Timeliness of corporate annual financial reporting in Greece

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/24080181 Timeliness of corporate annual financial reporting in Greece Article in European Accounting Review February 2006 DOI: 10.1080/09638180500252078 Source: RePEc CITATIONS 79 READS 1,356 2 authors: Stephen Owusu-Ansah Houston Baptist University 17 PUBLICATIONS 742 CITATIONS SEE PROFILE Stergios Leventis International Hellenic University 38 PUBLICATIONS 681 CITATIONS SEE PROFILE All content following this page was uploaded by Stephen Owusu-Ansah on 18 August 2014. The user has requested enhancement of the downloaded file.

European Accounting Review Vol. 15, No. 2, 273 287, 2006 RESEARCH NOTE Timeliness of Corporate Annual Financial Reporting in Greece STEPHEN OWUSU-ANSAH AND STERGIOS LEVENTIS The University of Texas Pan American, Edinburg, TX, USA and Lancaster University, UK ABSTRACT This paper reports on the results of an empirical investigation of the factors that affect timely annual financial reporting practices by 95 non-financial, group companies listed on the Athens Stock Exchange. A descriptive analysis indicates that 92% of the companies reported early (relative to the 161-day regulatory deadline), 3% reported on the 161st day and 5% reported late. A multivariate regression analysis suggests that large companies, service companies and companies audited by the former Big-5 audit firms have shorter final reporting lead-time. Our tests provide strong empirical evidence to suggest, however, that companies in the construction sector, companies whose audit reports were qualified and companies that had a greater proportion of their equity shares directly and indirectly held by insiders do not promptly release their audited financial statements. No empirical evidence was found in support of the monitoring cost theory. Policy implications of the results for the regulatory agency of the stock market are suggested. 1. Introduction This paper empirically investigates the effects of both company-specific and auditrelated factors on timely financial reporting practices of group companies listed on the Athens Stock Exchange (ASE). The results of our descriptive analysis indicate that 92% of the companies are prompt reporters. A multivariate regression analysis suggests that large companies, service companies and companies audited by the former Big-5 audit firms are prompt reporters. The results suggest, however, that companies in the construction sector, companies whose audit reports were qualified (auditors remarked in their reports) and companies that had a greater Correspondence Address: Stephen Owusu-Ansah, Department of Accounting & Business Law, College of Business Administration, The University of Texas Pan American, 1201 West University Drive, Edinburg, TX 78541-2999, USA. Fax: þ1 956 381 2407; Tel.: þ1 956 381 3385; E-mail: stephen@utpa.edu 0963-8180 Print=1468-4497 Online=06=020273 15 # 2006 European Accounting Association DOI: 10.1080/09638180500252078 Published by Routledge Journals, Taylor & Francis, on behalf of the EAA.

274 S. Owusu-Ansah and S. Leventis proportion of their equity shares directly and indirectly held by insiders (i.e. top management and directors) have longer final reporting lead-time. Further, the results do not provide any empirical evidence in support of the monitoring cost theory, which postulates that highly geared companies are timely reporters. Timely corporate financial reporting is an essential ingredient for a wellfunctioning capital market. Undue delay in releasing financial statements increases uncertainty associated with investment decisions (Ashton et al., 1987). Timeliness of accounting information has become an important issue now than ever before as a result of phenomenal changes in both modern technology and business practices worldwide. The importance of timely accounting information for operational reasons in general and, capital markets in particular, cannot be over-emphasized. Consequently, most professional and regulatory bodies of capital markets have taken measures to reduce delays in the release of financial statements (see, e.g. US Securities and Exchange Commission, 2002). Timely reporting in emerging markets is of particular importance since information in these markets is relatively limited and has a longer time lag (Errunza and Losq, 1985). Timely reporting will enhance decision-making and reduce information asymmetry in these markets. Hence, research into the determinants of timely reporting would be of much importance to regulators of emerging capital markets in formulating new policies to enhance the allocational efficiency of their markets. The predominant focus of prior studies has been on the developed markets in North America (see, e.g. Garsombke, 1981; Ashton et al., 1989; Bamber et al., 1993); in Europe (see, e.g. Frost and Pownall, 1994; Soltani, 2002); and in Oceania (see, e.g. Dyer and McHugh, 1975; Courtis, 1976; Carslaw and Kaplan, 1991). Some emerging markets have, however, been studied, including Bahrain by Abdulla (1996), China by Haw et al. (2000), Hong Kong by both Ng and Tai (1994) and Jaggi and Tsui (1999), Zimbabwe by Owusu-Ansah (2000), Bangladesh by Iman et al. (2001), Bangladesh, India and Pakistan by Ahmed (2003) and Greece by Leventis and Weetman (2004). While timely reporting practices of Greek companies might have been studied by Leventis and Weetman (2004), the present study complements their work by focusing on the pattern and determinants of timely reporting by group listed companies in Greece using 1999 data. Leventis and Weetman (2004), however, studied 1997 data of non-group companies. Moreover, the focus of their study is different from ours. Thus, they extended Owusu-Ansah (2000) by focusing on: (i) variables relating to discretionary disclosure theories of proprietary costs, information cost savings and news content, and (ii) the testing of these theories in a context different from that of the USA. Greece provides an appropriate context to investigate timeliness of corporate reporting. The Greek economy has witnessed tremendous growth since 1995 due to policies encouraging competition and market liberalization, as part of the government s commitment to enter the Euro zone, which materialized in 2001. The ASE has immensely contributed to the economic development of Greece through the provision of funds. The market s total capitalization has

