自 1990 年以來,全球各地發生多起風險管理不當、掏空資產、財報不實、炒作、操縱公司股票,而且造成重大影響的案例,這幾起地雷股事件均嚴重危及金融體質及投資信心,因此若能建構一個能事前或即時偵測的動態財務危機預警模型相形重要。本文以台灣股市中曾經發生的著名地雷股案例,分為「炒作股票」、「掏空、財報不實」公司,選取 15 個案例,及「一般」、「有題材」公司,選取 12 個案例,在不考慮基本面因素,而以個股股價為基礎,以價格泡沫形成的理論為架構,利用資產發生泡沫時在資產價格上所反應的動態性質所建構的改良式單根檢定及遞迴迴歸法,驗證股價是否因投機炒作造成股價緩步墊漲,形成如價格泡沫一般的輕微發散的動態過程趨勢。研究顯示,實證結果與實際發生情形符合,檢驗出投機炒作所造成的股價緩步墊漲與破滅時點,與該公司事件發生時點吻合,因而歸結出:以遞迴迴歸模型檢驗股票投機炒作的有效性,值得投資大眾與監理單位做參考。 How to construct an early warning system for financial distress or crises has always been an important issue for both the academics as well as in practice for the industry and governmental supervision. In the essay, we propose to employ the newly-developed econometric test for testing price bubbles in financial assets as a tool for detecting the early stage of speculative manipulations in equities. By using solely the informatioin form the price dynamics without any financial variables in ratios, the test has been shown to be informative in characterizing a locall upward trending increments in the price series. Thus we believe it shares some potential in identifying the existence of speculative manipulation of stock prices. For comparisons, we also examine the efftiveness of this approach by exploiting three other types of firms in the Taiwan Stock Exchange. Our results confirmed our conjectures and suggest that this method could be further considered by supervisors for disclosing relevant warning information about some equities.