ERP시스템 도입이 이익조정에 미치는 영향에 관한 실증연구

This study investigates empirically whether the implementation of an ERP system limits the scope of manipulation of financial data by managers and whether this, in turn, effectively reduces manipulation of earnings. As for the stage of empirical analysis, the process was carried out in the order of establishment of research hypotheses, elaboration of a test model, definition and measurement of variables, selection of sample companies, empirical analysis and interpretation of result. In the regression analysis, discretionary accruals is a dependent variable providing measures of earnings manipulation. To measure discretionary accruals, we resorted to the CFO model, the Rangan model, the modified time-series Jones model and the cross-sectional adaptation of the Jones model. For the regression analysis, ERP variable is a test variable. In addition to the test variable, tax benefit, debt-to-equity ratio and increase of capital are chosen as control variables because these variables are likely to affect the direction of discretionary accruals. The analysis in the case of positive discretionary accruals shows that the research hypothesis is supported under the Rangan model and the CFO model at the 5% significance level. On the other hand, the analysis in the case of negative discretionary accruals shows that the research hypothesis is partly supported under the CFO model and modified cross-sectional Jones model at the 10% significance level. In other words, statistically significant reduction in amounts of discretionary accruals is observed after the introduction of an ERP system. In conclusion, the practical implication of the result attained by this study is that in order to create a transparent climate in the korean business world, practical tools such as ERP system are no less important than legal and institutional devices, and that firms may be rightfully encouraged to adopt ERP systems.