Analysis of Factors Affecting Tax Avoidance and Firm Value

* Corresponding author, Email: 1kristikamayke@gmail.com, 2astit@perbanas.ac.id ABSTRACT The aim of this study is to examine and analyze factors affecting tax avoidance and fi rm value. The independent variables are audit quality and profi tability, while the dependent variable is fi rm value, and the intervening variable is tax avoidance. The sample used in this study is banking sector companies in Southeast Asia that listed in www.orbis.bvdinfo.com for the period 2014-2016. The data were analyzed using a multiple regression analysis with SPSS 22.0 For Windows Program and path analysis. The results show that audit quality has no infl uence on tax avoidance, but profi tability has infl uence to tax avoidance. Audit quality, profi tability, and tax avoidance have infl uence on fi rm value, while tax avoidance cannot mediate the infl uence of audit quality and profi tability on fi rm value.


INTRODUCTION
Tax is a mandatory contribution for individuals and entities to the state that is coercive and based on applicable laws. Taz payment is not only limited to responsibility, but the right of every citizen to participate in the national development. Rewards from the tax payment cannot be directly felt by citizens, but the tax is used by the state to realize people's prosperity (Mardiasmo, 2016: 3).
However, the reward that cannot be directly felt by taxpayers causes many people and entities to carry out tax avoidance. Tax avoidance has come to the attention of every country, especially tax avoidance carried out by corporate taxpayers. The banking sector has been growing rapidly lately and the Financial Services Authority suspects that it is vulnerable to reducing the quality of fi nancial statements.
The banking industry is currently inseparable from the issue of tax avoidance.
The data obtained show that in 2015 European countries were shocked by a taxation case involving HSBC in Switzerland. In this case, HSBC allegedly helped wealthy customers avoid taxes by offering aggressive schemes to reduce taxes in their home countries, especially Europe. In addition, in 2017 the Financial Services Authority of Indonesia (OJK) in coordination with the Indonesian Financial Transaction Reports and Analysis Centre or INTRAC (PPATK) followed up on a case of fund transfers totaling IDR 18.9 trillion or 1.4 billion US dollars from Guernsey at Standard Chartered Bank in Singapore which occurred at the end of 2015 (Bambang, 2017).
The difference in interests between investors and company management tends to encourage tax avoidance. Tax avoidance is carried out by companies to allocate taxable income. In addition, the banking sector has the pressure to always increase profi ts so that when banks can generate high profi ts, they will carry out profi t engineering to avoid taxes.
Tax avoidance can be minimized by the auditors' role. If they are independent and competent, they can achieve the audit quality. De Angelo (1981) argues that the quality of auditors depends on the size of the Public Accounting Firm (KAP). Large Public Accounting Firms tend to be more independent. Arry (2017) states that when companies are audited by Public Accounting Firms affi liated with the big four accounting fi rm, they will be increasingly diffi cult to carry out aggressive tax policies. Meanwhile, Fitri & Tridahus (2015) state that audit quality does not affect tax avoidance. In addition, audit quality is a factor that can affect company value. The results of the research conducted by Assidi et al (2016) and Afza & Nasir (2014) show that audit quality has an effect on company value. On the contrary, research conducted by Dodik (2013) proves that audit quality does not affect company value.
Profi tability is one of the factors that can infl uence tax avoidance. Profi tability is the company's ability to generate profi ts through the use of resources it has, such as cash, capital, number of branches, number of employees, and sales activities (Sofyan, 2015: 305). Fitri & Tridahus (2015) state that companies that have high profi tability have the opportunity to position themselves in tax planning that reduces the amount of tax liability burden, while Moses & Nur (2017) state that profi tability does not affect tax avoidance. In addition, profi tability is also thought to have an infl uence on the company value. Research conducted by Bayu & Panji's (2015) shows that profi tability has an effect on company value. On the contrary, research conducted by Ni Putu & Made (2017) states that profi tability does not affect the value of the company.
Tax Avoidance is one of the factors that can infl uence company value. According to Assidi et al (2016), Amalia & Catur (2014) and Chen et al (2016), tax avoidance has a negative effect on fi rm value. On the contrary, research conducted by Anita (2016) provides evidence that tax avoidance does not affect the value of the company.
Based on the above description, the researchers were motivated to conduct further testing entitled "Analysis of the Factors that Infl uence Tax Avoidance and Firm Value".

