The effect of IFRS adoption on the value relevance of accounting information: evidence from South Korea

This study investigates whether the value relevance of accounting information was changed after IFRS adoption in South Korea. Related prior studies have found mixed empirical evidence depending on research methodologies or research periods. Moreover, the effect of IFRS adoption on value relevance can be different between Korean stock markets (KSE and KOSDAQ) because they have different characteristics. Also, the main financial statements reported by Korean firms had changed from individual financial statements to consolidated financial statements after IFRS adoption. Thus, this study analyzes the effect of IFRS adoption on the value relevance of individual and consolidated accounting numbers expanding research periods (5 years before and after IFRS adoption) and comparing changes in explanatory powers of Ohlson (1995) model on each listing market. The empirical results indicate that the value relevance of Korean listed firms generally decreased after IFRS adoption. However, the value relevance of KSE listed firms decreased, while the value relevance of KOSDAQ listed firms increased after IFRS adoption. In addition, it was found that the effects of IFRS adoption on value relevance of individual and consolidated financial information were different depending on listed markets. This implies that different level of demand for information environment may induce differential effects of IFRS adoption on value relevance.


INTRODUCTION
Under Korean GAAP, all Korean firms had to disclose their individual financial statements within 90 days after the closing date of fiscal year prior to disclose consolidated financial statements. Accordingly, most of accounting information users mainly used (timely) individual financial statements rather than consolidated financial statements in making economic decision.
South Korea (hereafter, Korea) had developed its own accounting standards (Korean GAAP) and applied it to all firms including listed and unlisted firms until 2010. Korean GAAP had characteristics that they were developed by rule-based approach, which allowed managers' discretion less than IFRS (International Financial Reporting Standards) applying principle-based approach and the main financial statements disclosed under Korean GAAP were not consolidated financial statements, but individual financial statements 1 . Those different characteristics of Korean GAAP had been a major factor, which devaluated accounting transparency of Korean firms and accordingly brought about so-called 'Korea Discount'.
Hence, Korea has fully adopted IFRS since 2011 for enhancing international coordination of accounting standards, accounting transpar-ency, and eventually the usefulness of accounting information as a solution for 'Korea discount' 2 . If IFRS had these positive effects, stakeholders, as well as firms, would enjoy various benefits by applying IFRS compared with applying local accounting standards (Korean GAAP). And one of the most important benefits could be the increase of value relevance of accounting information (hereafter, value relevance) due to improving usefulness and transparency of accounting numbers.
However, prior studies that investigated the effect of IFRS adoption on value relevance addressed mixed empirical results. For example, Suadiye (2012), Daske et al. (2008) and Barth et al. (2008) reported that IFRS adoption had positive effects on value relevance, while Clarkson et al. (2011), Horton and Serafeim (2010), Hung and Subramanyam (2007) presented mixed effects or found no clear evidence of IFRS adoption on value relevance. Relevant researches in Korea also reported mixed empirical results depending on their research methodology and research periods. Ji (2013) found positive effect of IFRS on value relevance, while Choi (2013) and Choi et al. (2013) found little evidence on improving value relevance. Y. Kim and K. Kim (2015) and Park (2016) also reported insignificant effect and positive significant effect of IFRS on value relevance in different analysis period.
These results can be attributed to a possibility that firms' financial status and performance might not be properly reflected in the accounting information under IFRS, because it's too complex to apply and allows managers' discretion on accounting choices more than local GAAP. For example, standard for financial instruments (IAS 39) had been replaced with IFRS 9, since it is too complex and obscure to interpret and apply. Besides, the comparability and consistency of accounting data under IFRS may be lowered when managers selectively apply fair value accounting. So, it is empirical question if value relevance has improved after IFRS adoption.
Recently, Kwon et al. (2017) analyzed various accounting qualities including value relevance and presented improvement of value relevance after IFRS adoption. However, the result might be influenced not only by adoption of new accounting standards (IFRS) but also by corresponding efforts to higher demand for better information environment (Kwon et al., 2017). So, it is inferred that IFRS adoption would have both positive and negative effect on value relevance depending on how appropriately it is applied. IFRS is principle-based accounting standard in which managements' accounting decision plays more important role than Korean GAAP (K-GAAP), which is rule-based standard. This can lead to either more managers' discretionary accounting choice (Ahmed et  The empirical results show that value relevance of total listed firms (total samples) in Korea had decreased after IFRS adoption. However, separating total sample firms into two group according to their listed market, value relevance of KSE-listed firms has weakened after IFRS adoption, whereas value relevance of KOSDAQ-listed firms has improved. Further, consolidated accounting information have more positive effects on value relevance for KOSDAQ listed firms and less negative effects for KSE listed firms. These empirical results imply that the IFRS adoption resulted in both positive and negative impacts on value relevance depending on information environment level, which in turn indicate improvement of the comparability of accounting information between two stock markets in Korea.
This study contributes that it provides empirical evidences that different level of demand for information environment can causes different IFRS adoption effect in terms of value relevance. Moreover, this study investigated relatively long-term effect of IFRS adoption itself, as well as the changing effect of main financial statements on value relevance in Korea. Therefore, results of this study provide further understanding on effects of IFRS adoption to governmental organizations and financial regulators and academics.  (2010) investigated IFRS adoption case of UK firms by testing the relationship between net income adjustments information from local GAAP to IFRS and price reaction in the stock market. The study found that positive adjustment from local GAAP net income to IFRS net income resulted in significant market response before the disclosure, as well as after the disclosure, but negative adjustment brought about significant market response only after the disclosure. According to these findings, Horton and Serafeim (2010) insisted that only negative adjustment served as a new information for investors.

