Predictive Value of Stock Message Board Sentiments

Posted: 22 Nov 2006 Last revised: 11 Nov 2008

See all articles by Bin Gu

Bin Gu

Arizona State University (ASU) - Department of Information Systems

Prabhudev Konana

University of Texas at Austin - Department of Information, Risk and Operations Management

Alex Liu

University of Texas at Austin - College of Engineering

Balaji Rajagopalan

Oakland University - School of Business Administration

Joydeep Ghosh

University of Texas at Austin - College of Engineering

Date Written: November 10, 2006

Abstract

Online brokerages and message board providers have introduced community stock sentiments to influence investors in their trading decisions. Community stock sentiments reflect the average opinions of investors to either buy (bullish) or sell (bearish). However, it is unclear if these sentiments have informational value. This paper studies the predictive power of message board sentiments over future abnormal stock returns. The results show that message board sentiments have a negative but insignificant impact on stock abnormal returns in the subsequent two weeks. However, the negative impacts become both statistically and economically significant in the third week. Moreover, sentiments of investors who have correctly predicted recent stock movements are more negatively associated with future stock movements. The findings are consistent with psychological theories suggesting investors overreact to news and those who happen to predict correctly in the past are more likely to overreact. Based on the analysis, a trading strategy that involves buying stocks with low sentiments while selling stocks with high sentiments was implemented. The results indicated weekly returns ranging from 0.44% to 0.66%.

[Note: We discovered an error in coding in the earlier paper with the same title. The abstract and paper have been updated.]

Keywords: Investor sentiments, efficient markets, virtual community, stock message boards

Suggested Citation

Gu, Bin and Konana, Prabhudev and Liu, Alex and Rajagopalan, Balaji and Ghosh, Joydeep, Predictive Value of Stock Message Board Sentiments (November 10, 2006). McCombs Research Paper No. IROM-11-06. Available at SSRN: https://ssrn.com/abstract=946498

Bin Gu (Contact Author)

Arizona State University (ASU) - Department of Information Systems ( email )

Tempe, AZ
United States

Prabhudev Konana

University of Texas at Austin - Department of Information, Risk and Operations Management ( email )

CBA 5.202
Austin, TX 78712
United States

Alex Liu

University of Texas at Austin - College of Engineering ( email )

1 University Station
Austin, TX 78712-1179
United States

Balaji Rajagopalan

Oakland University - School of Business Administration ( email )

Rochester, MI 48309-4401
United States

Joydeep Ghosh

University of Texas at Austin - College of Engineering ( email )

1 University Station
Austin, TX 78712-1179
United States

Register to save articles to
your library

Register

Paper statistics

Abstract Views
3,817
PlumX Metrics