THEORETICAL FRAMEWORK AND HYPOTHESIS Agency Theory
Agency theory was proposed by William H. Meckling and Michael C. Jensen in 1976. It explains the relationship between business owners as the principals and business management as the agents (Jensen & Meckling, 1976). Company managers, as parties who have direct access to company information, sometimes have more information than external parties of the company. Such a condition is called information asymmetry or agency problems.
The method to reduce information asymmetry is by increasing supervision. To improve supervision, the principals are required to spend costs, or referred to as agency costs. Agency costs can be in the form of giving a salary or bonus to agents, paying to public accountants, or providing adequate control system. The existence of agency costs is expected to be able to make the CEO work optimally and pay attention to honesty (Romanus, 2016: 229).
The relationship between agency theory and this research is that when a company has different interests from external parties, the company will try to achieve its objectives. When the company gets high profi ts, the company will try to engineer or manipulate fi nancial statements so that the taxes paid can be smaller. In addition, audit quality can infl uence companies to disclose accurate information. This is because the auditor has an interest in ensuring that the company, in making fi nancial statements, must comply with the applicable rules and regulations. When the quality of the audit increases, it will reduce the company's efforts to carry out tax avoidance.

Signaling Theory
Signaling theory is a theory that explains the importance of information made by a company for external parties in order these parties can make decisions to invest. Information provided by the company is necessary for related parties because the information explains the company's condition in the past, present and future regarding the continuity of the company (Brigham & Houston, 2011: 184-186).
Information disclosed by the company is a signal to investors. The investors will make decisions based on the information provided, and therefore, investors need accurate, relevant, complete and timely information. The information issued by the company will cause positive signals or negative signals from stakeholders.
The relationship between signaling theory and this research is that when the company is audited by competent and independent auditors, the fi nancial statements produced by the company are more accurate. Therefore, this will affect the market's reaction to give a positive signal. In addition, the high profi ts that can be generated by the company will provide a positive signal for investors to make investment decisions because with high profi ts, investors will get a bigger dividend. When a company carries out tax avoidance, the external party will give a negative signal so that the company's value will decrease.
Audit quality is an opportunity for market valuation whether the fi nancial statements or annual reports contain material misstatements. The output of an audit is independent verifi cation related to fi nancial statements presented by the management of the company equipped with auditor opinion in accordance with the dimensions of quality, because the auditor has the task of verifying the fi nancial data that have been presented by the management of the company (De Angelo, 1981).
Tax Avoidance is the engineering of tax affairs that is still in the framework of taxation provisions. Taxpayers are said to carry out tax avoidance when the taxpayers engage in engineering in order that the tax burden can be reduced as low as possible by utilizing applicable regulations but different from the objectives of lawmakers (Erly, 2016: 7-9).
Profi tability is the company's ability to generate profi ts through the use of resources owned, such as cash, capital, number of branches, number of employees, and sales activities (Sofyan, 2015: 305). The increase in the amount of profi ts generated by the company will provide an overview related to the prospect of the company's performance in carrying out its operational activities.
Firm value is the market value of the company. It is said to be market value because the value of the company can provide prosperity to the shareholders maximally if the company's stock price increases (Harmono, 2014: 50). The increase in the company's stock price will provide profi ts and prosperity for the shareholders.

The Effect of Audit Quality on Tax Avoidance
Accurate information disclosure is one of the requirements for transparency. At present transparency in terms of taxation is increasingly demanded by public authorities because the information disclosed will affect stakeholders in making decisions. The higher the audit quality, the higher the transparency of tax-related information disclosure. When audit quality increases, the company tends not to do earning management related to tax interests. Research conducted by Arry (2017) shows that audit quality has an effect on tax avoidance. H1: Audit quality has an effect on tax avoidance.

The Effect of Profi tability on Tax Avoidance
Profi tability describes the company's performance in managing its resources effectively. Profi t is an important factor in determining the amount of effective tax rate payments. High profi t can affect the company in carrying out tax avoidance actions so that many companies tend to carry out earnings management so that the tax paid is smaller than the amount of tax that should be paid. Research conducted by Fitri & Tridahus (2015) states that profi tability has an effect on tax avoidance. H2: Profi tability has an effect on tax avoidance.

