本研究旨在探討台灣股票市場中，長期陷入經營困境後成功扭轉局勢、由虧轉盈之「轉機股」是否具備超額投資報酬，並進一步剖析影響異常報酬之關鍵因素。本研究以台灣上市櫃全產業（排除金融業）公司為篩選標的，量化定義轉機股為「連續四個年度每股盈餘（EPS）小於零，且第五個年度達到損益兩平或開始獲利」之企業，由2006年起到2024年，建立19個年度投資組合；為了避免倖存者偏誤（Survivorship Bias），同時篩選TEJ已下市櫃資料庫中，在研究周期內仍可以在市場正常買賣的公司，上市櫃以及已下市櫃公司共計篩選出396個轉機股。實證方法採用事件研究法，計算虧轉盈資訊公告前後的累積異常報酬（CAR）與買進持有異常報酬（BHAR），並建立三組情境稍有不同的迴歸模型檢定財務困境嚴重程度（Z-score）、自由資產、股東權益報酬率（ROE）、股價淨值比、公司規模、董監持股比例以及股權集中度變動等變數對超額報酬之影響。 研究結果顯示，台灣股市確實存在跨越「EPS = 0」的門檻效應，轉機股在盈餘資訊公告月後 3 個月內（研究窗期 (1，3)）以及公告月前2個月到公告月（研究窗期 (-2，0)）存在顯著大於零的超額報酬，有明顯的市場過度反應效應與盈餘慣性現象（PEAD），PEAD現象約莫就三個月左右。若未經進一步策略選擇轉機股而長期持有至一年半或兩年，投資組合將面臨大幅度累積虧損之風險。 在關鍵影響因素方面，實證指出低股價淨值比與小公司規模是延長轉機股獲利天期與提升超額報酬的核心基本面因子，比較特別的是，財務困境嚴重程度Z-score的反向影響相當明顯，這三個變量的顯著性、趨勢性與穩定性在三組迴歸分析結果的交叉比對中都較好，此實證結果顯示，Z-score較低財務風險稍高、股淨比較低估值較低且規模較小的企業，投資報酬率較高。本研究發現，轉機股的超額報酬主要受到風險溢價與價值低估的共同驅動，而透過股價淨值比、公司規模以及Z-score幾個要素的篩選，確實可以提高投資勝率與報酬率。;This study investigates whether "turnaround stocks"—defined as firms transitioning from chronic financial distress to profitability—generate abnormal investment returns in the Taiwan stock market， and analyzes the key determinants influencing these excess returns. Utilizing an empirical sample of listed， OTC， and delisted non-financial firms from 2006 to 2024， we construct 19 annual portfolios. Turnaround stocks are quantitatively designated as firms experiencing negative earnings per share (EPS) for four consecutive fiscal years followed by a breakeven or profitable fifth year. Delisted firms from the TEJ database are explicitly incorporated to eliminate survivorship bias， yielding a total sample of 396 turnaround stocks. Methodologically， an event study approach is deployed to compute cumulative abnormal returns (CAR) and buy-and-hold abnormal returns (BHAR) surrounding earnings reversal announcements. Three econometric regression specifications with distinct situational scenarios are further constructed to examine the impacts of financial distress severity (Altman′s Z-score)， free assets， ROE， price-to-book (P/B) ratio， firm size， insider ownership， and ownership concentration changes on these abnormal returns. The empirical findings reveal a distinct threshold effect surrounding the "EPS = 0" boundary in the Taiwanese equity market. Turnaround stocks exhibit statistically significant positive abnormal returns in event windows (1， 3) and (-2， 0)， providing clear evidence of a market overreaction effect and a post-earnings-announcement drift (PEAD) phenomenon that persists for approximately three months. Conversely， holding these turnaround portfolios long-term for 1.5 to 2 years without strategic fundamental screening exposes investors to severe cumulative losses. Regarding key determinants， a low P/B ratio and small firm size serve as core fundamental factors extending holding profitability and maximizing excess returns. Notably， the inverse impact of the Z-score is highly pronounced; the significance， directional consistency， and robustness of these three variables are strongly validated through robust cross-comparisons across all regression specifications. This evidence demonstrates that firms with lower Z-scores (higher financial risk)， lower P/B ratios (market undervaluation)， and smaller market capitalizations yield superior investment performance. This study concludes that turnaround stock abnormal returns are jointly driven by risk premia and market undervaluation， and screening via P/B ratio， firm size， and Altman′s Z-score effectively optimizes investment win rates and return performance.
