“The determinant of transfer pricing in Indonesian multinational companies: Moderation effect of tax expenses”

In calculating the transfer price of a transaction for goods, services, intangible assets, or financial transactions, a corporation has a policy known as transfer pricing. Due to its widespread abuse, transfer pricing is frequently associated with negative connota- tions. For example, this practice manipulates prices so that it has the potential to harm state revenues. This study uses tax expenses as a moderating variable to evaluate how intangible assets, debt covenants, and bonus systems affect the company’s decisions to use transfer pricing. This paper uses quantitative research approach with multiple linear regression analysis. The data used are panel data, consisting of cross-section data from 23 international manufacturing businesses on the Indonesian Stock Exchange, and time-series data from 2017 to 2019. Based on the tests, only the debt covenant variable significantly positively affects the transfer pricing action (sig. 0.000). In contrast, the intangible asset and the bonus mechanism variables are insignificant for transfer pricing. Furthermore, tax charges cannot mitigate the impact of intangible assets on transfer pricing decisions. However, tax charges may be able to mitigate the debt covenant in a way that makes the company’s decision to use transfer pricing stronger (sig. 0.024). Additionally, the bonus mechanism may be negatively moderated by tax expenses, weakening the company’s decisions to use transfer pricing (sig. 0.045). namely a special relationship at home and abroad. The parent company will manipulate the price by setting the price of goods or services traded to its subsidiaries or branches either at a discount or a premium to the market rate. A domestic holding company can also establish a business in a nation with a low tax rate, thereby entering into fictitious transactions.


INTRODUCTION
One of the tax planning concerns that has generated debate is transfer pricing. In terms of taxes, transfer pricing refers to a price strategy used in transactions between parties with particular relationships, namely a special relationship at home and abroad. The parent company will manipulate the price by setting the price of goods or services traded to its subsidiaries or branches either at a discount or a premium to the market rate. A domestic holding company can also establish a business in a nation with a low tax rate, thereby entering into fictitious transactions.
In Indonesia, the Procedure for Implementing the Transfer Pricing Agreement is governed by Regulation Number 22/PMK.03/2020 of the Minister of Finance of the Republic of Indonesia (Advance Pricing Agreement). The Directorate General of Taxes (DGT) may re-stipulate income subject to taxation for taxpayers with unique relationships. A unique relationship can occur if (i) a taxpayer owns at least 25% of the direct or indirect equity of another taxpayer; (ii) a taxpayer directly or indirectly controls another taxpayer or two or more taxpayers are under the same control; (iii) the presence of a blood relationship, straight line link, and/or a relationship to one degree's side. Several cases of transfer pricing have occurred in Indonesia. The PT. Asian Agri Group case is one illustration that occurred in early 2013. This was due to the disclosure of tax evasion by the Asian Agri Group in 2006 through transfer pricing. The Asian Agri Group's method is to sell Asian Agri's crude palm oil products to overseas affiliates at below-market prices to resell to original buyers at high prices. This method is employed to lessen the tax burden. In addition, foreign companies partnering with Asian Agri appear to be partially fictitious.
Various reasons can trigger companies to make transfer pricing decisions. These factors can come from non-financial and financial sources. Financial factors can emerge when a company wants to earn high profits to meet its operations, the existence of a high tax burden, and debt in the company. Managers look for loopholes using transfer pricing schemes. Meanwhile, non-financial factors stem from differences in the interests of owners and managers of a company. The owner wants to get the maximum profit, while the manager wants to get the highest return. Factors such as intangible assets, debt covenants, bonus mechanisms, and tax costs contribute to transfer pricing. Previous studies had mixed findings. For example, Firmansyah and Yunidar (2020) and Fadhilah (2018) found that the influence of intangible assets on transfer pricing has a significant positive direction. Meanwhile, Anh et al. (2018) and Khasanah and Suryarini (2020) stated that intangible assets significantly negatively affect transfer pricing.
In contrast, Anisyah (2018) noted that intangible assets have no impact on transfer pricing. Regarding debt covenants, Nuradila and Wibowo (2018) showed that debt covenants significantly improve transfer pricing. Meanwhile, Sari and Mubarok (2018) also revealed that debt covenants are detrimental to transfer pricing. This contrasts with Richardson & Lanis (2013), who stated that debt covenants have an absolute value in transfer pricing.
