The Impact of IFRS on Value Relevance of Accounting Information: Evidence from the Indonesian Stock Exchange

This study investigates the relationships between several accounting variables, International Financial Reporting Standards (IFRS) adoption, and stock market prices in Indonesia. The variables of interest include lagged price, book value per share (BVPS), earnings per share (EPS), market capitalization, Revenue, and price-to-earnings (P/E) ratio. We apply multiple regression analysis to examine the influences of these factors on stock prices. Our preliminary findings suggest that EPS and BVPS have a significant positive association with market prices, aligning with existing literature and highlighting the importance of these measures for investors. Additionally, our results indicate that IFRS adoption improves the value relevance of accounting information in the Indonesian market. We also explore potential size-related variations in the impact of IFRS adoption on the value relevance of accounting information. This study contributes to the ongoing debate on the effectiveness of IFRS and provides insights to investors, policymakers, and practitioners about the factors influencing stock prices in Indonesia.


INTRODUCTION
Accounting, as the "language of business," is integral to the functioning of every institution.
It serves as a fundamental tool for financial accountability, decision-making, and control.Over time, as economies have globalized and businesses have expanded beyond national boundaries, the need for a common language in accounting has become increasingly critical.
The International Financial Reporting Standards (IFRS), as a global set of accounting standards, are designed to meet this need.Through providing a "common accounting language," IFRS enhances the comparability and transparency of financial statements across different jurisdictions, facilitating international investment and economic growth.Moreover, the adoption of IFRS promotes good governance by enhancing the quality and reliability of financial reporting.
The application of a standardized accounting framework can contribute to increased accountability and transparency, which are key pillars of good governance in any institution, whether public or private.This has implications not just for corporations but also for other institutions including governments and non-profit organizations.Good governance is not only critical at a macro level, such as for institutions or countries, but it is also crucial at a micro level, such as for individual corporations, non-profit organizations, and even smaller parts of economic activity.Good governance involves various factors such as accountability, transparency, responsiveness, rule of law, effectiveness, efficiency, participatory and consensus-oriented behavior (World Bank, 1992).
One of the ways in which good governance manifests at the organizational or individual level is through the practice of sound financial management and reporting.Accounting and financial reporting play a vital role in achieving good governance.They provide crucial information to stakeholders, aiding them in decision-making and holding the management accountable for their actions (Eccles and Youmans, 2016).
The adoption of IFRS enhances the quality and comparability of financial reporting across organizations, which in turn contributes to good governance.For instance, Ball (2006) argues that high-quality accounting standards like IFRS can lead to higher-quality financial reporting.This improves transparency, makes the organization's actions and performance clearer to stakeholders, and increases the organization's accountability.Moreover, the adoption of IFRS may also facilitate more efficient and effective resource allocation within organizations.As Armstrong et al. (2010) posit, the improved comparability of financial information following IFRS adoption could enhance the efficiency of capital allocation, contributing to more effective decision-making at both the organizational and individual levels.However, the effects of IFRS adoption are complex and can vary significantly across different contexts.The impact of IFRS on the value relevance of accounting information -that is, the usefulness of accounting information in reflecting a company's true economic value -is one such area of divergent findings.
In a series of studies conducted in different countries, the implications of IFRS adoption for value relevance have been extensively explored.For example, research conducted in Iraq by

