散戶投資人處分效果之研究－－考量公司市場價值下之實證結果

A Study on Disposition Effect of Individual Investors: Empirical Findings Taking into Account Market Valuations

Shefrin and Statman(1985)所提出之處分效果(Disposition Effect)，乃指投資者對於獲利的股票有賣出太快的傾向，對於有損失的股票則顯現持有過久的現象。本研究參照Weber and Camerer(1998)之處分係數，以散戶投資人之融資賣出資料代表其處分行為，實證結果顯示台灣股市之散戶投資人確實存有處分效果。此外，本研究進一步分析投資者心理效果之差異性，即於研究樣本中，選取上市股票市值比重最大與最小者各30家，依報酬率區間、多空走勢、月份等不同屬性加以區隔，作為差異性分析的依據。研究發現，以前述屬性劃分的分析結果，皆顯著存有處分效果。然而加入報酬率區間考量後，則前一週報酬率介於0％～±2％（不含0％）時，處分效果不顯著。至於本研究依序以規模效應、不同走勢期間以及月份效果所提出之假說，除了春節期間（一、二月份）與年終（十二月份）處分效果小於其他月份之驗證無法獲得充分支持外，其餘假說在不同屬性下，實證結果大多得到支持。

The disposition effect proposed by Shefrin and Statman (1985) suggests that investors tend to hold losing stocks too long and to sell winning stocks too early. We employ the disposition coefficient from Weber and Camerer (1998), and use data on the redemption of margin purchase to represent the trading behavior of individual investors in Taiwan’s stock markets. We try to find whether the individual investors in Taiwan’s stock markets indeed exhibit the disposition effect. Moreover, we analyze the different psychological effects of investors by selecting the biggest and smallest 30 listed companies amongst our sample data and categorizing them according to the range of their rates of return, bullish or bearish market, and, the month of the year. The literature has seen two main streams in the study of disposition effect. One involves comparison of unusual trading volumes, which might have positive relationship with the disposition effect. The other is based on realized gains and realized losses to test the disposition effect such as models introduced by Odean (1998) and Weber and Carmerer (1998). Most of previous studies of the disposition effect, however, use Odean’s (1998) model rather than that of Weber and Camerer (1998). In view of this, we use the disposition coefficient (α) of Weber and Camerer (1998) in this study to examine individual investors’ behaviors in bull and bear markets. We propose four hypotheses for our empirical study. The first hypothesis assumes that individual investors do show the disposition effect (α> 0). The second hypothesis supposes that investors tend to show stronger disposition effect for the stocks of small companies than for those of large companies. The third hypothesis presumes that the disposition effect in a bullish market is stronger than in a bearish market. The final hypothesis supposes that the disposition effects in the period of Chinese New Year vacation (January and February) and at the end of the year (December) are significantly less obvious than those otherwise. In other words, we nest the models of Odean (1998) and Weber and Camerer (1998) to test individual investors’ disposition effects via empirical studies taking into account market valuations. Generally, we examine the disposition effect in Taiwan’s stock markets from the perspective of behavioral finance. From the results of our empirical study, we find that the disposition effect does exist within each category except when the rate of return for the previous week falls between 0% and ±2% (0% excluded). Also, we cannot find evidence of significant difference in the disposition effect during the Chinese New Year period (January and February) to those in the rest of the year. Similarly, investors do not tend to show significantly different disposition effects at the end of the year (December) to those during the remaining months.