Ohlson (1995)與Feltham and Ohlson(1995,1996)曾使用各種線性資訊模式的假設,推導出公司價值係決定於帳面價值、異常盈餘、投資,及其他資訊四者。本文目的即在從實證角度,探討線性資訊模型理論限制的合理性,以及該評價模式(在台灣)的適用性。 本文以我國股票上市公司民國79年第2季至89年第3季為樣本。針對個別公司之時間序列聯立迴歸模式用Seemingly Unrelated Regressions (SUR)方法做測試。 實證結果發現: (1)整體而言,線性資訊模式在我國適用狀況比美國為佳;例如,就線性資訊模型而言,Myers研究顯示:異常盈餘持續性限制(0< <1) 獲得支持,但保守參數限制(0< <1)、(0< <1)不獲得支持。本文則發現異常盈餘持續性限制(0< <1)與帳面價值保守參數限制(0< <1)均獲得支持;資本支出保守參數(0< <1)不獲支持。(2)就評價模式而言,Myers研究顯示異常盈餘、帳面價值、資本支出以及其他資訊(以截距項替代)對股價之解釋能力達50%。本研究則顯示上述四項變數對股價之解釋能力僅為39%,因此評價模式在我國股市之適用性劣於美國。(3)經以線性資訊模型迴歸係數代入公式所求得的估計價值與真實價值之比例,平均值大多小於1,此與Myers(1999)之結果相同。主要是因為替代其他資訊的截距項大都為負的,導致所設算的價值EV值較小,第二,市場有捕捉到公司未來預期的異常盈餘資料,但會計資料無法捕捉到此一資料。(4)各行業之保守係數有顯著差異,其中水泥窯製業之保守參數最大。固定資產佔總資產的比例與保守係數之間為負相關,此結果與預期假設不符。
Ohlson (1995)and Feltham and Ohlson (1995,1996)develops model of a firm’s intrinsic value as it relates to contemporaneous and future earnings、book value、dividends、investment、or other informations, based on linear information models (LIMS).This study provides an empirical assessment of the necessity and reasonability of using LIMS of the time series of residual incomes, and residual income valuation model proposed by Ohlson (1995)and Feltham and Ohlson (1995,1996) in firms’ listed on the Taiwan Stock Exchange (hereafter TAIEX). We document that the empirical results generally support the residual income valuation model. Specifically, LIMS assumptions, including parameters of abnormal earnings (RI)’ persistence and conservatism, are necessary and reasonable in TAIEX. We also find that accounting-based valuation models fits better for TAIEX than America’ companies. Third, by comparing the implied price coefficients to coeddicients from a price regression, We also find that the models imply ineffidient weightings on the accounting numbers. Furthermore, contrary to our hypothesis, fixed assets ratio is negatively related to coefficient of conservatism.