Clarkson et al. (2011) analyzed the effects of IFRS adoption in fourteen Australian and European
countries and showed that value relevance did not significantly change after IFRS adoption. Specifically, using the traditional linear regression model, the value relevance of equity and net income increased after IFRS adoption for the code law countries, while value relevance of the common law countries decreased after IFRS adoption. However, using non-linear regression model, there was no significant change in value relevance of accounting information after IFRS adoption. Kargin (2013) reported that the value relevance of equity was improved after IFRS adoption, whereas the value relevance of net income was not significantly changed after IFRS adoption, consistent with Hung (1) and Equation (2) on pre-IFRS period and post-IFRS period, respectively, and compare adjusted R-squares from the estimated models for each period 5 . Especially, this study uses data from both individual and consolidated financial statements data for two reasons. First, it allows to find whether changes in value relevance are incurred from the change of main financial statements. Second, it enables to find 4 This study establishes null hypothesis due to the different results of related prior studies. 5 We additionally applied non-linear model following Clarkson et al. (2011), which included interaction term between EPS and BPS into the model and the results were similar among the models. We only present results of linear model analysis, because results of non-linear model did not indicate biased results from significant heteroscedasticity. Instead, we present heteroscedasticity-consistent t-value in results of the main analysis.
whether changes in value relevance are consistent regardless of financial statement types.
According to Ohlson (1995), the price of firm's equity is determined by earnings, book value of equity, and other information. The higher value relevance of accounting information, the lower importance of other information. As a result, explanatory power (adjusted R-square) of the regression model increases. Therefore, following our null hypothesis, adjusted R-squares from the research models should not be different between pre-IFRS period and post-IFRS period.    Table 3 and Table 4 provide whether the value relevance of accounting information had changed after IFRS adoption in Korea. Table 3 and Table 4 present test results of Equation (1) and Equation (2), respectively.

Results of hypothesis test
In Table 3   Notes: *, *** represent significance at 10 and 1 percent levels, respectively. The t-values were calculated by heteroscedasticityconsistent standard errors. Chow presents Chow's test for structural breakpoints of 6341 (total sample), 3,535 (KOSDAQ), 2,806 (KSE), respectively. Ind. and Con. denote individual and consolidated financial statements, respectively.
listed firms had strengthened after IFRS adoption, but for KSE listed firms, the value relevance had weakened after IFRS adoption. Finally, Table 5 which presents annual time-series trend of adjusted R 2 provides more detailed evidence on those results. In Table 5, adjusted R 2 s from Equation (1) and Equation (2)  Additionally, adjusted R 2 s from consolidated financial statements show similar pattern of mitigated gap between two listed market subsamples. However, it is interesting that the gaps of adjusted R 2 s between individual and consolidated financial statements during IFRS period are inconsistent for KODAQ listed firms but con- Combining results of Table 3, Table 4 and

CONCLUSION
Prior studies which examined effect of IFRS adoption on the value relevance of accounting information provided mixed results. This may, at least in part, come from that prior researches used relatively shortterm research periods or didn't consider information environment levels. Thus, this study investigated the effect of IFRS adoption on the value relevance in Korea complementing limitations of prior studies. Specifically, this study analyzed change of the value relevance by comparing the explanatory power of research models estimated for 5 year periods before and after IFRS adoption, respectively. Further, we divided total sample firms into two groups (KSE and KOSDAQ listed firms) and compared the value relevance using both individual and consolidated accounting information. We performed the robustness test utilizing industry-fixed effect model as well.
Empirical results show that the value relevance of accounting information of listed firms in Korea decreased after the introduction of IFRS. However, we found that fall-off of total value relevance is largely attributed to KSE listed firms because the value relevance of KOSDAQ listed firms increased after the introduction of IFRS, while the value relevance of KSE listed firms decreased. Further, we found different effect of IFRS adoption in line with value relevance depending on types of financial statements. Consolidated accounting information of KSE firms showed more value relevance than individual accounting information, but lower value relevance than pre-IFRS period. Gaps of value relevance between two accounting information for KOSDAQ firms were inconsistent and showed lower value relevance than pre-IFRS period. These differentiated effects of IFRS adoption on value relevance may indicate that there is enhancement of comparability of accounting information under IFRS.
This study has contributions that it provides further understanding on the relationship between IFRS adoption and value relevance. Especially, this study expanded prior studies on the value relevance by expanding research periods and comparing changes in the value relevance by listing markets. Also, this study contributes that it specifically analyzed whether the changes of value relevance after IFRS adoption were due to the introduction of IFRS itself or the change of reported financial statements in case of South Korea.