財務比率資訊內含之實證研究

Abstract

以前的財務會計研究大部份都使用簡單迴歸分析來探討盈餘和股價的關係，因此這些研究可能會有遺漏重要自變數的問題。為了彌補以前研究報告的不足，本黨研究使用多元迴歸分析和等級偏相關分析來探討盈餘與非盈餘財務比率的資訊內容。兩種統計分析一致顯示：盈餘比率擁有最顯著的資訊內含。多元迴歸分析結果顯示：盈餘比率不但可預測股價調整的方向，更可以預測調整的幅度。但是個別的非盈餘財務比率則不具有資訊內含，即使所有的非盈餘財務比率一起考慮，他們也沒有提供額外的資訊內含。但是等級偏相關分析結果則顯示：非盈餘財務比率，不論是個別考慮或整體考慮，確實都擁有顯著的資訊內含。二種統計方法結果的不一致可能來自於這二種統計方法不同的基本假設。

Previous financial accounting research investigating the earnings response models employed the simple regression analysis which might have committed variable omission problems. This study used multiple regression and ranked partial correlation analyses to investigate the information content of both earnings and non-earnings financial ratios. Earnings were consistently found to have possessed the greatest information content using the either analysis. The multiple regression provided evidence that not only the direction of earnings` impact could be predicted, but also its magnitude. But the multiple regression analysis found non-earnings ratios individually to have no incremental value. The partial F-test was used to examine whether or not non-earnings variables, taken as a whole, had incremental content beyond earnings. It failed to find supporting evidence. However, ranked partial correlation analysis found that non-earnings ratios, individually and jointly, possessed information content. The inconsistent results seemed to arise from the different underlying assumptions of each statistical procedure.