The effects of audit client tenure, audit lag, opinion shopping, liquidity ratio, and leverage to the going concern audit opinion

Purpose – The going concern audit opinion is an audit opinion issued by an auditor to evaluate the company’s ability in maintaining the business continuity. The purpose of this paper is to discover the effects of audit client tenure, audit lag, opinion shopping, liquidity ratio and leverage on the going concern audit opinion. Design/methodology/approach – The study used secondary data obtained from financial reports and independent audit reports published by Indonesian Stock Exchange (ISE) as well as Indonesian Capital Market Directory. Besides, the population of the study included manufacturing companies registered in ISE from 2009 to 2013. Further, the present study applied purposive sampling technique which resulted in 16 companies used as the sample of the study. Then the hypothesis was examined by applying logistic regression. Findings – Results of the hypothesis examination indicated that the variables of opinion shopping and leverage affected the going concern audit opinion, whereas the variables of audit client tenure, audit lag and liquidity ratio did not affect the going concern audit opinion. Originality/value – Results of the hypothesis examination indicated that the variables of opinion shopping and leverage affected the going concern audit opinion, whereas the variables of audit client tenure, audit lag and liquidity ratio did not affect the going concern audit opinion.


Introduction
A financial statement is one of the media for the management to communicate the financial information to the stakeholders to assess the company's performance. A financial statement must have complete or comprehensive information in disclosing all facts done by the company during one period. The main purpose of the audit is to provide a sufficient trust that the financial statement has been reasonably presented according to the Generally Accepted Accounting Principles (Boynton et al., 2006). The result of the financial statement audit is an auditor opinion that is regarded as a public trust symbol to the information accountability presented in a financial statement (Aprilia, 2012).
In the ongoing process, the auditor is urged to evaluate a business continuity and a company's ability in maintaining the business for a limitless period. Going concern is an assumption which obliges the economic entity to have operational and financial abilities in maintaining their business continuity (Purba, 2009). According to Carson et al., an auditor is accountable for evaluating whether there is a substantial doubt about the company's ability in sustaining their business continuity for a reasonable period of time. The going concern audit opinion is a modification audit in which, regarding the auditor's consideration, it has a significant inability or uncertainty to the continuity of a company in running their operation (The Institute of Indonesian Public Accountant, 2011). The importance of going concern matters shows that it is needed to find out the factors influencing the going concern audit opinion in the relationship between auditor and client. According to Carson et al., the characteristics of the relationship between auditor and client encompassed audit client tenure, audit lag and opinion shopping.
Audit client tenure is the length of audit engagement between an auditor and a client (Knechel and Vanstrelen, 2007). The length of audit client tenure can improve the auditor's competence in making an auditing decision according to the auditor's ongoing knowledge, which is in line with the audit engagement. Nevertheless, due to the engagement length of audit client tenure, the auditor's independence can weaken because it will grow a relationship between the management and the auditor, whereas the short audit client tenure will weaken the auditor's competence because the auditor lacks knowledge about the company's environment in the early years of doing the audit.
Audit lag is the number of days between the end date of the financial statement and the issuance date of the auditing report (Ryu and Roh, 2007). Carson et al. asserted that the going concern audit opinion was more frequently met when the opinion issuance was late, whereas, according to , audit lag does not affect the going concern audit opinion in which the audit length does not influence the issuance of the going concern audit opinion.
According to The Security Exchange Commission (SEC), opinion shopping is an activity to look for an auditor with the goal of supporting the accounting treatment proposed by the management to achieve the goal of the company reporting although it may cause the report to be less successful. There are several factors that motivate a manager to conduct an opinion shopping. One of them is a willingness to achieve the target as well as the needs to maintain the continuity of the business. On one hand, the opinion shopping does not affect the going concern audit opinion in which it means the auditor's independence is not affected even though he/she is threatened to lose clients, should he/she provide the going concern audit opinion ( Praptitorini and Januarti, 2011). On the other hand, Lennox (2000) asserted that the opinion shopping influenced the going concern audit opinion.
The going concern audit opinion is also related to the financial situation of a company, in this case, the liquidity ratio and the leverage level. The liquidity ratio aims to measure the company's ability to fulfill the current liabilities (Weston and Brigham, 2001;Masyitoh and Adhariani, 2010). The smaller the liquidity of a company, it shows the struggle of the company in paying out the liabilities. Therefore, an auditor will possibly provide an audit opinion with going concern. On one hand, Januarti and Fitrianasari (2008) discovered that the liquidity ratio influenced the determination of the audit opinion with the paragraph of going concern. On the other hand, Masyitoh and Adhariani (2010) found out that the liquidity did not affect the issuance of going concern opinion by the auditor.
