An international study of the relation between book-tax conformity and the value relevance of earnings components.

Yoon, Sung Wook

INTRODUCTION

International factors in accounting information are becoming more important due to growth in international capital markets (Meek and Thomas, 2003). As the economies of different countries have become more closely linked, factors that affect a firm's performance in one country may impact the economies of other countries. In today's globalized environment, international accounting research can help us understand how current financial accounting data relate to future performance and overall firm value. Prior research (Alford et al. 1993, Joos and Lang 1994, Guenther and Young 2000, Ali and Hwang 2000, Ball et al. 2000, Herrmann et al. 2000, Hung 2001, Young and Guenther 2003, Ball et al. 2003, Leuz et al. 2003, and Haw et al. 2004) suggests that the value relevance of a country's financial reporting may be affected by various accounting practices and tax rules because each country has different political, social, and economic systems.

Since the primary objective of tax rules is not to satisfy the information needs of capital market participants, the value relevance of financial reports in countries with high tax-book conformity may be compromised (Ali and Hwang, 2000). Additionally, because firms have competing incentives between tax planning and financial reporting, accounting information in countries with a strong link between tax and financial accounting may be less value relevant. As Guenther and Young (2000, p. 58) state, "firms have strong incentives to reduce taxes, and if financial and tax accounting must conform, tax considerations may dominate other considerations. Hence, financial accounting information may be less likely to reflect underlying economic events when firms attempt to minimize taxes." Ali and Hwang (2000, p. 2) argue that, "requiring book-tax conformity provides incentives to reduce taxes by reporting systemically lower profits, thereby undermining the value relevance of financial reporting." If financial statements are to provide a true and fair view of performance, financial accounting should not be influenced by tax considerations. Hence, if a country's financial accounting principles are strongly aligned with tax rules, its financial accounting information is likely to be less value relevant.

Many prior studies (e.g., Alford et al. 1993; Ali and Hwang 2000; Guenther and Young 2000; Young and Guenther 2003) have found that the degree of book-tax conformity is negatively associated with the value relevance of a firm's earnings for financial reporting. However, Hung (2001) does not find evidence to support a relation between book-tax conformity and value relevance of earnings, which is contrary to the findings of Ali and Hwang (2000) and Young and Guenther (2003). Given these contrary findings, the relation between book-tax conformity and the value relevance of earnings remains unresolved. This study adds to existing literature by providing evidence regarding the relation between book-tax conformity and the value relevance of earnings across countries. It is not clear why book-tax conformity is related to value relevance of accounting earnings. The reason may be either that high book-tax conformity causes earnings to be less value relevant or that high book-tax conformity is related to an information and regulation environment that results in limited association between accounting information and stock price. One way to address this issue is to examine the value relevance of earnings components. A Special Report issued by the Financial Accounting Standard Board (FASB) suggests that disaggregating earnings components would be a useful format for reporting income (FASB, 1998, para. 4.4). Prior research (e.g., Lipe 1986; Ohlson and Penman 1992; Lev and Thiagarajan 1993; Abarbanell and Bushee 1997) also indicates that earnings components contain useful information in the U.S. market; however, to-date there has been little to no investigation of the relation between the value relevance of earnings components and country-specific factors.

Hence, this study extends prior research by examining the relation between the value relevance of earnings components and country-specific factors, especially focusing on book-tax conformity. The results of my paper provide evidence regarding how book-tax conformity is related to the value relevance of accounting earnings. I examine whether earnings components in countries with high book-tax conformity are valued differently than in countries with low book-tax conformity, focusing on changes in three earnings components: sales, depreciation, and tax expense. I evaluate the value relevance of earnings and earnings components by coefficients and explanatory power in return-earnings regressions for each country. I examine whether there are differences in value relevance of earnings and earnings components between counties with high book-tax conformity versus countries with low book-tax conformity. Prior studies focus on relations between the value relevance of earnings and book-tax conformity using country-level tests. My study is based on a firm-level test and is not a replication of previous studies. The remainder of this paper is organized as follows: Section 2 develops the hypotheses; Section 3 describes data and samples; Section 4 discusses the regression models used to test the hypotheses; Section 5 presents the results of empirical tests and Section 6 offers conclusion.

HYPHOTHESES

Book-tax conformity has been used in many previous international accounting studies to evaluate the value relevance of accounting numbers. Young and Guenther (2003, p. 556) argue that, "the financial tax conformity index may be a more comprehensive measure of the value relevance of financial reporting than the disclosure index, capturing features beyond disclosure practices such as accounting measurement differences and the external audit function." Some studies find that book-tax conformity is related to the value relevance of firms' accounting earnings. However, question still remains as to how book-tax conformity is related to the value relevance of accounting earnings. The answer may be either that book-tax conformity causes earnings to be less value relevant or that book-tax conformity is associated with an accounting environment that results in low quality earnings, such as each country's legal origin. Prior research (e.g., Lipe 1986; Lev and Thiagarajan 1993; Abarbanell and Bushee 1997) finds that earnings components have incremental information content relative to that contained in earnings. Hence, I examine the relation between the value relevance of earnings components and country-specific factors, especially focusing on book-tax conformity.

Similar to the value relevance of aggregated earnings, I expect the value relevance of firms' sales component to be lesser in countries with a high level of conformity between financial and tax accounting. In countries with high book-tax conformity, some portion of sales revenue is likely to be recognized on a cash basis in firms' books since tax rules in many countries still follow the cash basis instead of the accrual basis. Because accrual accounting is likely to provide better information that reflects underlying economic substance than cash based accounting, the information in sales components based on accrual accounting is likely to be more relevant. I examine whether the valuation implication of the sales component of earnings is lower in countries with high

book-tax conformity than in countries with low book-tax conformity. Hence, Hypothesis 1 is:

[H.sub.1]: Changes in the sales component of net income are valued more in countries with low book-tax conformity than in countries with high book-tax conformity.

I also investigate whether the depreciation component are more value relevant in countries with low book-tax conformity than in countries with high book-tax conformity. Firms are generally required to use depreciation methods and useful lives as dictated by each country's tax authorities. Thus, tax depreciation leaves less discretion to the company because the tax rules are generally intended to lead to more uniform application of the tax system. However, because tax rules do not necessarily represent underlying economic events, depreciation methods and useful lives prescribed by the country's tax authorities are less likely to reflect various firms' activities related to the use of depreciable assets. Therefore, I expect information in the depreciation component to be less value relevant in countries with high book-tax conformity. My second hypothesis is:

[H.sub.2]: Changes in the depreciation component of net income are valued more in countries with low book-tax conformity than in countries with high book-tax conformity.

If accounting earnings are more value-relevant in countries with low book-tax conformity than in those with high book-tax conformity, tax expense, which is based on accounting earnings, should also be more value-relevant in countries with low book-tax conformity than in those with high book-tax conformity. Hence, another hypothesis is:

[H.sub.3a]: Changes in the tax component of earnings are valued more in countries with low book-tax conformity than in countries with high book-tax conformity.

However, when firms confront high effective tax rates, they face stronger incentives to minimize their taxable income. In general, firms with higher effective tax rates pay taxes at higher marginal tax rates and are more likely to attempt to reduce their tax expenses, as compared to firms with low effective tax rates. In countries with low book-tax conformity, firms with high effective tax rates are more likely to minimize taxable income while maximizing their reported income using accruals. For example, some of the largest and most profitable US firms did not pay taxes for several years in the 1990s due to excessive use of option-based compensation.