According to the bonus hypothesis idea, managers who select bonus plan policies will determine various strategies in accounting procedures so that reported earnings will experience changes from future periods to current periods. Nazihah et al. (2019) found that the bonus system positively impacts transfer pricing. With the high profits earned by a company, owners consider directors to have succeeded in carrying out their responsibilities well; thus, company owners will give bonuses to directors. This is different from Nuradila, & Wibowo (2018), who found that the bonus mechanism had no bearing on the choice of transfer pricing. Including the tax burden variable as a moderating variable, this study differs from earlier transfer pricing studies. The tax burden was selected as the moderating variable because it would push businesses to make transfer pricing decisions that would lead them to hunt for loopholes.

LITERATURE REVIEW
The relationship between the agent, who serves as the company's executor, and the principal, who is the owner, is described by agency theory (Jensen & Meckling, 1976). Differences in interests between owners and managers cause information asymmetry between the two parties leading to agency problems. One of the agency conflicts that arise is transfer pricing using intangible assets. Intangible assets have difficult to measure and identify characteristics but have future economic benefits. Valuing intangible assets at fair prices is also not easy. Thus, agents tend to organize their profits in-to intangible assets. High intangibles will definitely attract investors because a company has high profits from copyrights, royalties, and franchises, so that business would perform well in the community's eyes and generate more money. Fadhilah (2018) and Novira et al. (2020) indicated that intangible assets have a favorable impact. Therefore, based on agency theory, intangible assets are expected to have a favorable impact on transfer pricing decisions.
The positive accounting theory posits three theories, which were initiated by Watts and Zimmerman (1990). Specifically, the following three hypotheses apply (1) bonus plans, (2) debt covenants, and (3) political costs. Positive accounting theory states that a company with a large debt contract will look for loopholes to comply with the debt contract rules set to increase profits. One of them is through transfer pricing. As a result, managers tend to report more earnings and assets to reduce the renegotiation costs of debt contracts. Nuradila and Wibowo (2018) revealed that debt covenants have a considerable beneficial impact on the transfer price. This study also strengthens the results of Pramana (2014) and Richardson et al. (2013), who used a debt-to-equity ratio proxy. This study suggests that debt covenant favors transfer pricing decisions based on the debt covenant hypothesis theory.
The positive accounting bonus plan hypothesis theory explains the presence of a bonus mechanism given to managers (Watts & Zimmerman, 1990). The existence of bonuses in the company makes managers report company profits that are not under the situation. The higher the net profit of a company, the more reputable the board of directors appears to the business owner. Transfer pricing is one method that can be utilized.  showed that the bonus mechanism variable positively affects transfer pricing decisions. This study predicts that the bonus mechanism has a favorable impact on transfer pricing decisions based on the bonus plan hypothesis.
Based on agency theory, agents want to maximize their profits for business sustainability. Intangible assets are assets that have economic benefits but cannot be measured. Companies can use intangible assets to ease the burden of taxes. Dudar et al. (2015) suggested that related companies can share intangible assets with corporate affiliates in tax haven countries and then get rewards. Royalties from companies with high state tax rates and the amortization charging method can be used as deductible expenses for income in business activities. Article 6 paragraph (1) letter of the Income Tax Law also states that research and development costs, which are intangible assets, can be charged as deduction costs to calculate the amount of taxable income (PKP) in Indonesia. Based on the agency theory, this paper predicts that tax expense strengthens the relationship between intangible assets and transfer pricing.
The debt contract hypothesis theory explains that companies frequently use accounting techniques that allow them to modify earnings following their objectives. For example, with a high tax burden in a company, it will try to reduce taxes to get more profits, one of which is through transfer pricing. Borrowing debt from related parties to increase income tax deductions, Law Number 36 the Year 2008 paragraph 1 article 6 concerning income tax reveals that interest is an expense that can reduce taxable income. Therefore, a company will consider using debt to finance its business as a tax management measure to reduce the amount of tax the business has to pay without violating the applicable tax regulations. Thus, the company's debt will be utilized by managers to lessen the tax burden on a corporation by increasing interest costs so that company's profits can increase. The higher the profit, the higher the creditor's trust in the loan funds. Based on the debt contract hypothesis theory, this paper predicts that transfer pricing and debt covenant have a stronger relationship due to tax expense.
The bonus plan hypothesis theory explains that companies that choose bonus policies tend to take advantage of accounting methods that maximize profits. The agent wants to get an appreciation in the form of the highest reward in each period. A manager will attempt to report the net profit as much as feasible if the award is derived from it. A tax expense in a company will prevent a company from implementing the bonus mechanism using transfer pricing. A company will use other methods to increase its net income, such as increasing promotions, making new products, and carrying out tax planning by statutory regulation. Apart from that, if a company has a high tax expense, the company's profit before tax is also high. Consequently, using transfer pricing techniques to achieve bonus sharing has little to no impact on the business. This study makes a prediction based on the bonus plan hypothesis that tax expense diminishes the link between the bonus mechanism and transfer pricing.
