THE EFFECT OF SALES GROWTH AND FIXED ASSET INTENSITY ON TAX AVOIDANCE MODERATED BY INSTITUTIONAL OWNERSHIP

0


INTRODUCTION
Until now, tax revenue still plays a role as the largest contributor to state revenue, which is part of the structure of the Indonesian State Budget (Wardani & Nugrahanto, 2022). Tax is a form of mandatory contribution made by individuals and entities to the state with a coercive nature and does not provide direct rewards but is used for state purposes for the country's development process. Especially for Indonesia as a developing country that always tries to improve national development for the welfare of the community. Taxes are one of the sources of national development funds and public contributions to the state for national development (Oktaviyani & Munandar, 2017). The statement is stated in the Law of the Republic of Indonesia Number 28 of 2007. Based on data obtained from the Central Statistics Agency, in the last three years tax revenue has dominated the source of state financial revenue as stated in the table below. The huge role of taxes in the eyes of the state causes the state to continue to strive to develop regulations that can support the growth of state tax revenue. Various systems were also developed in order to facilitate the process of reporting and paying taxes and accommodate the needs of taxpayers. Of course, this aims to encourage taxpayers to be more obedient to taxes so that tax revenues are higher and the country's growth is more advanced.
Unlike countries that consider taxes as profitable, taxpayers actually see taxes as an obligation that can reduce company assets for the benefit of the state (Irawati et al., 2020). Taxes are considered a burden that can reduce the profits and income of economic activity actors (Nindita & Budi, 2022). The perspective of taxpayers has led to the rampant practice of tax avoidance among business people who aim to reduce the amount of tax burden that will be paid to the state. Companies will try to find weaknesses in existing regulations and then take appropriate business steps to reduce the tax burden. These measures are known as tax avoidance measures (Rahmawati et al., 2016).
Similar tax avoidance practices have also occurred in Indonesia, namely in the company PT Bentoel Internasional Investama. The British American Tobacco (BAT)-owned company is claimed to have made an economic contribution to offset the enormous health costs. As reported by the Tax Justice Network Institute on (Kontan.co.id, 2019) that BAT has diverted most of its revenue out of Indonesia using two ways, namely: through intra-company loans between 2013 and 2015 and the second way through repayments back to the UK for royalty, fee and service payments. Bentoel took many loans from his Dutch affiliate Rothmans Far East BV on the grounds of financing bank debt and paying for the purchase of machinery and equipment. In addition, Bentoel also makes payments back to the UK for royalties, fees and IT costs at a rate of US $ 19.7 million per year. Both of these significantly exacerbated Bentoel's losses in Indonesia. In fact, these two costs are equivalent to 80% of the losses suffered by Bentoel before 2016.
Tax avoidance practices are generally initiated by company managers to meet the expectations of shareholders. Managers try to get the largest profit figures in the financial statements because the greater the profit reported in the financial statements, the greater the reciprocity that managers will receive from company owners. Management no longer carries out tax responsibilities as it should because their personal interests take precedence over the public interest, especially the public. Seeing how the state is so ambitious in its efforts to increase tax revenues to support state growth due to the need for greater state costs and financial support from taxpayers is one way out that is expected to save the country from the economic crisis. A much different perspective can be found from the side of companies that are now facing an onslaught of economic problems and are faced with the possibility of an increasingly real recession. Various efforts are made by the company in order to maintain existence in the business world. Both statements will encourage more tax avoidance cases in the future. Until now, tax avoidance is still a problem that is growing every day and is the concern of economic actors in the world, especially Indonesia. This then prompted the author to conduct further research on tax avoidance. Tax avoidance is driven by various factors, including the amount of company debt, high profits, and an increase in sales received (Arinda &;Dwimulyadi, 2019). In essence, every company certainly has the same goal, which is to pursue maximum profits, especially in companies that are experiencing an increase in sales. The company will certainly try to keep sales growth always leading to positive and better things. When the company has high sales, of course, the profit received will also increase so that the tax responsibility owned becomes greater (Nugraha & Mulyani, 2019). Companies with rapid sales growth certainly need large financial support to fund this growth. This causes companies to tend to carry out tax avoidance to reduce the amount of tax burden that must be paid to the state and increase profits to be allocated to efforts to increase company investment. Based on the results of research from (Nugraha & Mulyani, 2019) it is stated that sales growth or sales growth has a positive effect on tax avoidance. Similar results were also obtained from research (Januari & Suardikha, 2019), namely that sales growth has a positive effect on tax avoidance. Another case with research conducted by (Irawati et al., 2020) which concluded that sales growth has a significant negative effect on tax avoidance. However, the results of the study (Astuti et al., 2020) shows the opposite, namely that sales growth has no effect on tax avoidance. Various inconsistencies arising from the results of previous research cause sales growth to need further investigation.
Taxes as the largest contributor to state tax revenue to date cause research on taxes is always interesting to investigate. The amount of fiscal effort in dispelling tax avoidance practices encouraged researchers to conduct this study. This research is expected to provide input to the Directorate General of Taxes in assessing which parts must be considered by the government in preventing tax avoidance practices that can harm the state. The results of the study can also be used by external parties in order to analyze financial statements more thoroughly and not solely trust existing financial statements because there may be other behaviors carried out by managers and underlying the financial statements.
Based on the background previously described, the researcher aims to conduct a study entitled "The Effect of Sales Growth and Fixed Asset Intensity on Tax Avoidance with Institutional Ownership as a Moderation Variable".

