The Value Relevance of Accounting Information under ... - SSRN papers

International Research Journal of Finance and Economics ISSN 1450-2887 Issue 76 (2011) © EuroJournals Publishing, Inc. 2011 http://www.internationalresearchjournaloffinanceandeconomics.com

The Value Relevance of Accounting Information under Greek and International Financial Reporting Standards: The Influence of Firm – Specific Characteristics Papadatos P. Konstantinos Department of Business Administration University of Patras, University Campus, Patras, Greece E-mail: [email protected] Tel: +30-2610-996269; Fax: +30-2610-996269 Bellas P. Athanasios Professor of Accounting, Department of Business Administration University of Patras, University Campus, Patras, Greece E-mail: [email protected] Tel: +30-2610-996269; Fax: +30-2610-996269 Abstract The aim of the present study is to compare the value relevance of accounting information under International Financial Reporting Accounting standards (IFRS) and Greek Accounting Standards (GAS) and to investigate whether the results are influenced from firm specific characteristics. Firstly, it is examined how the mandatory application of IFRS affected the relative and incremental value relevance of book value and net income in Greece. Secondly, it is investigated whether the size of the companies and their level of fixed assets affect the value relevance of accounting information. The results show that both firm size and fixed assets become significant factors, implying that the consequences of the mandatory transition to IFRS may not be the same for all firms.

Keywords: International Financial Reporting Standards (IFRS), Greek Accounting Standards (GAS), Greek firms, Mandatory adoption, Value relevance, Relative value relevance, Incremental value relevance JEL Classifications Codes: G14, G15, M41

1. Introduction One of the reasons that led EU to the mandatory application of IFRS was to improve the quality of financial statements. Consequently, many empirical studies investigate whether the adoption of new accounting standards affected the quality of provided accounting information. One of the most common methods of examination of accounting information quality, even indirectly, is value relevance. In this context, the present study examines the value relevance of accounting information in Greece under IFRS and GAS. However in order to interpret the consequences of the mandatory adoption of IFRS, the study was extended by exploring whether value relevance is affected by firm

Electronic copy available at: http://ssrn.com/abstract=1963369

International Research Journal of Finance and Economics - Issue 76 (2011)

7

specific characteristics, hence, whether the consequences from IFRS adoption are the same for all firms. The motivation for our study was the different philosophy and orientation of IFRS compared with that of GAS, since the domestic accounting standards in Greece are stakeholder-oriented while IFRS are shareholder-oriented. Specifically, given that GAS are based on the French-German accounting model which emphasizes the protection of creditors and tax transparency, apply exclusively the method of historical cost. On the other hand, since IFRS are based on Anglo-Saxon accounting principles target on the protection of investors and introduce the method of fair value accounting. Therefore, the mandatory adoption of IFRS in Greece is expected to have a significant impact on financial statements. Many studies have compared the value relevance of IFRS with various domestic accounting standards. However, most of them examined companies which applied IFRS voluntarily (e.g. Harris and Muller, 1999; Jennings et al., 2004; Barth et al., 2006; Gassen and Sellhorn, 2006; Hung and Subramanyam, 2007; Barth et al., 2008). Nevertheless, the international bibliography (e.g. Tarca, 2004; Ashbaugh, 2001; El–Gazzar et al., 1999; Gassen and Sellhorn, 2006; Dumontier and Raffournier, 1998; Murphy, 1999) has clearly shown that firms which implemented voluntarily the new accounting standards, have specific characteristics and different motives from others. Consequently, the effects of the voluntary adoption of IFRS cannot be generalized for firms that were obliged to adopt the new standards (Soderstrom and Sun, 2007). In this context, the present research compares the value relevance IFRS with that of GAS, by examining firms which applied IFRS mandatory. In addition, the study investigates whether the value relevance of accounting information is affected by firm specific characteristics; namely, the size and the fixed assets. At a first level, Bellas et al. (2007), via small sample of companies, investigated how IFRS affected the financial statements in Greece. In accordance with theory, they found that the prudence principle is apparent in GAS, which is not the case in IFRS as they are balance sheet oriented most probably due to the introduction of fair values. However, in order to receive more information and lead to safer conclusions, the research of Bellas et al. (2007) was expanded. Given that empirical researches with small samples may lead to statistically insignificant results, the present study investigates the value relevance (relative and incremental) of accounting information under GAS and IFRS, by using as a sample the majority of listed companies in the Athens Stock Exchange (ASE) at the first year of application of IFRS (2005). Furthermore, in order to answer the question whether the consequences of IFRS are the same for all firms in Greek market, it is examined whether the size of the companies and the levels of fixed assets differentiate the results. In order to track the effects from the accounting transition from GAS to IFRS, the accounting information under GAS were compared directly with that under IFRS, for the same sample of companies and for the same year. According to IFRS, all firms that adopt the accounting standards of IASB for the first time should republish their financial statements of the previous year under IFRS, for comparability reasons. Therefore, the vast majority of Greek listed companies that applied IFRS for the first time in 2005 were obliged to restate the financial statements of 2004 under IASB’s standards. Consequently, the year 2004 was the unique financial year that financial statements were provided under both standards (GAS and IFRS). Thus, for the year 2004, three different empirical investigations were conducted. The first examined the relative and incremental value relevance of accounting information under GAS and IFRS. The second investigated whether the results are affected from the size of companies and the last one, whether the results are affected from the level of fixed assets. The remainder of this study is organized as follows: Part 2 provides a detailed description of the bibliography with regard to the value relevance of IAS. The data and the sample companies that were used are presented in Part 3. Part 4 analyzes the methodology that was applied. The results of the study are shown in Part 5. Finally, the conclusions are recorded in Part 6.