| 65 |
The Impact of IFRS on Value Relevance of Accounting Information: Evidence from the Indonesian Stock Exchange Imanuel Wahyu Christanto, Fuad Fuad Salman (2013) and in Nigeria by Umoren & Enang (2015) found a significant improvement in the value relevance of earnings and book value after IFRS adoption.Similarly, Perera & Thrikawala (2012) conducted a study in Sri Lanka and concluded that the adoption of IFRS enhanced the quality of financial reporting and hence the value relevance of financial statement items.However, these findings are not universal.A study conducted in India by Kumar and Visvanathan (2013) showed a decline in the value relevance of accounting information following IFRS adoption.
This underscores the potential influence of other factors, such as institutional and country-level characteristics, in shaping the impact of IFRS adoption.These varied findings suggest that the relationship between IFRS adoption and value relevance is complex and context-dependent.This underscores the need for further investigation into this relationship in different national contexts.
This research aims to contribute to this line of inquiry by examining the impact of IFRS adoption on the value relevance of accounting information in the context of Indonesia, a country with its unique institutional environment and challenges in financial reporting.
The IDX serves as a critical nexus of financial and economic activity, and its influence extends well into the broader governance landscape of Indonesia (Susanto, 2017).The enhanced accountability and transparency fostered by IFRS have far-reaching implications, not just for firms listed on the IDX, but for other sectors and entities within the economy (Zeghal & Mhedhbi, 2006;Gordon et al., 2012).For instance, the increased transparency and accountability exhibited by IDX-listed firms following the adoption of IFRS serve as a model for other entities, potentially catalyzing widespread improvements in governance practices (Choi & Meek, 2008).This ripple effect can enhance public trust (Daske et al., 2008), promote more efficient allocation of resources (Leuz & Verrecchia, 2000), and facilitate sustainable economic growth (Ramanna & Sletten, 2014).
Therefore, studying the impact of IFRS adoption within the IDX context can provide valuable insights into its broader implications for governance, accountability, and the value relevance of accounting information in Indonesia.
However, the actual impact of IFRS adoption on the value relevance of accounting information in the IDX context remains empirically underexplored, thus providing the motivation for the current study.By focusing on firms listed on the IDX, this study aims to shed light on the broader implications of IFRS adoption for good governance, accountability, and the value relevance of accounting information in Indonesia.

EPS Relationship to Market Prices
The Value Relevance Theory posits that accounting information such as EPS significantly influences market prices (Ohlson, 1995).Under the umbrella of this theory, Liu et al. (2019) conducted a quantitative study examining Chinese listed firms.They used regression analysis with EPS as the independent variable and market prices as the dependent variable.The study provided empirical evidence supporting the positive association between a firm's EPS and its market value (Liu, Yao, Hu, & Liu, 2019).Concurrently, the Institutional Theory, which acknowledges the role of formal rules in shaping firm behavior, also underlines the importance of EPS as financial metrics directly impact market perception (Scott, 2008).
H 1 : EPS is positively associated with market prices in Indonesia BVPS Relationship to Market Prices Ohlson's Residual Income Valuation Model (1995) posits that BVPS, an intrinsic value measure, significantly impacts market prices.Frankel and Lee (2018) used this model in their empirical research.Using a large sample of US listed firms and multiple regression analyses, they found that BVPS was a significant determinant of market prices.Simultaneously, according to the Signalling Theory, firms with higher BVPS might be signalling strong financial health, potentially driving up their market prices (Spence, 2002).
H 2 : BVPS is positively associated with market prices in Indonesia

IFRS Adoption Enhances the Value Relevance of Accounting Information
IFRS adoption, under the Institutional Theory, is posited to enhance the quality and value relevance of financial reporting.Barth et al. (2017) tested this in a large cross-country study involving firms from 26 countries that adopted IFRS.Using panel data regression analysis, they found that the quality and value relevance of accounting information improved post-IFRS adoption.Nevertheless, the effects of such regulatory compliance could differ due to variances in enforcement and interpretation across jurisdictions (Impact of Regulation and Compliance).
H 3 : IFRS adoption enhances the value relevance of accounting information for companies listed on the

Market Capitalization Relationship to Market Price
According to the Efficient Market Hypothesis, all publicly available information, including a firm's Market Cap, is reflected in its stock prices (Fama, 1970).Fama & French (2015), in a comprehensive empirical study, analyzed the relationship between firm's Market Cap and its stock prices using multiple regression analyses, confirming the positive association.In terms of Behavioral Finance, investors might perceive firms with larger Market Cap as less risky, driving their stock prices up.

