as a measuring tool in assessing companies.(Rahman & Khairunnisa, 2016) stated that Tobin's Q indicator provides an overview of management performance in managing the assets owned by the company and can also illustrate the potential for company growth.Previous research found that Tobin's Q, which proxies company value, has a significant positive influence on stock returns (Kurniadi et al., 2022;Rahman & Khairunnisa, 2016).

METHOD
This research uses panel data regression of property and real estate sector companies listed on the IDX.The data used in this research are financial reports obtained from the official IDX website (www.idx.co.id) as well as the company's official website.The data used is limited to the period 1 January 2016 to 31 December 2019.The selection of research samples was carried out by using purposive sampling on criteria: registered on the IDX, published financial reports during the research period, and free float ratios greater and/or equal to 7.5%.The total sample used in this research was 172 samples.
The data processing method used in this research includes outlier tests, descriptive statistical analyses, selection of panel data regression models, Gaus Markov theorem tests, and hypothesis tests.The outlier tests use data conversion for each variable into a score standardized (Z-Score) with a maximum limit of 2.58.Descriptive statistical analysis uses the minimum, maximum, average, and standard deviation values of each variable.Astarina et al. (2019) who define it stock return as the investors receive from the difference in stock prices at the close of the stock market during the observation and reduced by the previous year's closing market price divided by the previous year's closing market price.The closing market price used is the closing market price when the financial statements are published.The choice of market closing price refers to signal theory, that investors will respond to the condition of a company based on signals sent by the company shown by financial reports.The closing price of shares when the financial statements were published the previous year Furthermore, earnings management in this research is proxied using discretionary accrual which is depicted through the Modified Jones Model.This model uses the difference between the total accrual and non-discretionary accrual.According to Dechow (1994), the Modified Jones Model has the highest level of accuracy among other earnings management detection models.Kusuma (2020) states that the Modified Jones Model is used in calculating discretionary accrual carried out with the following steps: 1. Total Accrual (FACINGit) is calculated using: TAt = NIt -CFOt 2. Accrual value which is estimated using the OLS regression equation (Ordinary Least Square): TAt/At-1 =  1 (1/At-1) +  2 (ΔREVt/At-1) +  2 (PPEt/At-1) + e 3. Non discretionary accrual value which is calculated using the formulation: NDAt =  1 (1/At-1) +  2 (ΔRevt/At-1-ΔRect/At-1) +  2 (PPEt/At-1) 4. Determine the value of accrual discretionary which is an indicator of earnings management with the equation: DAt = TAt/At-1 -NDAt Information: Tat : Total value of company accruals at the end of year t.In this research, the dividend policy variable refers to Wijaya (2017), Fitri (2018), Retnaningrum and Haryanto (2018) which uses a proxy Dividend Payout Ratio (DPR) to describe the company's dividend policy.According to Astarina et al. (2019), DPR is the ratio between the dividends to be distributed and the net profit owned by the company.The formula for calculating DPR is as follows: The free float ratio used in this research refers to Suwandi (2020), namely the number of shares circulating in the community with ownership by the community below 5% per individual at the end of the year.However, in this research, modifications are taking into account IDX provisions that the minimum number of shares circulating in the public reaches 7.5%.The calculation of the free float ratio used in this research, is as follows: The profitability control variable in this study uses Return On Equity (ROE).According to Aryanti et al. ( 2016), ROE is the level of profitability of a company by taking into account the level of profit generated based on the amount of equity owned by the company.ROE measurement uses the following ratio: Meanwhile, the company value in this research uses Tobin's Q ratio proxy.Tobin's Q looks at the company value by calculating a combination of market capitalization, liabilities, and assets owned by the company.According to Rahman and Khairunnisa (2016), Tobin's Q can be an indicator of a company's growth potential.Tobin's Q measurement uses the following ratio:

RESULT AND DISCUSSION
Based on data, descriptive statistical analysis shows that on average companies in the property and real estate sector have a negative value of stock return or capital loss of -10.4%.Meanwhile, earnings management has an average value of 0.0249, which indicates that companies in the property and real estate sectors are indicated to increase their company profits.In the dividend policy data, it was found that the sample data had negative values.This indicates that the companies in the property and real estate sector distribute dividends even though they experienced losses during the period.The free float ratio variable has an average value of 33.81%, indicating that on average companies in the property and real estate sector own 33.81% of company shares traded in the public.
As between variables other than the dependent variable with a value limit of 0.90.Based on the results, the correlation between variables other than the dependent variable was below 0.90.This shows that there is no multicollinearity problem between variables other than the dependent variable.
The heteroscedasticity test in this study uses a Glazes test by carrying out a regression on existing variables other than the dependent variable on the absolute value of the residual and comparing it with the existing alpha (α) value.The α value used is 0.05.Based on the results, the residual absolute value of each variable was above 0.05 which indicates that there was no heteroscedasticity problem in the research model.
According to Ghozali (2017), the more appropriate autocorrelation test to use in research with sample sizes above 100 is the Breusch -Godfrey Test (BG Test).The autocorrelation test was carried out using the Breusch-Godfrey serial correlation method by comparing the probability value chi-square with the alpha value (α).The α value used is 0.05.Based on the results, the probability value chi-square exceeds the α value means there is no autocorrelation problem.
The next stage is the panel data regression test which consists of a simultaneous significance test and a partial significance test, then the test continues with the coefficient of determination.The research model used is the Fixed Effect Model (FEM).The results of the panel data regression tests are shown in Table 2.  2 shows that the probability value of the research model is 0.000086 or below 5%.It means that earnings management, dividend policy, and free float ratios simultaneously have a significant effect on stock return in the property and real estate sector for the period 2016 to 2019.In addition, Table 2 shows that the Adjusted R-squared value was 0.364909 or around 36.5%.It means, the research model can explain 36.5% of the dependent variable, while the remaining 64.5% is explained by variables that are not included in the research variables.Table 2 also shows that the regression equation model used in this research is as follows: Rit = -0.43666-1.352279 DAit -0.221536 DPRit -0,003272 Ffit + 2,081839 ROEit + 0,241416 Tobinit + error Based on the partial significance test, Table 2 shows that only earnings management (DAit) has a significant effect on the dependent variable, whether the other two (2) independent variables are dividend policy (DPRit) and free float ratio (Ffit) has no significant effect on the dependent variable.The control variable is profitability (ROEit) and company value (Tobinit), the results of this study show that both variables have a significant effect on the dependent variable.Significant results on the control variables indicate that the independent variables used in the research have constant values and are protected from the influence of external factors examined in the research.
The results of this study show that earnings management has a significant negative effect on stock returns in the property and real estate sector from 2016 to 2019.The results prove that the indications of earnings management as proxied by the Modified Jones Model are relevant to investor decision-making.
These results are in line with agency theory.In agency theory, the agent's interest causes management to hide information from shareholders regarding the company's future performance and prospects (Godfrey, 2010).Healy (1985) states that differences in information held by management and shareholders cause information asymmetry.One of the information asymmetries is moral hazard, when shareholders do not comprehensively know the condition of the company, while company management can manipulate the company's financial reports (Scott, 2015).Jensen and Meckling (1976) stated that agency theory could raise agency costs, including residual loss.
Referring to the result, earning management as an information asymmetry is a negative signal for investors.As Spence (1973), found that the signals can be an illustration of providing information to investors regarding the company's future performance.In this research, the information in the form of an earnings management indication which is calculated using the modified Jones model is a negative signal for investors.When a company's financial report is published and investors think there are indications of earnings management, investors will try to stay away from the company.https://www.ilomata.org/index.php/ijtcThis can cause a decline in stock return.The results of this research are in line with Istiqomah and Adhariani (2017) and Salim and Rusman (2019).
In the dividend policy variable tests, the results do not support the research hypothesis.The results show that the dividend policy proxied by the Dividend Payout Ratio (DPR) does not have a significant influence on stock return.Thus, the results of this study are not in line with Salim and Rusman (2019), Astarina et al. (2019), andZulfrida (2021).However, the results support Effendi and Hermanto's ( 2017) and (Dewi & Yudowati, 2020) results.
There are several indications why the results of the research carried out differ from the initial research hypothesis.The first indication is the difference in sectors used in the research sample.Astarina et al. (2019) and Gunawan (2019) conducted research in the manufacturing sector which showed that dividend policy has a significant positive influence on stock returns.Meanwhile, Pratiwi (2015) and Yuliana ( 2020) who conducted research in the consumer goods sector found that dividend policy had a significant negative effect on stock returns.However, research by Luthfiyanto and Isynuwardhana (2019), and Dewi and Yudowati (2020) in the mining sector shows the same results as the results of this study that dividend policy does not have a significant effect on stock return.
The different effects of dividend policy in different business sectors are confirmed by the suitability of research results in the property and real estate sectors in the previous period.Some studies on the property and real estate sector were conducted by Effendi and Hermanto (2017) from 2011 to 2014, and Widiarini and Dillak (2019) from 2013-2017 period.These two studies show that dividend policy as proxied by the DPR does not have a significant effect on stock return.
Different results indication to be due to differences in business sectors can be explained by the client effect theory.Baker and Powell (1999) show that one of investors' preferences regarding dividend policy is the sector that is the object of investment.In brief, the property and real estate sector in Indonesia does not consider dividends for investors' decisions.It can be seen from the amount of net profit distributed in the form of dividends.
Based on descriptive statistical data, the average DPR value in this study was 14.01%.Several studies show that differences in the amount of dividends distributed also influence the research results.(Gunawan, 2019) and Saputri (2018) conclude that a dividend policy with an average DPR value above 35% has a significant positive influence on stock return.In another study, Widiarini and Dillak (2019) concluded that dividend policy does not have a significant effect on stock returns.In this research, the average DPR used in the research was 22%.
Apart from the business sector, differences in the amount of dividends distributed are also influenced by the company's life cycle.Gup and Agrrawal (1996) stated that the company would distribute fewer dividends during the expansion period.The company will save its net profit to increase the company's value in the future.In this case, companies in the property and real estate sector are in the expansion phase.The company is in the stage of growth and development so it requires more assets in its operations.Saving the current year's profits and the previous year's profits from being distributed as dividends is important to finance the company's operational development.https://www.ilomata.org/index.php/ijtc The other results that do not support the research hypothesis are also found in the free float ratio.This study shows that the free float ratio of property and real estate sector companies in Indonesia does not have a significant influence on stock.Thus, the results of this study contradict the research results of Wang and Xu (2004) and Bostanci and Kilic (2010) which stated that the ratio free float has a significant influence on return share.However, the results of this study support Çalişkan and Keresteci's research (2013).
The different results to the research hypothesis are thought to occur due to differences in the proxies used.Wang and Xu's (2004) show that the free float ratio proxy used monthly data, while in the research of Bostanci and Kilic (2010) and Çalişkan and Keresteci̇ Oğlu (2013) uses a sample in the form of daily data.In this research, the sample used is annual data, is an indication of research results that are different from previous research.
The different results on the free float ratio variable may arise due to the company's action.The Free float ratio variables may respond to corporate actions.Corporate actions that cause changes in free float ratios is the issuance of new shares (right issue) as well as share buybacks (buyback).These corporate actions generally only occur during certain periods or do not occur continuously in the long term.The proportion of changes in the free float ratio in the research period can be seen in Figure 2. Figure 2 shows that 50% of the samples do not have experienced in changes the free float ratio consecutively during the research period.This indicates that some companies do not carry out corporate actions every year.Corporate actions can cause changes in the free float ratio.Figure 2 also shows that there are 18% of companies that have not experienced a change in the free float ratio for 4 years.It may be an indication of the differences in research results compared to the research hypothesis.