RESEARCH METHOD Research design
This research is a type of quantitative research. The quantitative method used in this study consists of testing the effect of sales growth and fixed asset intensity on tax avoidance with institutional ownership as a moderation variable.
This research design uses a causalexplanatory design. In this study, a causal design was used to test the relationship between independent variables, namely sales growth and fixed asset intensity, to the dependent variable, namely tax avoidance with a moderating effect of institutional ownership.
The type of data used in this study is in the form of secondary data originating from the Indonesia Stock Exchange (IDX) through the www.idx.co.id page. The data used comes from the company's financial statements.

Population and Sample Selection Techniques
The population of this study is manufacturing companies listed on the Indonesia Stock Exchange (IDX) in the 2018-2020 period. The reason for choosing the population in the form of manufacturing companies is because until now manufacturing companies are still considered as one of the largest tax-contributing industrial sectors in Indonesia. Sample selection by purposive sampling method. Purposive sampling is a method of selecting samples by making several considerations so that the data obtained can be more representative of this study. The 135 observations consist of 45 manufacturing companies from three different sub-sectors, all of which were listed consecutively from 2018 to 2020 on the Indonesia Stock Exchange (IDX). The selection of 45 companies is based on several criteria needed in obtaining the necessary data and reflecting the appropriate research subjects in order to obtain the expected research results.

