王泰昌劉嘉雯臺灣大學：徐政緯2007-11-282018-06-292007-11-282018-06-292004http://ntur.lib.ntu.edu.tw//handle/246246/61778關係企業合併財務報表之編製自民國八十八年起實施，提供投資者制定決策另一參考依據，基於有關關係企業合併財務報表、母子公司合併財務報表與母公司財務報表之研究甚少，因此本研究探討三者間之資訊內涵，期望能找出決策有用性相對較佳之財務報表。故本研究擬以線性多元迴歸模式，比較三種財務報表之財務比率指標對股票超額報酬之解釋能力，作為三種財務報表孰具相對有用性之依據。 本研究所建構之多元迴歸模型，定義應變數為股票超額報酬，計算上採市場調整模式（以年資料計算）和市場模式（以日資料計算）兩方法，自變數則是參考Lev & Thiagarajan（1993）研究中所定義之財務比率指標，選取可由TEJ擷取的7項指標加上簽證會計師事務所指標，作為本研究模型一之自變數；另外以我國證期會於證券發行人財務報告編製準則中要求揭露之財務比率選取15項，作為本研究模型二之自變數。資料取自台灣經濟新報社資料庫（TEJ），篩選民國89年至91年有同時編製三種財務報表之上市上櫃公司作為研究對象。迴歸測試首先以判定係數作出初步判斷，復以JA-test作統計上之檢定，並試著找出有解釋能力之解釋變數。 實證結果如下： 一、在模型一中若以市場調整模式計算超額報酬時，發現關係企業合併財務報表、母子公司合併財務報表與母公司財務報表，三者財務比指標對股票超額報酬之解釋能力，並無相對較佳之模式；在模型一中若以市場模式計算累積超額報酬時，其結果亦同。 二、在模型二中若以市場調整模式計算超額報酬時，僅發現關係企業合併財務報表所建構之模型優於母子公司合併財務報表所建構之模型，且具有統計上之顯著性，即僅有假說三成立；在模型二中若以市場模式計算累積超額報酬時，發現關係企業合併財務報表所建構之模型優於母子公司合併財務報表及母公司財務報表所建構之模型，且具有統計上之顯著性，即假說一和假說三成立。 三、僅有在模型二的模式下且股票超額報酬以市場模式計算時，可以找到解釋變數中有EPS和IBTR能有效解釋股票超額報酬，其餘情況下皆無任何一個解釋變數可以有效解釋股票超額報酬。 四、實證結果因模型不同和股票超額報酬計算方式不同，而出現不同之結果。Consolidated financial statements of affiliated corporations have started to be compiled since 1999, and it can provide investors with another reference when they want to make decisions. This study probes into information content about three kinds of financial statements: Consolidated financial statement of affiliated corporations, consolidated financial statement, and separate financial statement of parent corporations because related researches are very deficient. This study hopes to find out the best financial statement in decision usefulness. This study uses linear multiple regression analysis to compare the explanatory power of financial ratios to stock excess returns among three kinds of financial statements. The conclusion of this study could be drew inferences about which kind of financial statement has better decision usefulness. The multiple regression models in this study define dependent variable as stock excess returns, which are calculated in two methods: the market-adjusted model (using annual data) and the market model (using daily data). This study defines independent variables as seven financial ratios and signals chosen from TEJ and CPA firm signal in accordance with the research of Lev & Thiagarajan (1993), and these financial ratios and signals are the independent variables of model 1 in this study. Besides, this study also chooses fifteen financial ratios based on the regulations of SEC in Taiwan as the independent variables of model 2. The data are gathered from TEJ, and the sample is restricted to the public corporations that had issued the three kinds of financial statements simultaneously from 2000 to 2001. This study uses multiple regression analysis. At first we use coefficient of multiple determination to make initiated judgement; then we use JA-test to do statistical estimates. Finally we try to find out independent variables that have explanatory power. The research findings are as follows: I. If stock excess returns are calculated based on market-adjusted model in model 1, none of the three kinds of financial statements is the best about explanatory power of financial ratios and signals to stock excess returns. If cumulative abnormal returns are calculated based on market model in model 1, the results are the same. II. If stock excess returns are calculated based on market-adjusted model in model 2, the result is that the model established by consolidated financial statements of affiliated corporations is more excellent than the model established by consolidated financial statements and it is statistically significant. That means hypothesis 3 is tenable. If cumulative abnormal returns are calculated based on market model in model 2, the result is that the model established by consolidated financial statements of affiliated corporations is more excellent than the model established by consolidated financial statements and separate financial statements of parent corporations and it is statistically significant. That means hypothesis 1 and 3 are tenable. Ⅲ. Only if cumulative abnormal returns are calculated based on market model in model 2, among these independent variables, EPS and IBTR can explain stock excess returns effectively. None of these independent variables has explanatory power in other situations. Ⅳ. The empirical results are different because of discrepancy in models and calculation methods of stock excess returns.第一章 緒論 1 第一節 研究動機與目的 1 第二節 研究流程與架構 5 第二章 文獻探討 8 第一節 合併財務報表之相關規章法令 8 第二節 合併財務報表資訊內涵之實證研究 24 第三節 基本分析與超額報酬之實證研究 34 第三章 研究方法 47 第一節 研究假說之建立 47 第二節 研究設計與變數衡量 51 第三節 實證方式與檢定方式 70 第四節 資料篩選與樣本特性 77 第四章 實證研究結果 84 第一節 自變數之基本敘述統計 84 第二節 多元迴歸模型檢定與分析 93 第三節 個別迴歸參數檢定與分析 103 第四節 JA -test檢定結果 106 第五章 研究結論、限制與建議 115 第一節 研究結論 115 第二節 研究限制與建議 119 參考文獻 121639098 bytesapplication/pdfen-US財務報表股票超額報酬財務比率financial ratiosCARfinancial statementsotherhttp://ntur.lib.ntu.edu.tw/bitstream/246246/61778/1/ntu-93-R91722029-1.pdf