The Effect of Audit Quality on Company Value
Audit quality can be said as the conformity between audit results and auditing standards. Audit is said to be of high quality when it is carried out by competent and independent auditors. Auditors who work in large Accounting Firms will tend to be more independent of their clients. The higher the audit quality, the more accurate the information presented by the company to external parties. Therefore, the value of the company will increase when the audit quality increases. Research conducted by Afza & Nazir's (2014) states that audit quality has an effect on company value. H3: Audit quality has an effect on company value.

The Effect of Tax Avoidance on Company Value
Companies that carry out tax avoidance cannot be said to violate the law, but they will even obtain tax savings because they avoid actions that should have been taxed. Tax avoidance actions can increase or decrease company value. Research conducted by Amalia & Catur (2014) shows that tax avoidance has an effect on company value. H4: Tax Avoidance has an effect on Company Value

The Effect of Profi tability on Company Value
Profi tability is an important element to maintain the company in the short and long term. The increase in the amount of profi ts generated by the company will provide an overview related to the prospect of the company's performance in carrying out its operational activities. The amount of profi ts generated by the company will affect the value of the company, because investors will be interested in investing in the company. Research conducted by Bayu & Panji (2015) shows that profi tability affects the value of the company. H5: Profi tability has an effect on company value

The Effect of Audit Quality on Company Value through Tax Avoidance
High audit quality shows that the company's fi nancial statements are fair and good. The high fairness of the fi nancial statements indicates that the possibility of the company in carrying out tax avoidance is lower. Fairness of fi nancial statements will have an impact on increasing the valuation of shareholders and investors towards the company, so that the value of the company will increase. Research conducted by Julian & Raisa (2017) shows that tax avoidance is able to mediate the infl uence of audit quality on fi rm value. H6: Audit quality has an effect on company value through tax avoidance as an intervening variable.

The Effect of Profi tability on Company Value through Tax Avoidance
Profi tability is a refl ection of the amount of dividends that can be obtained by the company's shareholders. The more profi t that can be generated by the company, the more interested the shareholders in investing their capital in the company. It is undeniable that when company profi t increases, the company manager will take tax avoidance measures so that the tax paid is smaller than what should be paid. This is because profi t is a quantity that can determine how much the tax that must be paid by the company. The manager's decision to carry out tax avoidance will affect the decline in shareholders' assessment of the company, so that the value of the company will decrease. Research conducted by Julian & Raisa (2017) shows that tax avoidance is able to mediate the effect of profi tability on company value. H7: Profi tability has an effect on company value through tax avoidance as an intervening variable.
Based on the description above, the framework underlying the research hypotheses is demonstrated as in Figure 1.

RESEARCH METHOD Research Design
This research is a quantitative research with a causal associative approach. It used secondary data, obtained from the fi nancial statements of the banking sector companies in Southeast Asia downloaded through www.orbis.bvdinfo.com and company's website.

Research Limitation
The sample used in this study is only banking sector companies in Southeast Asia which got profi t and were listed on www. orbis.bvdinfo.com in 2014-2016.

Variable Identifi cation
The variables in this study are divided into three: independent variable, dependent variable, and intervening variable. The details of the identifi cation of variables in this study are as follows: a. Independent variable: audit quality and profi tability. b. Dependent variable: fi rm value. c. Intervening variable: tax avoidance.

Operational Defi nition and Measurement of Variables Audit Quality
In this study, audit quality is measured using a dummy variable that refers to the research conducted by Nuralifmida & Lulus (2012) where number 1 is for Acouting Firms affi liated with The Big Four and number 0 is for Accounting Firms affi liated with non-The Big Four.

Profi tabiity
In this study, profi tability is measured using the ratio of Return on Assets (ROA). ROA is used to measure the effectiveness of managing all available assets to generate profi ts. Referring to the research conducted by Fitri & Tridaus (2015), the formula of the ROA ratio is as follows: Profi t after Tax ROA = Total Assets

Tax Avoidance
In this study, tax avoidance is measured using CETR (Cash Effective Tax Ratio) which refers to the research conducted by Amalia & Catur (2014). CETR can be calculated using the following formula: Tax Payment CETR = Profi t before Tax

Firm Value
In this study, fi rm value is measured using Price to Book Value (PBV) which refers to the research conducted by Ni Putu & Made (2017). This ratio shows the company's ability to create a fi rm value from the funds invested by investors. The Price to Book Value (PBV) ratio can be calculated as follows: Market Price per Share PBV = Book Value per Share

Population, Sample, and Sampling Technique
The population in this study is banking sector companies in Southeast Asia listed at www.orbis.bvdinfo.com in 2014-2016, consisting of 167 companies per year. The sampling technique used in this study is a census or sample method that represents the population of each banking company in several countries in Southeast Asia.

