Audit firm rotation and audit quality: Comparison before vs after the elimination of audit firm rotation regulations in Indonesia

Abstract This paper examines whether the changes in audit rotation policy affect the relationship between audit firm rotation and audit quality in Indonesia. This study uses 2,403 observations from 371 firms listed on the IDX for the period 2010–2017. This study uses Ordinary Least Squares regression to test the hypothesis. We find that, in the mandatory audit rotation period (pre-period), there is no significant association between audit rotation and audit quality. Interestingly, we find a positive relationship between audit rotation and audit quality in the voluntary audit rotation (post-period). Our result implies that voluntary audit rotation without any coercive regulation is more effective in improving audit quality. This research provides an insight to policy makers that eliminating the regulation of audit firm rotation is an effective decision to enhance audit quality.


PUBLIC INTEREST STATEMENT
This paper analyzes audit regulation and its impact on audit quality. This regulation is considering the audit firm rotation, which is newly published in Indonesia. This study is essential because standard-setter and regulators need to measure whether their regulation is effective and have a positive impact on audit practice or not. This study compares the audit quality resulted from auditing practice before and after the regulation of audit firm rotation is applied. This study found that audit firm rotation doesn't affect audit quality. However, in the period after the regulation of audit firm rotation was abolished in Indonesia, there is a significant positive effect on audit quality. The audit quality is increasing after those phenomena. This result provides a reference to the regulator as well as management as the consideration in making decisions related to audit firm rotation.