Lagged Price Relationship to Market Prices
The Signalling Theory and Engle's ARCH model (1982) both suggest that past market prices can inform about future performance.Fan et al. (2021) empirically applied this model in a timeseries analysis of Chinese stock market data, which supported the predictive power of lagged prices.
H 5 : Lagged price is positively associated with current market prices in Indonesia.

Revenue Relationship to Market Prices
The Revenue Recognition Principle is fundamental to the Value Relevance Theory.Landsman & Peasnell (2020) tested this in a panel data study on US firms, confirming that recognized revenue significantly impacts a company's stock price.
H 6 : Revenue is positively associated with market prices in Indonesia Price-to-earnings ratio (P/E) Relationship to Market Prices Gordon's Dividend Discount Model (1962) suggests that P/E ratio can predict a company's future dividends, which in turn, can drive its market price.Damodaran (2012) empirically confirmed this in a cross-sectional study of US firms.
H 7 : P/E ratio is positively associated with market prices in Indonesia

METHOD, DATA, AND ANALYSIS Theoretical Basis and Empirical Model
Our model's development (Figure 1) is informed by relevant existing research in our field.
Central to this is Ohlson's (1995) linear valuation model, which posits that a company's market value is primarily influenced by its earnings and book value -two fundamental elements of accounting information.This provides us with a basic framework for our empirical model.
Further insight is derived from the work of Barth, Landsman, and Lang (2008), who investigated the relationship between International Financial Reporting Standards (IFRS) adoption and accounting quality.Their findings suggest that companies that apply IFRS demonstrate higher accounting quality compared to those that use non-US local standards, a consideration integral to our model.
Our model also builds on the pioneering empirical model by Ball and Brown (1968), which utilized accounting income figures to predict stock returns.Their model, grounded in the efficient market hypothesis, asserts that current earnings are a vital piece of information incorporated into a firm's stock price.

Jurnal Akuntansi & Perpajakan
Vol. 9 (1) 2023: 63-81 influenced by its earnings and book value -two fundamental elements of accounting information.This provides us with a basic framework for our empirical model.
Further insight is derived from the work of Barth, Landsman, and Lang (2008), who investigated the relationship between International Financial Reporting Standards (IFRS) adoption and accounting quality.Their findings suggest that companies that apply IFRS demonstrate higher accounting quality compared to those that use non-US local standards, a consideration integral to our model.
Our model also builds on the pioneering empirical model by Ball and Brown (1968), which utilized accounting income figures to predict stock returns.Their model, grounded in the efficient market hypothesis, asserts that current earnings are a vital piece of information incorporated into a firm's stock price.

Figure 1.
Illustrates the theoretical framework of our pre-IFRS and post-IFRS models.

Model Specification
The proposed model investigates the relationship between accounting information (EPS and BVPS), IFRS adoption, and market prices.It also explores the influence of firm characteristics, such as company size (measured by market cap), on these relationships.The model can be specified as: In the first model, the dependent variable, 'Price', is regressed on three independent variables: Earnings per Share (EPS), Book Value per Share (BVPS), and Market Capitalization (Market_Cap).The objective of this model is to explore the relationships between a company's stock price and its profitability (EPS), intrinsic value (BVPS), and size (Market_Cap).In summary, our proposed empirical model for this research integrates these fundamental theories and aims to evaluate the value relevance of accounting information and the effects of IFRS adoption.

Incorporation of the Difference-in-Differences (DiD) Method
To add depth to our analysis, we incorporate the Difference-in-Differences (DiD) method.This statistical technique measures the effect of a 'treatment' (the adoption of IFRS, in our case) on an 'outcome' (the price return), compared to a control group that did not receive the treatment (Wooldridge, 2002; Illustrates the theoretical framework of our pre-IFRS and post-IFRS models.