This study uses tax expenses as a moderating variable to evaluate how intangible assets, debt covenants, and bonus systems affect the company's decision to use transfer pricing. The hypotheses of this paper are: H1: Intangible assets positively affect transfer pricing decisions.
H2: Debt covenant positively affects transfer pricing decisions.
H3: Bonus mechanism positively affects transfer pricing decisions.
H4: Tax expenses strengthen the relationship between intangible assets and transfer pricing.
H5: Tax expenses improve the relationship between transfer pricing and debt covenant.
H6: Tax expenses weaken the relationship between the bonus system and transfer pricing.

METHODOLOGY
The methodology in this study employs a quantitative approach. A quantitative approach uses data in the form of numbers to analyze predetermined variables. Secondary data, or information gathered from previously published sources, were utilized in the paper. The data taken in this study are in the form of publications with time-series data for three years (2017 to 2019). Annual reports and financial statements from international industrial companies listed on the Indonesian Stock Exchange are among the data sources (IDX). The population of this study consists of all companies in the multinational manufacturing sector registered on the Indonesian Stock Exchange (IDX) between 2017 and 2019. In the total population, 23 companies met the sampling criteria, with 61 units of final analysis ( Table 1).
where Y -transfer pricing; α -constant; β1-β6regression coefficient; ZX1 -the Zscore value of intangible assets; ZX2 -the Zscore value of debt covenant; ZX3 -the Zscore value of bonus mechanism; |ZATP -TE| -the absolute value of score of intangible assets and tax expense; |ZDC -ZTE| -the absolute values or core debt covenant and tax expense; |ZMB -ZTE| -the absolute value of the bonus mechanism and tax expense Z-score; εthe error term.
This study consists of one dependent variable (Y), namely transfer pricing, three independent variables (X) in the form of debt covenant, bonus mechanisms, and intangible assets, and one moderating variable (Z), namely the tax expense. The operational definition of variables is displayed in Table 2.

RESULTS
This study employs documentation methods for data collection using IBM SPSS Version 21 assistance in descriptive statistical analysis and inferential statistics. Inferential statistics are performed by examining moderate regression using the absolute difference value test. The debt cove- The purpose of the multiple linear regression analysis test is to objectively test the hypotheses that two or more independent variables can predict the dependent variable or that there is a functional link between them. The dependent variable (Y) in this study is transfer pricing, which is regulated by the tax expense (X1), the debt covenant (X2), and the bonus mechanism (X3). Moderated regression analysis is used in this research model. The im-  Bonus Mechanism The mechanism for providing additional salary or remuneration from the business owner to the manager who has done the best job of running the business to achieve the desired target where the performance achievement is the basis for giving bonuses such as net profit earned in that period.  pact of the moderating variable, specifically the tax expense (Z), is examined using the moderated regression analysis test. The absolute difference value test is used to examine the interaction between the independent and moderating variables and how it affects the dependent variable, as indicated by the absolute difference value between the independent and moderating variables. Table  4 displays the findings of the absolute difference value test.
A partial test is utilized to examine each independent variable's impact on the dependent variable separately. The partial test in t-test research utilizes a significance level of 5%, or 0.05. If the significance value of the independent variable is less than 5%, or 0.05, it impacts the dependent variable. Based on Table 4, the following is an analysis of the t-test results and the Sig value: 1. The Z-Score (X1) regression result for intangible assets has a t-count of 0.789 and a significance level of 0.434, higher than alpha 0.05 (0.434 > 0.05) with a t-value of 0.78. Since intangible assets have no impact on transfer pricing, the first hypothesis (H1) is rejected.
2. The results of the debt covenant regression (Z-score (X2)) have a t-count of 8.046 with a significance of 0.000, which is smaller than alpha 0.05 (0.00 < 0.05). This means that the debt covenant variable affects transfer pricing, and the beta coefficient value shows a positive result, meaning that the debt covenant has a positive effect on transfer pricing. Thus, it can be concluded that the second hypothesis (H2) is accepted.
3. The Z-score (X3) regression result for the bonus mechanism has a t-count of -0.486 and a significance level of 0.641 above 0.641 alpha (0.641 > 0.05) with a t value larger than zero.