Electronic copy available at: http://ssrn.com/abstract=1963369

8


2. Literature Review As already mentioned, one of the most common methods of investigating the quality of accounting information is value relevance. This was firstly applied by Ball and Brown (1968). By examining the correlation of earnings with share returns, they concluded that high correlations are interpreted as a sign of accounting information of high quality. The analysis of Ball and Brown (1968) generated many studies that compared value relevance of accounting information with different accounting standards. Firstly, US GAAP was used as a benchmark and was compared with many domestic accounting standards (such as Amir et al., 1993; Chan and Seow, 1996). However, the last decade, due to the gradual global recognition of IFRS, many empirical investigations compare IFRS with domestic accounting standards. Many studies such us Eccher and Healy (2003), Lin and Chen (2005), Wu et al. (2005), Sami and Zhou (2004), Bao and Chow (1999) and Dan (2002) compared IFRS with Chinese Accounting Standards (CAS). Given that these studies did not apply exactly the same methodology their results are not directly comparable. However, it could be mentioned that Eccher and Healy (2003), Lin and Chen (2005), Wu et al. (2005) and Dan (2002) are not in favor IFRS. On the contrary, Sami and Zhou (2004) and Bao and Chow (1999) suggest that accounting information under IFRS record greater value relevance compared with that under CAS. However, the international bibliography did not only focus on the Chinese market. Many studies compared the value relevance of IFRS accounting information with that of US GAAP. One of them was the research of Harris and Muller (1999) in which they used a sample of 31 non-US companies, which adopted IFRS voluntarily, for the period 1992-1996. Although they used several models, their study gave ambiguous results. However, since they used a small sample of firms and the quality of IFRS was improved significantly after 1996, i.e. after the period of examination (19921996), it can be argued that their conclusions are not safe enough. On the contrary, the results of Gordon et al. (2007) were straightforward. By using a similar methodology in a sample of 83 non-US companies which apply IFRS mandatory for the period 2004-2005, they were led to the conclusion that US GAAP are qualitatively superior from IFRS. The difference in the value relevance of accounting information between US GAAP and IFRS was also confirmed by Barth et al. (2006). By comparing a sample of 428 companies that adopt IFRS voluntarily for the period 1990-2004 with an equal sample of US companies, they found that although the application of IFRS improves value relevance of accounting information, it is still less than the value relevance of US GAAP. The gradual global recognition of IFRS made many researchers to compare IFRS with many other domestic accounting standards, like German GAAP. Bartov et al. (2005) examined 416 German companies and investigated the value relevance of earnings for the period 1998-2000. Their results show that earnings under IFRS are more value relevant than earnings under German GAAP. According to the authors this results is explained by the fact that since IFRS are shareholder oriented provide accounting information with greater quality, compared with German GAAP which are stakeholder oriented. The comparison of earnings quality between IFRS and German GAAP was the main point of the analysis of Gassen and Sellhorn (2006), as well. By examining the period 1993-2004, their results are in favor of IFRS, hence in accordance with that of Bartov et al. (2005). A group of empirical investigations expanded the methodology by comparing IFRS with several accounting standards simultaneously for many countries and not exclusively for a single one. Although these studies have the advantage that their results are more representative regarding the consequences of IFRS, on the other hand they do not take into account specific country factors that may affect the results. For example the analysis of Barth et al. (2008), in which they compared a sample of 411 companies from 24 countries that applied IFRS voluntarily, with an equal, respective sample of companies that applied domestic accounting standards for the period 1990-2004. The results recorded that firms which adopted IFRS voluntarily, performed more value relevant accounting information. Jennings et al. (2004) also came to the same conclusion. Specifically, by using a sample of companies from 12 countries, in which taxation affects significantly financial statements, they


9

examined whether the voluntary accounting transition improved the quality of accounting information. Focusing on period 1991-2001, they concluded that voluntary application of IFRS improves significantly the value relevance of accounting information. Finally, Hung and Subramanyam (2007) and Bellas et al. (2007), although they examined the value relevance of accounting information as well, they adopted an innovative methodology, which is followed in the present study. Contrary to the above studies, they did not examine the extent in which alternative accounting numbers are differently valued by the investors. They actually used stock prices as the fundamental value of the firm and investigated whether the alternative measurements correlate with the information set used by investors in setting stock prices. Although Hung and Subramanyam (2007) examined the voluntary adoption of IFRS in Germany and Bellas et al. (2007) the mandatory adoption of IFRS in Greece, they recorded similar results. Specifically, both studies by using small samples, they found that book value, in opposition to net income, play a more significant role under IFRS, compared to that under German GAAP and GAS, respectively. However, they did not find that the new accounting standards improved the relative value relevance of accounting information. In general, the results of both studies confirmed the theory, according to which German GAAP and GAS are more conservative and IFRS more balance sheet oriented due to the fair value accounting.

3. Sample and Data The sample used in the present study is composed of companies listed in the ASE for the period 20042005 and they applied mandatory IFRS in 2005 for the first time. In order to collect the data, the ASE database was used. The requirement of IFRS_1 that all firms should republish their financial statements of the previous year under IFRS, gave us the opportunity to collect for each company financial statements under both standards (GAS and IFRS) for the year 2004. Specifically, the accounting information under GAS was extracted from the annual reports of 2004 and the accounting information under IFRS from the republished accounting figures of 2005 annual reports. However, from our sample we excluded companies that they did not have all the necessary data for our analysis. In addition, we used only firms which the end of their financial year was the 31th of December. Furthermore, in compliance with the majority of similar empirical studies, banks, insurance and financial service companies were excluded from the sample. Given the above limitation, a final sample of 233 observations was used for accounting figures under IFRS and GAS simultaneously. The aim of using such a big sample in a small market like Greece is to make the present study more reliable and explore the real differences between IFRS and GAS, something that would be less feasible under smaller samples. It is also noted that contrary to many other studies, the present research focus on firms which applied IFRS mandatory. Therefore the sample of our study is devoid of selection bias, which usually affects detrimentally studies on IFRS voluntary adoption (Soderstrom and Sun, 2007).

4. Methodology and Hypotheses As already mentioned, for a large sample of companies, data for both accounting standards were collected for the year 2004. This pioneer methodology gave us the ability to compare directly accounting figures under IFRS and GAS and therefore to check whether some of the theoretical differences of the above standards appear in practice. In order to answer this, the value relevance of accounting information provided by financial statements prepared under both standards was examined. The accounting figures examined were the book value from the balance sheet and the net income from the profit and loss account. According to international bibliography value relevance is defined as the ability of financial information derived from financial statements to reflect the value of the company when the latter is based on share prices. However, it has to be clarified that as Hung and Subramanyam (2007) and Bellas et al. (2007), the present study does not attempt to reveal whether alternative

10


accounting measures are evaluated differently by the market. Share prices are just used as a proxy for the fundamental value of the company. Hence, the level that alternative measures are related to the information set used by investors to determine share prices is examined (Barth et al., 2001). Like other similar studies, we used Price Models. Price Models allow the simultaneous examination of the combined value relevance coming from both balance sheet and profit and loss account. This advantage is important since it is quite possible that trade-off takes place between the value relevance of the book value and the net income. However, in order to deal with econometric problems that usually appear with such models the use of deflators was applied. In both models of our study, the number of shares outstanding was used as deflator. At the first of the two price models that were applied, the relative value relevance of accounting information under IFRS is compared with that under GAS. This model gives the opportunity to compare the ability of book value and net income, for each accounting standard separately, to reflect financial information incorporated in stock prices, when information from one accounting standard is available in the market. The theoretical background for this model is found in company valuation theory. In accordance with Ohlson (1995), the share price, as a proxy of the fundamental value of the firm, can be expressed as a linear equation where the book value and the net income are used as independent variables. Therefore, the first model examined had the following form: (Model 1) Pit = α + β* BVit + γ* NIit + eit Pit is the share price for the company i at the end of year t (2004), BVit is the book value of equity per share for the company i at the end of year t, NIit is the net income before taxes per share for company i at the end of year t. Like Hung and Subramanyam (2007), the book value and net income was that after the exclusion of minority interests. In addition, as net income we used the net income before taxes. Apart from model (1), two additional versions of Ohlson model were examined. In the first one, the book value was used as the only independent variable. This model had the following form: Pit = α + β* BVit + eit (Model 1a) The variables definitions in model (1a) are the same with model (1). Model (1a) gives the opportunity to examine only the effects of IFRS in the value relevance of balance sheet information. Moreover, a second version of Ohlson model was applied. In this one the net income was used as the only independent variable. This model had the following form: Pit = α + β* NIit + eit (Model 1b) The variables definitions in model (1b) are the same with the above models. Thus the relative value relevance was examined in three different versions and all the models (1, 1a and 1b), formulated twice. The first one under GAS and the second under IFRS. For each model, the results under GAS and IFRS were compared and they were examined whether recorded statistical significant differences between the independent variables coefficients and the Adjusted R-Squares indicators. The tests in coefficients are based on t-test and the tests in Adjusted R- Squares indicators are based on Vuong Tests (Vuong, 1989). Apart from the relative value relevance, it was investigated the incremental value relevance as well. In opposition to the first one, with the incremental value relevance it was examined to what extent the accounting figures under IFRS provide further information from that under GAS, when both set of accounting information are available simultaneously in the market. Thus, instead of using two different models, one for each accounting standard, one model that includes simultaneously accounting figures under both IFRS and GAS was applied. Specifically, the model includes accounting information under GAS and the adjustments of the same accounting information from GAS to IFRS. The model had the following form: (Model 2) Pit = α + β* BV_GASit + γ* BV_DIFit + δ* NI_GASit + ε* NI_DIFit + eit Pit is the share price of company i at the end of year t (2004), BV_GASit is the book value of equity per share for company i at the end of year t under the GAS, BV_DIFit is the book value of equity per share under the IFRS - book value of equity per share under the GAS, NI_GASit is the net