Moreover, as Dhaliwal et al. (2004) argue, tax expense can be used as a tool for earnings management. Under U.S. Generally Accepted Accounting Principles (GAAP), firms may be able to reduce tax expenses using permanent differences between financial and taxable income, income from foreign sources taxed at different rates, tax credits, and net operating loss carry forwards. U.S. firms can reduce tax expense when non-tax sources of earnings management are insufficient. Therefore, if a country's book-tax conformity is high, room for reducing tax expense will be lower and the value relevance of tax components of earnings may be higher. According to recent research by Schmidt (2006) that supports the results of Dhaliwal et al. (2004), an initial tax change component is more persistent for future tax changes than a revised tax change component. In addition, Guenther and Jones (2003) found that unexpected income resulting from a change in tax component is value relevant, but less value relevant than other unexpected earnings. I examine whether tax components of earnings are more value relevant in countries with high book-tax conformity than in countries with low book-tax conformity, when firms have high effective tax rates. My hypothesis is:

[H.sub.3b]: For firms with high effective tax rates, changes in the tax component of earnings are valued more in countries with high book-tax conformity than in countries with low book-tax conformity.

Similar to the value relevance of earnings, I expect the value relevance of firms' other components to be higher in countries with a low level of conformity between financial and tax accounting. My final hypothesis is:

[H.sub.4]: Changes in other components of earnings are valued more in countries with low book-tax conformity than in countries with high book-tax conformity.

In summary, changes in all components of earnings are expected to be valued more in countries with low book-tax conformity. However, tax components are expected to be valued more in countries with high book-tax conformity when firms confront high effective tax rates.

DATA AND SAMPLE SELECTION

All data are from the Compustat Global Vantage database from 1995 through 2003. Accounting information data are from the Global Industrial/Commercial files and stock price data are from the Global Issue files. I have imposed data restrictions on the sample, such as the availability of accounting variables and market price variables. I have also limited the sample to firms from countries with book-tax conformity indices derived in prior research. The book-tax conformity index is mainly developed by Hung (2001) and the index for Korea by Young and Guenther (2003). To control for potential outliers, observations in the highest 1% or the lowest 1% values of all the continuous variables were excluded. The sample selection procedures yielded 57,554 firm-year observations from 1995 to 2003 for twenty-two countries. Table 1 lists the distribution of firm-year observations by country and fiscal year. Total firm-year observations for the sample countries range from 35 (South Africa) to 22,488 (United States).

RESEARCH MODELS

I first compare the value relevance of earnings between countries with low book-tax conformity and countries with high book-tax conformity from the following regression models.

[CAR.sub.it] = [[alpha].sub.0] + [[alpha].sub.1][Delta][EARN.sub.it] + [[epsilon].sub.it] (1)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

[CAR.sub.it] : Compounded market adjusted stock return for 15 months for firm i for year t [EARN.sub.it]: Net Earnings (Data #32 in Global Industrial/Commercial files) for firm i for year t [Delta][EARN.sub.it]: Change in Net Earnings for firm i from year t-1 to year t

Similar to Ali and Hwang (2000) and Hung (2001), market adjusted stock returns ([AR.sub.it]) are compounded for fifteen months, ending three months after the fiscal year-end. Since the Global Issue files contain only monthly price data, I have calculated each firm's returns using the monthly price adjusted for stock splits and dividends. Similarly, each country's monthly market returns were calculated using each country's stock market index. Monthly market adjusted returns were calculated by subtracting each country's market returns from individual firm returns. Compound market adjusted stock returns ([CAR.sub.it]) are calculated by multiplying monthly market adjusted returns for fifteen months ending three months after the fiscal year-end. To control for firm size, earnings variables are scaled by the market value of the firm at the end of the previous year.

As Easton and Harris (1991) and Kothari and Zimmerman (1995) argue, errors-in-variable problems can be avoided by including a levels variable in the regression. Furthermore, since earnings indicate the performance of a firm during a specific period, it would be appropriate to include an earnings level variable as well as a changes variable in the regression models. To test how book-tax conformity affects the value relevance of earnings across countries, the following firmlevel pooled regression models are estimated:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (3)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)

[CAR.sub.it] : Compounded market adjusted stock return for 15 months for firm i for year t [EARN.sub.it] : Net Earnings (Data #32 in Global Industrial/Commercial files) for firm i for year t [Delta][EARN.sub.it] : Change in Net Earnings for firm i from year t-1 to year t [BTCONF.sub.j] : Book-Tax conformity Index for country j

I examine whether changes in net earnings and book-tax conformity are value relevant for stock returns using Model (3). The independent variable is the book-tax conformity index. I use the book-tax conformity index developed by Hung (2001), who aggregates the average consensus estimate of the proximity of tax and financial accounting systems with other tax related indicators such as deferred taxes and accelerated depreciation methods. Hung (2001) developed the Tax-Book Conformity Index using the following criteria and weight: (1) average consensus estimate of the relation between tax and financial reporting (60%); (2) do deferred taxes exist? (20%); (3) does legal form dominate substance? (5%); (4) is additional accelerated depreciation allowed? (5%); (5) do amortization periods depend on tax laws? (5%); and (6) does lease capitalization depend on tax law? (5%). Higher index therefore indicates a stronger link between tax and financial reporting. Consistent with prior studies (Ali and Hwang 2000; Hung 2001; Young and Guenther 2003), I also use a high/low scheme to classify the resulting numbers from the book-tax conformity calculation. Hung (2001) argues that it is reasonable to assign the book-tax conformity index as a binary variable rather than a continuous one due to extreme values, finding that the results are qualitatively the same regardless of continuous or binary book-tax conformity measures. I code the sample countries as 0 for low conformity and 1 for high conformity of financial and tax accounting, based on the classification presented by Hung (2001). Table 2 presents some accounting standards related to book-tax conformity and the Book-Tax Conformity Index in each country described above.

An interaction term of the changes in net earnings and book-tax conformity is included. By observing this interaction term, I test whether earnings are less value relevant in countries with high book-tax conformity. To show how book-tax conformity is related to the value relevance of accounting earnings, I examine the relation between the value relevance of earnings components and book-tax conformity. I evaluate the value relevance of changes in sales, depreciation, tax expense, and other earnings components.

I use the following procedure to calculate the value relevance of earnings components from the stock return--earnings components regressions for each country. The changes in earnings are decomposed as follows:

[Delta][EARN.sub.it] = [Delta][SALES.sub.it] - [Delta][DEP.sub.it] - [Delta][TAX.sub.it] - [Delta][OTHER.sub.it] (5) [Delta][EARN.sub.it] : Change in Net Earnings (Data #32 in Global Industrial/Commercial files) for firm i for year t [Delta][SALES.sub.it] : Change in Sales (Data #1 in Global Industrial/Commercial files) for firm i for year t [Delta][DEP.sub.it] : Change in Depreciation (Data #12 in Global Industrial/Commercial files) for firm i for year t [Delta][TAX.sub.it] : Change in Tax Expenses (Data #23 in Global Industrial/Commercial files) for firm i for year t [Delta][OTHER.sub.it] : Change in Other Earnings Components for firm i for year t

To control for firm size, all accounting variables are scaled by the market value of the firm at the end of the previous year. Then, the following regression models are estimated for countries with low book-tax conformity and countries with high book-tax conformity:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (6)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (7)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (8)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (9)

[Delta][OTHERS.sub.it] : Change in Other Earnings Components except Sales for firm i for year t [Delta][OTHERD.sub.it] : Change in Other Earnings Components except Depreciation for firm i for year t [Delta][PTI.sub.it] : Change in Pre-Tax Income ((Data #23 in Global Industrial/Commercial files) for firm i for year t

To test Hypothesis 1, earnings are decomposed into 'sales and other components' in countries with both low and high book-tax conformity. Earnings are also decomposed into 'depreciation expenses and other components' and 'pretax income and tax component' to test Hypotheses 2 and 3a. Regression models are estimated as follows:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (10)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (11)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (12)

In the regression Model (10), the coefficient of interaction variable [w.sub.4] represents the different effect of sales component changes for countries with high book-tax conformity from countries with low book-tax conformity. Similarly, the coefficients f4 and l4 in Models (11) and (12) represent the differences in the sensitivity of CAR to the changes in depreciation and tax components respectively between countries with high book-tax conformity and countries with low book-tax conformity. Yearly regressions are also estimated to compare mean coefficients of each earnings component between countries with low book-tax conformity and countries with high book-tax conformity. From yearly regression models, I compare explanatory power, measured by adjusted [R.sup.2], between two groups to examine whether the value relevance of earnings and earnings components is generally higher in countries with low book-tax conformity.