Besides, the leverage measures how far the company's assets are funded by debts (Riyanto, 2001). According Weston and Brigham (2001, p. 138), the leverage ratio aims to measure how far the company's financial needs are covered by loans. The bigger the company's assets are covered by loans, the more dependent the company to the loans in running their activities. Furthermore, the company should bear the bigger debts and interests. Ohlson (1980) asserted that the leverage had an influence in predicting the bankruptcy. Lennox (2000) also discovered that a company would frequently receive a going concern audit opinion when the leverage level was high. Nevertheless, Januarti and Fitrianasari (2008) argued that the leverage did not affect the going concern audit opinion.
The present study contributes to improving the previous studies in accordance with the influence of audit client tenure, audit lag, opinion shopping, liquidity ratio and leverage to the going concern audit opinion of a company. Results of the study indicated that the variables of the opinion shopping and the leverage affected the going concern audit opinion, whereas the variables of the audit client tenure, the audit lag and the liquidity ratio did not affect the going concern audit opinion.
The next section of the paper will present the literature review and the hypothesis development, the methods of study, the findings and discussions, and the last section consisting of conclusion, limitations and suggestions of study.
Literature review and hypothesis development Agency theory According to Astuti (2012), an agency problem will appear when a conflict of interest occurs between principals and agent. In this case, principals include the shareholders, whereas agent is the management. The relationship between the shareholders and the management can lead to the asymmetry of information. The shareholders want a decision resulting in high profits or the increase in the investment value, whereas the management wants an adequate compensation revenue of the work that the employees have done. Because of the difference in the desires, an independent third party is needed as a mediator between the principals and the agent.
An auditor is a party considered able to bridge the interests of the principals and the agent in managing the company's finance, so that the auditor has a monitoring function over the work done by a manager through a financial statement as well as considering the continuity of the company's business in running their business activities (going concern). The auditor's accountability encompasses providing the assurance service in the form of an evaluation of the financial statement made by the agent about the reasonableness of the financial statement. This evaluation finally results in an audit opinion. The audit opinion given by the auditor can be a measurement for the principals to assess the agent's performance in managing the company's business activities.
The going concern audit opinion According to PSA number 29 SA Section 208 (quoted by Agoes, 2012), the auditor's opinion is grouped into five types: unqualified opinion, unqualified opinion with explanatory language, qualified opinion, adverse opinion and disclaimer of opinion. The going concern audit opinion is a modification audit in which, in the auditor's consideration, there is a significant inability or uncertainty of the company's continuity in running their operation (The Institute of Indonesian Public Accountant, 2011). The auditor can provide a going concern audit opinion if he/she finds a condition or an incident during the auditing process that leads to the auditor's doubt about the sustainability of a company (Astuti, 2012).
Audit client tenure to the going concern audit opinion Audit client tenure is the length of audit engagement between an auditor and a client (Knechel and Vanstrelen, 2007), whereas, according to Januarti and Fitrianasari (2008), audit client tenure is the number of years when kantor akuntan public (KAP) conducts an audit engagement to the same company. A long audit engagement can cause the decreasing KAP independence because it can grow a relationship between the management and the auditor, whereas a short audit engagement can weaken the auditor's competence because he/she lacks knowledge about the company environment in the early years of doing the audit (Knechel and Vanstrelen, 2007): H1. Audit client tenure affects the going concern audit opinion.
Audit lag to the going concern audit opinion Audit lag is the number of days between the end date of the financial statement and the audit report issuance date (Ryu and Roh, 2007), whereas, based on Lee and Jahng (2008), audit report lag is a time period between the end of the company's fiscal year and the audit report date. Carson et al. indicated a possibility of delay of the issued opinion because the auditor tended to do tests and the management might do a long negotiation when the business uncertainty was discovered. The auditor delays issuing an opinion with the hope that the management can solve the problem, so that they can avoid the going concern audit opinion: H2. Audit lag affects the going concern audit opinion.
Opinion shopping to the going concern audit opinion Opinion shopping is defined by SEC as an activity to search for an auditor that is willing to support the accounting treatment proposed by the management to attain the goal of the company reporting. The company usually does an auditor shift to avoid receiving the going concern audit opinion (Alfasa, 2013). Several factors motivating the manager to conduct an opinion shopping include the willingness to fulfill the target as well as the needs to maintain the business sustainability (Praptitorini and Januarti, 2011): H3. Opinion shopping affects the acceptance of the going concern audit opinion.
The liquidity ration to the going concern audit opinion A company liquidity is defined as a company's ability to carry out their current liabilities (Munawir, 2001). The liquidity ratio aims to measure the company's ability in fulfilling their current liabilities (Weston and Brigham, 2001). In this circumstance, the smaller the liquidity of a company, it shows that the company only has few current assets to fulfill their accountability in repaying the debts whoe maturity is less than a year. Thus, the auditor will probably provide a going concern audit opinion: H4. The liquidity ratio affects the going concern audit opinion.
Leverage to the going concern audit opinion The leverage ratio aims to measure how far the company's financial needs are covered by loans (Weston and Brigham, 2001;Riyanto, 2001). The liability that is getting bigger shows the company's failure to repay the loans that are getting higher. Consequently, the auditor will possibly provide a going concern audit opinion ( Figure 1): H5. Leverage affectis the going concern audit opinion.