Introduction
Audit quality has been questioned since the Enron Corp and WorldCom Inc. cases. Enron Corp is the largest energy company in the United States, having a very competitive executive council to create a new type of company energy. After several years of international and domestic expansion, Enron Corp had billions of dollars of debt, but this was hidden from shareholders through partnerships with other companies, accounting fraud, and illegal loans. Another case is from WorldCom Inc., the second-largest telecommunications company in the United States, which was running overcapacity and experiencing poor financials due to lack of demand of the agreement to pay lane leases to other providers in exchange for access in 1999. In 2000, the obligation to pay for the right to use cables became a problem. However, WorldCom Inc., under its chief executive Bernie Ebbers, used aggressive earnings management techniques to hide the problem by reducing its expenditure and increasing its income. Meanwhile, audit firm Arthur Andersen, who acted as auditor of the two companies, claimed that his failure to recognize the two scandals was due to executives hiding information from them. However, because this didn't just happen once, the audit opinion containing Arthur Andersen's signature lowers investor confidence (Handley-Schachler & Li, 2006).  (Presiden Republik Indonesia, 2015)., which only contained restrictions on the provision of general audit services on an entity's financial statements by a public accountant.
The limitation on the provision of general audit services on financial statements of an entity by an audit firm is also expected to bring client equity between large, medium, and small firms. However, the practice in Indonesia shows that, if an audit firm has reached its tenure limits, several audit firms tend to change their audit partner structure to retain clients, even though they have to pay more (Febrianto, Dini, Audina, Yuskar, & Juita, 2017). Therefore, regulations in Indonesia regarding restrictions on general audit services on financial statements by audit firms were abolished with the issuance of Government Regulation of the Republic of Indonesia Number 20 of 2015 concerning Practices of Public Accountants.
Study related to the audit rotation and audit quality is essential to do since audit quality is an assurance of financial statement quality. Moreover, this study seeks to compare the audit quality pre and post the abolition of audit rotation regulation. The result will be useful for both practitioners and regulators to examine the effectiveness of the regulation to make a better decision in the future. Previous research related to audit firm rotation and audit quality has been carried out in various countries, such as Australia (Jackson, Moldrich, & Roebuck, 2008), Jordan (Al-Thuneibat, Al Issa, & Ata Baker, 2011), Indonesia (Siregar, Amarullah, Wibowo, & Anggraita, 2012), Italy (Cameran et al., 2014)), South Korea (Choi, Lim, & Mali, 2017;Kim, Lee, & Lee, 2015;Kwon, Lim, & Simnett, 2014), Malaysia (Bamahros & Wan-Hussin, 2015), Oman (Baatwah, 2016), and Iran (Azizkhani, Daghani, & Shailer, 2018). Previous studies show that the rotation of audit firms influences audit quality (Al-Thuneibat et al., 2011;Azizkhani et al., 2018;Bamahros & Wan-Hussin, 2015;Jackson et al., 2008). Previous studies also show that there is a difference in audit quality between the rotation of audit firms in the mandatory and voluntary period (Choi et al., 2017;Colusi et al., 2014;Kwon et al., 2014;Siregar et al., 2012). However, another paper shows that audit firm rotation does not influence audit quality (Baatwah, 2016). This research extends the previous literature and gives contribution specifically by comparing between pre and post period of the abolition of audit firm regulation with Indonesia's specific setting. There are several ways to measure audit quality, such as from their audit opinion or audit report aggressiveness (Ocak, 2018), misstatement (Azizkhani et al., 2018), or discretionary accrual model (Choi et al., 2017). The measurements for audit quality in this study refer to the research of Choi et al. (2017) by using a discretionary accrual model (Kothari, Leone, & Wasley, 2005). Audit quality in the study of Azizkhani et al. (2018) used a proxy of misstatements in the company's financial statements and identified as auditor errors by IACPA. It was not used in this study because of differences in policies used and the limitations of data. This study uses discretionary accruals as a proxy for audit quality because it provides an indication of management intervention in reporting earnings (Al-Thuneibat et al., 2011).
This study used 2,403 observations from 371 companies listed on the IDX for the period 2010-2017. The analytical model used in this study is Ordinary Least Squares regression. This study found that companies with audit firm rotation had no effect on audit quality. However, in the period after rotation of audit firm regulations were abolished in Indonesia, there is a significant positive effect on audit quality. Audit quality is higher for companies which use audit firms in the period after the rotation regulation of audit firms was abolished in Indonesia rather than before.
The results of this study make a contribution, both theoretically and practically. Theoretically, this research can be used as a means to increase knowledge and insight into the rotational influence of audit firms in the period before and after the rotation of audit firms was abolished in Indonesia on audit quality, and can contribute in developing further research in related fields. Practically, it can be used as a consideration in making decisions regarding the rotation of the audit firm in the period after the rotation regulation of audit firms was abolished in Indonesia. This research consists of: Part 1, introduction; Part 2, literature review and hypothesis development; Part 3, a description of samples and variables; Part 4 which models the analysis and presentation of the main results; and Part 5 is the conclusion of the study.

Institutional setting
The obligation to do audit firm rotation in Indonesia was previously regulated in Article 3 Paragraph 1 of the Minister of Finance Regulation Number: 17/PMK.01/2008 concerning Public Accountant Services, which is obliged to rotate if an audit firm has provided general audit services for financial statements of an entity for six (6) consecutive financial years. However, the regulation of rotation obligations of the audit firm was abolished with the issuance of the Republic of Indonesia Government Regulation Number 20 of 2015 concerning the Practice of Public Accountants. This is because the purpose of the regulation was not fulfilled, namely for even distribution of clients. This is because in practice in Indonesia, if an audit firm has reached the specified limit, there are those that change their local identity by changing the structure of the partnership but still retain their foreign affiliates, in order to retain clients even though they have to pay more (Febrianto et al., 2017).