For Hypotheses [3.sub.b], I test whether the changes in tax expenses of the firms with high effective tax rates (firms within the first quintile on effective tax rates) are more value relevant in countries with strong book-tax conformity than in countries with weak conformity, by comparing explanatory powers and the coefficient of tax components between two groups of countries.

I also examine whether decomposing earnings into various components increases value relevance on stock returns in countries with high book-tax conformity and countries with low book-tax conformity. Vuong (1989) proposed a simple likelihood-ratio based statistic, testing the null hypothesis that competing models are equally close to the true data generating process against the alternative hypothesis that one model is closer. Vuong's test statistics demonstrate whether regression models for earnings components have statistically significant incremental explanatory power compared with the models for earnings.

EMPIRICAL RESULTS

Table 3 reports the Pearson correlations between dependent variable (CAR) and independent variables, such as changes in earnings, sales, depreciation, tax expenses, and other earnings components, demonstrating that earnings and other earnings components variables are highly correlated with each other both in countries with strong book-tax alignment and in countries with weak book-tax alignment. Moreover, all independent variables are also highly correlated with compounded market adjusted returns, which are the main dependent variable for the tests.

Table 5 (Panel A) compares the value relevance of earnings between countries with high book-tax conformity and that of countries with low book-tax conformity. From the return-earnings regression models, I compare coefficients of changes in earnings and level of earnings and also compare explanatory powers in return-earnings regressions between two groups of countries. Coefficients of changes in earnings are higher in countries with weak book-tax alignment, but coefficients of level of earnings are higher in countries with strong book-tax alignment. The explanatory powers, measured by adjusted [R.sup.2], of return-earnings regressions are higher in countries with low book-tax conformity for both Model (1) and Model (2).

Table 5 (Panel B) presents results from firm-level pooled regression models estimated to test whether book-tax conformity is related to the value relevance of firms' accounting earnings. The inclusion of interaction terms in the regression models tests the effects of earnings changes on the returns of companies in countries with different levels of book-tax conformity. In Model (3), the coefficient of interaction variables represents the different effect of earnings changes for countries with high book-tax conformity from countries with low book-tax conformity. Consistent with prediction, the coefficient of interaction variables is significantly negative.

In Model (4), I include level of earnings as well as changes in earnings. As in Model (3), interaction terms are used to test different effects of book-tax conformity on returns across countries. Two coefficients of interaction terms, [phi]4 and [phi]5, show the difference between the effects of earnings changes and earnings levels on stock returns for countries with high book-tax conformity and those for countries with low book-tax conformity. As predicted, coefficients of interaction terms [phi]4 are significantly negative, which indicates that the value relevance of changes in earnings is higher in countries with low book-tax conformity. However, coefficient [phi]5 is not statistically significant. When I use the compounded market adjusted returns based on average monthly returns of sample firms in each country, both the coefficients of interaction terms for the change of earnings ([phi]4) and coefficients of interaction terms for the level of earning ([phi]5) are significantly negative. Overall, I find that the value relevance of changes in earnings is negatively associated with the degree of book-tax conformity across countries.

Table 6 (Panel A) compares the value relevance of earnings components form Model (6) to Model (9) between countries with high book-tax conformity and countries with low book-tax conformity. As predicted in Hypothesis 1, the coefficient of changes in sales and explanatory power of the regression are higher in countries with low book-tax conformity. The explanatory powers of regression models for depreciation and tax expenses are also higher in countries with weak alignment of book-tax conformity. As also predicted in Hypothesis 2, coefficient of the changes in depreciation is negatively related to stock return perhaps because expenses reduce net earnings. However, the coefficient of changes in depreciation is less negative in countries low book-tax conformity perhaps because the negative effect of book expense on stock returns may be mitigated by tax savings from depreciation due to accelerated depreciation and shorter useful life used in computing tax depreciation in countries with weak alignment between tax and financial reporting.

Table 6 (Panel B) presents results from firm-level pooled regression models, Model (10) through Model (12), to test Hypothesis 1, 2, 3a, and 4. Results show that the coefficient of changes in sales is significantly positive, consistent with the prediction. The coefficient of changes in tax expense is also significantly positive, perhaps because investors recognize that more profitable firms probably pay more taxes. The interaction variables of book-tax conformity and changes in each earnings components variable show the effect of book-tax conformity on the value relevance of the components. In Model (10), the coefficient of the interaction variable of book-tax conformity and changes in sales is significantly negative, which is consistent with the prediction in Hypothesis 1 that changes in the sales components are valued more in countries with low book-tax conformity than in countries with high book-tax conformity.

Consistent with Hypothesis 2, the interaction coefficient of book-tax conformity index and changes in deprecation is also significantly negative, perhaps indicating that change in the depreciation component of net income is valued more in countries with low book-tax conformity than in countries with high book-tax conformity. When earnings are decomposed into tax expenses and pre-tax income as in Model (12), the interaction coefficient of book-tax conformity index and changes in tax component is significantly negative, which supports Hypothesis 3a, indicating that changes in tax component of earnings are valued less in countries with high book-tax conformity than in countries with low book-tax conformity. Consistent with expectations, the interaction coefficient of book-tax conformity index and changes in pre-tax income is also significantly negative.

In summary, this study finds evidence that sales, depreciation component, and tax component measured by regression coefficients, are more value relevant in countries with low book-tax conformity. When the value relevance of earnings components is measured by explanatory powers of regression models, I also find that sales, depreciations, and tax components are valued more in countries with weak alignment between financial and tax accounting. Figure 1 describes explanatory powers of earnings components models between countries with low book-tax conformity and high book-tax conformity.

