A study on audit report timeliness: The Macedonian Stock Exchange

The International Accounting Standards (IAS) were the first worldwide accounting principles delivered by the International Accounting Standards Committee (IASC), established in 1973. The main objective then, as it is today, was to facilitate business comparisons across the borders while increasing financial reporting trustworthiness and promoting investments.

The globally adopted reporting standards promote transparency allowing financial statement users to make informed investment decisions. In this regard, some authors even emphasize that IAS contribute to a more stable world. Moreover, the conformity with these unique standards reduces the reporting and governance expenses for internationally dispersed entities whose subsidiaries operate in multiple legislation milieus. As the International Accounting Standards Board (IASB) took over IASC's role, significant progress has been made toward the goal of setting universal international accounting principles of undisputable quality.

The International Financial Reporting Standards (IFRS) have since been embraced by the European Union, leaving the United States, Japan, and China as the only major non-IFRS-mandated capital markets. As of July 2020, 166 states call for the implementation of IFRS for domestic publicly listed enterprises [IFRS, n.d.]. The American Accounting Standards Board—The Financial Accounting Standards Board (FASB) and the IASB are working together to advance and converge IFRS and the US Generally Accepted Accounting Principles (US GAAP). Nevertheless, even though the IASB and FASB jointly issue accounting standards, the process of coming together is still ongoing, partly due to the enacting of the Dodd-Frank Wall Street Reform and the Consumer Protection Act and its intricateness.

In North Macedonia, the IAS were introduced in 2004 with the adoption of the Rulebook on accounting [2015]. Article 469 stipulates that every large and medium-sized entity, the entities determined by law, entities performing banking activities, insurance activities, entities listed on the stock exchange, and entities whose financial reports form part of the consolidated statements of the aforementioned entities need to comply with the IFRS published in the “Official Gazette of the Republic of Macedonia”. The Amendment to the Rulebook on Accounting determines the differences between the accounting standards that refer to large enterprises and the standards for small- and medium-sized entities (SMEs). Following the implementation of the amendments to the Law on Trade Companies [2019], mandatory use of IAS for SMEs is introduced in North Macedonia. The IFRS for SMEs are completely independent standards whose use is expected to provide higher quality in financial reporting and also to simplify bookkeeping operations in small and medium firms.

In the Republic of North Macedonia, the Law on Audit [2015] regulates the role and status of the audit activity. The law states that there is a statutory and contractual audit. Statutory audit means mandatory examination, procedures, checks, and assurances on the annual accounts and/or financial statements or consolidated financial statements performed in conformity with the law, IFAC's International Standards on Auditing, and other relevant publications of IFAC that are accepted and published in the Republic of North Macedonia to express an opinion, conclusion, or assurance regarding their veracity, impartiality, and compliance with the reporting standards.

The law also determines who can perform the audit, i.e., only a person holding a certificate issued by the Institute of Certified Auditors of North Macedonia and registered in the Registry of Certified Auditors. The law also defines the audit company as an entity registered in accordance with the Law on Trade Companies and the Registry of Audit companies and holding a license to operate from the Audit Promotion and Supervision Council.

In addition to the certificate issued by the Institute of Certified Auditors to professionals in the audit business, other conditions regulated by law are required. That is, a license for certified auditor is issued to a person who has obtained the auditor certificate and has at least 3 years of experience in auditing, of which 2 years under the supervision of a certified auditor. In this regard, the law does not allow a person who does not have the status of certified auditor and who is not a member of the Institute to perform auditing services in North Macedonia or to present himself/herself as a certified auditor or use any label or description that will create the impression that the person is a certified auditor.

The amendments to the Law on Audit [2015] were issued within the time frame covered by this study (i.e., 2014–1017) but have no effect on the study results before and after the new Law issue date. Namely, the amendments do not affect the legally prescribed time limit for the publication of the annual report of entities listed on the MSE as this is within the scope of the Law on securities [2005]. That is, the legally prescribed time limit for listed entities is 4 months [Law on securities, 2005].