Timeliness of Corporate Annual Financial Reporting in Greece 275 been increasing rapidly since 1995 due to new and seasoned issues of shares. The greatest increase occurred in 1999 when the total value of listed companies reached E184,600 million, an increase of 195% over that of 1998 (ASE, 2000). An influx of international investors (notably from France, Germany, the UK and USA) to the ASE in recent years has been phenomenal. 1 These developments call for further improvements in market transparency, and in corporate financial reporting and governance systems. Consequently, the government charged the market regulator, the Hellenic Capital Markets Commission (HCMC), to examine corporate governance issues, including provision of information and timeliness of financial reporting, and to make appropriate recommendations. The outcome of our investigation would provide a valuable input for the work of the commission. The rest of the paper proceeds as follows. We present the regulatory framework for timely reporting in the next section. We then discuss the factors that influence timely reporting in Section 3. In Section 4, we describe the research design, and report the empirical results in Section 5. We conclude the paper in Section 6. 2. The Regulatory Framework for Timely Reporting The reporting obligations of Greek listed companies relating to timeliness of annual financial statements are found in two regulatory sources issued by the Greek government: (i) the Tax Law and (ii) the Company Law (Law 2190/ 20). The Tax Law, which is very influential and generally concerned with accounting disclosure issues, is the Presidential Decree 186/92 (also known as Code of Books and Records). The Presidential Decree 186/92 requires companies to have a period of 12 calendar months as their legal fiscal (financial) year, which should end either on 30 June or 31 December (Article 26). Law 2190/20 prescribes publication requirements relating to annual financial statements in Greece. According to a sub-section of Law 2190/20, Article 43b, a company should publish its annual financial statements and submit a copy to the Ministry of Trade at least 20 days before the company s annual general meeting (AGM). An AGM is required by Article 25a to be held within six months (i.e. 181 days) after a company s financial year-end. Implicitly, Greek companies are expected to publish their annual financial statements by the 161st day after their financial year-end. According to paragraph 2 of Article 26, annual financial statements of a company should be published in both a political and financial daily newspaper that are nationally circulated. However, if a company is not located in the prefecture of Athens, then its annual financial statements can be published in a local newspaper (paragraph 2 of Article 26). The publication of the annual financial statements in a political or financial newspaper, whichever comes first, signifies the date on which the financial statements of a company are deemed to have become publicly available for the first time. 2 It is after this publication date that listed companies in Greece will respond to any request from the general public for a copy of their audited annual financial statements.

276 S. Owusu-Ansah and S. Leventis After approval of the annual financial statements by shareholders at the AGM, companies are required to submit a copy of the approved financial statements along with a copy of the proceedings of the meeting to the Ministry of Trade (Perakis, 1992). Following the submission of the approved financial statements to the Ministry of Trade, companies are required to publish a full set of balance sheet, profit and loss account, and a table of appropriation of profits in the government gazette. Apart from these requirements under Presidential Decree 186/92 and Law 2190/20, which apply to all companies in Greece, there are no specific annual timely financial reporting requirements solely for listed companies in Greece. 3. Factors Influencing Timely Reporting To better understand how Greek companies respond to the timely reporting requirements, it is necessary to relate their timely reporting practices to certain factors. While there may be many factors, company-specific and audit-related ones have been proposed and tested in prior studies as being particularly important. Company-specific factors are those that enable management of a company either to produce more timely financial statements or to reduce costs associated with undue delay in reporting (Owusu-Ansah, 2000). Such factors include company size, profitability, gearing, financial condition, industry type and ownership structure. 3 In contrast, audit-related factors are those that enable auditors to carry out (or inhibit them from carrying out) audit assignments, and issue assurance reports promptly (Owusu-Ansah, 2000). Examples of audit-related factors are auditor type, presence of extraordinary and/or contingent items, audit opinion, month of financial year-end and complexity of operations. In this study, we investigate the effects of a subset of both company-specific and audit-related factors that are relevant to the Greek socio-economic environment, and for which data were available. The company-specific factors are company size (SIZE), gearing (GEAR), insider equity shareholding (EQOS) and industry type (INDT). The audit-related factors tested are the number of remarks made by auditors in their reports (RMAK) 4 and auditor type (AUDT). The data on each factor were computed from the audited financial statements of the companies. 4. Research Design Sample and Sampling Design There were 294 companies listed on the ASE as at 31 December 1999 (ASE, 2001). We chose our sample on the basis of the following criteria. First, we eliminated 41 companies that were listed for the first time in 1999. Second, we excluded 45 financial companies because such companies are subject to additional regulations, which may impact on their reporting behaviour relative to the general population. Third, we de-selected 16 cross-nationally listed nonfinancial companies to control for any extraneous factors on their reporting