RESULT AND DISCUSSION Description of the object of study
Based on the selection results using predetermined criteria, a total of 135 companies were obtained as data in this study. In processing data and tabulating data that has been collected, researchers use Microsoft excel 2010 and Economic Views software (EViews) version 12. Data description in the form of mean value, maximum, minimum, and standard deviation value from the data pentabulation results. This study succeeded in obtaining the results of descriptive statistical tests as follows:  Table 1 above shows the results of descriptive statistical tests on this study where the study used 135 observations. The observations were taken from a sample of manufacturing companies in 2018-2020 registered and listed on the IDX. Descriptive statistical tests were carried out on the variables used, namely tax avoidance as the dependent variable (Y), sales growth and fixed asset intensity as independent variables (X1 and X2), followed by moderation variables in the form of institutional ownership (M) and equipped with two control variables, namely company size and leverage. This study uses two proxies of tax avoidance because the purpose of this study is to compare the effect received by tax avoidance through conforming tax and non-conforming tax approaches. For tax conforming proxies, current ETR is used as one of the proxies that is considered to best describe the position of tax avoidance in the company. Non-conforming tax itself is represented by a TaxOCF proxy that compares taxes paid with operating cash flow. Table  1 presents the two proxies together. From the results of the chow test above, equation 1 (without moderation) and equation 2 (with moderation) found results that are both below the alpha value of 5% (0.05) so it is concluded that H0 from both equations is rejected so that the fixed effect model is considered more recommended to be used. The same results are also obtained from the results of the chow test using non-conforming tax avoidance variables where the value of prob. cross-section F shows the number 0.0000 or prob<0.05 so the more recommended model is the fixed effect model.   Unlike the results of the hausman test on conforming tax avoidance, the test results on nonconforming tax avoidance show a prob value. 0.0241 in equation 1 where the result of prob<0.05 and the value of equation 2 is 0.1238 so that prob.>0.05. Seeing this difference, H0 is considered rejected so that the best model is the same fixed effect model as used in conforming tax avoidance. The reason is because the results of the chow and hausman tests have produced the same model, namely the fixed effect model, the model selection test has been stopped and produced a fixed effect model as the best model so that in order to obtain the best results, the same model is used in both studies.

Classical Assumption Test a. Multicollinearity Test
In the multicollinearity test, it is expected that variables can be tested regarding the presence or absence of entanglement between the variables used. This test will test the correlation of each independent variable used in this study. According to (Ghozali, 2018), several causes of correlation between independent variables can be found, including: (i) sampling method as a method of collecting data; (ii) the emergence of constraints on models and populations; (iii) model selection specifications; (iv) the number of independent variables greater than the observations used. In order to be declared free of multicollinearity symptoms, the correlation value found between independent variables must be lower than 0.8. This study has also conducted a multicollinearity test with the following results: The results of multicollinearity testing in this study can be seen in table 7 above. For the variable sales growth against fixed asset intensity and vice versa between fixed asset intensity on growth is indicated by a value of 0.0628 which is smaller than 0.8. Therefore, there is no problem of multicollinearity between the two.

b. c test
The heteroscedasticity test is used to test for variance inequalities from residuals that can occur between observations. Researchers used the whit test in determining the degree of heteroscedasticity of this study. The test was performed on two equations both those that use moderation and those that do not use moderation. The criteria used in testing are the magnitude of the probability value of the heteroscedasticity test results. If the value of prob.>0.05 or alpha then the model is declared free from symptoms of heteroscedasticity.
Here are the details of the test results that were successfully obtained.  (4) 0.0000 Source: author's preparation with EViews 12. Table 8 shows the results that the variable conforming tax avoidance is free from heteroscedasticity. This can be seen from the magnitude of the probability value of both equation 1 which is 0.4136 and equation 2 of 0.3309 which are both values greater than alpha (5%). The results of the heteroscedasticity test on non-conforming tax avoidance test variables were also declared free from heteroscedasticity. Tested from the probability value of the two equations used that have yielded a value greater than 0.05 (5%). Equation 1 shows a value of 0.0929 and equation 2 with a value of 0.1421. Both test results showed that both models used by this study were tested heteroscedasticity free.
From Table 10 above, the probability number (F-statistic) of 0.016331 is found, the result is smaller than the alpha value of 5%. Therefore, it can be concluded that sales growth and fixed asset intensity as independent variables in this study significantly affect together on the dependent variable tax avoidance.