In addition to book-tax conformity, the legal environment of each country may have influenced the value relevance of earnings components. Therefore, I add a legal environment control variable, developed by prior literature (e.g., Hung 2001), to the regression models. After controlling legal environment variable, the results for changes in earnings and changes in sales are not significantly different from the main results. However, after controlling for the legal environment variable, the interaction coefficient of book-tax conformity index and changes in tax expenses as well as interaction coefficient of changes in depreciation are positively related to stock returns. When I use the compounded market adjusted returns based on average of monthly returns of sample firms in each country, the results are similar to the main results. Because firms with higher effective tax rates pay taxes at higher marginal tax rates and are more likely to attempt to reduce their tax expenses, as compared to firms with low effective tax rates, I test whether the changes in tax expenses of the firms with high effective tax rates are more value relevant in countries with strong book-tax conformity than in countries with weak conformity. Table 7 (Panel A) compares value relevance of changes in earnings and changes in tax expenses for the firms with high effective tax rates (11,508 firm-year sample within first quintile on effective tax rates) between countries with weak book-tax alignment and those with strong book-tax alignment. Similar to the results from the total 57,554 firm-year sample, explanatory power of stock returns-earnings regression is higher in countries with low book-tax conformity. However, explanatory power of stock returns-tax expenses regression is higher in countries with high book-tax conformity, which supports Hypothesis [3.sub.b]. Table 7 (Panel B) presents the results from firm-level pooled regression models, to test Hypothesis [3.sub.b]. In Model (12), the coefficient of the interaction of book-tax conformity and changes in tax expenses is significantly negative, which is consistent with the prediction in Hypothesis [3.sub.b] that changes in tax components are valued less in countries with low book-tax conformity than in countries with high book-tax conformity.

[FIGURE 1 OMITTED]

Figure 2 describes explanatory powers of earnings and tax components models for the firms with high effective tax rates between countries with low book-tax conformity and high book-tax conformity.

[FIGURE 2 OMITTED]

I use panel data containing two kinds of information in both cross-sectional data and time-series data. Cross-sectional information reflects the differences among countries, and the time-series, or within-country information, reflects the changes within countries over time, for a period of ten years. The ordinary multiple regression techniques used on panel data in the previous sections may not be optimal. The estimates of coefficients derived from regression may be subject to omitted variable bias--a problem that arises when there is some unknown variable or variables that cannot be controlled for that affect the dependent variable. To control for omitted variables that differ among countries but are constant over time, fixed effects regression are also estimated. Results from the regression models after controlling for fixed effects are similar to the main results.

Another way to address fixed effects is by estimating yearly regressions. To examine whether there exists differences in the value relevance of earnings and earnings components between countries with low book-tax conformity and countries with high book-tax conformity, I compare mean coefficients from yearly regressions following the approach of Fama and Macbeth (1974). Table 8 (Panel A) presents the mean of coefficients from yearly regressions for countries with low book-tax conformity and the mean coefficient from yearly regressions for countries with high book-tax countries for a period of ten years.

Consistent with results in Table 5, the mean coefficient of changes in earnings is slightly higher in countries with low book-tax conformity. Mean coefficients of changes in sales, depreciation components, and tax components are significantly higher in countries with low book-tax conformity than in those with high book-tax conformity. Interestingly, however, the mean coefficient of changes in the depreciation component is positively related to stock returns in countries with low book-tax conformity. This positive sign can be interpreted as a result of investors valuing firms whose depreciation expenses increase in countries with low book-tax conformity as related to the firm's higher level of investment in property and equipment, which results in higher depreciation expenses.

Using a methodology similar to the approach of Fama and Macbeth (1974), I examine whether the explanatory powers of earnings components models are statistically higher in countries with low book-tax conformity than those in countries with high book-tax conformity. Table 8 (Panel B) compares the means of explanatory powers of yearly regression models between the two groups. Explanatory powers of changes in earnings model and changes in sales are higher in countries with low book-tax alignment than in countries with high book-tax conformity. However, explanatory powers of changes in depreciation and taxes models are not statistically different between the two groups, although explanatory powers are slightly higher in countries with low book-tax alignment than in countries with high book-tax conformity.

I also test whether decomposing earnings into various components increases value relevance on stock returns and how they affect countries with high book-tax conformity versus countries with low book-tax conformity. In Table 9, Vuong's Z-values indicate that explanatory powers, measured by [R.sup.2], significantly increase when earnings are decomposed into 'sales and other components' or 'taxes and pre-tax income,' meaning that regression models for earnings components have statistically incremental explanatory power as compared with the models for earnings in both countries with strong book-tax alignment and countries with weak alignment.

CONCLUSION

This study examines how book-tax conformity affects the value relevance of earnings across countries. I find evidence that changes in earnings are less value relevant in countries with high book-tax conformity than in those with low book-tax conformity, consistent with prior research.

I measure the value relevance of earnings components using the coefficients and explanatory power in return-earnings component regressions. Using a binary variable of book-tax conformity developed in prior research, I investigate whether the value relevance of changes in sales, depreciation, tax expenses, and other earnings components are associated with book-tax conformity across countries.

As hypothesized, empirical evidence reveals that sales, depreciation, and tax expenses components, measured by regression coefficients, are more value relevant in countries with low book-tax conformity. When the value relevance of earnings and earnings components are measured in terms of explanatory powers of regression models, changes in earnings and changes in sales are valued more in countries with low book-tax alignment than in countries with high book-tax conformity. Explanatory powers of the changes in depreciation and tax expenses models are slightly higher in countries with weak alignment between financial accounting, although those explanatory powers are not statistically different between the two groups of countries.

I also examine whether changes in tax expenses of the firms with high effective tax rates are more value relevant in countries with high book-tax conformity and find that, for the firms with high effective tax rates, explanatory power of stock returns-tax expenses regression is higher in countries with strong book-tax alignment. One possible explanation of this result is that firms with higher effective tax rates are more likely to attempt to reduce their tax expenses in countries with weak book-tax conformity, resulting in less value-relevant information reflecting in stock returns. Finally, I find evidence that decomposing earnings into various components increases value relevance on stock returns in both countries with high book-tax conformity and countries with low book-tax conformity.

ENDNOTES

This paper is based in part on my dissertation at the Leeds School of Business at the University of Colorado. I would like to thank the members of my dissertation committee for their guidance: Philip Shane, John Jacob, Thomas Buchman, Sanjai Bhagat, and especially David Guenther (Chair). I am also grateful to Naomi Soderstrom, Steve Rock, Srinivasan Rangan, David Weber, Veronda Willis, and anonymous referees for helpful comments and suggestions.

REFERENCES

Abarbanell, J. & B. Bushee, (1997), Fundamental Analysis, Future Earnings, and Stock Prices, Journal of Accounting Research 35(1): 1-24.

Alford, A., L. Jones, R. Leftwich & M. Zmijewski., (1993), The Relative Informativeness of Accounting Disclosures in Different Countries, Journal of Accounting Research 31(3): 183-223.

Ali, A. & L. Hwang, (2000), Country-Specific Factors Related to Financial Reporting and the Value Relevance of Accounting Data, Journal of Accounting Research 38 (1):1-21.

Ball, R., S. P. Kothari, & A. Robin., (2000), The Effect of International Institutional Factors on Properties of Accounting Earnings, Journal of Accounting & Economics 29 (1):1-51

Ball, R., A. Robin, & S. Wu, (2003), Incentives versus Standards: Properties of Accounting Income in Four East Asian Countries, Journal of Accounting and Economics 36 (1): 235-270

Barber, B. & D. Lyon, (1997), Detecting Long-Run Abnormal Stock Returns: The Empirical Power and Specification of Test Statistics, Journal of Financial Economics 43 (1997):341-372

Brown S., K. Lo, & T. Lys, Use of R2 in Accounting Research: Measuring Changes in Value Relevance over the Last Four Decades, Journal of Accounting & Economics 28(2): 83-115

Center for International Financial Analysis & Research, Inc.(CIFAR), (1993), International Accounting and Auditing Trends, 3rd edition, Princeton NJ.

Devinney, T., (2000), Implications of Tax Systems to Multinational Corporations in Asia Pacific, Australian Graduate School of Management: http://www.agsm.edu.au/~timdev/courses/resource/ IBSWebsite/taxation/index.htm

Dhaliwal, D., C. Gleason, & L. Mills, (2004), Last Chance Earnings Management: Using the Tax Expense to Achieve Earnings Targets, Contemporary Accounting Research 21 (2): 431-459.

