Note—Revising Forecasts of Accounting Earnings: A Comparison with the Box-Jenkins Method
通过实验比较了Theil最优线性修正技术对季度Box-Jenkins及其他简单模型预测的会计盈余进行修正的效果，发现修正后预测更准确，且简单外推模型优于Box-Jenkins模型。
The purpose of this study was to contribute to the literature concerning forecasting the time series of accounting earnings. To accomplish this objective an experiment was conducted to compare the performance of Theil's Optimal Linear Correction technique for revising quarterly Box-Jenkins and other naive model forecasts of accounting earnings against the unrevised forecasts. Several results of this study are of particular interest. First, the study indicated that the Watts-Griffin parsimonious model outperformed other firm specific Box-Jenkins models. Second, the Optimal Linear Correction produced revised forecasts that were uniformly more accurate than the original unadjusted forecasts. Finally, the naive extrapolative time series models outperformed Box-Jenkins forecasts of accounting earnings.