In light of this, it may be argued that the bonus mechanism has no impact on transfer pricing, and the third hypothesis (H3) is rejected.
4. According to the regression analysis, the interaction variable between intangible assets and tax expense (ABSATP TE) has a t-value of -1.713 and a significance value of 0.92. The ABSATP TE variable has a significance value greater than 0.05 in this condition, demonstrating that tax expenses do not moderate the relationship between intangible assets and transfer pricing. As a result, it can be concluded that tax expenses cannot moderate the impact of intangible assets on the transfer pricing decisions (H4 is rejected).

The regression results
show that the t-value is 2.319 with a significance value of 0.024 for the interaction variable between the debt covenant and the tax expense (ABSDC TE). On the other hand, the ABSADC TE variable has a significance value of less than 0.024 (0.024 0.05), demonstrating that the tax expense can moderate the relationship between debt covenant and transfer pricing. Additionally, the beta value indicates a positive direction, supporting the hypothesis that the tax expense can moderate the positive effect of debt covenants on transfer pricing decisions (H5).
6. According to the regression results for the ABSMB TE variable, which is the interaction between the bonus mechanism and the tax expense, the t-value is -2.052, and the significance level is 0.045. On the other hand, the ABSADC TE variable has a significance value greater than 0.045 (0.024 0.05), demonstrating that the tax expense can moderate the relationship between debt covenants and transfer pricing. Additionally, the beta value indicates a positive direction, supporting the conclusion that the tax expense can moderate the positive effect of debt covenants on transfer pricing decisions (H6).
The corrected R Square value, according to the results of the coefficient of determination, is 0.745. This value means that the variables can explain 74.50% of the transfer pricing variable in this study, namely intangible assets, debt covenant, bonus mechanism, and tax expense as the moderating variable. In comparison, the remaining 25.50 percent (100.00% -74.50%) is affected in this study by other factors. In conclusion, the partial test and determination coefficient test using IBM SPSS version 21 resulted in several conclusions on the six hypotheses proposed in this study (Table 5).

DISCUSSION
The first hypothesis is rejected based on the findings of the regression test, which indicate that the number is higher than the significance level of 0.05. The results of this study also strengthen the results of Muhammadi et al. (2016), Jafri and Mustikasari (2018), and Anisyah (2018), who found that intangible assets do not influence transfer pricing decisions. This condition explains why the sample company's intangible assets, which are calculated using logarithms, cannot affect the choice of transfer pricing. This indicates that the sample company's intangible assets are not considered when determining transfer pricing. Research data show that transfer pricing is very low while intangible assets are very high. Therefore, there is a different direction between the two. This is what causes rejection. The study results are reversed with agency theory, where managers tend to increase profits on intangible assets through transfer pricing. The effective internal control system is to blame for this research's absence of a company and the behavior of managers who tend to want to be more transparent in reporting their profits. It demonstrates that a company's substantial intangible assets have no bearing on its decision to do transfer pricing.
The second hypothesis has a significance level of less than 0.05, meaning that the debt covenant hypothesis positively affects accepted transfer pricing. Nuradila and Wibowo (2018) and Richardson et al. (2013) also support that debt covenants positively affect transfer pricing decisions. This result also strengthens the positive accounting theory, where a higher company's debt ratio will increase the risk of its deficiency in meeting its debt obligations. High debt rates will threaten a company with bankruptcy or financial difficulties if it is not immediately paid off. Hence, a company tries to find loopholes by choosing accounting techniques that can generate the most considerable profit. One of the accounting procedures for increasing profits and circumventing lender rules is transfer pricing ).
The third hypothesis shows that the bonus mechanisms affecting transfer pricing are rejected. This study supports Saraswati and Sujana (2017), who stated that the bonus mechanism variable does not affect transfer pricing practice. According to positive accounting theory, managers will employ accounting strategies to improve profits for bonus distribution, but this study contradicts that theory. This study proves that the bonus mechanism does not affect transfer pricing decisions to increase profits in sample companies. Distribution data show that it can be seen from the maximum profit value of PT ISSP (Steel Pipe Industry of Indonesia Tbk), which increased by 564% from the previous year. However, PT ISSP only carried out transfer pricing measured by RPTAL of 0.009 or 0.9%. Tax expenses cannot mitigate the impact of intangible assets on transfer pricing choices. The outcome goes against the agency theory, which states that the agent wants to maximize the profit while the principal wants to maximize the return. Management will carry out accounting procedures to minimize this tax expense. The tax expense variable cannot moderate the relationship between intangible assets and transfer pricing. It can be because managers choose other creative accounting procedures that do not pose a considerable risk. Not only do large companies also have high value in the eyes of the public, but they tend to maintain firm value. One is not reducing tax costs and maintaining a transparent presentation of intangible assets. A company will seek to increase profits in other ways to maintain the company value it has obtained from intangible assets because it is considered to be more profitable than carrying out transfer pricing, which will only aim to drink the tax expense (Khasanah & Suryarini, 2020). Therefore, a company will maintain its image by not using transfer pricing, and the multiplier effect of the public's eyes will be excellent, and many investors will arrive.
