Further Evidence on the Usefulness of Simulated Mergers
通过模拟合并真实独立公司来扩充多部门企业数据，验证该方法能否代表实际合并结果，对研究企业分部门财务预测的学者有参考价值。
Simulated mergers of existent autonomous firms have been proposed as a way of providing new and additional data for segmentation research (e.g., Silhan [1982], Hopwood, Newbold, and Silhan [1982] and Silhan [1983]). In effect, multisegment firms are created by hypothetically merging a sample of actual single-product companies into various nsegment combinations. This approach makes it possible to expand the available data beyond those firms actually reporting segmental data and to avoid various problems created by segment ambiguities, intersegment transfers, and common cost allocations. On the other hand, it is possible that the firms used as surrogate segments may not very well represent the reporting segments of actual multiproduct firms. Silhan [1983] provided some initial evidence suggesting that the empirical results obtained with simulated mergers are representative of the results obtained with actual mergers. In that study, quarterly sales and margins were used to generate ARIMA forecasts of consolidated (CN) and segmented (SG) earnings after taxes. The results appeared to confirm at least partially the results of Kinney [1971] and Collins [1976] which were based on reported annual data and a very different set of forecasting models.'