회계이익-세무이익 차이 정보의 재해석 - 이익지속성과 당기 및 차기시장 반응을 중심으로

It is usually accepted that accounting income cannot properly reflect management results according to the manager's discretionary accounting choices and that tax income does not reflect the manager's arbitrariness due to the strict application of tax law. Interest in tax income and BTD has widely expanded since companies of accounting fraud turned out to have very big BTD. Both level and change variables were used as independent variables to increase the explanatory power of a model represented by the adjusted coefficient of determination, but the explanatory power made only a small increase and the current level variables failed to provide enough explanations about the relationships among the three variables as they were divided into prior level variables and change variables. This study analyzed information value of the income persistence and current and next stock price responses of accounting income, and the constituents of accounting income. Accounting income is divided into BTD and tax income, BTD into temporary and permanent difference, and the current level variables of each constituent of accounting income into prior level variables and change variables. The purpose of this study was to investigate if there was information value in the income persistence of each variable of each constituent of accounting income, explanatory power for current excess return and explanatory power for next excess return and if there was any difference in information value if they was any. For that purpose, multiple regression analysis and Mishkin(1983) test were conducted with such independent variables as current level variables, change variables, current level variables and change variables, and prior level variables and change variables. The sample consists non-banking firms(maximum 512 firm-year) with a December fiscal year-end listed on the Korean Stock Exchange over 1999-2007. The data on tax income, temporary and permanent difference, are collected from notes to financial statements directly. And other firm-specific financial data are collected from KIS-VALUE database. Empirical results were as follows : first, change variables had positive correlations with income persistence but had lower explanatory power than current level variables; second, change variables had higher explanatory power for current excess return of accounting information than current level variables, and change permanent difference had huge positive correlations with that; third, it was proven that change temporary difference and change BTD had significant positive correlations with next excess return of accounting information; fourth, the research that used both change variables and current level variables as independent variables would not increase in explanatory power for income persistence and current and next excess return; fifth, a characteristic of information value was proven that change temporary difference and change BTD were underestimated in the capital market; and sixth, the coefficient of current change variables was the difference between the coefficient of prior change variables and that of prior level variables, which means extra attention should be paid to the interpretation of the results of a research that uses both change variables and current level variables as independent variables. The study thus made a contribution to the interpretation development of results of researches that use both current level variables and change variables as well as researches on the information value of BTD, temporary difference, and change variables.