自古以來，月亮的週期就和人類的行為息息相關。在台灣這塊慣於使用農曆的土地上，月亮週期更存在一定的影響。雖然目前科學尚無法證實月亮對人類的直接影響，但已有不少理論以間接的方式來說明月亮效應的發生。在行為財務學逐漸發展下，月亮和人類行為的關係逐漸引申至月亮週期和股價報酬之間的關係。本研究試圖以傳統理性財務資產訂價模型-公司特徵因子模型，以控制傳統代表理性決策的財務變數，並同時控制週一、週末及月份等時間因素，以求較少的時間干擾。採用縱橫資料方法(Panel Data Methods)以考慮最多的樣本，藉以檢驗台灣證券交易所上市公司之股價報酬，討論月亮效應所造就之非理性行為決策在台灣的影響。 本研究使用由台灣經濟新報資料庫(TEJ)所取得之股價日報酬資料。實證發現，滿月與新月在台灣的證券市場的確存在影響，且於所討論之民國84年1月1日至民國96年12月31日區間內，越接近滿月日則負面報酬影響越強，但新月區間影響相對較弱，且轉為正向報酬影響的速度也較快；另外發現滿月效應與產業的差異而有不同的表現，顯示產業間存有其他變數干擾月亮效應之影響；就台灣市場而言，月亮效應對於大市值公司的影響較顯著。本研究雖無法證實月亮對股價報酬的直接影響，但在統計上提供月亮週期影響股價報酬的證明，於此基礎之下，試圖推論在台灣股價報酬的月亮效應可能是由哪些間接因素所造成。 Since ancient times, the lunar cycles have been closely related to human behavior. In Taiwan, because Taiwanese are accustomed to use the lunar calendar, so there is a certain impact on people in Taiwan with the lunar cycles. The modern science is still unable to confirm how the moon impact on human, although there are many indirect ways illustrate that the effect of the lunar effect really occurred. Accompanied with the development of Behavioral Finance, the relationship between the moon and the human behavior are gradually extended to discuss about the lunar cycles and the stock return. This study attempted to use traditional financial assets pricing model, which control the company characteristics variables introduced by Fama and French (1992). And this study controlled some time variables (Monday, Weekend and Month) to avoid misinterpreting the effect of lunar cycle. In order to include more information of samples, this study use panel data methods. By discussing Taiwan listed companies’ stock return, this study tests the relationship between the lunar cycles and the abnormal stock returns resulted from irrational investment decision-making. The daily return data of this study was obtained from Taiwan Economic Journal’s (TEJ) database. This study finds that in full moon and new moon periods, there are some special effects on Taiwan security market. During January 1995 to December 2007, this study finds that when observed period is closer to full moon days the negative effects on stock returns is more significant. But the effects on returns in the new moon period are weaker than full moon period. And the effect is rapidly changed to positive return effect with the extension of new moon sample interval. Also, the moon has different effects across different industries. In the extension research of this study, lunar effects are significantly in company with large market value. Although we are unable to comfirm the direct effects of moon on stock return, the statistics results here are available to proof the effect of lunar cycles on stock returns. Under this basis, this study try to infer what kind of indirect effects of lunar cycle in Taiwan will affect stock returns.
