投資人情緒與股票報酬之研究

Abstract

我國證券市場個別投資者參與者眾，平均而言，其成交金額佔全體比重將近七成，過去實證研究認為個別交易者為雜訊交易者，對市場價格不具有系統性之影響力。有鑑於此，本研究針對我國證券市場交易現況，透過主成分分析法建構一投資人情緒指標，以探討投資人情緒對於橫斷面股票報酬績效之影響。此外，本研究進一步採用公司規模與風險、獲利能力與股利發放、有形資產與無形資產、成長機會及負債比率等公司特徵變數進行分析，以了解投資人情緒對股票報酬之影響是否因不同公司特徵而存有差異。 實證結果發現，投資人情緒指標對於股票報酬具有預測能力。就公司特徵而言 ，對投資人越難評價之公司投資組合，包含股票波動性高、公司規模較小、未具帳面獲利或未發放股利、有形資產比重較低、無形資產比重較高、負債比率較高等，其與相對應公司特徵變數之當期投資組合報酬率差異均與前期(期初)投資人情緒呈現負向之關係。在IPO超額報酬部分，若前期(期初)投資人情緒越高，則當期IPO公司投資組合之超額報酬越高。其次，我國投資人情緒對於股票報酬之影響在高投資人情緒期間有較強烈的反向回饋效果，此一發現可以提供投資人擬定投資決策時機之參考。

Accounting to statistics of Taiwan Stock Exchange, the trading scale of individual investors in Taiwan was up to 70%. Traditional financial theory has viewed individual investors as noise traders and they had no significant influence on the market price. This research examines how investor sentiment influences the cross-sectional stock returns in Taiwan stock market. Firstly, this study use principal component analysis to form a composite index of investor sentiment to capture the fluctuation of stock returns. Secondly, this study use investor sentiment as an explanatory variable in regression model to see if the influence of investor sentiment would change along with firms’ characteristics. Empirical results show the investor sentiment could predict stock returns in Taiwan. In terms of firms’ characteristics, the subsequent portfolio returns differences for those are even difficult to reasonably evaluate were negatively relative to investor sentiment of prior period, including firms of high volatility, small scale, unprofitability, non-dividend-paying, lower tangible assets, high R&D expenses and high debt ratio. Furthermore, the current period stock returns for IPO stocks would be positive with the prior period investor sentiment. Finally, the negatively relation between investor sentiment and subsequent stock returns was even stronger when the investor sentiment of prior period is high. These findings could be helpful for investors to make appropriate investing decisions.