When deciding where to invest, investors seek confidence and security in the potential allocation of their funds, i.e., they want stability and profitability in the realization of their business decisions. That is why the role of the auditor is invaluable for the business world. In that context, the key beneficiaries of the audited financial statements are the company managers, who evaluate company performance; the shareholders who, based on the audit report, make decisions about their investment in the company; banks in terms of liquidity of the enterprise, i.e., whether to extend financial support or not; the tax authorities in determining the tax liabilities; and other state institutions in terms of whether or not to undertake certain activities to control the enterprise scope of operations. Also, beneficiaries of the audit report are workers and trade unions in terms of negotiations with the company to increase or decrease wages, or change the number of employees. Audit reports are subject to elaboration by economic analysts in relation to statistical analyses for the purpose of adopting better economic policies. Hence, it can be concluded that the role of the external auditor is very important because the audit opinion refers to a wider range of users and whereby the negative consequences of an unreliable report may be consequential for the entire business world. These consequences were experienced in the past with the collapse of Enron, WorldCom, Tyco, Parmalat, Lehman Brothers, and, more recently, the Wirecard scandal.

The primary task of the auditor is to give assurance regarding the financial statements. Namely, the purpose of the annual account audit is to allow the auditor to assess whether the financial statements are prepared, from a material point of view, in compliance with the pertinent reporting legislation [Bozinovska-Lazarevska, 2011, p. 8]. In this regard, the auditor provides an answer and assurance to multiple financial statement users that the information is presented objectively and truthfully. Or, while accounting aims to identify, record, measure, classify, and summarize the business events and transactions of the enterprise for the purpose of enabling the management to compile and present the financial reports, the independent auditor checks and confirms the veracity and objectivity of the presented data based on the collected evidence.

The American Accounting Association (AAA) defines an audit as a “systematic process of collecting and assessing proof related to reports of economic events and results in to verify the compatibility between existing operating statements and pre-established criteria and to distribute the conclusions to stakeholders” [Bozinovska-Lazarevska, 2011, p. 4]. According to the International Federation of Accounting (IFAC), “the purpose of the audit is to increase the reliance of users in the presented financial information” [Bozinovska-Lazarevska, 2011, p. 5]. The audit opinion is important because it is a bridge of trust that connects companies and users of financial statements.

According to Indriyani and Supriyati [2012], timeliness is the time required to conduct an external audit of the annual accounts. That period is expressed as the number of days necessary for the publication of the audit opinion calculated from the accounting year-end until the publication date of the external audit report. Delays in the audit report lead to a reduction in the quality of the information communicated in the financial reports which increases the uncertainty of decisions based on the disclosed data. In his 2011 paper, Akle points out that timeliness means that financial reports ought to be disclosed when users rely on them to form opinions because information usefulness declines if it is not at one's disposal when necessary. Several US studies argue that delayed earnings’ releases have an unfavorable impact on stock prices [Chambers and Penman, 1984; Kross and Schroeder, 1984], thus inducing stakeholders to rely on other sources for information on company performance and triggering information leakages [Brown et al., 2011]. This feature becomes even more imperative in developing economies. Concurrently, Aktas and Kargin [2011] define timeliness as the days that elapse between the fiscal year-end and the date when listed entities disclose their annual report in compliance with legal regulations.

The time required to conduct an audit from the closing date of the books to the audit opinion issue date is called the “audit delay” according to Davies and Whittred [1980]. Dyer and McHugh [1975] define the audit lag as the period between the fiscal year-end and the audit report issue day. Iskandar and Trisnawati [2010] claim that the length of time needed to complete the audit—the timeliness of the audit opinion—will affect the time needed to publish the annual report. This study measures the timeliness as the days between 31 December, the legally set fiscal year-end in North Macedonia, and the annual report publication date, whereby the legally prescribed time limit is 4 months [Law on securities, 2005].