Easton, P. & T. Harris, (1991), Earnings as an Explanatory Variable for Returns, Journal of Accounting Research 29 (1): 19-36

Fama, E. & J. MacBeth., (1974), Tests of the Multi-period Two-Parameter Model, Journal of Financial Economics 1 (1): 43-66

Financial Accounting Standard Board, (1998), Special Report-Reporting Financial Performance: Current Developments and Future Directions, Financial Accounting Series

Guenther, D. & D. Jones, (2003), Valuation Implications of Changes in Firms' Effective Tax Rates, Working Paper. University of Colorado at Boulder.

Guenther, D. & D. Young, (2000), The Association between Financial Accounting Measures and Real Economic Activity: A Multinational Study, Journal of Accounting & Economics 29(1): 53-72

Haw, I., B. Hu, L. Hwang, & W. Wu, (2004), Ultimate Ownership, Income Management, and Legal and Extra-Legal Institutions, Journal of Accounting Research 42 (2); 423-462

Herrmann, D., T. Inoue, & W Thomas, (2000), The Persistence and Forecast Accuracy of Earnings Components in the USA and Japan, Journal of International Financial Management & Accounting 11 (1): 48-70

Hung, M., (2001), Accounting Standards and Value Relevance of Financial Statements: An International Analysis, Journal of Accounting & Economics 30 (3): 401-420

James, S., (2002), The Relationship between Accounting and Taxation, Working paper, University of Exeter

Joos, P. & M. Lang, (1994), The Effects of Accounting Diversity: Evidence from the European Union, Journal of Accounting Research 32(3), 141-168

Kothari, S. P. & J. Zimmerman, (1995), Price and Return Models, Journal of Accounting & Economics 20 (2), 155-192

Leuz, C., D. Nanda, & P. Wysocky, (2003), Earnings Management and Investor Protection: An International Comparison, Journal of Financial Economics 69, 505-527

Lev, B. & S. R. Thiagarajan, (1993), Fundamental Information Analysis, Journal of Accounting Research 31(2): 190-215.

Lipe, R.,(1986), The Information Contained in the Components of Earnings, Journal of Accounting Research 24 (3): 37-64

Lyon, J., B. Barber, & C. Tsai., (1999), Improved Methods for Tests of Long-Run Abnormal Stock Returns, The Journal of Finance 54 (1): 165-201

Meek G. & W. Thomas, (2003), A Review of Markets-Based International Accounting Research, Working Paper, Oklahoma State University and University of Oklahoma

Ohlson, J. & S. Penman, (1992), Disaggregated Accounting Data as Explanatory Variables for Returns, Journal of Accounting, Auditing & Finance, 7 (4): 553-573

Organization for Economic Co-operational and Development (OECD) Working Group on Accounting Standards, 1987, Accounting Standards Harmonization Report No.3: The Relation between Taxation and Financial Reporting, Paris: OECD

Porcano, T. & A. Tran., (1998), Relationship of Tax and Financial Accounting Rules in Anglo-Saxon Countries, The International Journal of Accounting 33 (4): 733-454

Schmidt, A., (2006), The Persistence, Forecasting Ability, and Valuation Implications of the Tax Change Component of Earnings, Accounting Review 81 (3): 589-616

Voung, Q., (1989), Likelihood Ratio Tests for Model Selection and Non-Nested Hypotheses, Econometrica 57 (2):307-333

Young, D. & D. Guenther, (2003), Financial Reporting Environments and International Capital Mobility, Journal of Accounting Research 41 (3): 553-579

Sung Wook Yoon, California State University, Northridge

Table 1: Distribution of firm-year observations
by country and fiscal year

Country 1995 1996 1997 1998 1999

Australia 34 147 159 197 196
Canada 350 394 384 389 377
Denmark 9 11 21 27 31
Hong Kong 39 61 64 87 84
Ireland 17 22 21 26 27
Netherlands 18 23 26 33 56
New Zealand 0 3 4 13 12
Norway 13 15 38 55 54
Singapore 96 130 146 172 179
South Africa 0 0 0 0 0
U. K. 464 675 792 866 851
U. S. 2,181 2,702 2,791 2,675 2,504
Low B-T conformity 3,221 4,183 4,446 4,540 4,371
countries total
Belgium 3 2 3 3 7
Finland 9 11 13 14 26
France 41 62 73 79 121
Germany 1 3 8 19 46
Italy 2 2 2 5 11
Japan 0 155 1,088 1,166 1,235
Korea 6 4 4 9 54
Spain 7 7 10 10 5
Sweden 9 14 34 75 95
Switzerland 2 7 10 21 35
High B-T conformity 80 267 1,245 1,401 1,635
countries total
Total 3,301 4,450 5,691 5,941 6,006

Country 2000 2001 2002 2003 Total

Australia 209 198 238 237 1,615
Canada 374 344 334 328 3,274
Denmark 30 30 47 51 257
Hong Kong 89 89 98 100 711
Ireland 24 21 22 18 198
Netherlands 78 112 104 112 562
New Zealand 13 9 8 14 76
Norway 55 62 80 74 446
Singapore 209 213 284 315 1,744
South Africa 0 1 1 33 35
U. K. 814 817 784 731 6,794
U. S. 2,512 2,476 2,370 2,277 22,488
Low B-T conformity 4,407 4,372 4,370 4,290 38,200
countries total
Belgium 17 27 32 37 131
Finland 33 56 58 69 289
France 187 317 379 373 1,632
Germany 101 211 283 238 910
Italy 11 4 8 7 52
Japan 2,342 2,753 2,833 2,826 14,398
Korea 104 131 158 164 634
Spain 4 3 4 4 54
Sweden 85 138 198 200 848
Switzerland 55 71 103 102 406
High B-T conformity 2,939 3,711 4,056 4,020 19,354
countries total
Total 7,346 8,083 8,426 8,310 57,554

Table 2. Book-tax conformity index and accounting standards
related book-tax conformity in each country

Country Deferred Depreciation Long-term
 Taxes (a) Policy (a) Financial
 Lease (a)

Countries with low B-T conformity

Australia Used Straight Line Capitalized
Canada Used Straight Line Capitalized
Denmark Used Straight Line w/ Capitalized
 excess dep.
Hong Kong Used Straight Line Capitalized
Ireland Used Straight Line Capitalized
Netherlands Used Straight Line w/ Capitalized
 excess dep.
New Zealand Used Straight Line Capitalized

Norway Paid As Straight Line w/ Not Disclosed
 Incurred excess dep.
Singapore Used Straight Line Capitalized
South Africa Used Straight Line w/ Capitalized
 excess dep.
U. K. Used Straight Line Capitalized
U. S. Used Straight Line Capitalized

Countries with high B-T conformity

Belgium Used Straight Line w/ Capitalized
 excess dep.
Finland Paid As Straight Line w/ Capitalized
 Incurred excess dep.
France Used Straight Line w/ Capitalized
 excess dep.
Germany Used by Mixed Dep. w/ Usually not
 some corp. excess dep. Disclosed
Italy Used Straight Line Not Disclosed

Japan Paid As Accelerated
 Incurred depreciation Mostly not
 Disclosed
Korea Used Straight Line w/ Capitalized or
 excess dep. Expensed
Spain Used Straight Line Not Disclosed