Test t (partial significant)
The summary of the results of the t test against equation 1 without moderation on the dependent variable conforming tax avoidance is as follows: In the table above we can see the results of multiple linear regression analysis based on fixed effect models. Looking at the results above, the first regression equation in the study can be made as follows: TaxAvo1 i,t= -0.410154 i,t + 0.061027 Growthi,t -0.384913 IAi,t + 0.008670 Sizei,t + 0.962061 Levi,t + ε1 i,t Description: TaxAvo: Tax Avoidance; Growth: sales growth; IA: intensity of fixed assets; Size: company size; Lev: Leverage.
In order to determine the effect of the independent variable on the dependent variable partially, a t test was conducted in this study. Here is the hypothesis that researchers used in the t-test: H0: sales growth and fixed asset intensity do not have a partial effect on tax avoidance. H1: sales growth and fixed asset intensity have a partial effect on tax avoidance.
The t test also uses 95% confidence so that when the probability value is below 0.05 then H0 is rejected while vice versa when the probability value is above 0.05 then H0 is accepted. Let's take a closer look at the position of the hypothesis testing results based on table 11 above: a) H1: Sales growth has a positive effect on tax avoidance The probability value resulting from testing through EViews 12 on the variable X1 sales growth is 0.4947 which is greater than the value of 5% so that H0 is accepted, so there is no partial influence between sales growth and tax avoidance. The coefficient is shown with a value of 0.061027 which means that the resulting influence is positive. The conclusion that can be drawn is that sales growth does not have a partial positive effect on tax avoidance. b) H2: Fixed asset intensity has a positive effect on tax avoidance The probability resulting from testing the variable X2, namely the intensity of fixed assets, is 0.2566 with a coefficient value of -0.384913. Both values indicate a position where the intensity of fixed assets is declared to have no partial effect on tax avoidance. A negative number on the coefficient indicates a negative nature, so the conclusion that can be made is that the intensity of fixed assets does not have a partial negative influence on tax avoidance. Therefore, H2 is rejected.

Adjusted R 2 Test (Coefficient of Multiple Determination)
The use of the coefficient of determination test is intended to gain an understanding of how much influence the independent variable can have on the dependent variable. The value used comes from the adjusted value of R 2 obtained from the results of data processing in the fixed effect model. Let's check the adjusted R 2 test results in the following table:  Table 11 shows that the adjusted value of R2 obtained from testing conforming tax avoidance using equation 1 produces a value of 0.200131 which means that sales growth and intensity of fixed assets as independent variables X1 and X2 are able to affect the tax avoidance variable or variable Y by 20.01% and the remaining 79.99% is explained using variables that are not used in this study.  Prob(F-statistic) 0.038413 Source: author's preparation with EViews 12. Table 13 above shows a probability value (F-statistic) of 0.038413 with a value below 5%. Therefore, it is concluded that the variables of sales growth, the intensity of fixed assets, along with the variables of moderation of institutional ownership can have a significant influence simultaneously on the variable of tax avoidance.

Test t (partial significant)
The summary of the results of the t test against equation 2 with the variable moderation of institutional ownership against the dependent variable conforming tax avoidance is as follows: Based on the above equation, the following is the hypothesis that the researcher used in the t test: H0: Sales growth and fixed asset intensity moderated by institutional ownership did not significantly affect tax avoidance. H1: Sales growth and fixed asset intensity moderated by institutional ownership significantly affected tax avoidance.
The alpha value used in this study is 0.05 or means that this study uses 95% confidence so that when the probability value is below 0.05 then H0 will be considered rejected while vice versa when the probability value is above 0.05 then H0 will then be accepted. The test results of this second equation produce answers to the researchers' hypotheses as follows: H3: Institutional ownership weakens the effect of sales growth on tax avoidance.
The probability value of this hypothesis is shown by 0.8089 or is above 0.05 besides that the test results also show a regression coefficient value of -0.076627 which means that the resulting relationship is true in the form of negative traits or institutional ownership is weakening the effect of sales growth on tax avoidance. The conclusion drawn from this test is that institutional ownership is not able to significantly weaken the influence that sales growth has on tax avoidance. H4: Institutional ownership weakens the effect of fixed asset intensity on tax avoidance Table 14 shows the probability value of institutional ownership to fixed asset intensity of 0.8650 with a coefficient value of 0.091518. This means that institutional ownership does not succeed in reinforcing the significant effect of fixed asset intensity on tax avoidance or it can be concluded that H4 is rejected.