Model Specification
The proposed model investigates the relationship between accounting information (EPS and BVPS), IFRS adoption, and market prices.It also explores the influence of firm characteristics, such as company size (measured by market cap), on these relationships.The model can be specified as: vides us with a basic framework for our empirical model.Further insight is derived from the work of Barth, Landsman, and Lang (2008), who investigated relationship between International Financial Reporting Standards (IFRS) adoption and accounting lity.Their findings suggest that companies that apply IFRS demonstrate higher accounting quality pared to those that use non-US local standards, a consideration integral to our model.
Our model also builds on the pioneering empirical model by Ball and Brown (1968), which utilized ounting income figures to predict stock returns.Their model, grounded in the efficient market othesis, asserts that current earnings are a vital piece of information incorporated into a firm's stock e.

Figure 1.
Illustrates the theoretical framework of our pre-IFRS and post-IFRS models.

del Specification
The proposed model investigates the relationship between accounting information (EPS and BVPS), S adoption, and market prices.It also explores the influence of firm characteristics, such as company (measured by market cap), on these relationships.The model can be specified as: In the first model, the dependent variable, 'Price', is regressed on three independent variables: nings per Share (EPS), Book Value per Share (BVPS), and Market Capitalization (Market_Cap).The ective of this model is to explore the relationships between a company's stock price and its profitability S), intrinsic value (BVPS), and size (Market_Cap).In summary, our proposed empirical model for this arch integrates these fundamental theories and aims to evaluate the value relevance of accounting rmation and the effects of IFRS adoption.

orporation of the Difference-in-Differences (DiD) Method
To add depth to our analysis, we incorporate the Difference-in-Differences (DiD) method.This istical technique measures the effect of a 'treatment' (the adoption of IFRS, in our case) on an 'outcome' price return), compared to a control group that did not receive the treatment (Wooldridge, 2002;

EPS BVPS Market Cap Price
In the first model, the dependent variable, 'Price', is regressed on three independent variables: Earnings per Share (EPS), Book Value per Share (BVPS), and Market Capitalization (Market_Cap).
The objective of this model is to explore the relationships between a company's stock price and its profitability (EPS), intrinsic value (BVPS), and size (Market_Cap).In summary, our proposed empirical model for this research integrates these fundamental theories and aims to evaluate the value relevance of accounting information and the effects of IFRS adoption.

Incorporation of the Difference-in-Differences (DiD) Method
To add depth to our analysis, we incorporate the Difference-in-Differences (DiD) method.
This statistical technique measures the effect of a 'treatment' (the adoption of IFRS, in our case) on an 'outcome' (the price return), compared to a control group that did not receive the treatment (Wooldridge, 2002;Angrist & Pischke, 2008).In addition to EPS and BVPS, this model introduces the Lagged Price (price of the stock in the previous period), the dummy variable IFRS (representing the period of International Financial Reporting Standards adoption), revenue, company size, and the price-to-earnings ratio (P/E).Here, b_1 to a_7 represent the respective changes in 'Price' for a one-unit increase in each of these variables, while other factors are held constant.
This model aims to examine the relationship between these additional factors and the stock Financial Reporting Standards adoption), revenue, company size, and the price-to-earnings ratio (P/E).
Here, β_1 to β_7 represent the respective changes in 'Price' for a one-unit increase in each of these variables, while other factors are held constant.This model aims to examine the relationship between these additional factors and the stock price.It allows you to test the impact of IFRS adoption on the stock price and understand the value relevance of the chosen accounting variables.The second model incorporates more factors into the analysis, including some that represent the temporal characteristics of the stock market and accounting standards.The model is as follows.The model specification for the DiD approach is as follows: Presents the theoretical framework of our DiD model.

Advanced Computational Approach
With the advent of advanced computational capabilities and data-driven insights, our research adopts a comprehensive approach that amalgamates linear, non-parametric, and machine learning models (Brynjolfsson, Hitt & Kim, 2011).This approach enables us to fully utilize our data and holistically evaluate the influence of IFRS adoption on the value relevance of accounting information.