Timeliness of Corporate Annual Financial Reporting in Greece 277 behaviour. Fourth, 72 non-group companies (i.e. those that are not required to publish consolidated financial statements) were excluded to ensure homogeneity. Fifth, 21 companies were excluded because of lack of data on some variables of interest. Lastly, we excluded four group companies having financial year-end other than 31 December because, as suggested in the literature, the month of financial year-end influences timely reporting behaviour. The final sample consists of 95 companies, representing about 32% of all companies listed on the market. Panel A of Table 1 reports the sampling design. Operationalization of Timeliness As in prior studies, we define timeliness as the number of days between a company s financial year-end and the day on which the company publicly releases its audited financial statements. As in Owusu-Ansah (2000), we prefer to use lead-time instead of delay, which is generally used in the literature, to denote timeliness because, if for example, a company releases its financial statements by regulatory deadline, then, it cannot be said that the company has delayed in releasing its financial statements. Therefore, we describe the number of days that elapses between a company s financial year-end and the date it publishes its financial statements as its final reporting lead-time (FRLT). We computed the FRLT for each company by counting the number of days that elapsed between its financial year-end and the date it published its financial statements for the first time in either a national daily newspaper (political or financial whichever is earlier) for companies located in Athens or a local newspaper for companies located in provinces other than Athens. 5 The Model To investigate the influence of the selected company-specific and audit-related factors on timely reporting behaviour of the companies in our sample, we estimated the following cross-sectional regression model with an ordinary least squares (OLS) technique: FRLT ¼ g 1 þ g 2 SIZE þ g 3 GEAR þ g 4 EQOS þ g 5 RMAK þ g 6 AUDT þ g 7 8 INDT þ 1 (1) where: FRLT ¼ number of days between financial year-end and the date of release of a company s audited annual financial statements; SIZE ¼ natural log of year-end total assets; GEAR ¼ ratio of long-term debt to equity; 6 EQOS ¼ proportion of equity shares directly or indirectly controlled by insiders;

278 S. Owusu-Ansah and S. Leventis RMAK ¼ number of remarks made in audit report; AUDT ¼ type of auditor indicator variable coded 1 if auditor is one of the former Big-5 audit firms, 0 otherwise; and INDT ¼ industry-type indicator variable. 7 Table 1. Summary of sample criteria and pattern of corporate reporting date Panel A: Summary of sample selection criteria Description No. of listed companies a Percentage of total population Companies with equity shares 294 100 listed on the ASE as of 31 December 1999 Deduct: Companies that listed for the first time in 1999 41 13.95 Companies in the banking, insurance, 45 15.31 investment and leasing businesses Non-financial companies with cross-national listing 16 5.44 Non-parent companies 72 24.49 Group companies lacking some data of interest 21 7.14 Group companies with financial year-end 4 1.36 other than December [ Group companies with usable data (the sample size) 95 32.31 Panel B: Pattern of publication dates of audited financial statements Lead-time of publication (FRLT) Frequency Percentage Cumulative percentage 0 30 0 0.00 0.00 31 40 1 1.05 1.05 41 50 0 0.00 1.05 51 60 27 28.42 29.47 61 70 3 3.16 32.63 71 80 3 3.16 35.79 81 90 5 5.26 41.05 91 110 0 0.00 41.05 111 120 3 3.16 44.21 121 130 3 3.16 47.37 131 140 8 8.42 55.79 141 150 17 17.90 73.69 151 160 17 17.90 91.59 161 (regulatory deadline) 3 3.16 94.75 162 170 3 3.16 97.91 171 180 1 1.05 98.96 181 190 1 1.05 100.00 Total 95 100.00 a Figures do not include companies whose shares were suspended from trading on the stock exchange.