11

income before taxes per share of company i at the end of year t under the GAS and NI_DIFit is the net income per share under the IFRS - net income per share under the GAS By examining the incremental value relevance it is evaluated the ability of accounting figures under IFRS to reflect information beyond the known accounting figures under GAS (Hung and Subramanyam, 2007; Lin and Chen, 2005; Holthausen and Watts, 2001). Consequently, it is investigated the value relevance of the adjustments of IFRS over the book value and net income under GAS, which are used as known. Like model (1), two additional versions of model (2) were examined. In the first one, only the accounting information regarding book value is used as independent variables. The model had the following form: Pit = α + β* BV_GASit + γ* BV_DIFit + eit (Model 2a) The variables definitions in model (2a), are the same with the model (2). In addition, a second version of the model was applied, in which only the information regarding net income is used as independent variables. The model had the following form: Pit = α + β* NI_GASit + γ* NI_DIFit + eit (Model 2b) The variables definitions in model (2b) are the same with the above models. The above models were applied to all observations of our sample (233). However, in order to receive more information about the consequences of the IFRS mandatory adoption in Greece, the research was expanded. Given that many studies (e.g. Dumontier and Maghraoui, 2007; Aljifri. and Khasharmeh 2006; Goodwin and Ahmed, 2006; Jones and Higgins, 2006; Floropoulos, 2006) recorded different results between small and large firms, we investigated whether firm size affects the results of the value relevance analysis. Aljifri and Khasharmeh (2006) found that firm size has a significant effect on the level of adoption of IFRS. Additionally, Jones and Higgins (2006) revealed that the extent of knowledge of IFRS is lower for smaller firms. Generally, the main argument of both studies was the limited ability of small firms to have access to the necessary accounting skills required for the implementation of IFRS, due to lack of resources. Given this limitation for small firms, AICD (2004) argued that smaller firms should be allowed more time to implement IFRS, since the cost of implementation (e.g. the cost of restatement of the past) is greater for these firms. Moreover, Goodwin and Ahmed, (2006) presented results that the transition to IFRS has not been onerous for small firms, since most of them are unaffected by the new accounting standards. Furthermore, Dumontier and Maghraoui (2007) concluded that IFRS do not increase the information content of small firms’ financial statements, but only for large firms. Due to the fore mentioned findings, we expect smaller firms in ASE to have limited access to the required specialized accounting skills needed for the IFRS implementation and this may led them to a less accurate application of IASB’s standards. Therefore, we expect value relevance analysis to record a lower impact of IFRS on small firms compared to large ones. Based on the aforementioned discussion we state the following hypothesis: H1: The impact of IFRS is lower for smaller firms. In this context, like Aljifri and Khasharmech (2006) and Goodwing and Ahmed (2006), we divided our sample into three subgroups (small, median and large firms) according to their size, as measured by their total assets. Thus, we compared the relative and incremental value relevance of accounting information under IFRS and GAS for both small and large firms. It should be noted that in order to lead to even safer conclusions, total sales were also used for the measurement of firm size. Their untabulated results were identical to those of total assets. Similarly we examined whether the level of fixed assets affects the results of the value relevance analysis. It is known that many IFRS are directly related to fixed assets and cause significant changes compared to GAS, like IAS_16, IAS_17, IAS_40 and IAS_36. In this context, Bellas et al. (2007) revealed that the accounting figures mostly affected by the mandatory application of IFRS in Greece are the tangible assets and the fixed assets. Given the strong emphasis of IFRS on fixed assets, we expect value relevance analysis to record significant deviations between firms with low level of fixed assets and firms with high level of fixed assets. Based on the above discussion we form our second hypothesis:

12


H2: The impact of IFRS varies between firms with low and high level of fixed assets. Therefore, we divided our sample into three subgroups (firms with low, medium and high level of fixed assets) according to the ratio fixed assets to total assets. Thus, we compared the relative and incremental value relevance under IFRS and GAS, for both firms with low and high level of fixed assets. For all the above models, it was tested whether autocorrelation, multicollinearity and heteroscedasticity affect the regressions results. The autocorrelation test was made by Breutch– Godfrey serial correlation LM Test, the multicollinearity test through the methodology of Ridge Regression and the White Test in order to control for heteroscedasticity in the errors term. The tests show that our results are not affected from autocorrelation, multicollinearity and heteroscedasticity. In addition, it is noted that the comparisons between GAS and IFRS were made for the same number of observations (each firm is represented by two observations, one with IFRS and one with GAS).

5. Results 5.1. The Value Relevance of Accounting Information Under GAS and IFRS This section presents the results of the value relevance investigation, for all the observations (233) of our sample. In our first basic model (model 1), it is examined the relative value relevance. As a first investigation of the correlations between the share prices and the accounting information, we examined the Pearson correlation coefficients, which are presented in Table 1. Table 1 shows that book value and net income relate positively to the share prices under both accounting standards. However, it is observed that the book value has higher correlation with the share prices under IFRS, in contrast to the net income which records higher correlations under GAS. Furthermore, in order to record the level of multicollinearity, we examined the Pearson correlation coefficients of the independent variables of model (1). The results show that the correlations between book value and net income are 69% for the GAS and 73% for the IFRS. In the past, similar studies with the same and even higher correlations coefficients, considered that no further analysis is necessary (e.g. Sami and Zhou, 2004; Hung and Subramanyam, 2004). Therefore, in accordance with international bibliography the model was applied without modifications. Nevertheless, the methodology of Ridge regression that we applied in regressions of the relative value relevance analysis showed that multicollinearity does not affect the results. Table 1: Pearson correlation coefficients on variables used in model (1):

Pit = α + β* BVit + γ* NIit + eit Panel A: GAS P BV NI Panel B: IFRS P

P 1,000 (0,000) 0,718*** (0,000) 0,840*** (0,000)

BV

NI

1,000 (0,000) 0,691*** (0,000)

1,000 (0,000)

1,000 (0,000) 1,000 BV 0,764*** (0,000) (0,000) 1,000 NI 0,818*** 0,734*** (0,000) (0,000) (0,000) Definitions, P: price per share at the end of the fiscal year t, BV: book value per share for firm i at the end of fiscal year t, NI: net income per share for firm i at the end of fiscal year t. *, **, *** statistically significant at 0.10, 0.05 and 0.01, respectively. P-values are in parenthesis