CONCLUSION
The results show, first, that the earnings management variable has a significant negative effect on stock return.These results fit Istiqomah andAdhariani (2017), andSalim andRusman (2019) as well as several cases of earnings management such as CIP, ARCP and MYRX.Second, in this research, the dividend policy variable does not have a significant influence on stock return.The results support the client effect theory, in which the existence of investor preferences regarding sector and time in looking at the amount of dividends that will be received and considered in making investment decisions.Apart from that, the DPR size for property and real estate sector companies is also relatively small.In this case, it is suspected that property and real estate sector companies are in a phase expansion so the dividend distribution is relatively small.This result also supports the company life cycle theory.Furthermore, in the third variable test, it is found that the free float ratio does not have a significant effect on stock return.In this case, there are differences between the research results and the research hypothesis.The discrepancy between the results and the research hypothesis is thought to be due to differences in time proxies and countries in the research sample.This research has several limitations, such as the sample size is limited to the property and real estate sector for the period 2016 to 2019 so it cannot be concluded about all companies listed on the Indonesian Stock Exchange.Apart from that, in this research, the free float ratio variable used is limited to using ratios and does not use further variables in the form of adjustment index free float ratio.Thus, in future research, the analysis can be carried out on differences in sectors and data proxies selected in the research period.For investors, the results of this research can be a reference for early identification of earnings management using existing financial report data.In terms of dividend policy, this research shows that dividend policy does not have a significant influence.However, considering the data in the research, investors can consider the pattern of dividend distribution amounts for each sector in making their investment decisions.This also applies to the free float ratio variables test which shows an insignificant effect on stock returns.However, research on the free float ratio in Indonesia is still limited.This is an opportunity for carrying out further research.

Figure
Figure 1.MYRX Share Price Source.Processed data companys' Pit : The closing price of shares when the financial statements are published P it-1 :

Figure 2 .
Figure 2. Changes in the company's free float ratio 2016-2019 Source: data processed

Table
The selection of the regression model determines the best model between Common Effect Model (CEM), Fixed Effect Model (FEM) and Random Effect Model (BRAKE).Gaus Markov theorem tests include the normality test, multicollinearity test, heteroscedasticity test, and autocorrelation test.Hypothesis tests are carried out by the feasibility tests of the model (F test) and partial significance test (T-test).The model developed in this research can be formulated as follows.Number of earnings management of company i at the end of year t.DPRit : Total dividends of company i at the end of year t.: ROE ratio of company i at the end of year t.Tobinit : Company value ratio i at the end of year t.E : error In this research, stock return refers to Ffit : Total free float of company i at the end of year t.https://www.ilomata.org/index.php/ijtcROEit a test of the research model, this study uses the Chow test, Hausman test, and Lagrange multiplier tests (lm test).Based on the results, it was found that the Chow test indicated Fixed Effect Model was more precisely chosen than the Common Effect Model.Apart from that, the results of the Hausman test also indicate that the Fixed Effect Model is more precisely chosen than the Random Effect Model.It shows that the Fixed Effect Model is the best model that can be used in this research so the Lagrange multiplier test (LM test) is not applied.The Gauss-Markov theorem test begins by carrying out a normality test.The normality test is carried out by comparing the probability values of Jarque Bera.In the results, the probability value of Jarque Bera is above 0.05.It means that the research does not have any normality problem for the research model used.The next test was carried out with a multicollinearity test.The multicollinearity test in this study used a correlation table Pearson where observations are made through a correlation matrix