Data Analysis Technique
Simple regression analysis and path analysis are used to test the factors that infl uence company value both directly and indirectly.
The multiple linear regression equation models used in this study are: The path analysis equation models used in this study are:

DATA ANALYSIS AND DISCUSSION Descriptive Statistical Analysis
Descriptive analysis aims to show an overview of the data used in the study. The description of the data can be seen from the mean value, maximum value, minimum value, standard deviation and frequency value in accordance with the research sample. Changes in stock prices that occur at any time caused the value of banking companies to experience fl uctuations in the study period. The average fi rm value decreased by 14.5% in 2014-2015 and increased by 106.45% in 2016. In 2014-2016 companies that could provide prosperity to shareholders below the average were 56.42% of the total sample of this study and companies that could provide prosperity to shareholders above the average were 43.58% of the sample of this study. This value shows that more than half of the banking sector companies in Southeast Asia that became the sample of this study had lower market value because the market price per share of the company was lower than the book value per share.
The increase in the amount of costs to be incurred by the company from year to year caused the profi ts generated by the company to decline in the study period. The average profi tability decreased by 0.12% in 2014-2105 and continued to decline in 2016 by 0.31%. In the study period, companies that had the ability to generate profi ts below the average were 73.98% of the research sample, and companies that had the ability to generate profi ts above the average were 26.02% of the sample of this study. This value shows that more than half of the banking companies in Southeast Asia that became the samples of this study had the ability to generate low profi ts because the profi ts generated were smaller than the total assets owned.
The company management effort to conduct tax avoidance has fl uctuated from year to year. The average CETR increased by 5.32% in 2014-2015. This value shows that in that year the practice of tax avoidance carried out by company management experienced a decline. In contrast, in 2015-2016 the average CETR decreased by 8.51%. This means that in 2015-2016 the practice of tax avoidance carried out by company management increased. In the study period, 79% of the samples were companies that carried out high tax avoidance practices because payment of corporate income tax was smaller than profi t before tax, while 21% of the samples were companies that carried out low tax avoidance practices because the payment of corporate income tax was greater than the profi t before tax.
The audit quality of banking sector companies in Southeast Asia during the study period increased from year to year. In 2014-2015 the average audit increased by 23.49% and in 2015-2016 it increased by 80.39%. Of the 319 data, 72.33% used the services of the Accounting Firms affi liated with the Big Four Accounting Firms, while 27.67% used the services of the Accounting Firms affi liated with non-the Big Four Accounting Firms. This means that the information received by investors is accurate information because more than half of the sample companies use the services of Accounting Firms affi liated with the Big Four Accounting Firms.

Classical Assumption Test Normality Test
The purpose of the normality test is to fi nd out whether the data regression model is normally distributed. Normality test was conducted using the Kolmogorov-Smirnov table. Data are said to be normally distributed if the normality test result gets a sig value more than or equal to 0.05. The value of Asymp. Sig.
(2-tailed) obtained in the Kolmogorov-Smirnov test is 0.150. This value is greater than 0.05 so that H0 is accepted. This means that the data in this study are normally distributed.

Autocorrelation Test
The purpose of the autocorrelation test is to test whether in the regression model there is a correlation between the variables used. The

Multicollinearity Test
The purpose of multicollinearity test is to test whether in the regression model there is a correlation between independent variables. A regression model is said to be good if it is free from multicollinearity problems. Data are said to have no multicollinearity if the VIF value is ≤ 10 and the Tolerance value is ≥ 1 0.1.
The multicollinearity test results show that the tolerance value of the variable of audit quality is 0.989, profi tability is 0.902, and tax avoidance is 0.912. The tolerance values are greater than 0.10. These values indicate that there are no cases of multicollinearity in the study. If viewed from the VIF value, the table above shows that the VIF value of the variable of audit quality is 1.012, profi tability is 1.109, and tax avoidance is 1.096. These values are less than 10. Based on this information, there are no cases of multicollinearity.