Timeliness depends on factors related to both the auditor and the audited company such as the time needed to prepare the financial statements, the board of directors decision to publish the financial statements, regulatory requirements for a minimum period between the Annual General Meeting date, and the reporting date as well as other factors related to the availability of a date or place for the Annual General Meeting [Ahmed and Karim, 2005]. According to Karim et al. [2006], the timeliness of the financial statements is influenced by two types of determinants—external determinants (like regulation, competition) and internal determinants (such as company features), and consists of three groups: audit procedure time frame, opinion publication time frame, and overall timeliness.

Wiwik [2006] argues that delays in the financial statements disclosure are very harmful for shareholders because they lead to market information asymmetry, thus leading to insider trading and innuendos that affect the stability of financial markets.

For this research, the authors define a linear regression model to assess the influence of explanatory variables on audit report timeliness (Model 1). Namely, in Model 1 the dependent variable LAG is the time period required for publishing the audit report. Variable LAG is calculated and expressed as the days from the accounting year-end for which the audit opinion is published till the opinion disclosure date.

Model 1: (1) LAG=α+β1 CR+β2 LEVER+β3 ROE+β4 LOSS+β5 LnASSETS+β6 BIG4+β7 QAR+β8 MANUF+n \matrix{{\rm LAG} \hfill & = \hfill & {\alpha + \beta 1\,{\rm{CR}} + \beta 2\,{\rm{LEVER}} + \beta 3\,{\rm{ROE}} + \beta 4\,{\rm{LOSS}} + \beta 5\,{\rm{LnASSETS}} + \beta 6\,{\rm{BIG}}4} \hfill \cr {} \hfill & {} \hfill & { +\, \beta 7\,{\rm{QAR}} + \beta 8\,{\rm{MANUF}} + n} \hfill \cr } where LAG—is a dependent variable, the timeliness of the audit opinion (in days), α—constant, CR—independent variable, liquidity, represented by current ratio (ratio between short-term assets and short-term liabilities), LEVER—independent variable, indebtedness represented as the ratio between liabilities and assets, ROE—independent variable, profitability represented as ROE—ratio between net profit and total equity, LOSS—independent variable, loss in the current year represented as a dichotomous variable with a value of 1 in case of current year loss and 0 otherwise, LnASSETS—explanatory variable, company size represented as natural logarithm of total assets, BIG4—independent dichotomous variable with value 1 in the case of the four major international auditing firms (Deloitte, KPMG, PricewaterhouseCoop, Ernst & Young) and 0 otherwise, QAR—explanatory dichotomous variable with value 1 in case of a modified opinion or a prominent issue that affects the audit opinion (qualified opinion, adverse opinion or disclaimer of opinion) and 0 otherwise, MANUF—independent dichotomous variable which is assigned the value 1 in case of a listed company with production as basic activity and 0 for service companies and companies in retail and wholesale, β1–β8—regression coefficients of the explanatory variables; and Ɛ—random error.

For the purposes of this research, we will transform the absolute value of assets into a natural logarithm of assets to better satisfy the assumption of an approximately linear relationship between variables modeled after authors such as Moalla [2017].

In this paper, the emphasis is on nonfinancial companies listed on the MSE. In that regard, the sample includes all publicly traded companies that publish their financial reports and audit opinions on the official website of MSE's electronic reporting system (https://www.seinet.com.mk). Data on total assets, liquidity, solvency, current-period loss, and sector were extracted from the financial reports of the companies. The identity of the auditor, the auditor opinion, and the publication date are part of the audit opinion which is published together with the annual report of the companies.

