Beliefs and decision making in asset markets: 3 essays in experimental economics

The experimental approach is used to study both the formation of beliefs and the process of decision making among individuals. Two environments are explored. In both of them a single individual faces nature and is asked to predict the outcome of a random variable. In the first environment, a single individual is asked to predict nature, simulated by a pre-programmed computer, generating a single outcome from a finite set, each period. In the second environment, a group of individuals trade a risky asset in an experimental asset market design. Each trader is asked to predict future prices. In the second environment, unlike the first one, the outcome of the market is a result of aggregate behavior of all traders.; In the first chapter, a learning model is introduced. The model captures the ability of individuals to recognize patterns in nature's previous actions and predict future outcomes as a continuation of the pattern they observed. The parameters of the model are estimated and the model is then compared to existing models. The proposed "Pattern Recognition" model is shown to fit the experimental data better than all other existing models.; In the second chapter, a model is suggested to estimate traders' beliefs in experimental asset markets. According to the model, individuals use previous trends in price changes and assume a repetition of the trends in the future. In addition, it is shown that traders' beliefs are biased as price predictors, although these beliefs are valuable to an observer who tries to predict future prices.; In the third chapter, experimental asset markets with a long time horizon (200 periods) are used to show that many results from experimental asset markets with short horizons (usually 15 periods) are not robust to a lengthening of the time horizon. The existence of multiple bubbles and crashes in experiments with a long horizon rules out explanations regarding the reasons and timing of the crash, such as the 'end-game' effect.