The debt covenant may be moderated by the tax charges, supporting the company's choice to use transfer pricing. A corporation wants to avoid this tax because it has a significant tax bill. Based on tax regulations, loan interest expenses can be deductible in an annual SPT. By setting the interest budget high so that the company's profits can rise, managers employ debt contracts to reduce the business's tax burden (Nuradila & Wibowo, 2018). The positive accounting theory of the debt agreement section supports the findings, holding that managers are more likely to utilize accounting reporting procedures that can boost profits with a transfer pricing scheme when the company's debt ratio is higher. The amount of transfer pricing used by the corporation in an effort to maximize profits increases as the company's debt ratio rises. Transferring profits to nations with low tax rates will boost it by reducing the tax expense. Companies can also borrow funds to calculate their taxes to increase interest expenses to affiliates. Therefore, the tax expense can be a reason why companies commit debt covenants with a transfer pricing scheme.
The bonus mechanism may be negatively moderated by tax costs, making the company's decision to use transfer pricing less strongly. The findings of this study are further supported by the existence of transfer pricing regulations. The Arm's Length Principle, or ALP, which governs transfer pricing (the principle of fairness and business practice), will discourage management from implementing a bonus mechanism to reduce tax expense with a transfer pricing scheme. This rule discusses the rules for taxpayers in transactions with related parties. Another factor is that if a firm's tax expense is high, its profit before tax is similarly large; therefore, bonus sharing through the use of transfer pricing procedures does not have a significant impact on a company. The bonus plan hypothesis, which explains why managers have a bonus mechanism, led to this conclusion. As a result, managers frequently employ accounting techniques to reflect variations in profitability from one period to the next (Watts & Zimmerman, 1990). In addition, companies will use safe accounting techniques to increase corporate profits, such as tax planning by the law.

CONCLUSION
This study uses tax expenses as a moderating variable to analyze the impact of intangible assets, debt covenants, and bonus systems on the company's decision to use transfer pricing. Intangible assets and bonus structures have no discernible impact on transfer pricing, according to multiple regression from 23 multinational manufacturing companies listed on the IDX from 2017 to 2019. However, debt covenants have a positive and discernible influence on transfer pricing. The use of moderating variables in this study demonstrates that tax expenses cannot mitigate the impact of intangible assets on transfer pricing decisions. However, tax charges can either favorably or negatively modify the debt covenant, boosting or weakening the company's decision to use transfer pricing. In addition, tax expenses can also positively or negatively moderate the bonus mechanism.
Future research can use proxy measurements such as research and development (R&D) costs because this study used the total log of intangible assets and has proven to have no effect. However, R&D is one component of intangible assets, so it is hoped that the results obtained will be more specific. From the descriptive statistical analysis of the transfer pricing variable, 80.4% is in the very low category. Other proxies for assessing transfer pricing factors, like related party transactions of sales and expenses, are anticipated to be used in future studies (RPTSE), which can be seen from the income statement. Moreover, they can change the object of the research sample, such as using sample companies that do not reveal good corporate governance where the sample is allegedly more likely to have high transfer pricing potential. It is expected that companies will disclose related parties more following PSAK No. 7 because many of the sample companies have not disclosed transactions with related parties (affiliates) containing identities and relationships with linked parties affiliation as well as relevant receivables and payables. For the government, it can be used as information and reference to improve supervision to suppress transfer pricing within a company.
Based on the conclusions, this paper is expected to contribute to developing taxation and management knowledge regarding corporate decisions in transfer pricing. There are suggestions and inputs in several matters for the government. First, the Directorate General of Taxes and the Directorate General of Customs and Excise under the Ministry of Finance are expected to increase cooperation with the Corruption Eradication Commission to monitor and minimize transfer pricing action. Second, companies must carefully follow the requirements for tax provision and consider tax planning to avoid breaking any applicable tax laws and incurring legal or administrative repercussions. The contribution of this study is that a company must consider the tax expenses in making decisions related to transfer pricing.