Timeliness of Corporate Annual Financial Reporting in Greece 279 5. Results and Discussion Descriptive Analysis Panel B of Table 1 presents the pattern of publication dates of audited annual financial statements of the companies. It is evident that about 92% of the companies reported earlier than the expected 161st day after a company s financial year-end. Three per cent released their audited financial statements exactly on the 161st day. About 5% of the ASE-listed companies reported late, with the maximum delay being 23 days. Leventis and Weetman (2004), however, found no evidence of late reporting among non-group ASE-listed companies. Of particular interest of the pattern of publication dates of the corporate audited annual financial statements is the concentration of many of the companies reporting within 51 60 days, and 141 160 days after their financial year-end. 8 A further analysis reveals the following: (1) Of the 27 companies that released their financial statements within 51 60 days (peaked head) after their financial year-end, 16 (59%) are in the manufacturing sector; eight (30%) are in the service sector; and three (11%) are in the construction sector; and (2) Of the 34 companies that publicly released their audited financial statements within 141 160 days (peaked tail) after their financial year-end, 10 (30%) are in the manufacturing sector; 11 (32%) are in the service sector; and 13 (38%) are in the construction sector. The 27 companies might have released their financial statements within 51 60 days after their financial year-end because Greek law allows companies that report within the first two months (i.e. 60 days) after financial year-end the dispensation of preparing a fourth (final) interim (quarterly) report. For the 34 companies that reported within 141 160 days after their financial year-end, the incentive might have been to avoid any possible sanctions that might be applied to them under Law 2190/20 if they report beyond the regulatory deadline. Table 2 reports summary statistics for the data on variables investigated in this paper. As is evident, it takes ASE-listed group companies 113 days, on average, to report to the public after the end of their financial year. The standard deviation for the FRLT variable is 44 days, suggesting considerable variability in timely reporting by the companies. The analysis also shows that while the shortest FRLT was as early as 34 days, the longest was as late as 183 days. A Shapiro Wilk s test suggests that the distribution of each continuous variable is not bell-shaped, except for SIZE and EQOS. As a result, we transformed the raw data for the FRLT, GEAR and RMAK variables. We selected a specific transformation procedure on the basis of the nature of the pre-transformation distribution of the variable in question, and the procedure s ability to correct the non-normality problem. Thus, we transformed both FRLT and GEAR by

Variable (expected sign) Table 2. Descriptive statistics, distributional characteristics, testing for normality and possible remedies (n ¼ 95) Possible remedies Summary statistics Shapiro Wilk s Min. Max. Mean Std. dev. normality test Z-Statistic Prob.. Z Description of distribution of raw data as per graph (not presented here) Transformation Significance after remedy FRLT (?) 34.00 183.00 113.26 44.00 5.647 0.0000 Heavily peaked at both Normal score a 0.9999 head and tail SIZE (2) 19.72 28.17 23.52 1.17 2.988 0.0014 Fairly normal distributed None needed GEAR (?) 0.00 1.43 0.16 0.24 6.959 0.0000 Positively skewed Normal score a 0.0199 EQOS (þ) 0.05 0.93 b 0.51 0.18 0.130 0.4483 Normally distributed None needed RMAK (þ) 0.00 10.00 3.14 2.29 3.128 0.0009 Positively skewed Square root 0.0140 AUDT (2) 0.00 1.00 0.18 0.39 INDT: SERV (2) 0.00 1.00 0.37 0.40 CONS (?) 0.00 1.00 0.21 0.49 MANU (?) 0.00 1.00 0.40 0.49 280 S. Owusu-Ansah and S. Leventis a The data were transformed by computing normal scores using Van der Waerden s transformation defined by the formula r/(w þ 1), where w is the sum of the case weight and r is the rank ranging from 1 to w. b This does not suggest an error in our data. This is because though Athens Stock Exchange (ASE)-listed companies are required to have at least 20% of their outstanding equity shares in the hands of public investors, there are two instances where deviations are allowed. First, large companies can have only 5% of theirs shares in the hands of the public if the shares have been allocated to at least 2,000 investors and none of whom controls more than 2%. Second, a deviation of the dispersion criterion may be allowed if the board of the ASE can ascertain that sufficient dispersion will take place in a short period through brokerage transactions (ASE, 2003, p. 64).

