本研究旨在探討我國上市(櫃)公司之每股盈餘對股價報酬在負債比率變化下是否存在縱橫平滑移轉效果，嘗試運用Gonza′lez, Teräsvirta and Dijk（2004, 2005）之縱橫平滑移轉迴歸模型，驗證每股盈餘對股價報酬在負債比率變化下存在縱橫平滑移轉效果，使得負債比率對股價報酬的影響變化呈現非線性關係，進而對於股價報酬受控制變數的影響進行評估與衡量。 實證結果發現，每股盈餘對股價報酬在負債比率變化下確實存在縱橫平滑移轉效果及非線性關係。研究中發現在負債比率為32.8553%時發生結構性變化，即產生一轉換區間，轉換速度為0.0058156，使得模型在轉換門檻值附近形成一平滑的轉換過程。 控制變數對股價報酬之影響方面，每股盈餘對股票報酬，在負債比率高於或低於32.8553%時，影響結果一致，皆呈現負相關，惟其影響程度不同，高於門檻值之影響程度較低於門檻值之影響程度大。股利率對股票報酬之影響結果亦相當一致，呈現顯著正相關，但當負債比率大於門檻值32.8553%時，股利率對股票報酬之影響程度降低。而移轉變數負債比率本身對股票報酬之影響，實證結果顯示兩者無顯著關係。因此，預計建議投資人擬定投資策略時，不能僅視每股盈餘很高，就盲目地大量長期投資，仍需多參考公司財務面的相關資訊（如：負債比率等），投資人可參考本文結論，對照自身的投資組合，擬定適當的投資策略。 The aim of this study is to investigate the non-linear relationship between earning per share and stock return under variations in level of the debt ratio. We examine the smooth transition effect of the EPS on the stock price by applying Gonza''lez, Teräsvirta and Dijk’s (2004, 2005) panel smooth transition regression model and using debt ratio as the transition variable. The confirmation of the nonlinear effect can further be accurately utilized for evaluating and measuring the impact of certain control variables considered on the stock return. We empirically find that there does exist a panel smooth transition effect on stock return when the level of debt ratio is 32.8553% with a fairly smoothed coefficient of the speed of transition as 0.0058156. The overall findings for those nonlinear effects of control variables on stock return are described as follows. The EPS shows the same negative impacts on the stock return no matter it is higher or lower than the threshold value of 32.8553% for the debt ratio. However, the magnitude of the impact is larger in the higher region of debt ratio. On the other hand, the impacts of dividend payout ratio on stock return are positive for both regions, whereas with relatively smaller impact under the higher debt ratio region. The other interesting finding is that the transition variable shows no significant impact on the stock return under both regions. Our empirically results provide valuable suggestions to the investors that one should not long his position merely based on the EPS. Whereas consulting the related financial information (e.g., debt ratio) of the firms and referring the conclusion of this study might be more appropriate when doing asset allocation and invest strategy setting.