Sample Selection and Data Collection
Our study utilizes a stratified random sampling of companies listed on the Indonesian Stock Exchange (IDX) between 2006 and 2020.This technique ensures diverse representation across different industries within the IDX, allowing us to analyze whether the impact of IFRS adoption varies by industry.Companies selected for the sample consistently report their financial information on the IDX within the given timeframe and provide key data such as earnings per share (EPS), book value per share (BVPS), price, market capitalization, Revenue, and P/E ratio.
The secondary data for our study, sourced from companies' annual reports through the Bloomberg terminal, are collected using the documentation method.This involves gathering relevant information from various sources like annual reports and other financial documents.Initially, 152 companies were selected that consistently published their financial reports during this timeframe, ensuring a comprehensive and reliable data set.The pool was further refined by eliminating outliers, which could compromise the accuracy of the model.Subsequent statistical tests were conducted to ensure compliance with classical assumptions of regression analysis-linearity, independence, homoscedasticity, and normality.This step aimed at further ensuring the validity of our results.Finally, a cluster analysis was performed to identify the most representative sample, enhancing the interpretability and applicability of our findings.This rigorous process resulted in a final sample of 97companies, establishing a robust foundation for our study.
In the data preprocessing stage, we begin by centering our variables: 'EPS', 'BVPS', 'Market Cap', 'Revenue', and 'P/E'.This reduces multicollinearity and improves the interpretability of our model coefficients (Schielzeth, 2010).Next, all data are shifted to be strictly positive, a necessary step for variables with negative or zero values.Lastly, a Box-Cox transformation is applied to align the data more closely with the normal distribution (Box & Cox, 1964).This transformation aids in fulfilling the assumptions of homoscedasticity and normality for subsequent analyses (Osborne, 2010).These preprocessing steps aim to establish a robust foundation for our analysis, enhancing the validity of our findings.Financial Reporting Standards adoption), revenue, company size, and the price-to-earnings ratio (P/E).

Results
Here, β_1 to β_7 represent the respective changes in 'Price' for a one-unit increase in each of these variables, while other factors are held constant.
This model aims to examine the relationship between these additional factors and the stock price.It allows you to test the impact of IFRS adoption on the stock price and understand the value relevance of the chosen accounting variables.The second model incorporates more factors into the analysis, including some that represent the temporal characteristics of the stock market and accounting standards.The model is as follows.The model specification for the DiD approach is as follows: Presents the theoretical framework of our DiD model.

Advanced Computational Approach
With the advent of advanced computational capabilities and data-driven insights, our research adopts a comprehensive approach that amalgamates linear, non-parametric, and machine learning models (Brynjolfsson, Hitt & Kim, 2011).This approach enables us to fully utilize our data and holistically evaluate the influence of IFRS adoption on the value relevance of accounting information.

Sample Selection and Data Collection
Our study utilizes a stratified random sampling of companies listed on the Indonesian Stock Exchange (IDX) between 2006 and 2020.This technique ensures diverse representation across different industries within the IDX, allowing us to analyze whether the impact of IFRS adoption varies by industry.Companies selected for the sample consistently report their financial information on the IDX within the given timeframe and provide key data such as earnings per share (EPS), book value per share (BVPS), price, market capitalization, Revenue, and P/E ratio.
The secondary data for our study, sourced from companies' annual reports through the Bloomberg terminal, are collected using the documentation method.This involves gathering relevant information from various sources like annual reports and other financial documents.Initially, 152 companies were selected that consistently published their financial reports during this timeframe, ensuring a comprehensive and reliable data set.The pool was further refined by eliminating outliers, which could compromise the accuracy of the model.Subsequent statistical tests were conducted to ensure compliance with classical assumptions of regression analysis-linearity, independence, homoscedasticity, and normality.This step aimed at further ensuring the validity of our results.Finally, a cluster analysis was performed to identify the most representative sample, enhancing the interpretability and applicability of our findings.This rigorous process resulted in a final sample of 97companies, establishing a robust foundation for our study.
In the data preprocessing stage, we begin by centering our variables: 'EPS', 'BVPS', 'Market Cap', 'Revenue', and 'P/E'.This reduces multicollinearity and improves the interpretability of our model coefficients (Schielzeth, 2010).Next, all data are shifted to be strictly positive, a necessary step for variables with negative or zero values.Lastly, a Box-Cox transformation is applied to align the data more closely with the normal distribution (Box & Cox, 1964).This transformation aids in fulfilling the assumptions of homoscedasticity and normality for subsequent analyses (Osborne, 2010).These preprocessing steps aim to establish a robust foundation for our analysis, enhancing the validity of our findings.
the stock in the previous period), the dummy variable IFRS (representing the period of International Financial Reporting Standards adoption), revenue, company size, and the price-to-earnings ratio (P/E).
Here, β_1 to β_7 represent the respective changes in 'Price' for a one-unit increase in each of these variables, while other factors are held constant.
This model aims to examine the relationship between these additional factors and the stock price.It allows you to test the impact of IFRS adoption on the stock price and understand the value relevance of the chosen accounting variables.The second model incorporates more factors into the analysis, including some that represent the temporal characteristics of the stock market and accounting standards.The model is as follows.The model specification for the DiD approach is as follows: Presents the theoretical framework of our DiD model.