Sweden Used Straight Line w/ Capitalized or
 excess dep. Expensed
Switzerland Used Straight Line Capitalized or
 Expensed

Country 'Hidden' non-equity B-T Conf.
 reserves usually (a) Index (b)

Countries with low B-T conformity

Australia Do not exist 0
Canada Do not exist 0
Denmark Exist and are separately 0
 reported
Hong Kong Do not exist 0
Ireland Not consistently disclosed 0
Netherlands As part of shareholders' 0
 equity
New Zealand As part of shareholders' 0
 equity
Norway Exist and are separately 0
 reported
Singapore Do not exist 0
South Africa As part of shareholders' 0
 equity
U. K. Do not exist 0
U. S. Do not exist 0

Countries with high B-T conformity

Belgium Exist and are separately 1
 reported
Finland Exist and are separately 1
 reported
France Exist and are separately 1
 reported
Germany Exist and are separately 1
 reported
Italy Exist and are separately 1
 reported
Japan Do not exist 1

Korea As part of shareholders' 1
 equity
Spain As part of shareholders' 1
 equity
Sweden Exist and are separately 1
 reported
Switzerland Exist and are separately 1
 reported

(a) Source: International Accounting and Auditing
Trends, 1993: CIFAR

(b) Book-Tax conformity Index show the degree of the
alignment between financial accounting and tax reporting.
The index is developed by Young and Guenther (2003) for
Korea and by Hung (2001) for all other countries. It
equals 1 for countries with high book-tax conformity
and equals 0 for countries with low book-tax conformity.

Table 3. Correlations between dependent variable and
independent variables for countries with low book-tax
conformity and countries with high book-tax conformity

 CAR EARN SALES DEP

CAR .22153 .12664 .02539
 (<.0001) (<.0001) (<.0001)
 .20306 .15799 -.09431
[Delta] EARN
 (<.0001) (<.0001) (<.0001)
[Delta] SALES .09502 .16752 .40770
 (<.0001) (<.0001) (<.0001)
[Delta] DEP -.03636 -.02395 .49295
 (<.0001) -.00090 (<.0001)
[Delta] TAX .14494 .33416 .26838 .12427
 (<.0001) (<.0001) (<.0001) (<.0001)
[Delta] OTHERS .06026 -.00797 .98450 .50426
 (<.0001) -.26740 (<.0001) (<.0001)
 CAR EARN SALES DEP
[Delta] OTHERD .18920 .96968 .28391 .22107
 (<.0001) (<.0001) (<.0001) (<.0001)
[Delta] PTI .21656 .94725 .23078 .02029
 (<.0001) (<.0001) (<.0001) -.00480

 TAX OTHERS OTHERD PTI

CAR .17926 .08021 .22635 .24093
 (<.0001) (<.0001) (<.0001) (<.0001)
 .33329 -.05621 .97688 .96938
[Delta] EARN
 (<.0001) (<.0001) (<.0001) (<.0001)
[Delta] SALES .26716 .97700 .24508 .20918
 (<.0001) (<.0001) (<.0001) (<.0001)
[Delta] DEP .08799 .43260 .12069 -.06157
 (<.0001) (<.0001) (<.0001) (<.0001)
[Delta] TAX .19816 .35123 .53661
 (<.0001) (<.0001) (<.0001)
[Delta] OTHERS .21277 .03685 .00216
 (<.0001) (<.0001) -.67280
 TAX OTHERS OTHERD PTI
[Delta] OTHERD .35636 .11548 .95338
 (<.0001) (<.0001) (<.0001)
[Delta] PTI .58385 .06558 .92904
 (<.0001) (<.0001) (<.0001)

Pearson correlation coefficients of variables for the
countries with low book-tax conformity (high book-tax
conformity) in the upper (lower) triangle; Two-tailed
p-values in parentheses.

Table 4. Univariate statistics of variables between countries
with low book-tax conformity and countries with high
book-tax conformity

 Countries with Low B-T
 Conformity

 Number Mean Std.
 of obs. dev.

Changes in Net 38,200 0.0220 0.2187
Earnings ([Delta] EARN)

Changes in Sales 38,200 0.2192 1.0239
([Delta] SALES)

Changes in 38,200 0.0083 0.0471
Depreciation([Delta] DEP)

Changes in Tax 38,200 0.0083 0.0589
Expense ([Delta] TAX)

Changes in Other Expense except Sales

([Delta] OTHERS) 38,200 0.1971 1.0127

Changes in Other Expense except

Depreciation 38,200 0.0303 0.2193
([Delta] OTHERD)

Changes in Pre-tax 38,200 0.0306 0.2481
Income ([Delta] PTI)

Compounded 38,200 0.0646 0.6240
Market Adjusted
Returns (CAR)

Compounded 37,187 -0.0090 0.4858
Market Adjusted
Returns II (CAR II)

 Countries with High B-T
 Conformity

 Number Mean Std.dev.
 of obs.

Changes in Net 19,354 0.0189 0.2084
Earnings ([Delta] EARN)

Changes in Sales 19,354 0.1306 1.1881
([Delta] SALES)

Changes in 19,354 0.0036 0.0522
Depreciation([Delta] DEP)

Changes in Tax 19,354 0.0078 0.0695
Expense ([Delta] TAX)

Changes in Other Expense except Sales

([Delta] OTHERS) 19,354 0.1118 1.1713

Changes in Other Expense except

Depreciation 19,354 0.0224 0.2136
([Delta] OTHERD)

Changes in Pre-tax 19,354 0.0272 0.2400
Income ([Delta] PTI)

Compounded 19,354 0.1386 0.5300
Market Adjusted
Returns (CAR)

Compounded 18,542 -0.0080 0.3998
Market Adjusted
Returns II (CAR II)

 Mean t-stat p-value
 Differ-
 ences

Changes in Net 0.0031 1.67 0.0954
Earnings ([Delta] EARN)

Changes in Sales 0.0886 9.28 <0.0001
([Delta] SALES)

Changes in 0.0047 11.00 <0.0001
Depreciation([Delta] DEP)

Changes in Tax 0.0005 0.79 0.4311
Expense ([Delta] TAX)

Changes in Other Expense except Sales

([Delta] OTHERS) 0.0854 9.06 <0.0001

Changes in Other Expense except

Depreciation 0.0079 4.12 <0.0001
([Delta] OTHERD)

Changes in Pre-tax 0.0035 1.60 0.1096
Income ([Delta] PTI)

Compounded -0.0740 -14.13 <0.0001
Market Adjusted
Returns (CAR)

Compounded -0.0010 -0.29 0.7691
Market Adjusted
Returns II (CAR II)

Variables in the above table are defines as follows;
[Delta] EARN : Change in Net Earnings (Data #32 in
 Global Industrial/Commercial files)
[Delta] SALES : Change in Sales (Data #1 in Global
 Industrial/Commercial files)
[Delta] DEP : Change in Depreciation (Data #12
 in Global Industrial/Commercial files)
[Delta] TAX : Change in Tax Expenses (Data #23
 in Global Industrial/Commercial files)
[Delta] OTHERS : Change in Other Earnings
 Components except Sales
[Delta] OTHERD : Change in Other Earnings
 Components except Depreciation
[Delta] PTI : Change in Pre-Tax Income ((Data #23
 in Global Industrial/Commercial files)
CAR : Compounded market adjusted stock return
 for 15 months
CAR II : Compounded market adjusted stock returns II,
calculated based on average of monthly returns of
sample firms in each country and used only for
the sensitivity tests.