Hypothesis development
There are some determinant factors of audit quality and audit firm rotation is considered to be one of them. Kim et al. (2015) argue that there are two different perspectives related to audit rotation. The first view is from the pro side, which supports the application of audit rotation. The supporters argue that audit firm rotation can prevent a long relationship between the auditor and his client to become a personal relationship. Besides which, audit firms may have a good effect on audit performance. The new auditor may have a different point of view and give new insight into the client's financial statements. They will be more skeptical to identify questionable accounting practices and detect material misstatements in financial statements. However, whenever the external auditor's characteristics prove to be high, the audit report publication date is discovered to be short (Fakhfakh Sakka & Jarboui, 2016). Meanwhile, another perspective from the opponents of audit firm rotation argues that audit firm rotation can increase initial costs through the changing of the auditor. A new auditor starts from the beginning to understand the companies, the management business process, and the industry of the client. This condition will make auditors more dependent on management estimates and representations during the initial year of audit involvement, which can lead to a possible bias. However, there are some different conditions in practice. For example, the cases of Enron Corp. and WorldCom Inc. involving audit firm Arthur Andersen prove that audit quality increases when the rotation of the audit firm is carried out. The existence of audit rotation can prevent a long relationship between the auditor and his client, which can reduce the independence of the auditor. Increasing the independence of auditors contributes to the improvement of the quality of the auditing services (Hoang Tien, Thuong, Minh Duc, & Hoang Yen, 2019). Auditor independence must be maintained, especially about their audit fees. Zhang, Hay, and Holm (2016) found that the amount of audit fees does not suggest any loss of independence. However, higher audit cost leads to more efficient audit services as compared to companies having less audit cost (Khan & Abdul Subhan, 2019). The results of some previous research support the auditor independence case (Al-Thuneibat et al., 2011;Azizkhani et al., 2018). Al-Thuneibat et al. (2011) show that audit quality deteriorates when the tenure of an audit firm is extended as a result of growth in the magnitude of discretionary accruals. Meanwhile, study by Azizkhani et al. (2018) shows that the possibility of misstatements identified in the financial statements is significantly lower in the first two years of the audit tenure compared to the longer tenure period. Thus, the hypothesis proposed is:

H1: Audit firm rotation is positively related to audit quality
The second hypothesis is built through the argument that, with the issuance of the Government Regulation of the Republic of Indonesia Number 20 of 2015 concerning the Practice of Public Accountants, to minimize the strategies carried out by several audit firms in dealing with audit firm rotation regulations, they change their local identity by changing the structure of partnerships while maintaining foreign affiliations, in order to retain clients, even though they have to pay more, in the period before the issuance of the Republic of Indonesia Government Regulation Number 20 of 2015 concerning the Practice of Public Accountants, due to weak law enforcement in Indonesia (Febrianto et al., 2017;Siregar et al., 2012). Furthermore, by minimizing the strategies carried out by the audit firm, it means that the auditor has carried out his responsibilities under the policeman theory, which are to find and prevent fraud (Hayes, Dassen, Schilder, & Wallage, 2005). Thus the hypothesis proposed is: H2: In the post-period of the abolition of audit firm rotation regulation, audit firm rotation is positively related to audit quality The third hypothesis is built through the argument that, before the audit firm rotation regulation was abolished in Indonesia, the rotation of the audit firm was forced to fulfill the regulatory obligations, so that there were audit firms that changed their local identity by changing the structure of the partnership, but still retained their foreign affiliation in order to be able to retain clients, even though they had to pay more (Febrianto et al., 2017). The strategy carried out by the audit firm is subsequently expected to lead to client equality and the prevention of long relationships between the auditor and his client changing into a personal relationship. Therefore, with the issuance of the Government Regulation of the Republic of Indonesia Number 20 of 2015 concerning the Practice of Public Accountants, it is expected to minimize the strategies carried out by the audit firm, because the rotation of audit firms can be done naturally without coercion from applicable regulations in Indonesia. This is supported by the results of research by Choi et al. (2017), which show that audit quality with coercion of audit firm rotation must be lower than the voluntary audit firm rotation. Thus the hypothesis proposed is: H3: Audit firm rotation in the post-period of the abolition of the audit firm has higher audit quality compared with the pre-period.