Advanced Computational Approach
With the advent of advanced computational capabilities and data-driven insights, our research adopts a comprehensive approach that amalgamates linear, non-parametric, and machine learning models (Brynjolfsson, Hitt & Kim, 2011).This approach enables us to fully utilize our data and holistically evaluate the influence of IFRS adoption on the value relevance of accounting information.

Sample Selection and Data Collection
Our study utilizes a stratified random sampling of companies listed on the Indonesian Stock Exchange (IDX) between 2006 and 2020.This technique ensures diverse representation across different industries within the IDX, allowing us to analyze whether the impact of IFRS adoption varies by industry.Companies selected for the sample consistently report their financial information on the IDX within the given timeframe and provide key data such as earnings per share (EPS), book value per share (BVPS), price, market capitalization, Revenue, and P/E ratio.
The secondary data for our study, sourced from companies' annual reports through the Bloomberg terminal, are collected using the documentation method.This involves gathering relevant information from various sources like annual reports and other financial documents.Initially, 152 companies were selected that consistently published their financial reports during this timeframe, ensuring a comprehensive and reliable data set.The pool was further refined by eliminating outliers, which could compromise the accuracy of the model.Subsequent statistical tests were conducted to ensure compliance with classical assumptions of regression analysis-linearity, independence, homoscedasticity, and normality.This step aimed at further ensuring the validity of our results.Finally, a cluster analysis was performed to identify the most representative sample, enhancing the interpretability and applicability of our findings.This rigorous process resulted in a final sample of 97companies, establishing a robust foundation for our study.
In the data preprocessing stage, we begin by centering our variables: 'EPS', 'BVPS', 'Market Cap', 'Revenue', and 'P/E'.This reduces multicollinearity and improves the interpretability of our model coefficients (Schielzeth, 2010).Next, all data are shifted to be strictly positive, a necessary step for variables with negative or zero values.Lastly, a Box-Cox transformation is applied to align the data more closely with the normal distribution (Box & Cox, 1964).This transformation aids in fulfilling the assumptions of homoscedasticity and normality for subsequent analyses (Osborne, 2010).These preprocessing steps aim to establish a robust foundation for our analysis, enhancing the validity of our findings.

Advanced Computational Approach
With the advent of advanced computational capabilities and data-driven insights, our research adopts a comprehensive approach that amalgamates linear, non-parametric, and machine learning models (Brynjolfsson, Hitt & Kim, 2011).This approach enables us to fully utilize our data and holistically evaluate the influence of IFRS adoption on the value relevance of accounting information.

