Explore publications.

Through a thematic analysis, we analyze the evolution of cognitive biases among eight participants (students) during a three-day trading simulation in a market that was perceived as unfavorable. The biases studied include overconfidence, representativeness, anchoring, herd behavior, availability and prospect theory. Prospect Theory, particularly through loss aversion, is strongly present and dominates the experience of most participants. It is perceived as ‘difficult’ even with virtual money. Some manage loss aversion with strict stop losses, while others, when confronted with losses, may take more risks in order to try to ‘recover’. Gains generate moderate satisfaction rather than euphoria. Virtual money seems to reduce the emotions connected with gains and losses for some participants. The Availability Bias is widespread. The participants focus on easily accessible information such as graphs, well-known company names, basic news and widely used indicators. Fundamental analysis is often rejected as too complex, time-consuming or irrelevant.  Anchoring Bias is variable. Some participants use specific limits as reference points. Representativeness bias is generally stable throughout the experiment. It involves using familiarity with companies as a decision-making criterion, or applying principles perceived as effective, such as using well-known technical indicators. Overconfidence varies considerably from one participant to another, influenced by previous experience and results achieved. It can be very low at the beginning, grow with confirmed intuition, or decrease in response to an unfavorable market. In the context of the study, Herd Behavior corresponds to a desire to find comfort in the group of participants in order to overcome a feeling of isolation, rather than a desire to copy behavior. The key findings of our research demonstrate a strong aversion to loss (prospect theory).   

Using a qualitative experimental design underpinned by elements of experimental finance, we analyze decision-making processes on stock markets, using a student population. The various results obtained, based on the tools used (open and closed questionnaires as well as trading journals), demonstrate the presence of certain biases (representativeness bias, rank bias, presence of heuristics and “all that glitters attracts”) and the influence of specific emotions (anger, fear, anxiety and sadness) in decision-making. The conclusions drawn from the analysis allow us to suggest that the development of biases and emotional pressure are strongly conditioned by the general context in which decision-making takes place, as well as by some socio-demographic characteristics of the participants (in particular, gender).