| Scope and Method of Study. This study compared accounting income measures produced by the temporal and current rate translation methods against a current (market) value income measure (control treatment). The superior translation method generated an accounting income signal that was the best predictor of the control treatment. A simulation model was used to generate financial data for subsidiaries located in Canada, France, West Germany, Japan, Switzerland, and the United Kingdom. Three industries with various proportions of inventory and fixed assets to total assets were simulated. Sensitivity analysis included using various debt-equity and dividend payout ratios.;Findings and Conclusions. Dunnett's procedure for multiple comparisons with a control treatment revealed that none of the accounting income measures were significantly different from the control treatment at the .05 level. However, the current rate method performed better than the temporal method using simple rankings of (1) mean prediction error and (2) mean absolute differences. While the rankings indicated the current rate method was the best method, the accounting treatment of translation adjustments was contextual with respect to the currency environment. In a depreciating U.S. dollar environment, the current rate method with translation adjustments included in income outperformed the other methods. In contrast, the current rate method with translation adjustments excluded from income performed the best in an appreciating U.S. dollar environment. Thus, this study offers no conclusive evidence for or against the recognition of translation adjustments in income. |