Heteroscedasticity Test
The purpose of heteroscedasticity test is to fi nd out whether in the regression model there is a similarity in the variance of the residuals between one observation and another. The regression model is said to be good if there is no heteroscedasticity. The Glejser test is done to detect whether in this study there is a cases of heteroscedasticity or not. Data are said to occur cases of heteroscedasticity if, statistically, there is independent variable that has a signifi cance value of less than 0.05.
The results of Glejser test show that the variable of profi tability has a signifi cance value of 0.036 which is smaller than 0.05, while the variables of audit quality and tax avoidance have a signifi cance value greater than 0.05. As there is variable that has signifi cance value of less than 0.05, it can be concluded that in this study there is a case of heteroscedasticity.

Results of Analysis and Discussion
Regression analysis aims to measure the strength of the relationship between two or more variables. Based on multiple linear regression analysis, the multiple linear regression models used are as follows: TA = 0.208 -0.018AQ + 4.757PROF + e FV = 1.343 + 0.394AQ + 5.136PROF -0.268TA + e Note: FV = Firm Value AQ = Audit Quality PROF = Profi tability TA = Tax Avoidance e = Error

Hypothesis Test F-Test
The F-test aims to fi nd out whether the regression model is fi t or not fi t. The F-value obtained is 15.236 and 29.405 with a signifi cance value of 0.000. Therefore, it can be concluded that H 0 is rejected because the signifi cance value is smaller than 0.05. This means that the regression model in this study is fi t and all independent variables simultaneously infl uence the dependent variable.

Coeffi cient of Determination (R 2 ) Test
The purpose of the determination coeffi cient (R2) test is basically to fi nd out the extent to which the ability of the regression model can explain the independent variables and dependent variable in a study. The greater the Adjusted R-Square value, the greater the ability of the regression model to explain the independent variables and the dependent variable.
The Adjusted R-Square value obtained is 0.082 or 8.2%. Adjusted R-Square value is used as a tool to measure the ability of the model in explaining the dependent variable. This value shows that audit quality and profi tability are able to explain tax avoidance by 8.2%, and there are other factors of 91.8%. The Adjusted R-Square value obtained is 0.21 or 21%. This value indicates that audit quality, profi tability, and tax avoidance can explain the fi rm value by 21%, and there are other factors of 79%.

The t-Test
The t-test aims to fi nd out the effect of each independent variable on the dependent variable. If the signifi cance level is ≤ 0.05, there is an effect of the independent variable on the dependent variable. The t-test results in this study are as follows: a. The First Hypothesis (H 1 ) Test The fi rst hypothesis test aims to fi nd out the effect of the variable of audit quality on tax avoidance. The t value obtained is -0.304 with a signifi cance level of 0.761. The signifi cance level of 0.761 is greater than 0.05. So it can be concluded that H 1 is rejected. This means that audit quality has no effect on tax avoidance.

b. The Second Hypothesis (H 2 ) Test
The second hypothesis test aims to fi nd out the effect of the variable of profi tability on tax avoidance. The value of t obtained is 5.513 with a signifi cance level of 0.000. The signifi cance level of 0.000 is smaller than 0.05. Therefore, it can be concluded that H 2 is accepted. This means that profi tability has an effect on tax avoidance. c. The Third Hypothesis (H 3 ) Test The third hypothesis test aims to fi nd out the effect of the variable of audit quality on fi rm value. The t value obtained is 6.426 with a signifi cance level of 0.000. The signifi cance level of 0.000 is smaller than 0.05. Therefore, it can be concluded that H 3 is accepted. This means that audit quality has an effect on fi rm value.

d. The Fourth Hypothesis (H 4 ) Test
The fourth hypothesis test aims to fi nd out the effect of the variable of tax avoidance on fi rm value. The t value obtained is -4.568 with a signifi cance level of 0.000. The signifi cance level of 0.000 is smaller than 0.05. So it can be concluded that H4 is accepted. This means that tax avoidance has an effect on fi rm value. e. The Fifth Hypothesis (H 5 ) Test The fi fth hypothesis test aims to fi nd out the effect of the variable of profi tability on fi rm value. The t value obtained is 5.447 with a signifi cance level of 0.000. The signifi cance level of 0.000 is smaller than 0.05. So it can be concluded that H5 is accepted. This means that profi tability has an effect on fi rm value.