Methods of study
Research approach, type and data sources The study in this final project applied an associative quantitative approach of causal relationship by conducting an experiment to the proposed hypotheses. The data of the study encompass the independent auditor reports and the companies' financial statements. The data sources of the study were from the annual financial statements of the manufacturing companies that were listed in Indonesian Stock Exchange (ISE/BEI) during the period of 2009-2013.

Population and sample
The population of the study included all manufacturing companies listed in ISE during the period of 2009-2013. Then the purposive sampling technique was applied to choose the sample of the study. Several criteria considered in taking the sample of the study encompassed the manufacturing companies that were listed consecutively during 2009-2013, the companies who did not do delisting during the research period of 2009-2013, the companies experiencing a negative net income for at least three years or a negative equity for a year during the observation period of 2009-2013, the companies which financial statements and audit reports were available in ISE and the companies which financial statements ended on December 31.

Variables of study
The independent variables of the present study are audit client tenure, audit lag, opinion shopping, liquidity ratio and leverage, whereas the dependent variable of the study is the going concern audit opinion.
The variable operational definition. The dependent variable of the study is the going concern audit opinion. This going concern audit opinion is an audit opinion issued by the auditor if there is a doubt about the entity's ability in maintaining the business sustainability. This variable was measured by applying a dummy variable where the going concern audit opinion was labeled code 1, whereas things which were not included into the going concern audit opinion (unqualified opinion) were labeled code 0.
The independent variables of the study were: (1) Audit client tenure: audit client tenure is the length of audit engagement between the auditor and the same client (Knechel and Vanstrelen, 2007), whereas Januarti defined audit client tenure as the time period of the engagement established between the public accountant office (KAP) and the same auditee. In particular, audit client tenure is measured by calculating the years when the same KAP partner conducts an engagement with the auditee.
(2) Audit lag: audit lag is the number of days between the end date of the financial statement and the issuance date of the audit report (Ryu and Roh, 2007). Besides, according to Lee and Jahng (2008), audit report lag is a time period between the end of the company's fiscal year and the audit report date. In this case, audit lag is measured by calculating the length of KAP in doing the audit from the end date of the financial statements till the issuance date of the audit report.

