企業年報之主要用途為協助使用者制定決策,因此企業年報之決策攸關性一直是受到關注的研究主題。企業年報中會計敘述性資訊較數字資訊更能提供與未來攸關或具有前瞻性之資訊,但關於企業年報的研究一直以來大多聚焦於財務報表的數字部分,而非文字資訊,忽略了敘述性會計資訊所隱含的資訊價值。 本研究旨在以可讀性指標及公式量化企業年報文字資訊之可讀性,探討年報可讀性與經營績效、盈餘持續性、以及其他財務變數之關聯性,以檢驗管理模糊假說(management obfuscation hypothesis)。 本研究係以荊溪昱(民84)發展之可讀性公式為基礎,將我國上市及上櫃公司民國88年至97年年報文字資訊量化為可讀性指標,並以Li(2008)之研究模型為基礎,分析我國企業年報可讀性與當期經營績效之關聯性,並進一步探討其與盈餘持續性之關聯性,依本研究分析結果顯示:1.年報可讀性度與當期經營績效呈正向關係,顯示當期經營績效較差之企業,其年報可讀性較績效高者差;2.年報可讀性與盈餘持續性呈正向關係,顯示企業對於持續性較短暫的好消息,企業會以較複雜之方式表達年報資訊。
The main purpose of annual reports is to help users make decisions. Hence, prior researches have focused on the decision relevance of annual reports for a long time. Compared with quantitative accounting data, textual information in annual reports contains much more forward-looking information, thus is more relevant to decision-making. However, most of prior studies focused on quantitative accounting data instead of textual information. This study attempts to test management obfuscation hypothesis by investigating the relationships between textual information in annual reports, performance, earnings persistence, and other financial variables. We use the readability formulas developed by Si-Yu Jing(1995)to calculate the readability of annual reports, and use the model of Li(2008)to investigate the relationships between annual reports readability and performance. Furthermore, we investigate the relationships between annual reports readability and earnings persistence. We find that annual reports issued by firms with lower earnings or less persistent positive earnings are harder to read.