Adjusted R2 Test (Coefficient of Multiple Determination)
Here are the details of the adjusted value of R 2 in equation 2 with the use of moderation variables:  Table 15 shows the adjusted value R2 of equation 2 produces a value of 0.172138 or can be interpreted as the amount of influence brought by the variable of sales growth, the intensity of fixed assets, along with its interaction with the moderation variable on the variable tax avoidance is 17.21% and the difference of 82.79% is explained using variables and other factors that have not been discussed in this study. Judging from table 16 above, it is seen that there is a probability value (F-statistic) of 0.000005, which is below the probability value of 0.05. So it can be concluded that sales growth and fixed asset intensity can significantly and simultaneously have an influence on tax avoidance.

Test t (partial significant)
Based on the results of previous model tests, in non-conforming tax avoidance research, fixed effect model results were also obtained as the data model to be used in this study. Here is a t-test table on non-conforming tax avoidance: Testing of this partial significance value uses a value of 0.05 as the basis for determining the test. The selection of 0.05 was due to the use of confidence by 95% in this study. So the conclusion is obtained if the probability value is below the alpha value then H0 is rejected and when the probability value is above the alpha value then H0 is accepted. a) H1: Sales growth has a positive effect on tax avoidance Looking at the probabtility value presented in Table 17, it was found that the value of 0.9776 was greater than the value of 0.05. This figure is obtained from the results of data processing using the FEM data model. Sales growth as variable X1 is considered unable to partially affect variable Y tax avoidance based on the results of the hypothesis test above. The value is 0.9776>0.05 then H0 is accepted. The value of the coefficient is indicated by a positive value of 0.001763 which means that the relationship between X1 and Y is positive. The conclusion of this hypothesis is that sales growth is not able to positively affect tax avoidance. b) H2: Fixed asset intensity has a positive effect on tax avoidance The variable intensity of fixed assets or variable X2 in this study displays a probability value of 0.0092 where the value is below the value of 0.05 so that it is concluded that H0 is rejected which means that there is a partial influence given by the variable intensity of fixed assets to the tax avoidance variable. The probability value is also followed by a coefficient value of 0.047459 which means that the existing relationship is positive. Therefore, based on these two values, it can be concluded that the intensity of fixed assets (X1) has a partial positive influence on tax avoidance (Y) so that H0 is rejected and H1 is accepted.

Adjusted R 2 Test (Coefficient of Multiple Determination)
The test is continued with the multiple coefficient of determination test where this test is expected to provide an overview of the amount of influence that can be given from the independent variable (sales growth and fixed asset intensity) to the dependent variable (tax avoidance). For the process of determining the results, we can see the adjusted value of R2 obtained in the previous test. The test results of equation 1 non-conforming tax avoidance are as follows:  Table 18 shows the adjusted value of R2 obtained from testing non-conforming tax avoidance equation 1 produces a value of 0.417169 which means that variables X1 and X2 in this study (sales growth and fixed asset intensity) affect the tax avoidance variable or variable Y by 41.72% and the remaining 58.28% can be influenced by other factors outside this study.  Table 19 above shows a probability number (F-statistic) of 0.000008 which means it is below 0.05. Therefore, it was concluded that the variables of sales growth, the intensity of fixed assets, along with the variables of moderation of institutional ownership were able to have a significant influence simultaneously on the variable of tax avoidance.