Path Analysis
Path analysis in regression is used to fi nd out the direct or indirect effects of the independent variables on the dependent variable. This analysis is a development of multiple linear regression analysis to estimate causality relationships between variables. The variable of tax avoidance can be regarded as intervening variable if the multiplication value of the indirect effects of the independent variable on the dependent variable is greater than that of the direct effect of the independent variable on the dependent variable. f. The Sixth Hypothesis (H 6 ) Test The sixth hypothesis test aims to fi nd out the effect of the variable of audit quality on fi rm value through tax avoidance as an intervening variable. Direct effect has a value of 0.394 while the indirect effect has a value of 0.004824. The value of direct effect is greater that the value of the indirect effect of the variable of audit quality on fi rm value. It can be concluded that H 6 is rejected and H 0 is accepted. This means that tax avoidance cannot function as an intervening variable.

The Seventh Hypothesis (H 7 ) Test
The seventh hypothesis test aims to fi nd out the effect of the variable of profi tability on fi rm value through tax avoidance as an intervening variable. Direct effect has a value of 5.136 while the indirect effect has a value of -1.274876. The value of direct infl uence is greater than the value of the indirect effect of the variable of profi tability on fi rm value. It can be concluded that H 7 is rejected and H 0 is accepted. This means that tax avoidance cannot function as an intervening variable.

Discussion The Effect of Audit Quality on Tax Avoidance
For investors, audit quality is one of the controls to minimize fi nancial statement manipulation, such as tax avoidance action. The audit quality is an audit conducted by Public Accounting Firm. The result of this study indicates that audit quality has no effect on tax avoidance. This means that the size of tax avoidance practices carried out by company management does not depend on the size of the Public Accounting Firm. The existing data show that the company's efforts to conduct tax avoidance remain high even though audit quality has increased in 2014-2016. This study cannot prove the existing theory, agency theory, which states that the way that can be used to reduce information asymmetry is agency costs, and one of the agency costs is payment to public accountants.
The results of this study are consistent with the results of previous research conducted by Fitri & Tridahus (2015) that audit quality has no effect on tax avoidance. However, the results of this study are not consistent with the results of the research conducted by Arry (2017), Kanagaretnam et al (2016) Nuralifmida & Lulus (2012.

The Effect of Profi tability on Tax Avoidance
Profi t is an important factor in determining the amount of effective tax rate payments. The results of this study indicate that profi tability has an effect on tax avoidance. This means that the size of tax avoidance depends on the profi ts generated by the company. Evidently in 2014-2015 it was found that the existing data showed that the average CETR increased. This means that tax avoidance actions decreases when the average profi tability decreases. The results of this study are in accordance with the agency theory. When a company has different interests from external parties, the company will try to achieve its objectives, such as tax avoidance or tax avoidance measures.
The results of this study support the previous research conducted by Fitri & Tridaus (2015) that profi tability has an effect on tax avoidance. However, the results of this study are not in accordance with the results of research conducted by Moses & Nur (2017) that profi tability has no effect on tax avoidance.

The Effect of Audit Quality on Firm Value
Audit quality can help investors get accurate information so that investors are sure that the information obtained is in accordance with the actual conditions of the company. Accurate information can make investors interested in investing their capital so that the value of the company will increase. This study provides results that audit quality has an effect on fi rm value. This means that the size of fi rm value depends on the audit quality. It is evident that in 2015-2016 the existing data showed that there was an increase in fi rm value when the average audit quality increased.
The results of this study can prove signaling theory that the better the audit quality, the better the fi rm value because the shareholders feel that the credibility of the company's fi nancial statements audited by Accounting Firms affi liated with the big four is more assured. This will affect the market reaction to give a positive signal to the company and make sure to invest in the company audited by the Big Four Accounting Firm.
The results of this study are in accordance with the previous research conducted by Assidi et al (2016), Afza & Nasir (2014), Sulong et al (2013) that audit quality has an effect on fi rm value. However, the results of this study are not in accordance with the research conducted by Anita (2016) that audit quality has no effect on fi rm value.