Sample and source of data
The population used in this study is companies listed on the Indonesia Stock Exchange (IDX) for the period 2010-2017. Then, from that population, a sample selection is carried out based on some criteria, such as the company is not categorized in SIC 6 code (based on the US SIC Code), published their Annual Report and audited Financial Reports for the period 2010-2017, and has the necessary data and information with the variables used in this study. Furthermore, based on financial data obtained from OSIRIS as well as non-financial data obtained manually through the Annual Report and Financial Report, which can be downloaded at www.idx.co.id, from 4,545 observations that became population, a sample of 2,403 observations was obtained. The sample selection proscess is provided in the Table 1.

Audit quality
Audit quality is a joint probability whereby the auditor can find and report violations in the client's accounting system (DeAngelo, 1981). The measurement of audit quality variables in this study refers to the research of Choi et al. (2017), by using the Kothari et al. model (2005) of discretionary accruals in the following ways: a Measurement of the company's total accruals The company's accrual total is calculated using the difference between current assets other than cash for the current period and the previous period, minus the difference between short-term liabilities for the current period and the previous period, minus depreciation and amortization.
b Measurement of non-discretionary accruals Kothari et al.'s model (2005) attempts to perfect the Jones (1991) model by adding return on assets (ROA) to control performance. This model argues that incorporating elements of return on assets (ROA) in the calculation of discretionary accruals will be able to minimize specification errors.
c Measurement of discretionary accruals Non-discretionary accruals (NDAC) are the fitted values of the Kothari et al. (2005) model equation, while discretionary accrual (DACC) is the value of the residuals (ε).

Measurement of audit quality
Audit quality is the negative value of the absolute discretionary accruals.

Audit firm rotation
Audit firm rotation is the replacement of an audit firm from the provision of general audit services to the financial statements of an entity. The measurement of audit firm rotation variables in this study is using dummy variables. Value of 1 (one) if the company rotates the audit firm and the value of 0 (zero) if the company does not rotate the audit firm.

The period after the audit firm rotation regulation was abolished in Indonesia
The period after the regulation of firm audit rotation was abolished in Indonesia began when the Republic of Indonesia Government Regulation Number 20 of 2015 was issued concerning the Practice of Public Accountants. Measurement of period variables after the regulation of audit firm rotation was abolished in Indonesia in this study uses a dummy variable; the value of 1 (one) for the period after regulation of the audit firm rotation was abolished in Indonesia and the value of 0 (zero) for the period before the audit firm rotation regulation was abolished in Indonesia. The complete variable measurements are provided in Table 2.