We assessed 99 enterprises over a 4-year period, from 2014 to 2017. This sample encompasses all publicly listed nonfinancial entities with available audited consolidated financial statements and a final audit opinion. The Macedonian financial market is incipient. During the selected period, the companies are subject to IFRS and the audit is compliant with IFAC's ISAs. Like most previous surveys, financial entities are eliminated from the sample as they abide by separate financial reporting rules. Table 1 illustrates the sample and the selection procedure.

The descriptive analysis of the dependent and the explanatory variables is presented in Tables 2 and 3.

Table 2 indicates that for 47% of the overall sample, a modified opinion was issued, which means a qualified opinion, adverse opinion, or disclaimer. Moreover, there is an unfavorable change in the audit opinions and a slight transition in favor of a modified opinion over the years. This claim is supported by descriptive statistics which indicate that unlike 2014 when 47% of the total 98 audit reports were unqualified, their share drops to 39% of the total 88 published opinions for 2017. These findings are similar in terms of modification rate to Tunisia as an emerging economy but are far worse than the rates of developed countries [Moalla, 2017].

Most of the remarks made by the auditors in the context of qualifications are related to customer receivables and investments available for sale. Namely, it is common practice for companies to show old receivable balances in the balance sheet without making a proper value adjustment in accordance with the requirements defined in IAS 39 Financial Instruments: Recognition and Measurement. As a consequence, there is an overestimation of receivables from customers and the financial result of the companies concerned. Furthermore, as per IAS 39 Financial Instruments, Recognition and Measurement, entities need to estimate the fair value of available-for-sale investments at the financial statement date according to their market price. In the absence of such an objective assessment and given the presentation of these securities at cost, there is an inadequate estimation of their value leading to an incorrect financial position and financial result.

Some of the companies in the examined sample recognize assets on which they do not have ownership, which would reduce the total assets and the capital that the companies show in the balance sheet for the period. In other companies, the auditor concludes that current liabilities exceed short-term assets, which calls into question the liquidity of the company. Also, several companies do not generate income from their core business activities but from other (one-time) income sources, primarily from the sale of tangible and intangible company assets, as well as revenues from the collection of written-off receivables, which calls into question the capability of the company to continue as a going concern.

Qualifications are oftentimes related to the lack of an estimate of the net realizable value of inventory. As a result, the firm does not present inventory at a lower than cost and net realizable value, which deviates from the requirements of IAS 2 Inventories. Thus, the financial result as well as the financial condition for the year that are presented in the annual accounts of these companies are overestimated.

In some of the modified opinions, the audit firm states that they are appointed after the inventory count date which is conducted at the very beginning of the new fiscal year or that the auditor has been excluded from the stock-taking all together. Thus, the auditor is not able to determine the initial amount of inventory. Because the initial inventories affect profit and cash flows, the auditor is unable to determine whether adjustments to the stated financial result and cash flows from operations are necessary.

Table 3 indicates that the average time required for preparation and publication of the audit report is 117.5 days or almost 4 months, which is in accordance with the legally prescribed time frame of 4 months from the calendar year-end [Law on securities, 2005]. A total of 162 reports from the overall sample were published with a delay, i.e. 43.9% of the analyzed reports took more than the legally prescribed 4 months to realize the audit and to disseminate the final product. For the publication of 8 of the indicated 162 reports, the preparation period takes over 6 months showing a noticeable delay. For 7 of the 8 disputed reports, a modification of opinion was issued, namely a qualified, an adverse opinion, or a disclaimer. Only one opinion is positive but contains a modification by applying an explanatory paragraph (emphasis of question). At the same time, Toplifikacija AD Skopje, one of the eight companies with a significant deviation in the deadline for publishing the audit opinion, is included in the empirical analysis only for the period 2014–2016 as a result of open bankruptcy proceedings against the company after a court hearing held on November 2, 2018.

