有關經理人為什麼執行股票購回的原因有很多，包括信號發射假說、股票替代假設與管理牙貿會主要假說等。然而目前文獻對股票購回解釋的主流仍屬「訊號理論」，而本文第一個目的就是要明真實購回比宣告是更可靠的訊號。在賽局工具的輔助下，可以挑述在股票市場資訊不對稱的環境中，存在股票真實購回是一個均衡狀態的條件。在一些特定的限制條件下，可以得到賽局的部分分離解，即均衡時，高回質公司且低財務因難成本的公司會選擇股票的真實購回；高品質且高財務困難成本公司與低品質公司將會拒絕股票的真實購回。本文第二個目是要驗理論而中所建立命題的實證意涵，即研究在股票真實購回誓股價的績效。平均而言，真實購回且低財務困難成本的公司展較高的異常績效。此實結果與模型的主張是一致的。

Mangers conduct open market repurchases for many different reasons, including signaling hypothesis, dividend substitution hypothesis, managerial opportunism hypothesis and so on. However, the most mentioned explanation for open market repurchases is the “signaling theory”. The first purpose of this paper is to show that actual repurchase is more reliable signal than its announcement. We use a game theory model to demonstrate that information asymmetries existing in the stock market can result in stock actual repurchase at the equilibrium. Under some specific circumstances, a partially separating equilibrium is derived. That is, the high-quality companies with low financial distress cost will choose stock actual repurchase, while the high-quality companies with high financial distress cost and the low-quality ones will refuse to repurchase stock. The second purpose of this paper is to examine the hypotheses implied in the proposition from our theoretical side. We investigate share price performance following actual share repurchase. On average, repurchasing firms with low financial distress cost do exhibit superior abnormal performance. The empirical results are in accordance with the suggestions in our model.