13


Table 2 presents the results of the relative value relevance. The results for each model include the coefficients of the independent variables with the corresponding p-values and the Adjusted R2 indicators. Starting with the Adjusted R2 indicators, in model (1a) it is observed that the explanatory power of IFRS is higher than that of GAS. However, their difference is not statistically significant. Conversely, in model (1b) the GAS present higher explanatory power than that of IFRS, but also their difference is insignificant. Finally, in the combined model where both book value and net income are included in the model (model 1), it is recorded that GAS have marginally higher explanatory power, although their difference is not significant too. Therefore, through the investigation of Adjusted R2 indicators it seems that the introduction of IFRS in Greece did not improve the relative value relevance of book value and net income, either separately or in combination. Next, we examined the coefficients of independent variables for each model separately. In model (1a), it is observed that the coefficients of the book value under both accounting standards are significant at the 1% level and their difference is not significant. In contrast, in model (1b) although the coefficients of net income under both standards are significant at the 1% level, the coefficient under GAS is higher than that of IFRS and their difference is significant at the 5% level. Higher net income coefficients under GAS are in accordance with the argument that earnings under GAS are smoother and therefore more stable than IFRS. In this context, as mentioned by Black (1993), the valuation model that includes only the net income as independent variable is based on the concept of permanent income and this justifies the higher coefficients under GAS. Finally, we examined the combined model that includes both book value and net income (model 1). Regarding the book value coefficients, it is observed that under both standards, they are significant at the 1% level. In addition, although the book value coefficient under IFRS is higher than that under GAS, their difference is significant only at the 11%. Conversely, the net income coefficients under both standards have the same statistical significance (1%), however the coefficient under GAS is higher than that of IFRS as their difference is significant at 2%. Table 2: Relative value relevance of book value and net income under GAS and IFRS

Pit = α + β* BVit + γ* NIit + eit BV MODEL (1a) Intercept BV R^2

(Model 1) NI MODEL (1b) Intercept NI R^2

BV & NI MODEL (1) Intercept BV NI

R^2 GAS coefficients 0,543 1,234*** 0,514 2,051*** 4,867*** 0,705 1,232*** 0,452*** 3,814*** 0,74 p - value 0,104 0,000 0,000 0,000 0,000 0,000 0,000 IFRS coefficients 0,641** 1,157*** 0,582 2,324*** 4,365*** 0,668 1,323*** 0,537*** 2,976*** 0,72 p - value 0,03 0,000 0,000 0,000 0,000 0,000 0,000 GAS-IFRS coefficients -0,098 0,077 -0,068 -0,273 0,037 -0,091 -0,085 0,502** 0,838** 0,02 p – value 0,395 0,456 0,404 0,312 0,05 0,192 0,2992 0,1122 0,02 0,38 Definitions, Pit: price per share at the end of the fiscal year t, BV it: book value per share for firm i at the end of fiscal year t, NI it: net income per share for firm i at the end of fiscal year t. The tests in coefficients are based on t-tests. The tests in adjusted R-squares are based on Vuong Tests. Two tailed p-values are used. *, **, *** statistically significant at 0.10, 0.05 and 0.01, respectively

Apart from the relative value relevance, with our second basic model we investigated the incremental value relevance. At first, we examine the Pearson correlation coefficients, which are presented in table 3. The results show that all the independent variables are positively and significantly correlated with the share prices. In order to measure the magnitude of multicollinearity, we examined the coefficients for the independent variables of model (2). Generally, the results show that the correlations in question are not very large. Nevertheless, as in the relative value relevance we applied the methodology of Ridge regression, which shows that multicollinearity does not affect the results.

14


Table 3: Pearson correlation coefficients on variables used in model (2): Pit = α + β* BV_GASit + γ* BV_DIFit + δ* NI_GASit + ε* NI_DIFit + eit P BV_GAS BV_DIF NI_GAS NI_DIF 1,000 0,000 1,000 BV_GAS 0,718*** (0,000) 0,000 1,000 BV_DIF 0,467*** 0,291*** (0,000) (0,000) 0,000 1,000 NI_GAS 0,840*** 0,691*** 0,469*** (0,000) (0,000) (0,000) 0,000 1,000 NI_DIF 0,211*** 0,238*** 0,124* 0,276** (0,001) (0,000) (0,060) (0,000) 0,000 Definitions, Pit: price per share for firm i at the end of the fiscal year t, BV_GASit: book value per share reported under GAS for firm i at the end of fiscal year t, NI _GASit: net income per share reported under GAS for firm i at the end of fiscal year t, BV_DIFit: difference between IFRS and GAS book value per share for firm i at the end of fiscal year t, NI_DIFit: difference between IFRS and GAS net income per share for firm i the end of fiscal year t. *, **, *** statistically significant at 0.10, 0.05 and 0.01, respectively P-values are in parenthesis P

Table 4 presents the results of the incremental value relevance. Each model includes the coefficients of the independent variables with the corresponding p-values and the Adjusted R2 indicators. Starting with model (2a), it is recorded that the coefficient of book value under GAS and the coefficient of the variable that measures the adjustments of IFRS to book value are both positive and significant at 1%. Therefore, the accounting information that expresses the adjustments of IFRS to the balance sheet is incrementally value relevant. A different picture is reflected in model (2b). While the coefficient of net income under GAS is positive and significant at the 1%, the coefficient of the adjustments of IFRS to net income is negative but not significant. In other words, unlike book value, the adjustments of IFRS to net income seem to be negative and not incrementally value relevant. Finally, we examined the combined model (2). Regarding the variables that express the accounting information under GAS, the results show that the coefficients of book value and net income are both positive and significant at the 1%. However, while the coefficient of the IFRS adjustments to book value is also positive and significant (1%), suggesting that such information is incrementally value relevant, the coefficient of the IFRS adjustments to net income is negative and statistically insignificant, suggesting that such information not only do not improve but may even impair value relevance. Table 4: Incremental value relevance of GAS adjustments to book value and net income Pit = α + β BV_GASit + γ BV_DIFit + δ NI_GASit + ε NI_DIFit + eit

(Model 2)

Intercept BV_GAS BV_DIF NI_GAS NI_DIF Adj R^2 BV MODEL coefficients 0,585 0,778** 1,093*** 1,512*** p - value 0,013 0,000 0,000 NI MODEL coefficients -0,575 0,704 2,026*** 4,904*** p - value 0,000 0,000 0,54 BV & NI MODEL coefficients -0,911 0,748 1,213*** 0,469*** 0,562*** 3,548*** p - value 0,000 0,000 0,006 0,000 0,295 Definitions, Pit: price per share for firm i at the end of the fiscal year t, BV_GASit: book value per share reported under GAS for firm i at the end of fiscal year t, NI _GASit: net income per share reported under GAS for firm i at the end of fiscal year t, BV_DIFit: the difference between IFRS and GAS book value per share for firm i at the end of fiscal year t, NI_DIFit: the difference between IFRS and GAS net income per share for firm i the end of fiscal year t. *, **, *** statistically significant at 0.10, 0.05 and 0.01, respectively

15

International Research Journal of Finance and Economics - Issue 76 (2011) 5.2. Value Relevance and firm Size (Testing H1)