Furthermore, Table 3 shows that ROE has an average value of 0.8%, which indicates a low average profitability of the analyzed firms, especially compared to other developing countries such as Tunisia, where the same parameter reaches double-digit values [Moalla, 2017]. In the global practice, investors often consider a ROE close to the S&P 500 average of 14% as an acceptable ratio, with values <10% considered unattractive. Logically, the values (high or low ROE rates) differ significantly from one industry to another, and the comparison will be more significant when this indicator is used to estimate companies belonging to similar sectors. In our case, negative values were observed in 86 out of 369 analyzed annual reports. The ROE values range from −238% to 175%, which indicate a large range in the ratio. The mean value of ROE and indebtedness highlights the financial risks and hardships of companies in North Macedonia and may interpret the audit lag.

Multiple linear regression relies on several key assumptions that needed to be tested. Primarily, a normal distribution of the dependent variable values is expected. The verification was done with the help of the Kolmogorov-Smirnov and Shapiro-Wilk tests. We observed an absence of normal distribution in the dependent variable because both tests determined that Sig. <5%. Since the normality assumption is not met, then the statistical base used in the linear regression should contain >20 observations for each independent variable. In our case, the additional requirement is satisfied given the fact that the database contains a total of 369 competent observations, which is well above the established threshold of 160 observations (i.e., over 20 observations for each of the eight explanatory variables).

Furthermore, there needs to be a linear relationship between dependent and explanatory variables. X–Y diagrams can show whether there is a linear or curvilinear relationship. The scatter plot test results are shown in Figure 1.

Figure 1

We notice that the assumption is not fulfilled in the part of the y-axis where five values, i.e., observations come out of the upper limit value (i.e., +3).

The individual inspection of the identified deviations indicates that this is not an error in the data but observations in which there is an obvious excess in the mean value of the independent variable—number of days necessary to conduct the audit. The financial statements of the companies reveal difficulties in terms of profitability, decline in sales revenues, and/or loss, which, together with the prominent views of the auditors, explain the pronounced delay in publishing the audited annual accounts of the three companies concerned.

Additional statistical analysis is necessary because data with large deviations can impair the regression accurateness and results. Cook's distance (commonly known as Cook's D) weights the effect of eliminating a case. There are different opinions as to which limit values should be used to mark points of pronounced impact. The diverse attitudes usually converge around a maximum limit value of 1, which is observed in our sample given the fact that both the upper and lower values of Cook's D are far <1 [Statistics Solutions, 2019]. As a result of these additional tests, we can conclude that the deviations in our statistical sample are quite acceptable, especially given the fact that we are not discussing errors but realistically explainable case values.

The next important condition is that of multivariate normality in all variables. Namely, multiple regression assumes that the residuals are normally distributed. The test results indicates that the multivariate normality condition values is met as is shown in Figure 2.

Figure 2

Furthermore, we need to verify that the multicollinearity is not present, which indicates an absence of correlation between the predictors. The Pearson correlation test results for Model 1 are displayed in Table 4.

Given the maximum recommended value of 0.7, i.e., >0.8 for severe multicollinearity between the predictors [Landau and Everitt, 2004, p. 116], multicollinearity issues are not detected. The highest value is 0.088 and refers to the correlation coefficient between the size of the audit company and the loss in the current year.

Aside from Pearson's correlation, the multicollinearity assumption was tested using variance inflation factor (VIF) analysis (see Table 6). According to the table, low values for VIF were found with a maximum value of 1.5 and a minimum value of 1.06 for the proposed model. VIF values of >10 are considered to indicate multicollinearity, whereas in feeble models, values exceeding 2.5 may raise concerns [Landau and Everitt, 2004, p. 116]. We conclude that the model contains no multicollinearity.

And finally, we checked for the fulfillment of the assumption of homoscedasticity. Figure 3 demonstrates that the residuals are equally distributed across the regression line [Statistics Solutions, 2019]:

Figure 3

Tables 5 and 6 provide data on the test results for our statistical sample.

According to the data in Table 5, we can clearly reject this null hypothesis (F (8,360) = 9.451, p < 0.001) and also reckon that at least one of the eight predictors is associated with the timeliness of the audit report. The R² result indicates that 17.4% of the audit delay variance can be accounted for by our model.

