摘要
本研究即首次利用Feltham and Ohlson（1995）線性模式及評價模式推導出適合符合實際情境的會計線性模式及評價模式，將其他資訊以產業進入障礙、市場佔有率、產業集中度作為衡量變數，來檢定商譽之決定因素 本研究預期之主要貢獻為：第一、以往國外文獻對商譽之研究，多以評價模式檢定購入商譽是否會反應於權益價值，本研究試圖以產業經濟學的觀點，推導符合實際情境的會計線性模式及評價模式，來檢驗商譽之決定因素。第二、國內研究對文獻對商譽之研究，著重於研究發展投資對公司商譽之影響(如劉正田，民國88年)，本研究為國內第一篇探討產業經濟學及保守會計對商譽影響之實證研究。第三、Feltham and Ohlson（1995）模式均假定其他資訊（other information）與股權評價有關，但並未將其他資訊做明確的定義。有關會計基礎評價模式之實證研究，亦將其他資訊視為殘差項，但此會使殘差項具序列相關特性，進而造成係數之估計偏誤。藍心梅(民國90年)將其他資訊視為常數項處理，本研究試圖將其他資訊以影響商譽之決定因素代入，並據此建構一個會計基礎評價模式。 本研究實證結果彙述如下： 一.檢測假說 1.產業集中度(HHIt)，產業進入障礙(PPEt/TAt)與投資報酬率(ROEt)正相關(即α3,α4>0)，無法獲得支持；而與未認列商譽(UNAt)正相關(β3,β4>0)，可獲得支持。 二.檢測假說 2.市場佔有率(MSt)與投資報酬率(ROEt)正相關(即α2>0)，獲得支持；而與未認列商譽(UNAt)正相關(即β2>0)，未獲得支持。 三.檢測假說 3.成長機會愈大，則α0愈大，α1愈小；β0愈大，β1愈小。均獲得支持。 四.將樣本資料分為高成長機(HPB)與低成長機會(LPB)二組，分別檢驗假說4a，負盈餘的公司，α0大，α1小；β0大，β1小，結果僅高成長機會的一組企業獲得支持。檢驗假說4b,發布壞消息的公司，α0大，α1小；β0大，β1小，結果僅β0大，β1小獲得支持。檢驗假說4C，業內收益比率比例高的公司，α0大，α1小；β0大，β1小。結果大部分不獲得支持。
並列摘要
Firstly using Feltham and Ohlson(1995) linear and valuation models, deduced are the accounting linear and valuation models suitable for actual circumstances at the paper. With the information like industrial penetration barrier, market share and industrial concentration as measurement variable, the decisive factors of goods reputation are inspected. The expected findings can be summarized as follows: Firstly, the studies on any reputation at foreign papers focused on the reaction on equity’s value through the valuation model inspection on purchasing goods reputation. However, this paper is trying to deduce the accounting linear model and valuation model corresponding to real situations from the view of industrial economics and then inspect the decisive factors. Secondly, domestic studies on goods reputation emphasized on the influence of research & development investment on goods reputation (e.g. Liu, Cheng-Tien, 1999). This paper is the first empirical study the influence of industrial economics and conservatism accounting on goods reputation. Finally, Feltham and Ohlson(1995) models both presumed other information and stock-holder valuation are related, but no precise definition for other information. Regarding the empirical study on accounting-based firm valuation model, the residual items are also treated as other information. But that will impose serial relative characteristics on the residual items and introduce estimate error at the coefficients. Lain, Shin-Mei(2001) takes other information as constant terms. This paper tries to include the other information as the decisive factors of goods reputation and build up an accounting-based valuation model. By utilization of linear model (empirical model-1) and valuation model (empirical model-2), this paper incorporates investment return rate at next term and current unrecognized goods reputation as individual variables, so as the current investment return rate and non-accounting information items(including market share, industrial concentration & industrial penetration barrier) as variables. Besides the discussion of the effects of accounting & non-accounting information to earning and goods reputation, it further determines the explanation capability and visibility of each variable. The empirical results are summarized as below: 1. Examine hypothesis 1: Positive relation (α3,α4>0) of Industry Concentration (HHIt), Entry Barrier (PPEt/TAt) and Investment Return Rate (ROEt) cannot be sustained, but(β2>0) confirmed at Unrecognized Goods Reputation (UNAt). 2. Examine hypothesis 2: Positive relation(α2>0) of Market share(MSt) and Investment Return Rate(ROEt) is sustained, but not for UNAt (β2>0). 3. Examine hypothesis 3: The more of growth opportunity, then the larger forα0 but smaller forα1; and larger forβ0 and smaller forβ1. They are all sustained. 4. Divide sample data into high growth possibility (HPB) and low growth possibility (LPB), two groups. Then examine hypothesis 4a. For firms with negative earning, larger forα0, smaller forα1; larger forβ0, smaller forβ1. The result for HPB firms is sustained. Examine hypothesis 4b, the firms with bad forecast, larger forα0, smaller forα1; larger forβ0, smaller forβ1. The result for largerβ0 and smallerβ1 is sustained. Examine hypothesis 4c, for firms with higher rate at operation income, larger forα0, smaller forα1; larger forβ0, smaller forβ1. The result cannot be sustained mostly.