Page Manager: Webmaster
Last update: 9/11/2012 3:13 PM
Authors |
Maria Andersson Martin Hedesström Tommy Gärling |
---|---|
Published in | Paper presented at the 10th meeting of the European Social Cognition Network (ESCON), Volterra, Italy. |
Publication year | 2008 |
Published at |
Department of Psychology |
Language | en |
Keywords | Herding, financial markets, accuracy, social influence processes |
Subject categories | Psychology |
Herding in financial markets refers to that investors influence each other when making investment decisions. Previous studies of herding have shown that a majority is more influential than a minority. This result is in line with theories of social influence, arguing that majorities in general are more influential but consistent minorities may also exert influence. The aim of this research is to investigate what factors may impact the level of influence from minorities and majorities in financial markets, focusing on accuracy and focal attention. Experiment 1 investigates the impact of the size of the herd (majority vs. minority) and the accuracy in its predictions. Participants’ task was to predict fictitious stock prices in 50 trials, conditional on information about the current price and predictions made by five fictitious other participants. Either four (majority) or two (minority) of the others’ predictions were correlated (rs > .95), and they made either random (uncorrelated with the price trend, rs < .20) or non-random predictions (correlated with the price trend, range of rs equal to .65 to .85). The results showed that the majority exerted more influence than did the minority, more when the majority made non-random than random predictions. The correlation and accuracy of predictions thus increased the majority influence, but not the minority influence. The aim of Experiment 2 was to investigate whether the majority influence would decrease if participants were instructed to focus on the others’ performance. Conditions in which the majority made random or non-random predictions were included. After a number of trials participants were requested to state either which of the others that made accurate predictions or which of the others that made predictions that were correlated. The results showed that focus on accuracy reduced majority influence.