T Test (partial significant)
The summary of the results of the t test against equation 2 with the variable moderation of institutional ownership to the dependent variable non-conforming tax avoidance is as follows:  Table 20 displays the results of multiple linear regression analysis based on equation 2 with the addition of institutional ownership variables as moderation variables. Based on the results of the t test above, the regression equation can be obtained as follows: TaxAvo2 i,t= -1.5538091 + 0.223196 Growthi,t + 0.579020 IAi,t + 0.056381 Sizei,t + 0.93171 Levi,t -0.323157 Ins*Growthi,t -0.085009 Ins*IAi,t + ε2,i,t Description: TaxAvo: Tax Avoidance; Growth: sales growth; IA: intensity of fixed assets; Size: company size; Lev: Leverage.
The test results of this second equation produce answers to the researchers' hypotheses as follows: a) H3: Institutional ownership weakens the effect of sales growth on tax avoidance.
The probability value of this hypothesis is shown by the number 0.1391 which is in a position of more than 0.05 with the test results of the regression coefficient of -0.323157 which means that the resulting relationship is negative or institutional ownership is weakening the effect of sales growth on tax avoidance. Therefore, it can be concluded that institutional ownership is not able to significantly weaken the effect of sales growth on tax avoidance. b) H4: Institutional ownership weakens the effect of fixed asset intensity on tax avoidance Table 20 can see the probability of institutional ownership to the intensity of fixed assets is at a probability value of 0.8176 accompanied by a coefficient value of -0.085009. The nature of the influence of institutional ownership on intensity is negative, which means that institutional ownership has a tendency to weaken the influence of fixed asset intensity on tax avoidance. Based on the two values above, the conclusion of hypothesis 4 is that institutional ownership has not succeeded in weakening the significant effect of fixed asset intensity on tax avoidance.

Discussion
After successfully testing with the results that have been expressed in the previous points, the results of hypothesis testing were obtained on hypotheses that had been prepared previously both on the types of conforming tax avoidance and non-conforming tax avoidance research. The following is a summary of all the results of the hypothesis test (t test) conducted on this study:

Accepted
Source: Author's work Looking at the results of the hypothesis test that has been summarized in Table 21 above, it can be described the discussion for each hypothesis used in this study as follows:

The Effect of Sales Growth on Tax Avoidance
Referring to the results of the hypothesis test which can also be seen in the summary in table 21, it can be seen that the sales growth variable as an independent variable produces 2 regression coefficient values and 2 different probability values from testing using TaxOCF proxies on conforming tax avoidance and those derived from Current ETR proxies for testing non-conforming tax avoidance . For testing with the TaxOcf proxy, a regression coefficient of 0.061027 was obtained, which means that the nature of the influence of this variable is positive. The same thing is obtained from the test results using the Current ETR proxy where the current of influence obtained is also positive, which is 0.001763. Both tests produce the same direction i.e. positive. To determine whether or not the hypothesis that has been built is accepted, the author refers to the probability value, namely 0.4974 for testing conforming tax avoidance and 0.9776 for non-conforming tax avoidance. The two values are in a greater position than 5%, so it shows that both in terms of conforming and non-conforming tax avoidance there is no significant effect of sales growth on tax avoidance in research conducted on manufacturing companies used as samples in the study.
Looking back at the agency theory used as the basis for this study, increased sales tend to encourage managers to do tax avoidance in order to maintain good performance of the company under the manager's managerial. The higher sales are synonymous with the higher profits that the company will report in the annual report as the basis for calculating its taxes, resulting in a mountain of tax burdens that must be borne by the company. In order to avoid this, it must be one of the manager's tasks to overcome these problems, tax avoidance tends to be the way to be used. It turns out that not all companies that experience an increase in sales choose to do tax avoidance. The reason is because with the greater sales, it will be accompanied by the greater the scale of the company's business and this development will inevitably cause an increase in the value of tax payments to the state (Arinda & Dwimulyani, 2018). Tax avoidance will be difficult to do considering the proportion generated in terms of company size will be unequal and not in accordance with the objectives of the company if the tax paid does not increase so that tax avoidance is not an option taken by many managers.

