Do Attention effect and news influence the buying behavior of individual investors in Taiwanese stock market?

本論文係研究關注效果是否會影響台灣地區股票市場個人投資人的買進賣出行為，研究架構與方法主要奠基於Barber與Odean在二零零三年發表的報告，並依據台灣股票市場特性略作修改。以吾人設定的關注效果變數與個人投資人在該個股的買進賣出行為作交叉分析後，研究結果顯示，具有下列幾種關注效果特性的股票，個人投資人較傾向淨買入該個股：該股票在前一日異常低的報酬、前日融資部位的大幅增加、前日融券部位的大幅減少、前日外資係淨買入、前日投信係淨買入、當日有較多報紙報導該股票相關新聞、當日該股票有新聞現報且股價呈現下跌。另外也發現，較不具經驗的投資人：例如未使用融資融券交易的投資人或每筆下單金額較低的投資人，更容易受到關注效果的影響。最後，投資組合的模擬研究發現，個人投資人購買的股票組合之報酬表現較他們所賣出的投資組合為差；但並無顯著的證據顯示受到關注效果所影響的買進行為將導致損失。This thesis is to study how the attention effects influence the net buying behavior of individual investors in Taiwan, the framework and methodology is mainly based on of Barber and Odean (2003), but some adjustments are made according to the characteristics of Taiwanese stock market. After cross-analyzing the attention effect variables of all stocks and the net buying behavior of individual investors on it, this research finds that attention-grabbing variables, such as: previous day extremely negative return, large increase of margin purchase position on previous day, large decrease of short sale position on previous day, previous day net buy position increase of foreign investors and investment trust companies, more newspapers coverage, and news release on negative return days will increase the net purchase of individual investors upon stocks with such characteristics. However, less sophisticated investors, for instance, investors of non-credit accounts or investors with lower volume per trade, tend to be more attention-driven. Finally, the portfolio simulation test finds that the portfolio of stocks individual investors buy generally underperforms the portfolio of stocks that they sell; but there’s no obvious evidence show that the attention-based buying behavior results in loss.1. Introduction 1 2. Related research 3 3. Data 5 3.1 Data of individual investors’ trading activities: 5 3.2 Sorting variables & Sorting methods 5 3.2.1 Return of previous day 6 3.2.2. Current day abnormal trading volume 6 3.2.3. Previous day incremental shares of margin purchase and short sale 7 3.2.4. Previous day net buy and net sale shares of major institutional investors of the previous day 7 3.2.5. Current day company news 8 4. Methodology 10 4.1. Order Imbalance calculation: 10 4.2. Asset Pricing Prediction: 10 5. Results 13 5.1. All the individual investors: 14 5.1.1. Previous day Return Sort 14 5.1.2. Current day Abnormal Volume Sort 14 5.1.3. Previous day Incremental shares of Margin Buy Sort & Previous day Incremental shares of Short Sale Sort 15 5.1.4. Previous day buying or selling positions of Foreign Investors Sort & Previous day Volume of Investment Trust Companies Sort & Previous day Volume of Dealers Sort 16 5.1.5. Current day news sort 16 5.2. Credit accounts vs. non-Credit accounts 17 5.3. Electronic accounts vs. non-electronic accounts 19 5.4. Female investors vs. male investors 20 5.5. Bear market period vs. Bull market period 20 5.6. Asset Pricing Prediction 21 5.6.1. Total individual investors 22 5.6.2. Non-credit accounts 23 5.6.3. Credit accounts 23 6. Conclusion 25 7. References 2