Data analysis technique
The software used for data analysis in this study is STATA 14.0. Winsorizing is performed before doing data analysis. Furthermore, data analysis techniques are in the form of descriptive statistics, Pearson correlation, independent t test, and multiple linear regression. The research model used in this study includes: 1. The research model used to analyze the influence of audit firm rotation and the period after regulation of audit firm rotation was abolished in Indonesia on audit quality 2. The research model used to analyze the effect of audit firm rotation on the period before and after the regulation of audit firm rotation was abolished in Indonesia against audit quality Description:   Table 3 presents the audit firm rotation carried out by each industry. From 2,403 observations that became the sample of the study, there was found 372 observations (15.48%) which carried out an audit firm rotation, with details of 14 observations obtained from the agriculture, forestry and fisheries industries (code SIC 0); 58 observations obtained from the mining industry (code SIC 1), 106 observations obtained from the construction industry (SIC code 2); 91 observations obtained from the manufacturing industry (SIC code 3); 39 observations obtained from the transportation, communication, and utility industries (SIC code 4); 24 observations obtained from wholesale and retail industry (SIC code 5); 37 observations were obtained from the service industry (SIC 7 code); and three observations were obtained from the health, legal, and education services and consulting industries (code SIC 8). Meanwhile, based on the ratio of companies rotating the audit firm against the total related industrial sector, the highest percentage is found in the service industry with SIC 7 code, which is 20.22%, and the lowest percentage is in the health, legal and education consulting industry with SIC 8 code, which is 8.33%. Table 4 presents the average value, middle value, highest value, and the lowest value for each variable. Based on Table 4, audit quality has an average value of −0.085; the middle value of −0.054; the highest value is −0.001, and the lowest value is −0.480, while the rotation of the audit firm has an average value of 0.155; the middle value of 0; the highest value of 1; and the lowest value is 0.  Table 5 presents the relationship between independent variables and dependent variables. Audit quality decreased when the company rotated audit firm (−0.078) but increased in the period after the rotation regulation of audit firm was abolished in Indonesia (0.098). The tenure of audit firm, large international audit firm (BIG 4), firm size, firm age, and inventory and receivable ratio to total assets have a relationship that is directly proportional to audit quality (0.130; 0.086; 0.076; 0.120; and 0.098, respectively), but market value to book value has a relationship that is inversely proportional to audit quality (−0.114). Table 6 presents a comparison of the average company rotating audit firm and the companies that do not rotate the audit firm on the given variable. Based on Table 6, companies that carry out audit firm rotation have lower average audit quality (AQ) compared to companies that do not rotate the audit firm (−0.102 < −0.082). This also applies to the tenure of audit firm (TENURE), large international audit firm 4 (BIG4), firm size (FSIZE), and firm age (FAGE) which have lower average value when the company rotates FIRM AUDIT (1,046 < 4,451; 0.172 < 0.460; 433,013 < 469,941; and 3,230 < 3,316). However, it does not apply to market value to book value (MTB) as well as inventory and receivables ratios to total assets (INVREC), which have a higher average value when the company rotates audit firm (2,675 > 2,436 and 0.266 > 0.256). Table 7 presents a comparison of the average period before and after the rotation of AUDIT FIRM was eliminated in Indonesia in the given variable. Based on Table 7, in the period after rotation of AUDIT FIRM regulation was abolished in Indonesia, the average audit quality (AQ) value was higher than in the period before the AUDIT FIRM rotation regulation was abolished in Indonesia (−0.074 > −0.093). This also applies to the tenure of audit firm (TENURE), firm size (FSIZE), and firm age (FAGE), which have a higher average value in the period after rotation of the audit firm regulations was abolished in Indonesia (4,938 > 3,235; 475,059 > 456,878; and 3,357 > 3,266). However, it does not apply to large international audit firms (BIG4), market value to book value (MTB), and inventory and receivable ratios to total assets (INVREC), which have a lower average value in the period after rotation of AUDIT FIRM abolished in Indonesia (0.415 < 0.416; 2,315 < 2,580; and 0.240 < 0.270). Table 8 presents the rotational effects of the audit firm and the period after AUDIT FIRM rotation regulation was abolished in Indonesia on audit quality in columns (1) and (2), while the AUDIT FIRM rotational influence in audit firm before and after rotation regulation was abolished in Indonesia on audit quality in columns (3) and (4). Column (1) and (3) present the regression results normally, while columns (2) and (4) present regression results robustly.  This study analyzes the rotational influence of audit firm and the period after AUDIT FIRM rotation regulation was abolished in Indonesia for audit quality by controlling the tenure AUDIT FIRM variable, international large audit firm, firm size, firm age, market value to book value, and inventory and receivable ratios of total assets using the regression model as follows:

Main analysis
Based on columns (1) and (2) in Table 8, the audit firm rotation (ROTATION) has a coefficient of −0.008 (t = −1.21 and t = −1.07), but it is not significant either with normal or robust regression. This means that the audit firm rotation does not affect audit quality, so the first hypothesis is rejected. Meanwhile, the period after the regulation of audit firm rotation was abolished in Indonesia (POST) has a coefficient of 0.014 (t = 3.31 and t = 3.44), with a significance level of 1%, both with normal and robust regression. This means that the period after the regulation of audit firm rotation was abolished in Indonesia has a significant positive effect on audit quality, so the second hypothesis is accepted. This result supports the argument that the period after regulation of audit firm rotation was abolished in Indonesia can make   (1) and (2) in Table 8 shows the relationship between audit firm rotation and the period after regulation of audit firm rotation was abolished in Indonesia with audit quality explained by 6.7% of the 2,403 observations that became the research sample.
This study analyzes the effect of audit firm rotation on the period before and after regulation of audit firm rotation was abolished in Indonesia on audit quality by controlling tenure audit firm variables, international BIG 4 audit firms, firm size, firm age, market value to book value, and inventory ratio and accounts receivable against total assets using the regression model as follows: Based on columns (3) and (4) in Table 8, the rotation of the audit firm in the period before the regulation of the audit firm rotation was abolished in Indonesia (ROTATION) has a coefficient of −0.019 (t = −2.65 and t = −2.08), with a significance level of 1% when using regression normally and 5% when using regression robustly. This means that the rotation of the audit firm in the period before the regulation of audit firm rotation was abolished in Indonesia has a significant negative effect on audit quality. Meanwhile, the rotation of the audit firm in the period after rotation of AUDIT FIRM regulations was abolished in Indonesia (ROTATION * POST) has a coefficient of 0.034 t statistics in parentheses *p < 0.1, **p < 0.05, ***p < 0.01 (t = 3.22 and t = 3.10), with a significance level of 1%, both with normal and robust regression. This means that the rotation of the audit firm in the period after the regulation of the audit firm rotation was abolished in Indonesia has a significant positive effect on audit quality. Through these two results, the third hypothesis is accepted, i.e. the rotation of the audit firm in the period after regulation of the audit firm rotation was abolished in Indonesia has a higher audit quality compared to the rotation of the audit firm in the period before the audit firm rotation was abolished in Indonesia. This result supports the argument that the rotation of the audit firm in the period after rotation of AUDIT FIRM was abolished in Indonesia can be done naturally without any coercion from prevailing regulations in Indonesia and does not need to carry out its own strategy in dealing with regulation of audit firm rotation. Furthermore, the value of r2 in columns (3) and (4) in Table 8 shows that the relationship between the audit firm rotation in the period before and after the rotation of audit firm was abolished in Indonesia with audit quality is explained by 7.1% of the 2,403 observations that were the research sample.

Conclusion
This study found that companies that carry out audit firm rotations have no effect on audit quality. However, the period after the regulation of audit firm rotation was abolished in Indonesia has a significant positive effect on audit quality. This shows that the period after regulation of audit firm rotation was abolished in Indonesia can make an audit firm adhere to the policeman theory and fulfill the expectation of the issuance of Government Regulation of the Republic of Indonesia Number 20 of 2015 concerning the Practice of Public Accountants, namely to minimize strategies conducted by audit firms in dealing with audit firm rotation regulations, in the period before the issuance of the Republic of Indonesia Government Regulation Number 20 of 2015 concerning the Practice of Public Accountants. Furthermore, the rotation of the audit firm in the period after regulation of the audit firm rotation was abolished in Indonesia has a higher audit quality compared to the rotation of the audit firm in the period before the regulation of the audit firm rotation was abolished in Indonesia. This shows that the rotation of the audit firm in the period after regulation of the audit firm rotation was abolished in Indonesia can be done naturally without coercion from applicable regulations in Indonesia and does not need to change their local identity by changing the structure of the partnership but still maintaining their foreign affiliation, if the company and the audit firm refuses to carry out an audit firm rotation, so that the expectation of increasing audit quality can be realized if the audit firm rotates voluntarily.
The limitation of this study is only using discretionary accruals as a proxy for audit quality. Future studies can use other proxies for audit quality or use multiple proxies for audit quality.