Table 5: (Panel A) Value relevance of earnings in countries with
low book-tax conformity and countries with high book-tax conformity

Dependent Variable = 15 months compound market-adjusted
return (CAR)

 Low B-T Conformity Countries

 Predicted Model (1) Model (2)
 sign

Intercept 0.05064 0.03646
 ?
 (<0.0001) (<0.0001)
Changes in Net 0.63204 0.48823
Earnings (EARN) (+)
 (<0.0001) (<0.0001)
Net Earnings (EARN) 0.40697
 (+)
 (<0.0001)
Adjusted-R2 0.0491 0.0664
Fixed Effects None None
No. of Observations 38,200 38,200

 High B-T Conformity Countries

 Model (1) Model (2)

Intercept 0.12887 0.12216

 (<0.0001) (<0.0001)
Changes in Net 0.51656 0.35672
Earnings (EARN)
 (<0.0001) (<0.0001)
Net Earnings (EARN) 0.44753

 (<0.0001)
Adjusted-R2 0.0412 0.0624
Fixed Effects None None
No. of Observations 19,354 19,354

Table 5: (Panel B) The effect of book-tax conformity on
the value relevance of earnings

 Dependent Variable = 15 months
 compound market-adjusted return (CAR)

 Predicted Model (3)
 sign
 Coeff. t-stat

Intercept ? 0.05064 16.98
Changes in Net (+) 0.63204 46.58
 Earnings ([Delta] EARN)
Net Earnings (EARN) (+)
B-T Conformity Index ? 0.07823 15.22
B-T Conformity x [Delta] EARN (-) - 0.11548 -4.78
B-T Conformity x [Delta] EARN (-)
Adjusted-[R.sup.2] 0.0502
Fixed Effects None
Number of Observations 57,554

 Predicted Model (4)
 sign
 p-value Coeff.

Intercept ? <0.0001 0.03646
Changes in Net (+) <0.0001 0.48823
 Earnings ([Delta] EARN)
Net Earnings (EARN) (+) 0.40697
B-T Conformity Index ? <0.0001 0.0857
B-T Conformity x [Delta] EARN (-) <0.0001 -0.13150
B-T Conformity x [Delta] EARN (-) 0.04057
Adjusted-[R.sup.2]
Fixed Effects
Number of Observations

 Predicted Model (4)
 sign
 t-stat p-value

Intercept ? 12.17 <0.0001
Changes in Net (+) 33.94 <0.0001
 Earnings ([Delta] EARN)
Net Earnings (EARN) (+) 27.99 <0.0001
B-T Conformity Index ? 16.17 <0.0001
B-T Conformity x [Delta] EARN (-) -5.07 <0.0001
B-T Conformity x [Delta] EARN (-) 1.45 0.1473
Adjusted-[R.sup.2] 0.0685
Fixed Effects None
Number of Observations 57,554

Table 6. (Panel A) Value relevance of earnings components in
countries with low book-tax conformity and countries with
high book-tax conformity

Dependent Variable = 15 months compound
market-adjusted return (CAR)

 Predicted Low B-T Conformity Countries
 sign
 Model (6) Model (7)
Changes in (-) -0.58967
Other
Components (<0.0001)
([Delta] OTHERS)

Changes in (+) 0.6446
Other
Components (<0.0001)
([Delta] OTHERD)

Changes in Pre- (+)
Tax Income
([Delta] PTI)

Changes in (-)
Other
Components
([Delta] OTHER)

Adjusted-[R.sup.2] 0.0576 0.0512
Fixed Effects None None
Number of 38,200 38,200
Observations

 Predicted Low B-T Conformity Countries
 sign
 Model (8) Model (9)
Changes in (-)
Other
Components
([Delta] OTHERS)

Changes in (+)
Other
Components
([Delta] OTHERD)

Changes in Pre- (+) 0.51123
Tax Income
([Delta] PTI) (<0.0001)

Changes in (-) -0.50576
Other
Components (<0.0001)
([Delta] OTHER)

Adjusted-[R.sup.2] 0.0615 0.0662
Fixed Effects None None
Number of 38,200 38,200
Observations

 Predicted High B-T Conformity Countries
 sign
 Model (6) Model (7)
Changes in (-) -0.48981
Other
Components (<0.0001)
([Delta] OTHERS)

Changes in (+) 0.5146
Other
Components (<0.0001)
([Delta] OTHERD)

Changes in Pre- (+)
Tax Income
([Delta] PTI)

Changes in (-)
Other
Components
([Delta] OTHER)

Adjusted-[R.sup.2] 0.0450 0.0421
Fixed Effects None None
Number of 19,354 19,354
Observations

 Predicted High B-T Conformity Countries
 sign
 Model (8) Model (9)
Changes in (-)
Other
Components
([Delta] OTHERS)

Changes in (+)
Other
Components
([Delta] OTHERD)

Changes in Pre- (+) 0.44202
Tax Income
([Delta] PTI) (<0.0001)

Changes in (-) -0.4065
Other
Components (<0.0001)
([Delta] OTHER)

Adjusted-[R.sup.2] 0.0473 0.0553
Fixed Effects None None
Number of 19,354 19,354
Observations

Table 6. (Panel B) The effect of book-tax conformity on the value
relevance of earnings components

Dependent Variable = 15 months compound
market-adjusted return (CAR)

 Model (10)
 Predicted Coeff. t-stat p-value
 sign

Intercept ? 0.03902 12.88 <0.0001
[Delta] SALES (+) 0.64695 47.79 <0.0001
[Delta] DEP (-)
[Delta] TAX (+)
[Delta] OTHERS (-) -0.58967 -43.08 <0.0001
[Delta] OTHERD (+)
[Delta] PTI (+)
BTCONF ? 0.0867 17 <0.0001
BTCONF x (-) -0.12914 -5.36 <0.0001
[Delta] SALES
BTCONF x (-)
[Delta] DEP
BTCONF x (-)
[Delta] TAX
BTCONF x (+) 0.09986 4.09 <0.0001
[Delta] OTHERS
BTCONF x (-)
[Delta] OTHERD
BTCONF x (-)
[Delta] PTI
Adjusted-R2 0.0575
Fixed Effects None
Number of 57,554
Observations

 Model (11)
 Coeff. t-stat p-value
 Predicted
 sign
 0.04522 14.92 <0.0001
Intercept ?
[Delta] SALES (+) -0.02597 -0.41 0.6821
[Delta] DEP (-)
[Delta] TAX (+)
[Delta] OTHERS (-) 0.6446 47.34 <0.0001
[Delta] OTHERD (+)
[Delta] PTI (+) 0.08482 16 <0.0001
BTCONF ?
BTCONF x (-)
[Delta] SALES -0.80863 -7.81 <0.0001
BTCONF x (-)
[Delta] DEP
BTCONF x (-)
[Delta] TAX
BTCONF x (+)
[Delta] OTHERS -0.12997 -5.37 <0.0001
BTCONF x (-)
[Delta] OTHERD
BTCONF x (-)
[Delta] PTI 0.0520
Adjusted-R2 None
Fixed Effects 57,554
Number of
Observations

 Model (12)
 Coeff. t-stat p-value
 Predicted
 sign
 0.04275 14.33 <0.0001
Intercept ?
[Delta] SALES (+)
[Delta] DEP (-) 0.74396 12.52 <0.0001
[Delta] TAX (+)
[Delta] OTHERS (-)
[Delta] OTHERD (+) 0.51123 36.27 <0.0001
[Delta] PTI (+) 0.08217 16.00 <0.0001
BTCONF ?
BTCONF x (-)
[Delta] SALES
BTCONF x (-)
[Delta] DEP -0.06921 -2.71 0.0067
BTCONF x (-)
[Delta] TAX
BTCONF x (+)
[Delta] OTHERS
BTCONF x (-)
[Delta] OTHERD -0.52972 -5.60 <0.0001
BTCONF x (-)
[Delta] PTI 0.0609
Adjusted-R2 None
Fixed Effects 57,554
Number of
Observations

The numbers in parenthesis represent p-value of coefficient.