The results shown in Table 6 assert that the regression coefficients are different from zero, thus indicating that the model is significant.

According to the presented data, it takes an additional 22.07 days to publish the audit opinion in case of a modified opinion (i.e., qualified opinion, negative opinion and disclaimer). Furthermore, if the assurance procedures are executed by one of the four major international audit firms, it takes 1.9 days less to publish the opinion. If a company is showing a loss in the current year, 9.2 additional days are required to publish the audited annual account. At the same time, if the audited company is involved with production, additional 8.6 days are required for publishing the audit opinion.

Furthermore, the results indicate that it takes 3.4 days less to publish the audit report when the size of the audited company increases by one unit provided that all other variables in the model remain unchanged. At the same time, it takes 0.7 days less to publish the audit report when the liquidity of the audited company increases by one unit, 0.1 more days if indebtedness increase by one additional unit, and 10 days less by increasing the profitability by one additional unit (again, provided that all other variables in the model remain unchanged).

The outcome of the regression for Model 1 indicates a positive and statistically significant association between the qualification and timeliness of the audit opinion (Sig <5%). Companies where auditors publish modified opinions are financially riskier, and it is more likely that it will take longer to shape the assurance opinion. This result confirms H7 that there is a positive correlation between the qualified audit opinion and the time required to disclose the audit report and is consistent with the findings of Bonsón-Ponte et al. [2008], Carslaw and Kaplan [1991], and Che-Ahmad and Abidin [2009]. Also positive and statistically significant is the relationship between the timeliness and companies in the manufacturing industry (Sig <5%). This finding confirms H8 and is in line with the results of Türel [2010], who finds that reporting lag is considerably longer in manufacturing entities. Correspondingly, Bamber, Bamber and Schoderbek [1993] and Carslaw and Kaplan [1991] found that the audit duration is shorter for entities in the financial sector due to minimal inventory and long-term assets. Likewise, given that Sig <5%, we can accept H1. This finding is consistent with Al-Ghanem and Hegazy [2011] and confirms that a longer period of time is required for auditing companies with lower current liquidity ratios than those with higher liquidity.

The size of the audited company is negatively and significantly correlated with timeliness expressed in number of days (Sig <5%), which confirms H5 that companies with larger assets complete the audit earlier than smaller companies. This finding is in conformity with multiple existing analyses [Davies and Whittred, 1980; Garssombke, 1981; Newton and Ashton, 1989; Carslaw and Kaplan, 1991; Abdulla, 1996; Afify, 2009] and can be interpreted by the need of large listed companies to disclose credible and timely information in order to earn the trust of investors and other stakeholders.

This finding detects a lack of a statistically significant correlation between profitability (ROE) and audit timeliness, although the relationship between these two variables is negative. Namely, Table 3 shows that the nonfinancial companies listed on the MSE are not exceedingly lucrative so this variable does not contradistinguish timely and untimely reports and might explain the lack of association.

The results presented in Table 6 indicate a direct correlation between current period loss and the audit delay. However, this ratio is not statistically significant (Sig >5%). The above data indicate that solvency difficulties do not lead to timeliness reduction but to an additional paragraph that emphasizes the issue. That is, the relationship between the two variables is positive (higher indebtedness leads to prolongation of the audit delay), but H2 is rejected due to the absence of a statistically significant correlation.

At the same time, in consistency with Carslaw and Kaplan [1991] and Davies and Whittred [1980], we cannot explain audit timeliness with audit quality (i.e., Big4 vs. other audit firms). Contrary to our expectations, the overall findings show that the auditor characteristics do not affect timeliness, while the characteristics of the audited company do. Although an inverse correlation is established between Big4 and the audit lag, H6 is rejected due to the absence of a statistically significant correlation (Sig >5%).