The effect of fixed asset intensity on tax avoidance
The hypothesis test conducted showed the value of the regression coefficient of the fixed asset intensity variable which differed between research on conforming and non-conforming tax avoidance. In testing the current ETR proxy, a positive result of 0.626765 was received, while testing using TaxOCF resulted in a negative value of -0.384913. This difference indicates a difference in the impact resulting from the same independent variable on 2 dependent variable proxies. The probability results received are also different where for conforming tax avoidance a result of 0.2566 or greater than 0.05 is found so that the hypothesis is rejected or there is no positive influence given by the intensity of fixed assets on tax avoidance. These results are relevant to previous results according to (Sundari & Aprilina, 2017) but are inversely proportional to research from (Purwanti & Sugiyarti, 2017;Rizky & Puspitasari, 2020) where fixed asset intensity has a significant positive influence on tax avoidance.
When viewed from the side of conforming tax avoidance, the intensity of fixed assets is considered not a red flag for the use of tax avoidance in a company. So far, the use of fixed assets as a way of reducing the tax burden is considered one of the safest and most effective ways. Where the size of the company is not necessarily accompanied by an increase in profits due to depreciation expenses that can suppress profits and improve tax avoidance practices. When related to the TaxOCF formula as a proxy that calculates the amount of practice from the point of spending money, the intensity of fixed assets is considered unable to show a significant effect considering that the intensity of fixed assets reduces the tax burden in terms of expenses rather than directly on money out. In addition, conforming tax avoidance focuses on looking at tax avoidance risks not only in terms of accounting income but also in terms of tax income, namely seeing whether there is a gap between the company's attitude and tax regulations (Satyadini, 2018). The high number of fixed assets cannot be used as a benchmark for tax avoidance practices considering that there is no gap between the increase in fixed assets and applicable tax regulations in Indonesia.
One interesting thing about the results of research with this variable is the difference in results obtained from the two proxies used. The results with the current ETR proxy show a probability value of 0.0092 is far below the alpha value of 0.05. This value shows that H2 is accepted, which means that the intensity of fixed assets positively has a significant effect on tax avoidance. This result becomes relevant to the research of (Purwanti & Sugiyarti, 2017;Rizky & Puspitasari, 2020) which was previously inversely proportional in terms of conforming tax avoidance. In terms of accounting income, the intensity of fixed assets is considered to play a role in tax avoidance practices carried out by companies as taxpayers. Because, the greater the fixed assets owned by the company, the greater the depreciation expense recognized and then will cause the effect of decreasing company profits and shrinking the value of the company's tax burden. This is in line with agency theory where management will do everything necessary in fighting for low taxable income which means increased profits for shareholders (Purwanti & Sugiyarti, 2017).Therefore, the intensity of fixed assets can still be accepted as one of the supporting factors in reviewing the risk of tax avoidance in a company.

Institutional Ownership in Moderation on the Effect of Sales Growth on Tax Avoidance
Based on the results of hypothesis tests that have been carried out both on conforming and non-conforming tax avoidance, similar results were obtained related to moderation carried out by institutional ownership on the effect of sales growth on tax avoidance. The coefficient values of both are negative, namely -0.076627 for the use of TaxOCF proxies and -0.323157 for current ETRs. This means that the influence generated by the moderation variable of institutional ownership is negative or tends to weaken the influence of the independent variable on the dependent variable. The probability value of institutional ownership is 0.8089>0.05 for the conforming side , which means institutional ownership cannot weaken the effect of sales growth on tax avoidance. The same was found in the non-conforming tax avoidance study with a probability of 0.1393>0.05 which means that both showed similarities in their inability to moderate the negative effect of variable X1 sales growth on variable Y tax avoidance. Similar to this study, research conducted by (Dewi & Sari, 2015;Tandean & Winnie, 2016;Windaryani & Jati, 2020) also found that institutional ownership has no effect on tax avoidance. Another thing with (Pramana & Wirakusuma, 2019) research which succeeded in finding the moderation effect of institutional ownership that is negative on tax aggressiveness. (Ristanti, 2022;Safitri & Damayanti, 2021) also received significant positive results in their research on institutional ownership as a moderating variable on the influence of independent variables on dependent variables.
The results showed that H3 was rejected on the hypothesis that had been built by the researcher. Institutional ownership is a form of Good Corporate Governance applied to several companies listed on the IDX in 2018-2020 as the population of the study. GCG basically aims to provide good supervision and management system for companies. In particular, institutional ownership is considered capable of suppressing conflicts that arise between management and shareholders due to the existence of institutions that constantly monitor the flow of company reports which makes it difficult for managers to carry out practices that are considered detrimental to the company's reputation. In fact, shareholders have almost the same goal, which is to obtain large profits and have a share in a profitable company. Therefore, institutional ownership may not necessarily be a benchmark for reducing the effect of sales growth on tax avoidance practices within the company.