The above analysis based on the total observations (233) of our sample. However, our research was expanded by examining whether the size of the companies affect the results. Thus, we compared the relative and incremental value relevance of accounting information under IFRS and GAS for small and large firms. Panel A of table 5 shows the results of the relative value relevance for small firms. By examining the Adjusted R2 indicators, as in the total sample, we observe that there are no significant differences between GAS and IFRS for any of the models [(1), (1a) and (1b)]. Next we investigated the coefficients of independent variables for each model separately. In model (1a) it is observed that, as for the total sample, the difference in book value coefficients between IFRS and GAS is statistically insignificant. Moreover, in model (1b) the net income coefficient under GAS is significantly higher than that of IFRS, although at the 7% and not at 5% as in the total sample. Different results are recorded in the combined model (1). Regarding the book value coefficients, it was found that the difference between GAS and IFRS is statistically insignificant. However, while for the total sample the book value coefficient under IFRS is higher than that under GAS, for the small firms the situation is reversed and the coefficient under GAS is higher. Nevertheless, their difference is also statistically insignificant. In addition, different results are recorded for the net income coefficients. Specifically, while for the total sample it was presented a statistically significant difference in favor of GAS, for small firms this result is not confirmed. However, the net income coefficient under GAS is statistically significant at the 1% level and that under IFRS only at the 10%. Table 5:

Small Vs Large Firms - Relative value relevance of book value and net income under GAS and IFRS (Model 1): P it = α + β BV it + γ NI it + e it

PANEL A: Small Firms Intercept GAS coefficients 0,658*** p – value 0,002 IFRS coefficients 0,704*** p – value 0,002 GAS-IFRS coefficients -0,046 p – value 0,382 PANEL B: Large Firms Intercept GAS coefficients p – value IFRS coefficients p – value GAS-IFRS coefficients p – value

BV MODEL BV

R^2

Intercept

NI MODEL NI

0,822*** 0,000

0,60

1,703*** 0,000

0,760*** 0,000

0,55

0,0062 0,478 BV MODEL BV

BV & NI MODEL BV NI

R^2

Intercept

R^2

3,038*** 0,000

0,116

0,450** 0,028

0,78*** 0,000

1,936*** 0,002

0,65

1,944*** 0,000

1,672** 0,049

0,039

0,621*** 0,006

0,742*** 0,000

0,96* 0,095

0,56

0,05 0,496

-0,241 0,259

1,366* 0,07

0,77 0,425

-0,171 0,301

0,038 0,143

0,976 0,382

0,09 0,39

R^2

Intercept

NI MODEL NI

R^2

Intercept

BV & NI MODEL BV NI

R^2

1,153 0,19

1,285*** 0,000

0,53

3,224*** 0,000

4,684*** 0,000

0,745

2,378*** 0,000

0,343** 0,025

3,894*** 0,000

0,76

1,291* 0,089

1,195*** 0,000

0,63

3,506*** 0,000

4,226*** 0,000

0,716

2,303*** 0,000

0,493*** 0,001

2,937*** 0,000

0,75

-0,138 0,368

0,09 0,500

-0,1 0,465

-0,282 0,299

0,458** 0,044

0,029 0,386

-0,075 0,295

-0,15 0,137

0,957** 0,028

0,01 0,47

Definitions, Pit: price per share at the end of the fiscal year t, BVit: book value per share for firm i at the end of fiscal year t, NIit: net income per share for firm i at the end of fiscal year t. The tests in coefficients are based on t-tests. The tests in adjusted R-squares are based on Vuong Tests. Two tailed p-values are used. *, **, *** statistically significant at 0.10, 0.05 and 0.01, respectively

Panel B of table 5 presents the results of the relative value relevance for large firms. Looking at the Adjusted R2 indicators it is observed that, as in small firms, there are no statistically significant differences between GAS and IFRS for any of the examined models. Next, we investigated the coefficients of independent variables for each model separately. In model (1a) the results are identical to those of small firms. Furthermore, in model (1b) the net income coefficient under GAS is higher than that under IFRS, although their difference is significant at 4% (compared to 7% for small firms).

16


Finally, in the combined model (1) the difference between the book value coefficients is found statistically insignificant. However, in contrast to small firms where the book value coefficient is higher under GAS, for large firms (as for the total sample) the book value coefficient is higher under IFRS. Regarding the net income coefficients, while for small firms there was no statistically significant difference, for large firms the net income coefficient under GAS is significantly higher than that under IFRS, at 3%. Apart from the relative value relevance, we investigated the incremental value relevance. The results for the small firms are presented in Panel A of table 6. In model (2a) the book value coefficient under GAS is positive and significant at the 1%. Regarding the coefficient of the adjustments of IFRS to book value, while for the total sample was positive and significant, for small firms is negative and insignificant. Similar results are recorded in model (2b). While the net income coefficient under GAS is positively significant at the 1%, the coefficient of IFRS adjustments to net income is negative and insignificant. Finally, we investigated the combined model (2). Regarding the variables that express the accounting information under GAS, the results show that the coefficients of book value and net income, as in the total sample, are both positive and significant at the 1%. However, the coefficients of the IFRS adjustments to book value and net income are both negative and insignificant. Therefore in contrast to the results for the total sample, the IFRS adjustments to book value and net income (separately or in combination) for small firms are incrementally value irrelevant and may even impair the value relevance of accounting information. Table 6:

Small Vs Large Firms - Incremental value relevance of GAS adjustments to book value and net income (Model 2):

Pit = α + β* BV_GASit + γ* BV_DIFit + δ* NI_GASit + ε* NI_DIFit + eit PANEL A: Small Firms Intercept BV_GAS BV_DIF NI_GAS NI_DIF Adj R^2 BV MODEL coefficients -0,078 0,599 0,665*** 0,822*** p - value 0,002 0,000 0,781 NI MODEL coefficients -0,852 0,11 1,697*** 2,939*** p - value 0,000 0,002 0,498 BV & NI MODEL coefficients -0,181 -0,502 0,645 0,461** 0,777*** 1,930*** p - value 0,026 0,000 0,495 0,002 0,528 PANEL B: Large Firms Intercept BV_GAS BV_DIF NI_GAS NI_DIF Adj R^2 BV MODEL coefficients 0,633 1,658** 1,077*** 1,755*** p - value 0,037 0,000 0,000 NI MODEL coefficients -1,015 0,743 3,18*** 4,758*** p - value 0,000 0,000 0,574 BV & NI MODEL coefficients -1,016 0,771 2,376*** 0,38** 0,759** 3,456*** p - value 0,000 0,012 0,026 0,000 0,555 Definitions, P it: price per share for firm i at the end of the fiscal year t, BV_GAS it: book value per share reported under GAS for firm i at the end of fiscal year t, NI _GAS it: net income per share reported under GAS for firm i at the end of fiscal year t, BV_DIF it: the difference between IFRS and GAS book value per share for firm i at the end of fiscal year t, NI_DIF it: the difference between IFRS and GAS net income per share for firm i the end of fiscal year t. *, **, *** statistically significant at 0.10, 0.05 and 0.01, respectively

The results of the incremental value relevance for large firms are presented in panel B of table 6. In model (2a) it is observed that book value coefficient under GAS is positive and statistically