Timeliness of Corporate Annual Financial Reporting in Greece 281 computing their normal scores using Van der Waerden s transformation defined by the formula r/(w þ 1), where w is the sum of the case weight and r is the rank ranging from 1 and w. Also, we square-root transformed the RMAK variable. The result of each transformation procedure is reported in Table 2. Table 3 reports the Pearson product moment correlation coefficient of the variables. None of the pairwise correlation coefficients exceeds the conventional threshold of 0.800 (see Gujarati, 1995, p. 335). Hence, multicollinearity is not a serious problem. Multivariate Analysis As Table 4 shows, the F-statistic of Model A is significantly different from zero, indicating that a subset of the independent variables does explain the variation in FRLT about its mean. The value of the adjusted R-squared indicates that only about 32% of the variation in FRLT is explained by the model. The coefficient estimates for EQOS, RMAK and AUDT are all statistically significant. The positive effect of EQOS, measured as the proportion of a company s outstanding equity shares held by its insiders (i.e. top management and directors), maybe due to the prevalence of share ownership concentration in Greece (Tsangarakis, 1996). Closely held companies are less likely to promptly report to the market because the insiders, who tend to be the majority shareholders of such companies, rarely sell their shares and as such would be less concerned about share prices than minority shareholders who are more likely to want to sell. Therefore, majority shareholders would show less interest in keeping the market informed. Another plausible reason could be that majority shareholders wield more power relative to that of minority shareholders, or maybe there is less protection for minority shareholders in Greece than in the UK and USA that majority shareholders Table 3. Pearson product moment correlation matrix (n ¼ 95) Variable FRLT SIZE GEAR EQOS RMAK AUDT FRLT 1.000 SIZE 20.092 1.000 GEAR 0.131 0.314 1.000 EQOS 0.254 20.065 0.030 1.000 RMAK 0.437 20.096 0.101 20.014 1.000 AUDT 20.332 0.250 20.021 0.016 20.176 1.000 INDT: a SERV 0.061 0.110 0.261 0.286 20.257 0.042 CONS 0.226 20.098 20.125 20.132 0.225 20.233 MANU 20.211 0.013 20.131 20.144 0.079 0.179 p, 0.05 (two-tail test); p, 0.001 (two-tail test). a The pair-wise correlation coefficients between the dummy variables (i.e. categories of the industry variable) are not reported because no meaningful inference can be drawn from them.

282 S. Owusu-Ansah and S. Leventis Table 4. OLS estimates of timely reporting regression model Model: FRLT ¼ g 1 þ g 2 SIZE þ g 3 GEAR þ g 4 EQOS þ g 5 RMAK þ g 6 AUDT þg 7-8 INDT þ 1 Model A Model B Coef. t-value VIF a Coef. t-value VIF a INTERCEPT(?) 21.504 20.81 23.361 21.77 SIZE (2) 20.002 0.00 1.22 0.076 0.97 1.25 GEAR (?) 0.064 0.65 1.25 0.024 0.27 1.25 EQOS (þ) 1.264 2.68 1.11 1.001 2.40 1.14 RMAK (þ) 0.490 4.26 1.17 0.571 5.28 1.17 AUDT (2) 20.566 22.50 1.16 20.598 22.87 1.15 INDT: b SERV (2) 20.307 21.54 1.42 20.306 21.74 1.44 CONS (?) 0.434 1.92 1.26 0.465 2.41 1.25 Model summary statistics: Number of observations 95 88 F-Statistic 7.24 9.07 Prob.. F 0.000 0.000 R-squared 0.368 0.442 Adjusted R-squared 0.317 0.394 a The rule of thumb is that there is evidence of collinearity problems if the variance inflation factor (VIF) of a variable exceeds 10 (Gujarati, 1995, p. 339). b As noted earlier, the manufacturing (MANU) category of the industry (INDT) variable is not included in the model because it serves as the base category, to which the service (SERV) and construction (CONS) categories are compared. are able to ignore the interests of the minority with impunity. 9 While the positive effect of EQOS is consistent with the results in Ashton et al. (1987), it does not corroborate those in Carslaw and Kaplan (1991), and Jaggi and Tsui (1999). The positive and significant coefficient on the RMAK variable indicates that it takes companies that received qualified audit reports longer to release their financial statements (p, 0.001). Plausible reasons for this result are the increase in time taken by auditors to complete the year-end audit and that spent in auditor client negotiations. 10 Leventis and Weetman (2004) reported similar results with non-group ASE-listed companies. The coefficient on AUDT is negative and statistically significant (p, 0.05), suggesting that companies audited by the former Big-5 audit firms have shorter FRLT than companies audited by local audit firms. The former Big-5 audit firms do have an incentive to differentiate their services from those of the local audit firms. Since the liberalization of the Greek audit service market in 1992, the former Big-5 audit firms operating in the country have been more aggressive in their marketing strategies to gain a foothold in the liberalized market, and to increase their market share (Citron and Manalis, 2001). In addition, Citron and Manalis (2001) found Greek companies sourcing international finance to have