Table 7: (Panel A) Value relevance of earnings components for
the firms with high ETRs in countries with low book-tax
conformity and countries with high book-tax conformity

 Dependent Variable = 15 months compound
 market-adjusted return (CAR)

 Predicted Low B-T Conformity Countries
 sign
 Model (1) Model (9) Tax Expense

Intercept 0.02724 0.00964 0.01683
 ?
 -0.0045 -0.3285 -0.0846
[Delta] EARN (+) 0.77378
 (<.0001)

[Delta] SALES (+) 0.64912
 (<.0001)

[Delta] DEP (-) -0.72188
 0

[Delta] TAX (+) 0.29927 1.57988
 -0.0878 (<.0001)

[Delta] OTHER (-) -1
 (<.0001)

Adjusted.- 0.05 0.0689 0.0368
 [R.sup.2]
No. of 3,656 3,656 3,656
 Observations.

 Dependent Variable = 15 months compound
 market-adjusted return (CAR)

 Predicted High B-T Conformity Countries
 sign
 Model (1) Model (9) Tax Expense

Intercept 0.12621 0.11255 0.11551
 ?
 (<.0001) (<.0001) (<.0001)
[Delta] EARN (+) 0.51582
 (<.0001)

[Delta] SALES (+) 0.1794
 (<.0001)

[Delta] DEP (-) -1.06819
 (<.0001)

[Delta] TAX (+) 1.06976 1.43237
 (<.0001) (<.0001)

[Delta] OTHER (-) -0.14796
 -0.0016

Adjusted.- 0.0207 0.0512 0.0438
 [R.sup.2]
No. of 7,852 7,852 7,852
 Observations.

Table 7: (Panel B) The effect of book-tax conformity
components for high ETR firms

 Dependent Variable = 15 months compound
 market-adjusted return (CAR)

 Model (3)
 Predicted Coeff. t-stat p-value
 sign

Intercept ? 0.02724 3.26 0.0011
 EARN (+) 0.77378 15.96 <.0001
 TAX (+)
 PTI (+)
BTCONF ? 0.09898 9.75 <.0001
BTCONF x [Delta] EARN (-) -0.25796 -3.97 <.0001
BTCONF x [Delta] TAX (+)
BTCONF x [Delta] PTI (-)
Adjusted-R2 0.0414
Fixed Effects None
Number of Observations 11,508

 Dependent Variable = 15 months compound
 market-adjusted return (CAR)

 Model (12)
 Predicted Coeff. t-stat p-value
 sign

Intercept ? 0.01673 2.00 0.0453
 EARN (+)
 TAX (+) 0.45154 3.03 0.0024
 PTI (+) 0.60097 11.8 <.0001
BTCONF ? 0.09751 9.59 <.0001
BTCONF x [Delta] EARN (-)
BTCONF x [Delta] TAX (+) 0.59172 2.94 0.0032
BTCONF x [Delta] PTI (-) -0.40986 -5.59 <.0001
Adjusted-R2 0.0617
Fixed Effects None
Number of Observations 11,508

Table 8: (Panel A) Mean of coefficients from yearly regressions

Regression Low B-T High B-T
Coefficient Conformity Conformity
 Countries Countries

 Mean Mean

[Delta] Earnings 0.6318 0.4088

[Delta] Sales 0.0948 0.0332

[Delta] Depreciation 0.5550 -0.1800

[Delta] Tax 1.9656 1.1815

(Panel B) Mean of explanatory powers of yearly regressions

Mean of [R.sup.2] Low B-T High B-T
of Conformity Conformity
Model for Countries Countries

 Mean Mean

[Delta] Earnings 0.0541 0.0333

[Delta] Sales 0.0271 0

[Delta] Depreciation 0.0061 0.0052

[Delta] Tax 0.0423 0.0322

Regression
Coefficient

 Differences t-value

[Delta] Earnings 0.223 3.18
 (0.0058)

[Delta] Sales 0.0615 4
 (0.0012)

[Delta] Depreciation 0.7348 2.47
 (0.0252)

[Delta] Tax 0.7841 3.49
 (0.0031)

(Panel B) Mean of explanatory powers of yearly regressions

Mean of [R.sup.2]
of
Model for

 Differences t-value

[Delta] Earnings 0.0208 1.93
 (0.0710)

[Delta] Sales 0 3
 (0.0139)

[Delta] Depreciation 0 0
 (0.8423)

[Delta] Tax 0.0101 0.85
 (0.4090)

Table 9: Explanatory powers of earnings components models

 Intercept [Delta] [Delta] [Delta]
 Net Sales Depreciation
 Earnings

For Countries with High Book-Tax Conformity

Model for 0.12887 0.51656
[Delta]Earnings (<0.001) (<0.001)

Model for 0.12571 0.51781
[Delta]Sales (<0.001) (<0.001)

Model for 0.13004 -0.83460
[Delta]Depreciation (<0.001) (<0.001)

Model for 0.12492
[Delta]Tax (<0.001)
Expense

For Countries with Low Book-Tax Conformity

Model for 0.05064 0.63204
[Delta]Earnings (<0.001) (<0.001)

Model for 0.03902 0.64695
[Delta]Sales (<0.001) (<0.001)

Model for 0.04522 -0.02597
[Delta]Depreciation (<0.001) (0.6962)

Model for 0.04275
[Delta]Tax (<0.001)
Expense

 [Delta] [Delta] Adjusted Differences
 Tax Other [R.sup.2] (a) in
 Components [R.sup.2]

For Countries with High Book-Tax Conformity

Model for 0.0412
[Delta]Earnings

Model for -0.48981 0.0450 0.0038
[Delta]Sales (<0.001)

Model for 0.51464 0.0421 0.0009
[Delta]Depreciation (<0.001)

Model for 0.21425 0.44202 0.0473 0.0061
[Delta]Tax (0.0012) (<0.001)
Expense

For Countries with Low Book-Tax Conformity

Model for 0.0491
[Delta]Earnings

Model for -0.58967 0.0576 0.0085
[Delta]Sales (<0.001)

Model for 0.64460 0.0512 0.0021
[Delta]Depreciation (<0.001)

Model for 0.74396 0.51123 0.0615 0.0124
[Delta]Tax (<0.001) (<0.001)
Expense

 Vuong's Sample
 Z-Value size

For Countries with High Book-Tax Conformity

Model for 19,354
[Delta]Earnings

Model for 3.26 19,354
[Delta]Sales (0.0011)

Model for 1.92 19,354
[Delta]Depreciation (0.0550)

Model for 3.44 19,354
[Delta]Tax (0.0006)
Expense

For Countries with Low Book-Tax Conformity

Model for 38,200
[Delta]Earnings

Model for 7.41 38,200
[Delta]Sales (<0.001)

Model for 3.45 38,200
[Delta]Depreciation (0.0006)

Model for 7.83 38,200
[Delta]Tax (<0.001)
Expense

The numbers in parentheses represent p-values.

(a) Differences between R2of the models for each earnings
component and R2of models for earnings