Sample Selection and Data Collection
Our study utilizes a stratified random sampling of companies listed on the Indonesian Stock Exchange (IDX) between 2006 and 2020.This technique ensures diverse representation across different industries within the IDX, allowing us to analyze whether the impact of IFRS adoption varies by industry.Companies selected for the sample consistently report their financial information on the IDX within the given timeframe and provide key data such as earnings per share (EPS), book value per share (BVPS), price, market capitalization, Revenue, and P/E ratio.
The secondary data for our study, sourced from companies' annual reports through the Bloomberg terminal, are collected using the documentation method.This involves gathering relevant information from various sources like annual reports and other financial documents.
Initially, 152 companies were selected that consistently published their financial reports during this timeframe, ensuring a comprehensive and reliable data set.The pool was further refined by eliminating outliers, which could compromise the accuracy of the model.Subsequent statistical tests were conducted to ensure compliance with classical assumptions of regression analysislinearity, independence, homoscedasticity, and normality.This step aimed at further ensuring the validity of our results.Finally, a cluster analysis was performed to identify the most representative sample, enhancing the interpretability and applicability of our findings.This rigorous process resulted in a final sample of 97companies, establishing a robust foundation for our study.
The descriptive statistics present a snapshot of the financial metrics for large corporations.On average, firms had a Book Value Per Share (BVPS) of 1.35 billion, although this varied significantly with a standard deviation of 2.29 billion, and a range from -1.27 billion to 26.47 billion, indicating a mix of firms with negative and positive equity.The Earnings per Share (EPS) showed a similar pattern, with a mean of 204 million and a broad spread, as the values ranged from -1.52 billion to 5.65 billion.
When it comes to stock prices, the firms under consideration showed a mean price of 3.4 billion.Yet, there was a high degree of dispersion, with prices as low as 40 million and as high as 98.4 billion.This trend of wide-ranging values also extended to revenues, which averaged around 11.2 trillion, but spanned from 22.4 million to a staggering 239 trillion.The Market Capitalization (Market Cap), a measure of a firm's size, averaged at 24.1 trillion.But, the standard deviation of 67.7 trillion and the range from 107 million to 684 trillion revealed significant differences in the size of firms within the sample.Lastly, the Price-to-Earnings (P/E) ratio, a common indicator of market valuation, averaged around 26.76 million across the firms.But, just like the other metrics, this ratio showed substantial variation, ranging from as low as 0.6 to as high as 1.26 billion.

| 71 |
The Impact of IFRS on Value Relevance of Accounting Information: Evidence from the Indonesian Stock Exchange Imanuel Wahyu Christanto, Fuad Fuad Table 2 presents statistical diagnostics for the pre-IFRS, post-IFRS, and Difference-in-Differences (DiD) models used in the study.The average residual, or unstandardized residual, is nearly the same for all models, indicating a similar average difference between actual and predicted values across the models.The Durbin-Watson statistic, which tests for autocorrelation in residuals, suggests no major autocorrelation problems in any of the models since all values are close to 2.
The independence tests (Lagrange Multiplier and F-test) show mixed results.For the pre-IFRS model, there is no evidence to reject the null hypothesis of independent error terms.However, for the post-IFRS and DiD models, the results suggest that the errors may not be independent, indicating potential autocorrelation issues.Homoscedasticity tests, which check if the variance of the errors is constant, suggest potential heteroscedasticity in the pre-IFRS model but not in the post-IFRS and DiD models.
Normality tests indicate non-normality of residuals in the pre-IFRS and post-IFRS models, while the DiD model's residuals appear to follow a normal distribution.The skewness and kurtosis values suggest that the distribution of residuals is right-skewed and "heavy-tailed" for the pre-IFRS and post-IFRS models, but approximately symmetrical and "light-tailed" for the DiD model.The adjusted R^2 score also increases across the periods, mirroring the behavior of the R^2 score.Adjusted R^2 accounts for the number of predictors in the model and can sometimes be a more accurate measure of the goodness of fit, especially when there are many predictors.
The Mean Squared Error (MSE) represents the average of the squares of the differences between actual and estimated values, essentially measuring prediction error.Interestingly, the MSE is highest in the post-IFRS period but decreases in the DiD period, suggesting the model's predictive accuracy improves over time.
Lastly, the Mean CV Score appears to refer to cross-validation score, a resampling procedure used to evaluate machine learning models on a limited data sample.Lower values are desirable as they indicate a model that generalizes well.The value increases from the pre-IFRS to post-IFRS periods but decreases in the DiD period.
In summary, this table suggests that the Random Forest model performs well across all three periods, with improving performance from the pre-IFRS period to the DiD period.These results support the efficacy of the Random Forest method in capturing complex patterns in this financial dataset and the potential benefits of the IFRS adoption.The pre-IFRS model tested the influence of BVPS, EPS, and Market Cap on a dependent variable (not explicitly mentioned).The degrees of freedom (df) is 1 for each variable, suggesting one category was examined for each.The F statistic, which is the ratio of the variance between groups to the variance within groups, indicates that BVPS (F=51.785,p<0.00000000198) and EPS (F=14.826,p=0.000314) significantly predict the dependent variable.However, Market Cap (F=1.508,p=0.225) does not have a statistically significant relationship.
In the post-IFRS model, again BVPS (F=42.164,p<0.000000003697) and EPS (F=29.210,p=0.0000004726) significantly predict the dependent variable.Similar to the pre-IFRS model, Market Cap (F=1.454,p=0.231) does not show a significant relationship.
In the DiD model, Lagged_Price (F=6.920,p=0.027),EPS (F=27.375,p=0.001), and Market Cap (F=8.576,p=0.017) are statistically significant predictors.The BVPS, Revenue, and P/E have p-values near or above 0.05, indicating a potential relationship that may not be strong or consistent enough to be considered statistically significant under conventional thresholds.
In summary, the variables BVPS and EPS are consistent, significant predictors across both the pre-and post-IFRS models.In the DiD model, Lagged_Price, EPS, and Market Cap are significant, while other factors show less significant relationships.This could suggest that the influence of certain factors on the dependent variable may have changed after the implementation of IFRS.[4,479.984, 25,500]

