Confirmation Bias in Financial Markets: Impact on Trading Behavior

This thesis studies the existence of confirmation bias in financial markets using real trading data and survey results. With this unique data set, I am able to observe whether and how confirmation bias may affect trading behaviors of investors in financial markets. Survey respondents are asked to forecast which one out of six listed firms will achieve the highest return next month. Some respondents were also asked to provide justification when forecasting. The sample used for this experiment comprised of student traders enrolled in a semester-long course who were surveyed prior to company contract trading at the Iowa Electronic Markets (IEM).This thesis examines whether forecasting unconsciously altered their trading behavior, resulting in additional confidence towards their initial forecast, which would indicate signs of confirmation bias. I find that traders that forecasted Microsoft to earn the highest return also ended up with higher net aggregate positions in the Microsoft contract, traded at the IEM. This finding is consistent with confirmation bias affecting investment decisions of traders in my sample. The main limitations of this study are the small sample size and differences in the popularity across firms traded at IEM that could bias the main results.