À繫¿¬±¸ Á¦ ±Ç È£ (2012³â 9¿ù)
Asian Review of Financial Research, Vol., No..
pp.930~977
pp.930~977
Investor Sentiment from Internet Message Postings and Predictability of Stock Returns
Dongcheol Kim Korea University Business School
Soon-Ho Kim Korea University Business School
There has been interest in the literature on whether investor sentiment as expressed in messages posted on Internet message boards has predictive power for stock returns. To study this issue, we use more than 32 million messages on 91 firms posted on the Yahoo! Finance message board in the period January 2005 to December 2010. What distinguishes our study is the use of sentiment information explicitly revealed by retail investors for individual firms and a longer sample period relative to other studies that use similar sentiment information source. As a proxy for investor sentiment, we use investor sentiment indexes constructed from sentiment explicitly revealed by retail investors and as classified by a machine learning classification algorithm. In intertemporal and cross-sectional regression analyses, we find no evidence that investor sentiment forecasts future stock returns at either the aggregate or individual firm level. Rather, we find evidence that investor sentiment is positively affected by prior stock price performance. We also find no evidence that investor sentiment from Internet postings has predictability for volatility and trading volume. We find no significant predictive ability for retail investor sentiment for the direction of the next period¡¯s stock price movement across demographic characteristics such as gender, age, and professionality.
Dongcheol Kim
Soon-Ho Kim
There has been interest in the literature on whether investor sentiment as expressed in messages posted on Internet message boards has predictive power for stock returns. To study this issue, we use more than 32 million messages on 91 firms posted on the Yahoo! Finance message board in the period January 2005 to December 2010. What distinguishes our study is the use of sentiment information explicitly revealed by retail investors for individual firms and a longer sample period relative to other studies that use similar sentiment information source. As a proxy for investor sentiment, we use investor sentiment indexes constructed from sentiment explicitly revealed by retail investors and as classified by a machine learning classification algorithm. In intertemporal and cross-sectional regression analyses, we find no evidence that investor sentiment forecasts future stock returns at either the aggregate or individual firm level. Rather, we find evidence that investor sentiment is positively affected by prior stock price performance. We also find no evidence that investor sentiment from Internet postings has predictability for volatility and trading volume. We find no significant predictive ability for retail investor sentiment for the direction of the next period¡¯s stock price movement across demographic characteristics such as gender, age, and professionality.
[ 07327 ] 67-8, Yeouinaru-ro, Yeongdeungpo-gu, Seoul, Korea TEL. +82-2-2003-9921 FAX. +82-2-2003-9979 E-mail. office@korfin.org COPYRIGHT(C) SINCE 1987 KOREAN FINANCE ASSOCIATION. All rights reserved.