H 4 :
Market cap is positively associated with market prices in Indonesia | 67 | The Impact of IFRS on Value Relevance of Accounting Information: Evidence from the Indonesian Stock Exchange Imanuel Wahyu Christanto, Fuad Fuad price.It allows you to test the impact of IFRS adoption on the stock price and understand the value relevance of the chosen accounting variables.The second model incorporates more factors into the analysis, including some that represent the temporal characteristics of the stock market and accounting standards.The model is as follows.The model specification for the DiD approach is as follows: | 69 | The Impact of IFRS on Value Relevance of Accounting Information: Evidence from the Indonesian Stock Exchange Imanuel Wahyu Christanto, Fuad Fuad

Figure 2 .
Figure 2. Presents the theoretical framework of our DiD model.

Table 3 .
Multicollinearity Test rule of thumb is that if VIF is 1 then there is no multicollinearity, if VIF is between 1 and 5 there is moderate multicollinearity, and if VIF is above 5 then there is high multicollinearity.This analysis indicates that multicollinearity is present to a moderate degree for most variables in the regression models.This isn't necessarily problematic but could affect the reliability of some of the findings if not addressed A

Table 4 .
Spearman Rank TestIn the post-IFRS period, correlations between the market price and EPS, BVPS, and Market Cap are slightly lower but remain significant for EPS and BVPS, with p-values effectively at zero.
In summary, these results provide empirical evidence supporting the theory of value relevance of accounting information, with EPS and BVPS showing significant correlations with |

73 | The Impact of IFRS on Value Relevance of Accounting Information: Evidence from the Indonesian Stock Exchange Imanuel
Wahyu Christanto, Fuad Fuad market price across all periods.The effect of IFRS adoption also manifests in the correlations, particularly visible in the DiD period, which adds a new dimension to the analysis.

Table 5 .
Random Forest TestThe R^2 score, or coefficient of determination, indicates the proportion of the variance in the dependent variable that is predictable from the independent variables.Here, all three periods show high R^2 scores, suggesting that the Random Forest model is highly predictive of the dependent variable.The R^2 scores increase over the periods, with the highest being in the DiD period at 0.9894498209130984, indicating nearly 99% of the variance in the dependent variable is predictable by the model, a very strong result.

Table 8 .
Statistical t-test