Timeliness of Corporate Annual Financial Reporting in Greece 283 preference for the services of the former Big-5 audit firms as a means to signal to foreign investors of their high-quality financial statements. While the coefficient estimates of EQOS, RMAK and AUDT are statistically significant, those of SIZE, GEAR, SERV and CONS are not. The SIZE coefficient is negative but statistically not significant (p ¼ 0.499). The negative effect of SIZE on timely reporting in Greece is consistent with the prediction in the literature, implying that large Greek companies are prompt reporters compared to their smaller counterparts. This result corroborates the conclusions in prior studies (e.g. Dyer and McHugh, 1975; Ng and Tai, 1994; Abdulla, 1996; Owusu-Ansah, 2000). Though the positive effect of GEAR is not statistically significant (p ¼ 0.520), it is consistent with the results reported by Owusu-Ansah (2000). This result suggests that audit of debt capital is more time consuming than that of equity capital, and as a consequence, highly geared companies are more likely to report late (Carslaw and Kaplan, 1991). Stated differently, our data do not support the monitoring cost theory, which speculates that debtholders usually require highly geared companies to report timely and at a certain frequency so as to monitor their interest. None of the coefficients on the industry categories is statistically significant (for SERV, p ¼ 0.063; and for CONS, p ¼ 0.058). However, as expected, the service sector (SERV) has a negative relationship with timely reporting. This result may be explained by the fact that service companies, like financial companies, generally do not hold much inventory and as such, are not engaged in the annual time-consuming task of counting and valuing inventory. Perhaps, the non-involvement in the annual counting of inventory enables them to timely report to the public than their counterparts in the manufacturing sector. Courtis (1976), on the contrary, found service companies in New Zealand to be slow reporters. Though we were unable to speculate how companies in the construction sector behave regarding timely reporting, the effect of the CONS variable is positive, suggesting that construction companies are late reporters compared to their counterparts in the manufacturing sector (MANU). Regression Diagnostics To ascertain whether our results are robust, we performed several post-estimation diagnostic tests on the model. First, we tested for the effect of heteroscedasticity on our results with the procedure suggested by Cook and Weisberg (1983). The results of this procedure suggest we should not reject the null hypothesis of homoscedasticity (x 2 -statistic ¼ 1.26, p ¼ 0.2618 [two-sided]), implying that heteroscedasticity has not contaminated our results. Second, because the variables in our model exhibit linear correlation; though not serious (see Table 3); we computed variance inflation factor (VIF) for each independent variable. The VIF is more powerful in detecting collinearity problems than the pairwise correlation matrix. The VIFs of the variables, which are reported in Table 4, suggest that collinearity is not a problem. Third, we

284 S. Owusu-Ansah and S. Leventis graphed the studentized residuals of the model. The Probability-Plot (P-P) procedure (not shown here) indicates that the distribution of the studentized residuals is fairly normal. Fourth, we tested for the effect of outlying observations on our results with Cook s distance procedure (Cook, 1977). This procedure identifies seven observations to be exerting disproportionate influence on the results. So, we eliminated the outlying observations and re-estimated the model. The estimates of this regression (Model B) are also reported in Table 4. While the results are qualitatively similar to those of Model A, there are four exceptions worth noting: (i) the SIZE effect is now positive yet statistically not significant, (ii) the level of significance of the coefficient on EQOS fell from 0.004 to 0.009, (iii) the adjusted R- squared increased from about 0.32 (for Model A) to about 0.39 (for Model B), and (iv) the level of significance of the coefficient on CONS increased from 0.058 to 0.018. 6. Conclusions This paper empirically investigates the influence of both company-specific and audit-related factors on timely annual financial reporting practices by 95 group companies listed on the Athens Stock Exchange (ASE) as of 1999. A descriptive analysis indicates that 92% of the companies took less than the maximum time allowed after their financial year-end to release their audited financial statements. Five per cent of the companies exceeded the regulatory deadline (161st day after a company s financial year-end). The remaining 3% reported on the 161st day. A multivariate regression analysis shows that large companies, service companies and companies audited by the former Big-5 audit firms are prompt reporters. Our analysis shows, however, that construction companies, companies whose audit reports were qualified (audit reports contained auditors remarks) and companies that had a greater proportion of their equity capital directly or indirectly held by insiders (i.e. top management and directors) do not promptly release their audited financial statements to the public after their financial year-end. Also, we found no empirical evidence in support of the monitoring cost theory, which postulates that highly geared companies promptly report to the market. While these conclusions are consistent with prior studies, they should be considered in the light of these limitations. First, the results may not be generalizable to financial, non-group listed companies and companies that are cross-nationally listed because such companies were excluded from our sample. Future research may study timely reporting behaviour of these companies, particularly financial companies. Second, our conclusions are based on cross-sectional data. A longitudinal study may provide a complete picture of how the factors investigated in this study affect timely reporting over time. Future research may explore this possibility. Third, because the ASE crashed in 2000 our results may not be generalizable to periods subsequent to 1999. A future study may provide some insight into whether the institutional changes introduced after the market had