The Effect of Tax Avoidance on Firm Value
Tax avoidance can affect investors in making investment decisions so that when the tax avoidance conducted by the company is high, it will reduce the interest of investors to invest so that the fi rm value will decline. The results of this study indicate that tax avoidance has an infl uence on fi rm value. This means that the size of fi rm value depends on tax avoidance. It is evident that in 2014-2016 the existing data showed a decrease in the fi rm value when the company's efforts to conduct tax avoidance increased. The results of this study can prove a signaling theory that explains the importance of information disclosed by the company to external parties and that the information is a signal to investors. When companies manipulate the fi nancial statements related to the taxation, of course the information obtained by investors is not in accordance with the actual state of the company, so investors give negative signals to the company.
The results of this study are in accordance with previous research conducted by Assidi et al (2016), Chen et al (2016), and Amalia & Catur (2014) that tax avoidance has an infl uence on fi rm value. However, the results of this study are not in accordance with the research conducted by Anita (2016) that tax avoidance has no effect on fi rm value.

The Effect of Profi tability on Firm Value
The increase in the amount of profi ts generated by the company will provide an overview related to the prospect of the company's performance in carrying out its operational activities. This can lead to the investors' prosperity because the returns obtained are high, and fi nally the fi rm value will increase. The results of this study indicate that profi tability has an infl uence on fi rm value. This means that the size of the fi rm value depends on the profi ts that can be generated by the company. It is evident that in 2014-2015 the data showed a decline in the fi rm value when the average profi tability decreased.
The results of this research support signaling theory. When a company cannot manage resources effectively, the company will suffer the greater costs. The amount of costs incurred will decrease their profi ts, thus causing investors not interested in investing in the company. They think that the lower the company's profi t, the lower the return that the investors will get.
The results of this research support the previous research conducted by Bayu & Panji (2015) that profi tability has an effect on fi rm value. However, the results of this research are not in accordance with the results of research conducted by Ni Putu and Made (2017) that profi tability has no effect on fi rm value.

The Effect of Audit Quality on Firm Value through Tax Avoidance as an Intervening Variable
Firm value can be seen from the company's stock price so that the company's stock price can either go up or down. It is when the company manager does or does not carry out tax avoidance and when the company is audited by Public Accounting Firm affi liated with the Big Four or Public Accounting Firm affi liated with non-the Big Four. The results of this study indicate that tax avoidance is not able to mediate the effect of audit quality on fi rm value. This means that the size of the company management effort to do tax avoidance has no impact on the infl uence of audit quality on the fi rm value. The existing data show that when the average audit quality and fi rm value increases, the average CETR decreases, which means that the tax avoidance practice increases. In addition, variations in CETR data are fairly large or the data are heterogeneous so that the distribution of data is classifi ed as poor. So, in this study, the variable of tax avoidance does not fulfi ll the requirements as mediating variable or intervening variable. The results of this study are not in accordance with the previous research conducted by Julian & Pratiwi (2017) that tax aggressiveness mediates the relationship between audit quality and fi rm value.

The Effect of Profi tability on Firm Value through Tax Avoidance as Intervening Variable
Profi tability can infl uence investors to invest in a company so that the company's high profi tability ratios can attract them. In addition to profi tability, fi rm value can be infl uenced by tax avoidance. Tax avoidance is done by companies in order that they can pay taxes smaller than they should be. The results of this study indicate that tax avoidance is not able to mediate the effect of profi tability on fi rm value. The existing data show that when the average profi tability and fi rm value decreases, the average CETR fl uctuates, which means that the practice of tax avoidance experienced fl uctuation during the study period. In addition, variations in CETR data are fairly large or the data are heterogeneous so that the distribution of data is classifi ed as poor. So, in this study, the variable of tax avoidance does not fulfi ll the requirements as mediating variable or intervening variable. The results of this study are not in accordance with the previous research conducted by Julian & Pratiwi (2017) that tax aggressiveness mediates the relationship between profi tability and fi rm value.