17

significant at the 1%. However, the coefficient of the adjustments of IFRS to book value, in contrast to small firms, is positive and significant at the 1%. Similar results with that of small firms are recorded in model (2b). While the net income coefficient under GAS is positive and significant (at 1%), the coefficient of IFRS adjustments to net income is negative but not significant. Completing the analysis, we examined the combined model (2). Regarding the variables that express the accounting information under GAS, again the results show that are both positive and significant at the 1%. The coefficient of the IFRS adjustments to book value, while for small firms was negative and insignificant, for large firms is positive and significant at the 2%. On the contrary, the coefficient of the IFRS adjustments to net income, as for the total sample and small firms, is negative and insignificant. Therefore, the above results seem to confirm H1. In particular, the incremental value relevance analysis shows that the IFRS adjustments to accounting information are affected by firm size. While the IFRS adjustments to net income (separately or in combination) seems to be incrementally value irrelevant regardless firm size, the IFRS adjustments to book value are affected significantly by firm size. Specifically, while for small firms are negative and incrementally value irrelevant, for large firms are incrementally value relevant and improve the value relevance of accounting information (separately or in combination). 5.3. Value Relevance and Fixed Assets (Testing H2) In this section we examined whether the level of fixed assets of the companies affect the results. Thus, we compared the relative and incremental value relevance under IFRS and GAS, for the firms with low and high level of fixed assets. Panel A of table 7 shows the results of the relative value relevance for firms with low level of fixed assets. From the examination of the Adjusted R2 indicators, it is observed that, there are insignificant differences between GAS and IFRS for all the models. Next, we investigated the coefficients of independent variables for each model separately. In model (1a) it is recorded that, as for the total sample, the difference in book value coefficients between IFRS and GAS is statistically insignificant. In model (1b), the net income coefficients under GAS is significantly higher than that of IFRS, although at the 8% and not at 5% as in the total sample. Finally, we examined the combined model (1). Regarding the book value coefficients, it is observed that, in contrast to the total sample, the coefficient under GAS is higher than that under IFRS and their difference is significant at the 10%. This result in favor of GAS is also reflected by the fact that the book value coefficient under GAS is statistically significant at the 3% and that under IFRS at the 6,5%. The results for the net income coefficients seem consistent with those initially reported for the total sample, since the coefficient under GAS is higher than that under IFRS, at the 3%. The above results were compared with that for firms with high level of fixed assets (panel B of table 7). The examination of Adjusted R2 indicators recorded that again the differences between GAS and IFRS (for all the models) are statistically insignificant. Next, we examined the coefficients of the independent variables. As previously, in model (1a) the difference between the book value coefficients is statistically insignificant. In model (1b) the net income coefficient under GAS is higher than that under IFRS and their difference is significant at the 6%. Finally, in the combined model (model 1), in contrast to firms with low fixed assets, the book value coefficient under IFRS is greater than that under GAS, although their difference is insignificant. Regarding the net income coefficients, in contrast to firms with low fixed assets their difference is not statistically significant.

18 Table 7:

International Research Journal of Finance and Economics - Issue 76 (2011) Firms with low level of fixed assets Vs Firms with high level of fixed assets - Relative value relevance of book value and net income under GAS and IFRS (Model 1): Pit = α + β* BVit + γ* NIit + eit

PANEL A: Firms with Low level of fixed assets BV MODEL NI MODEL BV & NI MODEL Intercept BV R^2 Intercept NI R^2 Intercept BV NI R^2 GAS coefficients 0,693 1,407*** 0,70 1,424*** 5,346*** 0,86 1,136*** 0,297** 4,479*** 0,872 p – value 0,23 0,000 0,000 0,000 0,004 0,027 0,000 IFRS coefficients 0,908* 1,248*** 0,73 1,669*** 5,005*** 0,87 1,435*** 0,235* 4,242*** 0,873 p – value 0,094 0,000 0,000 0,000 0,000 0,065 0,000 GAS-IFRS coefficients -0,215 0,159 -0,03 -0,245 -0,01 -0,299 0,341* 0,062* 0,237** -0,001 p – value 0,383 0,470 0,475 0,244 0,078 0,510 0,364 0,10 0,033 0,436 PANEL B: Firms with High level of fixed assets BV MODEL NI MODEL BV & NI MODEL BV R^2 Intercept NI R^2 Intercept BV NI R^2 Intercept GAS coefficients 1,87** 0,794*** 0,15 3,471*** 3,339*** 0,38 1,905*** 0,542*** 3,017*** 0,445 p – value 0,021 0,000 0,000 0,000 0,004 0,003 0,000 IFRS coefficients 0,609 1,179*** 0,34 3,701*** 2,549*** 0,30 1,065* 0,899*** 1,836*** 0,472 p – value 0,396 0,000 0,000 0,000 0,10 0,000 0,000 GAS-I FRS coefficients 1,261 -0,385 -0,19 -0,23 0,08 0,84 -0,357 1,181 -0,027 0,79* p – value 0,385 0,453 0,49 0,259 0,062 0,432 0,326 0,112 0,385 0,435 Definitions, P it: price per share at the end of the fiscal year t, BV it: book value per share for firm i at the end of fiscal year t, NI it: net income per share for firm i at the end of fiscal year t. The tests in coefficients are based on t-tests. The tests in adjusted R-squares are based on Vuong Tests. Two tailed p-values are used. *, **, *** statistically significant at 0.10, 0.05 and 0.01, respectively

The results of the incremental value relevance for the firms with low level of fixed assets are presented in panel A of table 8. In model (2a), as for the total sample, both the book value coefficient under GAS and the coefficient that measures IFRS adjustments to book value are positive and significant at the 1%. Similarly, in model (2b) the net income coefficient under GAS is positively significant at the 1%. However, the coefficient of IFRS adjustments to net income, while for the total sample was negative and statistically insignificant, for these firms is positive and significant at the 4%. Finally, in the combined model (2) the variables that express the accounting information under GAS, as for the total sample, are both positive and significant at the 2% and 1%, respectively. On the other hand, completely different results are presented for the variables which measure the IFRS adjustments to accounting information. The coefficient of the IFRS adjustments to book value, while for the total sample was positively significant (at the 1%), for the firms with low fixed assets is negatively significant (at the 1%). Conversely, the coefficient of the IFRS adjustment to net income, while for the total sample was negative and insignificant, for the firms with low fixed assets is positive and significant (at the 1%). The results of the incremental value relevance analysis show that for the firms with low levels of fixed assets the IFRS adjustments to the accounting information are different from those recorded for the total sample. While the IFRS adjustments to net income seem to be incrementally value relevant and improve (separately or in combination) the value relevance, the adjustments to book value are incrementally value relevant but lead to ambiguous results regarding whether they improve or impair the value relevance of accounting information.

International Research Journal of Finance and Economics - Issue 76 (2011) Table 8:

19

Firms with low level of fixed assets Vs Firms with high level of fixed assets - Incremental value relevance of GAS adjustments to book value and net income (Model 2):

Pit = α + β* BV_GASit + γ* BV_DIFit + δ* NI_GASit + ε* NI_DIFit + eit PANEL A: Firms with Low level of fixed assets Intercept BV_GAS BV_DIF NI_GAS NI_DIF Adj R^2 BV MODEL coefficients 0,738 1,215** 1,052*** 2,496*** p - value 0,034 0,000 0,002 NI MODEL coefficients 0,87 1,561*** 5,163*** 3,098** p - value 0,000 0,000 0,042 BV & NI MODEL coefficients 0,892 0,99*** 0,292** -2,19*** 5,503*** 4,066*** p - value 0,008 0,018 0,002 0,000 0,005 PANEL B: Firms with High level of fixed assets Intercept BV_GAS BV_DIF NI_GAS NI_DIF Adj R^2 BV MODEL coefficients 0,902 0,373 1,063*** 1,892*** p - value 0,204 0,000 0,000 NI MODEL coefficients 0,413 3,299*** 3,823*** -3,649** p - value 0,000 0,000 0,03 BV & NI MODEL coefficients 0,836 0,585 0,849*** 1,174*** 2,967*** -3,758** p - value 0,162 0,000 0,000 0,000 0,013 Definitions, P it: price per share for firm i at the end of the fiscal year t, BV_GAS it: book value per share reported under GAS for firm i at the end of fiscal year t, NI _GAS it: net income per share reported under GAS for firm i at the end of fiscal year t, BV_DIF it: the difference between IFRS and GAS book value per share for firm i at the end of fiscal year t, NI_DIF it: the difference between IFRS and GAS net income per share for firm i the end of fiscal year t. *, **, *** statistically significant at 0.10, 0.05 and 0.01, respectively