Audit client tenure
Audit lag Opinion shopping The going concern audit opinion Liquidity ratio Leverage Figure 1.

Conceptual framework
(3) Opinion shopping: opinion shopping is defined as an activity searching for an auditor who is willing to support an accounting treatment proposed by the management to achieve the company reporting goal. This variable is measured by changing the auditor because the company received a going concern audit opinion in the previous year, whereas code 0 is given to the company in a mandatory way and to others.
(4) Liquidity ratio: the company liquidity is defined as a company's ability to fulfill their current liabilities or to analyze and interpret a company's current financial position (Munawir, 2001). This variable is measured by quick ratio which can well illustrate the liquidity level. Quick ratio is formulated in the following: QuickRatio ¼ The total of current assets-Stock ð Þ =Account Payable: (5) Leverage: leverage ratio aims to measure how far the company's financial needs are covered by loans (Weston and Brigham, 2001, p. 138). This variable is measured by debt ratio that is a ratio showing a proportion between the liabilities and the company's entire wealth. Leverage is formulated in the following: Debt to Total Asset Ratio ¼ The Total of Liabilities=The Total of Assets: Metode analisis data The data analysis method. The data analysis method applied in the present study included the logistic regression equation, the goodness of fit test, the overall model fit test, the determinant coefficient test, the model classification test and the hypothesis test. The level of significance in the hypothesis test was 5 percent.

Findings and discussions
The description of the object of the study According to the criteria in Table I, there were 80 companies meeting the sample criteria of the study.

The descriptive statistical analysis
The description of results of the study explains the minimum value, the maximum value, the mean, and the deviation standard or the frequency distribution from each variable of the study. The going concern audit opinion. Table II shows the frequency distribution of the going concern audit opinion and the non-going concern audit opinion every year. research period, the sample of companies receiving the going concern audit opinion were 56, out of 80 companies. It indicated that the manufacturing companies receiving the going concern audit opinion were 70 percent. Audit client tenure. Audit client tenure was calculated from the year of study backward to the year when the company employed the same partner. Table III showed the value of audit client tenure for a year, and for a maximum of six years. The average of audit client tenure The going concern opinion Year Non-going concern Going concern Total was 2.38 years with the deviation standard of 1.286. The average of audit client tenure in the companies receiving the going concern audit opinion had the average of audit engagement with the same KAP for 2.30 years and for a maximum of 6 years, whereas the audit client tenure in the companies receiving the non-going concern audit opinion has the average of audit engagement with the same KAP for 2.54 years and for the maximum of 5 years. Audit lag. The descriptive statistics in Table III indicated that the time needed by KAP to complete the Audit Report from the end date of the financial statements was, on average, 84.78 days, the minimum duration needed was 48 days and the maximum duration was 177 days, with the deviation standard of 16.545. The average of audit lag in the companies receiving the going concern audit opinion was 85.27. The minimum time needed was 48 days, and the maximum was 177 days with the deviation standard of 19.039, whereas audit lag in the companies receiving the non-going concern audit opinion had the duration of 83.62 days, the minimum duration needed was 60 days and the maximum was 102 days, with the deviation standard of 8.454.
Opinion shopping. Table IV showed that during the research period, the number of companies doing opinion shopping and obtaining the non-going concern audit opinion was 1, whereas the number of companies doing opinion shopping and receiving the going concern audit opinion were 17. Besides, the number of companies not doing opinion shopping and receiving the non-going concern audit opinion were 23, whereas the number of companies not doing opinion shopping and receiving the going concern audit opinion were 39. Therefore, the total number of companies doing opinion shopping were 18 samples, and the total number of companies not doing opinion shopping were 62 samples.
Liquidity ratio. The descriptive statistics in Table III indicated that the average of the company liquidity level was 0.8919 with the deviation standard of 1.84874 and the minimum and maximum liquidity levels were 0.08 and 10.51, respectively. The liquidity level in the companies receiving the going concern audit opinion was averagely 1.0045 with the deviation standard of 2.18309 and the minimum liquidity level of 0.08, and the maximum of 10.51, whereas the liquidity level of the companies receiving the non-going concern audit opinion was, on average, 0.6291 with the deviation standard of 0.48982 and the minimum and maximum liquidity levels of 0.09 and 2.08, respectively.
Leverage. The descriptive statistics in Table III indicated that the average level of the company leverage was 1.1266 with the deviation standard of 0.71561 and the minimum leverage level of 0.21 as well as the maximum leverage level of 2.99. The leverage level on the companies receiving the going concern audit opinion was, on average, 1.2647 with the deviation standard of 0.80855 and the minimum and maximum leverage levels of 0.21 and 2.99, respectively, whereas the leverage level of the companies receiving the non-going concern audit opinion was 0.8044 with the deviation standard of 0.20221 and the minimum and maximum leverage levels of 0.4 and 1.13, respectively.

