Positive and Negative Bubbles With Asymmetric Extrapolative Beliefs and a Dynamic Trading Population

ABSTRACT We propose a dynamic asset pricing model that features asymmetric extrapolative beliefs and a dynamic trading population to investigate the behavioral drivers of asset bubbles. In the model, extrapolators place more weight on past negative returns than on positive ones when forming their beliefs about future returns, and the trading population varies as investors enter or exit the market in response to past performance. Exit is defined as a complete withdrawal from the market, while previously nonparticipating investors enter through an uncertain process, with an entry probability reflecting this uncertainty. The model generates classical price bubbles with high trading volume as well as negative bubbles following a crash. The results show that the interaction between asymmetric extrapolative beliefs and a dynamic trading population plays a critical role in bubble formation. Moreover, when extrapolators have prospect theory preferences rather than constant absolute risk aversion preference, the bubble effects become more pronounced.