The results of the incremental value relevance for the firms with high fixed assets are presented in panel B of table 8. In model (2a) it is observed that the results are similar with those recorded for firms with low level of fixed assets. Specifically, the book value coefficient under GAS and the coefficient of the IFRS adjustments to book value are both positively significant at the 1%. In model (2b), the results differ significantly from those reported above. While the net income coefficient under GAS is positive and significant (at the 1%), that of the IFRS adjustments to net income is negative and significant (at the 3%). Finally, it was examined the combined model (model 2). Regarding the variables that express the accounting information under GAS, again are both positive and significant at the 1%. The coefficient of the IFRS adjustments to book value, while for the firms with low fixed assets was negative and significant, for these firms is positive and significant at the 1%. Conversely, the coefficient of the IFRS adjustment to net income, while for the firms with low fixed assets was positively significant, for the firms with high fixed assets is negative and significant (at the 1%). Consequently, the above results seem to confirm H2. Particularly, the results of the incremental value relevance analysis show clearly that the IFRS adjustments to accounting information are affected by the level of fixed assets. While for the firms with high fixed assets the IFRS adjustments to book value seem to be incrementally value relevant and improve the value relevance of accounting information, for the firms with low level of fixed assets ambiguous results are recorded. On the other hand, while for the firms with low fixed assets the IFRS adjustment to net income seem to be incrementally value relevant and improve the value relevance of accounting information, for the firms

20


with high levels of fixed assets are also incrementally value relevant but impair the value relevance of accounting information.

6. Conclusions In the present study it was investigated the value relevance of accounting information provided by financial statements when prepared under GAS and IFRS. Firstly, we examined all the observations of our sample. The results of the relative value relevance show that the introduction of IFRS does not seem to improve the relative value relevance of book value and net income, either separately or in combination. Consequently, neither GAS nor IFRS could be argued to be better at explaining market values. However, in a combined valuation model that includes both accounting information, while net income play a more significant role in valuation under GAS than under IFRS, book value, although records a pattern in favor of IFRS, it is not confirmed statistically whether is more important than that under GAS. In addition, the examination of incremental value relevance recorded that while the adjustments of IFRS to book value are positively statistical significant and therefore incrementally value relevant, the adjustments of IFRS to net income are negative and incrementally value irrelevant. Generally, the results of our analysis show that IFRS, by eliminating the income smoothing orientation of GAS and emphasizing on the fair value accounting, seems to introduce a greater transitory component to income, which reduces its persistence. Specifically, the recorded results show that net income under IFRS is less value relevant than that under IFRS and the adjustments of IFRS to net income are incrementally value irrelevant. On the other hand, the adjustments of IFRS to book value are positive and incrementally value relevant. However, the examination of relative value relevance did not make clear whether book value under IFRS is more important than that under GAS. Therefore, the examination of the total sample of our study confirms that the accounting standards of IASB are balance sheet oriented and lay emphasis on fair value accounting, while GAS on income smoothing. Given that GAS emphasize on creditors and taxation protection (stakeholder oriented), the results of our study could be compared with that of empirical studies, which applied similar methodology in countries with similar accounting system. The research of Hung and Subramanyam (2007) for the German market is an international empirical investigation, which applied a comparable methodology to our study. Although our study used a larger sample of companies, examined the consequences of mandatory and not voluntary adoption and applied a slight different methodology, the results of both studies are (indirect) comparable, since Greece and Germany have common accounting tradition and orientation. In general terms, the results of Hung and Subramanyam (2007) are in accordance with that of our study and this advocates that the results of our research, even with reservations, could be expanded in countries with similar accounting system. Also, we investigated whether the results are affected by the size of the firms and the level of fixed assets. The results and mainly the incremental value relevance, show that there are significant deviations between small and large firms. Specifically, it was recorded a lower impact of IFRS on small firms compared to large ones. This implies that the consequences of IFRS mandatory adoption are not the same for all sizes of firms. Specifically, the results for the total sample of our study regarding book value are the same with that for large firms, but not for small firms. Therefore, it is generated the question for what reason small firms present different results. The answer that probably justifies the above result is that small firms may have not applied IFRS properly for the first year of their adoption. Generally, the issue of the correct application of IFRS has been raised many times in the literature, since the quality of financial statements is not only related to the quality of accounting standards, but also to the way that accounting standards are applied by the firms. Regarding Greece, given the wrong mentality that often dominates the accounting treatments and many other issues, it is quite possible that the financial statements prepared for the first time under IFRS had significant shortcomings. The results of our study support that aspect of view,


21

however it is concluded that false or at least less accurate application of IFRS is more likely to occur in small firms. In order to be established whether smaller firms applied IFRS correctly, it should be noted that the IASB’s standards cover a wide range of issues, they are complicated and many times their comprehension is difficult, at least in relation to GAS. Therefore, taking into account that IFRS was something new for Greek firms, it is clear that the accurate application of IFRS requires the appropriate infrastructure and the acquisition of specialized knowledge, skills and professional advice. In other words, IFRS require significant changes in a firm, such as changes in information systems, hiring new or training the existing staff, the realization of which has a significant cost. Thus, given that smaller firms have fewer resources, it is likely that they have limited access to required specialized accounting skills needed for the application of IFRS. Consequently, smaller firms probably were in a difficult position, regarding the application of IFRS and this may led them to a less accurate application of IASB’s standards. The difficulty for smaller firms to adopt IFRS is not recorded for the first time. Jones and Higgins (2006) found that larger firms, mainly due to greater availability of resources, have more knowledge about the new standards. Therefore, it is quite likely large (small) firms due to their greater (less) knowledge about IFRS, to apply them with more (less) precision. Regarding Greece, Floropoulos (2006), found that large and medium-sized firms, in contrast to small, tend to comply with the requirements of IFRS at a greater extent. The above analysis examined the possibility that small firms did not apply IFRS properly during the accounting transition. However, it should be noted that this may not be a temporary phenomenon that occurred only in the first year, but may still exists today. Nevertheless, the burning issue of the accurate implementation of IFRS requires further empirical investigation and analysis. The value relevance of accounting information regarding the level of fixed assets, show that the level of fixed assets is an additional factor that affects significantly the results. For companies with high level of fixed assets, the results are generally in line with that for the total sample. On the contrary, our findings regarding firms with low levels of fixed assets are diametrically opposite to those of the total sample. Specifically, the very different results of the relative and incremental value relevance regarding book value show that some of the theoretical characteristics of IFRS are not captured with the same intensity in firms with low levels of fixed assets. In particular, the emphasis of IFRS on the balance sheet and their orientation on fair value accounting, do not seem to be recorded in practice for these firms. Generally, these results can be attributed to the great importance that IFRS give to fixed assets. Mainly IAS_16, but also IAS_40, IAS_36, IAS_17, IAS_38 and IAS_20 cause large changes in the accounting treatment of fixed assets, compared to GAS. Therefore, for the companies with low levels of fixed assets there is less room for IFRS to act and this might lead those firms to present different results. It can be argued that the above results extend the bibliography since the mandatory transition to IFRS, in combination with the level of fixed assets, is examined for the first time. However, these results give the opportunity for further empirical investigations. Concluding our analysis, it could be argued that the introduction of IFRS affected the financial statements of listed companies in the ASE, significantly. However, as it is clear from our results, the consequences of IFRS implementation are not the same for all firms, since they are affected by firm specific characteristics.