Timeliness of Corporate Annual Financial Reporting in Greece 285 crashed did alter the timely reporting behaviour of group listed companies in Greece in the years subsequent to 1999. Fourth, our study did not investigate the reaction of the market to information released. Research shows that timeliness is associated with market reaction to information released (e.g. Givoly and Palmon, 1982). Future studies may examine the relationship between timeliness and information content of Greek financial statements and the market reactions they evoke. Notwithstanding the above limitations, the results have two policy implications for the Greek market regulator, the Hellenic Capital Markets Commission (HCMC). First, the HCMC may have to assess the adequacy of the 161-day reporting requirement, given the influx of international investors to the ASE. International investors demand more timely information, and to meet their information needs, the 161-day reporting requirement may have to be shortened. Second, it may have to come out with a timely reporting requirement specifically meant for ASE-listed companies in addition to the existing requirements imposed by the Company Law (Law 2190/20) and Presidential Decree 186/92, which apply to all companies in the country. Acknowledgements We appreciate the assistance provided by officers of the Athens Stock Exchange in data gathering. We thank Elizabeth Dietz, Gouranga Ganguli and Tejinder Sethi for their comments on an earlier draft of this paper. We also thank George Alifantis of Ernst & Young in Athens for responding to our inquires. Further, we thank the anonymous reviewers of the paper, the editor, Professor Kari Lukka, and the associate editor, Professor Markus Granlund, for their useful comments and encouragement during the review process. Notes 1 International investors, who are known to have demand for quality financial information comparable to the levels in their home countries, own about 30% of the listed shares on the ASE. 2 There is no requirement for preliminary announcement or press release of earnings ahead of publication dates in Greece. 3 As noted by Owusu-Ansah (2000), the dividing line between what constitutes a companyspecific factor and audit-related factor is not clear-cut as portrayed here. Certain companyspecific factors, such as company size and gearing can also be classified as audit-related factors (see Simnett et al., 1995, p. 5). 4 It is a customary practice in Greece for auditors to make remarks in their audit reports in accordance with the Hellenic Auditing Standards and Professional Ethics (The Institute of Certified Auditors of Greece, 1999, p. 180). These remarks would be considered in the UK and USA as audit qualifications. 5 It is difficult for an individual to keep track of publication of dates of financial statements in Greece because there are numerous newspapers in circulation in the country. In view of this, the ASE keeps a record of the dates of first appearance of financial statements of listed companies in Greek newspapers. The ASE gave us access to this database, which we are grateful.

286 S. Owusu-Ansah and S. Leventis 6 An anonymous reviewer drew our attention to the fact that this conventional measure of GEAR may not be appropriate in Greece, as some Greek companies use short-term debts to finance long-term assets. While this might be the case, we did not change our measurement, as the opinion of this reviewer is only anecdotal. Future research should look further into this issue. 7 Although the listed companies on the ASE are classified into 12 major sectors, for statistical purposes the 95 companies in our sample were broadly grouped into three categories: (i) service ([SERV] consisting of 35 companies), (ii) construction ([CONS] consisting of 20 companies) and (iii) manufacturing ([MANU] consisting of 40 companies). However, we included k 2 1 dummy variables in equation (1), where k is the number of categories. Thus, to avoid the dummy variable trap problem the MANU category is omitted to serve as a base (Gujarati, 1995, p. 504). The coefficients for the remaining categories (SERV and CONS) measure the extent to which they differ from this base. 8 A histogram of the raw data on the FRLT variable (i.e. the untransformed data on date of publication of annual reports), not shown here, shows that it is heavily peaked at both 51 60 days (head) and 141 160 days (tail). 9 This explanation is due to an anonymous reviewer. The interests of minority shareholders are now protected in Greece by the enactment of the Corporate Governance Act of 2002 but not during the year of the study, 1999. 10 Interviews conducted by the second author of this paper with auditors in Greece suggest that auditor client accounting method choice negotiations occur regularly and delay completion of audit engagements. Such negotiations cover issues, such as the number of remarks, the content and the style of the qualification. Many of the auditors also expressed the view that a trade-off between their legal liability and their future re-appointment by client companies takes place when the issue of remarks is discussed. References Abdulla, J. Y. A. (1996) The timeliness of Bahraini annual reports, in: T. S. Doupnik and S. B. Salter (Eds) Advances in International Accounting, Vol. 9, pp. 73 88 (Greenwich, CT: JAI Press). Ahmed, K. (2003) The timeliness of corporate reporting: a comparative study of South Asia, in: T. S. Doupnik and S. B. Salter (Eds) Advances in International Accounting, Vol. 16, pp. 17 43 (Amsterdam: Elsevier). Ashton, R. H. et al. (1987) An empirical analysis of audit delay, Journal of Accounting Research, 25(2), Autumn, pp. 275 292. Ashton, R. H. et al. (1989) Audit delay and the timeliness of corporate reporting, Contemporary Accounting Research, 5(2), pp. 657 673. Athens Stock Exchange (2000) Annual Report (Athens: Athens Stock Exchange S.A.). Athens Stock Exchange (2001) Fact Book (Athens: Athens Stock Exchange S.A.). Athens Stock Exchange (2003) Fact Book (Athens: Athens Stock Exchange S.A.). Bamber, E. M. et al. (1993) Audit structure and other determinants of audit report lag: an empirical analysis, Auditing: A Journal of Practice and Theory, 12(1), pp. 1 23. Carslaw, C. A. P. N. and Kaplan, S. E. (1991) An examination of audit delay: further evidence from New Zealand, Accounting and Business Research, 22(85), pp. 21 32. Citron, D. B. and Manalis, G. (2001) The international firms as new entrants to the statutory audit market: an empirical analysis of auditor selection in Greece, 1993 to 1997, European Accounting Review, 10(3), pp. 439 459. Cook, R. D. (1977) Detection of influential observations in linear regression, Technometrics, 19(1), February, pp. 15 18. Cook, R. D. and Weisberg, S. (1983) Diagnostics for heteroscedasticity in regression, Biometrika, 70(1), April, pp. 1 10. Courtis, J. K. (1976) Relations between timeliness of corporate reporting and corporate attributes, Accounting and Business Research, 6(25), pp. 45 56.

