本研究探討了投資者是否能夠利用財務報表資訊來找出未來超額報酬可能性更高的投資標的。我的方法將財務報表資訊彙整「預測獲利增長分數」，這個分數預測了一年後超額報酬的可能性。進一步利用預測獲利增長分數的六個財務報表比率對股票超額報酬進行回歸分析，以取得回歸模型並執行選股交易策略。我發現，預測獲利增長分數並不完全適用於台灣股市；六個財務報表比率對股票報酬進行回歸模型也不能完全幫助現在和未來的投資者更有效地進行選股，因此回歸模型也無法百分之百提高獲利預測和股票推薦準確性。在過去的6年樣本期間，預測獲利增長分數最佳投組平均每年實現了19.31%的超額報酬，回歸分析模型最佳投組平均每年實現了8.03%的超額報酬。;This study investigates whether investors can utilize financial statement information to identify investment targets with a higher likelihood of future abnormal returns. The approach involves summarizing the six financial statement information signals into a "predicted earnings increase score”, which predicts the likelihood of future abnormal returns one year ahead. Furthermore, I perform a regression analysis on abnormal stock returns by utilizing six financial statement information signals and implement a stock selection trading strategy. The findings indicate that the predicted earnings increase score is not entirely applicable to the Taiwanese stock market. The regression model utilizing the six financial statement information signals also does not significantly enhance the effectiveness of stock selection for both current and future investors, thereby failing to substantially improve profit forecasting accuracy and stock recommendation precision. Over the past six years, the hedge portfolio consisting of the best-performing investment targets based on the predicted earnings increase score achieves an average annual abnormal return of 19.31%, while the best-performing hedge portfolio based on the regression analysis model attains an average annual abnormal return of 8.03%.