References [1]

Aljifri, K. and H., Khasharmeh, 2006. “An Investigation into the Suitability of the International Accounting Standards to the United Arab Emirates Environment”, International Business Review, 15, pp. 505-526

22


[2]

Amir, E., T., Harris and E., Venutti, 1993. “A comparison of the value-relevance of US versus non-US GAAP accounting measures using form 20-F reconciliations”, Journal of Accounting Research, 31 (Supplement), pp. 230-264 Ashbaugh, H., 2001. “Non–US firms’ accounting standards choices”, Journal of Accounting and Public Policy, 20 (2), pp. 129–153 Ball, R. and P., Brown, 1968. “Empirical evaluation of accounting income numbers”, Journal of Accounting Research, pp. 159–178 Bao, B. and L., Chow, 1999. “The usefulness of earnings and book value for equity valuation in emerging capital markets: Evidence from the listed companies in the People’s Republic of China”, Journal of International Financial Management and Accounting, 10 (2), pp. 85–104 Barth, M.E., W.H., Beaver and W.R., Landsman, 2001. “The relevance of the value relevance literature for financial accounting standard setting: Another view”, Journal of Accounting and Economics, 31 (1–3), pp. 77–104. Barth, M.E., W.R., Landsman and M.H., Lang, 2008. “International accounting standards and accounting quality”, Journal of Accounting Research, 46 (3), pp. 467- 498 Barth, M.E., W.R., Landsman, M.H., Lang and C.D., Williams, 2006. “Accounting quality: international accounting standards and US GAAP”, Working Paper, University of North Carolina and Stanford University Bartov, E., S., Goldberdg and M., Kim, 2005. “Comparative value relevance among German, U.S. and international accounting standards: A German stock market perspective”, Journal of Accounting, Auditing & Finance, 20 (2), pp. 95–119 Bellas, A., K., Toudas and K., Papadatos, 2007. “What international accounting standards (IAS) bring about to the financial statements of Greek listed companies? The case of the Athens stock exchange”, Spoudai: Journal of Economics and Business, 57 (3), (September–December), pp. 54-77 Black, F., 1993. “Choosing Accounting Rules”, Accounting Horizons, 7, pp. 1-17 Chan, K. and G., Seow, 1996. “The association between stock returns and foreign GAAP earnings versus earnings adjusted to U.S. GAAP”, Journal of Accounting and Economics, 21 (1), pp. 139–159 Dan, H., 2002. “The usefulness of financial statements under Chinese-GAAP vs IAS: Evidence from the Shangai stock exchange in PRC”, Kobe Economic and Business Review, 48, pp. 1–25 Dumontier, P. and B., Raffournier, 1998. “Why firms comply voluntary with IAS: An empirical analysis with Swiss data”, Journal of International Financial Management and Accounting, 9 (3), pp. 216-245 Eccher, E. and P., Healy, 2003. “The role of international accounting standards in transitional economies: A Study of the People’s Republic of China”, Working paper, Massachusetts Institute of Technology El - Gazzar, M. S., M.P., Finn and R., Jacob, 1999. “An empirical investigation of multinational firms’ compliance with International Accounting Standards”, The International Journal of Accounting, 34 (2), pp. 239-248 Floropoulos, I., 2006. “IFRS. First time users: Some empirical evidence from Greek companies”, Spoudai: Journal of Economics and Business, 56 (3), pp. 39-70 Gassen, J. and T., Sellhorn, 2006. “Applying IFRS in Germany – determinants and consequences”, Betriebswirtschaftliche Forschung und Praxis, 58 (4), pp. 365–386 Goodwin, J. and K., Ahmed, 2006. “The Impact of International Financial Reporting Standards: Does the Size Matter?”, Managerial Auditing Journal, 21 (5), pp. 460-475 Gordon, E.A., B.N., Jorgensen and C.L., Linthicum, 2007. “Are IFRS – U.S. GAAP Reconciliations Informative?”, 30th Annual Congress of European Accounting Association, 25 – 27 April, Lisbon – Portugal.

[3] [4] [5]

[6]

[7] [8]

[9]

[10]

[11] [12]

[13] [14]

[15]

[16]

[17] [18] [19] [20]

International Research Journal of Finance and Economics - Issue 76 (2011) [21]

[22]

[23]

[24]

[25] [26]

[27] [28] [29]

[30] [31]

[32] [33]

23

Harris, M.S. and A.K., Muller, 1999. “The market valuation of IAS versus US - GAAP accounting measures using form 20-F reconciliations”, Journal of Accounting and Economics, 26 (1-3), pp. 285–312 Holthausen, R.W. and R.L., Watts, 2001. “The relevance of the value–relevance literature for financial accounting standard setting”, Journal of Accounting and Economics, 31 (1-3), pp. 3– 75 Hung, M. and K.R., Subramanyam, 2007. “Financial statement effects of adopting international accounting standards: The case of Germany”, Review of Accounting Studies, 12 (4), pp. 623657 Jennings, R., W. Mayew, and S., Tse, 2004. “Do international accounting standards increase the timeliness and value-relevance of financial statement disclosure?”, Working Paper, Department of Accounting McCombs School of Business, University of Texas at Austin Jones, S. and A.D., Higgins, 2006. “Australia’s switch to international financial reporting standards: a perspective from account prepares”, Accounting and Finance, 46, pp. 629-652 Lin, Z.J. and F., Chen, 2005. “Value relevance of international accounting standards harmonization: Evidence from A - and B - share markets in China”, Journal of International Accounting, Auditing and Taxation, 14, pp. 79–103 Murphy, B.Α., 1999. “Firm characteristics of Swiss companies that utilize International Accounting Standards”, The International Journal of Accounting, 34 (1), pp. 121–131 Ohlson, J.A., 1995. “Earnings, book values and dividends in equity valuation”, Contemporary Accounting Research, 11 (2), pp. 661–688. Sami, H. and H., Zhou, 2004. “A comparison of value relevance of accounting information in the different segments of the Chinese stock market”, The International Journal of Accounting, 39, pp. 403–427 Soderstrom, S. and K., Sun, 2007. “IFRS adoption and accounting quality: A review”, European Accounting Review, 16 (4), pp. 675-702 Tarca, A., 2004. “International convergence of accounting practices: Choosing between IAS and US GAAP”, Journal of International Financial Management and Accounting, 15 (1), pp. 60–91 Vuong, Q., 1989. “Likelihood ratio tests for model selection and non-nested hypotheses”, Econometrica, 57, pp. 307–333 Wu, S.H., M., Koo, and T., Kao, 2005. “Comparing the value–relevance of accounting information in China: Standards and factors effects”, Working paper, Tainan University of Technology