Timeliness of Corporate Annual Financial Reporting in Greece 287 Dyer, J. C. and McHugh, A. J. (1975) The timeliness of the Australian annual report, Journal of Accounting Research, 13(3), pp. 204 219. Errunza, V. R. and Losq, E. (1985) The behaviour of stock prices on LDC markets, Journal of Banking and Finance, 9(4), pp. 561 575. Frost, C. A. and Pownall, G. (1994) Accounting disclosure practices in the United States and the United Kingdom, Journal of Accounting Research, 32(1), pp. 75 102. Garsombke, H. P. (1981) Timeliness of corporate financial disclosure, in: J. K. Courtis (Ed.) Communication via Annual Reports, AFM Exploratory Series No. 11, pp. 204 218 (Armidale, NSW: University of New England). Givoly, D. and Palmon, D. (1982) Timeliness of annual earnings announcements: some empirical evidence, Accounting Review, 57(3), pp. 485 508. Gujarati, D. N. (1995) Basic Econometrics, 3rd edn (New York: McGraw-Hill). Haw, I. M. et al. (2000) Timeliness of annual report releases and market reaction to earnings announcements in an emerging capital market: the case of China, Journal of International Financial Management and Accounting, 11(2), pp. 108 131. Iman et al. (2001) Association of audit delay and audit firms international links: evidence from Bangladesh, Managerial Auditing Journal, 16(3), pp. 129 133. Jaggi, B. and Tsui, J. (1999) Determinants of audit report lag: further evidence from Hong Kong, Accounting and Business Research, 30(1), Winter, pp. 17 28. Leventis, S. and Weetman, P. (2004) Timeliness of financial reporting: applicability of disclosure theories in an emerging capital market, Accounting and Business Research, 34(1), pp. 43 56. Ng, P. P. H. and Tai, B. Y. K. (1994) An empirical examination of the determinations of audit delay in Hong Kong, British Accounting Research, 26(1), March, pp. 43 59. Owusu-Ansah, S. (2000) Timeliness of corporate financial reporting in emerging capital markets: empirical evidence from the Zimbabwe Stock Exchange, Accounting and Business Research, 30(3), Summer, pp. 241 254. Perakis, E. (1992) The Law of Societe Anonyme (2190/20) (Athens: Nomiki Vivliothini Athina) (in Greek). Securities and Exchange Commission (US) (2002) SEC to propose new corporate disclosure rules. Available at: http://www.sec.gov/news/headlines/corpdiscrules.htm (accessed: November 2003). Simnett, R. et al. (1995) The determinant of audit delay, in: P. M. J. Reckers (Ed.) Advances in Accounting, Vol. 13, pp. 1 20 (Greenwich, CT: JAI Press). Soltani, B. (2002) Timeliness of corporate and audit reports: some empirical evidence in the French context, International Journal of Accounting, 37, pp. 215 246. The Institute of Certified Auditors of Greece (SOE) (1999) Hellenic Auditing Standards and Professional Ethics (Athens: SOE). Tsangarakis, N. V. (1996) Shareholder wealth effects of equity issues in emerging markets: evidence from rights offering in Greece, Financial Management, 25(3), Autumn, pp. 21 32.

View publication stats