Institutional Ownership in Moderation on the Effect of Fixed Asset Intensity on Tax Avoidance
This study obtained results that showed the influence of institutional ownership moderation variables on the independent variable of fixed asset intensity in conforming tax avoidance of 0.091518 where the interaction was in the form of positive interactions. The result is also equipped with a probability value of 0.8650>0.05 so that this result is contrary to H4 so that the hypothesis is rejected which means institutional ownership is considered unable to weaken the effect of fixed asset intensity on tax avoidance. For research on non-conforming tax avoidance, a coefficient value of -0.085009 was obtained, which means that the nature of the influence of institutional ownership on the effect of fixed asset intensity on tax avoidance is negative, slightly different from the results obtained from the conforming side , which is positive. In terms of probability, the value obtained is 0.8176>0.05, which means that the hypothesis is also rejected from this side.
Although institutional ownership is considered capable of violently stopping management behavior in an effort to maintain company profits, it has not been able to be proven tangibly through this study. Decisions from institutional shareholders do not only focus on maintaining reputation and systems but must be understood that the welfare of shareholders remains paramount. Therefore, institutional ownership is not necessarily able to definitively reduce tax avoidance in a company.

CONCLUSION
Based on all the results of the tests carried out, the conclusions conveyed are as follows: 1. Sales growth does not significantly affect tax avoidance. This result is obtained both from the conforming and non-conforming sides, which means that hypothesis 1 (H1) that has been built is declared rejected. The hypothesis previously built was that sales growth had a significant positive influence on tax avoidance. The rejection of the hypothesis was motivated by the results of the t test which showed probability values of 0.4974 and 0.9776 so that H1 was rejected and it could be said that sales growth did not have a significant effect on tax avoidance.
2. Fixed asset intensity significantly affects tax avoidance in terms of non-conforming but is not able to significantly affect in terms of conforming tax. These results are obtained based on different probability values between conforming and non-conforming tax avoidance tests. The probabilities are 0.2566>0.05 (conformin) and 0.0092<0.05 (non-conforming). Researchers believe that this is due to differences in the point of view of conforming and non-conforming tax where the intensity of fixed assets has a significant positive influence on tax avoidance when viewed from the accounting income side. However, it becomes less significant when viewed in terms of tax income or money outflow.
3. Institutional ownership cannot weaken the effect of sales growth on tax avoidance. Therefore, hypothesis 3 of this study which states that institutional ownership can weaken the effect of sales growth on tax avoidance is rejected. The decision was taken based on the results of the t test with a coefficient value of -0.076627 with a probability of 0.8089 in terms of conforming tax and a coefficient result of -0.323157 with a probability of 0.1391. Furthermore, it is certain that institutional ownership is unable to weaken the effect of sales growth on tax avoidance.
4. Institutional ownership is unable to weaken the effect of fixed asset intensity on tax avoidance. Based on the results of the t test in the data model, a coefficient value of 0.091518 and a probability of 0.8650 were obtained in terms of conforming tax, which means institutional ownership cannot significantly weaken the effect of fixed asset intensity on tax avoidance. The same is also obtained from non-conforming tax testing where the coefficient value is -0.085009 and the probability is 0.8176. From the value submitted, it can be seen that both are above the value of 5%, then H4 becomes rejected.