CONCLUSION, LIMITATION, AND SUGGESTION Conclusion
This study aims to test and analyze the factors that infl uence tax avoidance and fi rm value. The study population is banking sector companies registered at www.orbis.bvdinfo.com in 2014-2016. The method of selecting samples in this study is a sequential method that produces data after outliers as many as 319 data from the study period 2014-2016. This study uses secondary data in the form of the company's fi nancial and annual reports obtained from several sources including www.idx.co.id, www.investing.com and the company's website. After conducting classical assumption test and path analysis, the results of hypotheses testing in this study are as follows: The results of the fi rst hypothesis testing show that audit quality has no effect on tax avoidance. The size of tax avoidance practices carried out by company management does not depend on the size of the Public Accounting Firm. Data existing in the study period 2014-2016 showed an unstable tax avoidance action when the average audit quality continued to increase.
The results of the second hypothesis testing shows that profi tability has an effect on tax avoidance. The size of tax avoidance depends on the profi tability of the company. Data existing in the study period 2014 -2016 showed that tax avoidance actions had decreased when the average profi tability had decreased.
The third hypothesis testing shows that audit quality has an effect on fi rm value. When a company has good audit quality, the information obtained by investors is accurate and investors will be interested in investing in the company so that the company's stock price will increase and the fi rm value will also increase. Data existing in the study period 2014 -2016 showed that there was an increase in the fi rm value when the average audit quality increased.
The fourth hypothesis testing shows that tax avoidance has an effect on fi rm value. The greater the effort of company management in carrying out tax avoidance measures, the lower the fi rm value is. Data existing in the study 2014-2016 showed that the fi rm value decreased when the company's management effort in conducting tax avoidance increased.
The fi fth hypothesis testing shows that profi tability has an effect on fi rm value. The size of fi rm value depends on the profi ts that can be generated by the company. Data existing in the study period 2014-2016 showed that the fi rm value decreased when the average profi tability decreased.
The sixth hypothesis testing shows that tax avoidance is not able to mediate the infl uence of audit quality on fi rm value. The size of tax avoidance has no impact on the effect of audit quality on fi rm value. Data existing in the study period 2014 -2016 showed that when audit quality and fi rm value increased, the average CETR decreased, which mean that tax avoidance practices increases. In addition, the average variation of CETR data was large or the data were heterogeneous so that the distribution of data was classifi ed as poor, so tax avoidance did not fulfi ll the requirements as an intervening variable.
The seventh hypothesis testing shows that tax avoidance is not able to mediate the effect of profi tability on fi rm value. The size of tax avoidance does not have an impact on the effect of profi tability on fi rm value. Data existing in the study period showed that when profi tability and company value decreased, the company's management efforts in conducting tax avoidance fl uctuated. In addition, the average variation of CETR data was large or the data were heterogeneous so that the distribution of data was classifi ed as poor, so tax avoidance did not fulfi ll the requirements as an intervening variable.

Limitation
Although the researchers have tried to design and develop this study in such a way, the researchers realize that there are still some limitations to be addressed. These limitations include 21 annual reports of companies that do not use international language so that the data cannot be used as research samples. In addition, there were 51 annual reports that were not informative, such as no data related to the company's cash fl ow statement and stock prices so that the data cannot be used as a research sample, there are 7 annual reports of companies that are not available so that researchers have to delete the companies from the study sample, there are 42 companies that suffer losses so researchers must eliminate the data from the study sample, and there are 61 data outliers in this study so that the research result is not as optimal as expected by researchers.
In addition, the researchers only accessed data from www.investing.com to obtain information about stock market prices and the number of shares outstanding. For that reason, the information obtained may be incomplete. The results of the coeffi cient of determination (R2) test show the value of 0.082 or very small, so that the ability of the model to explain the dependent variable is still low. In this study, audit quality was measured using Accounting Firms affi liated with the Big Four Accounting Firm, whereas in other countries, such as Malaysia and Thailand, Public Accounting Firms such as Pricewaterhouse Cooper, Ernst & Young, and KPMG are the big four Accounting Firms. In this study, tax payment information was obtained from cash fl ow statement, where the tax payment describes all taxes paid by the company. Therefore, the information may not show the payment of corporate income tax.

Suggestion
The limitations such as the researchers should use a test tool other than SPSS to minimize outliers. The next, the researchers should develop their research by adding independent variables other than audit quality and profi tability because tax avoidance and fi rm value are infl uenced by several factors, such as internal and external factors. These factors include family relations and political relations. In addition, it is expected that the researchers add more sample because the sample represents the research population so that when the number of samples gets bigger, the level of generalization will increase.
In addition, the further study should provide research limitations, such as determining that the fi nancial statements used are in international language, accessing data from various sources related to stock market prices and the number of shares outstanding so that the information obtained is complete, and using the Current Effective Tax Ratio to measure the variable of tax avoidance so that the payment of taxes can show corporate income tax.