The model analysis and the hypothesis test
The logistic regression analysis. The hypothesis test in the present study applied the logistic regression analysis technique because the dependent variable in the study was binary or The going concern opinion Non-going concern Going concern Total Non-opinion shopping Count  23  39  62  Opinion shopping  Count  1  17  18  Total  Count  24  56  80   Table IV.
The frequency distribution of the going concern opinion dichotomy and the independent variable was the combination of non-metric and metric. The data analysis was done by using the SPSS software, version 18.0. Before conducting the hypothesis test, the study conducted a logistic regression model test.
To assess the feasibility of the regression model, a goodness of fit test was conducted. From the goodness of fit test, it was discovered that the χ 2 was 7.424 with the significance level of 0.492. The significance level, which was more than 0.05, indicated that there was no difference between the prediction of the logistic regression model and the value of observation result. This test concluded that the model was feasible and acceptable.
Besides that, an overall model fit test was also conducted to assess whether the model hypothesized fitted the data. The overall model fit test was conducted by comparing the value of −2 initial Log Likelihood (Block number ¼ 0) with the value of −2 last Log Likelihood (Block number ¼ 1). Block Number 0 indicated that the value of − Log Likelihood was 97.738, whereas in Block Number 1, the value of -2 Log Likelihood showed the value of 80.022. The value reduction of 17.716 with the significance level of 0.003 showed that the model hypothesized fitted the data. Based on this test, the regression equations of audit client tenure, audit lag, opinion shopping, liquidity ratio and leverage could be applied.
The determinant coefficient. The determinant coefficient showed the value of Nagelkerke R 2 of 0.282. It means the independent variables (audit client tenure, audit lag, opinion shopping, liquidity ratio and leverage) could affect the dependent variable (the going concern audit opinion) for 28.2 percent, whereas the remaining percentage for 71.8 percent affected other factors outside the variables being studied.
The classification test. The classification test was applied to clarify the illustration of the logistic regression model accuracy with the observation results (Table V ).
The table indicated that from 24 samples receiving the non-going concern audit opinion, only 6 samples or 25 percent could be predicted precisely by the logistic regression model, and 18 samples could not be predicted precisely by the mode, whereas, from 56 samples receiving the going concern audit opinion, 48 samples or 85.7 percent could be predicted precisely by the logistic regression model, and 8 samples could not be predicted precisely by the model. Based on the results, 54 out of 80 samples, or 67.5 percent, were able to be precisely predicted by the logistic regression model.
Hypothesis test. Hypothesis test was conducted partially or individually on each variable by applying Wald test. The hypothesis test was assessed by using the logistic regression on the significance level of (α) 5 percent. The hypothesis was accepted if the significance value is less than 5 percent. If the significance value is more than 5 percent, the hypothesis is refused.
Table VI indicated that the audit client tenure had a significance level of 0.676 (greater than α), the audit lag had a significance level of 0.752 (greater than α), the opinion shopping had a significance level of 0.023 (smaller than α), the liquidity had a significance level of 0.877 (greater than α), and the leverage had a significance level of 0.021 (less than α). Based on the hypothesis test above, a logistic regression model could be acquired as in the following:

Discussions
The influence of audit client tenure to the going concern audit opinion According to the results of hypothesis test, the significance level of the logistic regression was 0.676. It indicated that audit client tenure did not affect the going concern audit opinion; thus, H1 was refused. This result was suitable with the study conducted by Knechel and Vanstrelen (2007) in which they discovered that there was no relationship between audit client tenure and the going concern audit opinion. It indicated that the auditor's independence was not affected by the length of audit engagement between the auditor and the client. Therefore, the auditor kept giving the going concern audit opinion to the companies which they doubted to be able to maintain their business sustainability without considering the length of audit engagement between the auditor and the client.
The influence of audit lag to the going concern audit opinion Audit lag affects the going concern audit opinion because the auditor tends to spend more time to audit a problematic company; in this case, they take time to meet the management for a certain period of time when the company may possibly receive the going concern audit opinion (Ryu and Roh, 2007). However, according to the results of hypothesis test, the significance level with the logistic regression of 0.752 indicated that audit client tenure did not affect the going concern audit opinion. Consequently, H2 was refused. This result showed that not all companies receiving the going concern audit opinion experienced a longer auditing process compared to the companies receiving a non-going concern audit opinion.
Pengaruh opinion shopping terhadap opini audit going concern Based on Lennox (2000), opinion shopping affects the going concern audit opinion in which it means the companies will possibly receive the going concern audit opinion if an auditor change frequently happens after the companies receive the going concern audit opinion. That statement is supported by the results of the hypothesis test indicated by the significance level of the logistic regression for 0.023. It means that the opinion shopping affected the going concern audit opinion. Thus, H3 was accepted. This result indicated that the auditor continued to act professionally and maintain his/her independence in doing the auditing process. The new auditor continued to provide a going concern audit opinion based on the real condition of the companies.
The influence of liquidity ratio to the going concern audit opinion According to Januarti and Fitrianasari (2008), the liquidity ratio has an influence in determining the audit opinion with the language of going concern. Nevertheless, based on the hypothesis test, the significance level with the logistic regression of 0.877 indicated that the liquidity level did not affect the going concern audit opinion; therefore, H4 was refused. The result of the study was in line with the statement of Masyitoh and Adhariani (2010) in which they asserted that the liquidity did not affect the issuance of the going concern opinion by the auditor. Further, this result showed that not all companies receiving the going concern audit opinion had low liquidity level compared to the companies receiving the non-going concern audit opinion.
The influence of leverage to the going concern audit opinion According to Ohlson (1980), leverage significantly affected the bankruptcy prediction. Based on the results of the hypothesis test, the significance level with the logistic regression of 0.021 indicated that the leverage affected the going concern audit opinion. Thus, H5 was accepted. This result of the study was suitable with a study by Lennox (2000), who found out that companies tended to frequently receive the going concern audit opinion when their leverage level was high. Because the assets used by the companies to run their operational activities are mostly covered by debts, the companies tend to depend on debts in running their business activities. Consequently, the companies bear a huge debt in which it can cause the company to not be able to afford repaying the debts.

Conclusions and limitations of study
The present study indicated that the opinion shopping and the leverage affected the going concern audit opinion. It showed that the auditor tended to give a going concern audit opinion to the companies applying the opinion shopping and having a high leverage level, whereas audit client tenure, audit lag and liquidity ratio did not affect the going concern audit opinion. Consequently, the auditor tended not to provide a going concern audit opinion over the company's audit client tenure, audit lag and liquidity ratio.
The present study has a limitation that can be considered in future studies. In this circumstance, the limitation was the variable of opinion shopping was only measured by considering the auditor change when the companies obtained a going concern audit opinion in the previous year. In this case, the present study did not predict the opinion which the companies would possibly receive when changing the auditor. The variables employed in the study were also limited in which the value of determinant coefficient of Nagelkerke R 2 , which was 28.2 percent, indicated that there were other factors outside the variables being studied for 71.8 percent. Thus, future research is strongly advised to add more variables in identifying the relationship between the opinion shopping and the going concern.