Agents’ Rationality
Recently, there has been a wide range of literature dealing with the rationality of investors. These studies try to examine how investors in financial markets - both professional and individual investors -- actually behave.
Many researches in the area demonstrate investor behavior that is difficult to reconcile with rationality or with predictions of modern finance models. In one example, Benartzi and Thaler analyzed retirement savings plans for the heuristic bias. Each savings plan offers a fixed number of investment options that varies across firms. Benartzi and Thaler found that some individuals spread their savings evenly across the investment alternatives and do not take into account the riskiness of the investment options (1995, p.90)As a result, it can be argued that the asset allocation of individuals is influenced by the percentage of stock funds offered. The higher the number of stock funds, the higher the allocation to equities a finding that is difficult to reconcile with agents’ rationality (Glaser, Noth, & Weber, 2003, p. 7).Another aspect of irrational behavior is that the market behavior of investors is influenced by framing. The behavior of market participants’ changes depending on the framing of gains and losses (Glaser et al., 2003, p. 7). Traders are willing to pay more for assets if they have a short position at the beginning of a trading period compared to situations with a long position, even though the expected value of both portfolios is the same. In the first case, trading is driven by loss aversion, whereas in the second case, diversification is the main reason for trading (Glaser et al., 2003, p. 7).
Furthermore, agents’ rationality requires that all available information is evaluated using Bayes’ rule. However, if investors use specific heuristics which put too much weight on recent information, this systematic bias has an impact not only on the price reaction to new information but also on the price reaction afterwards when this error becomes obvious. In their 1997 paper, Barberis, Shleifer, and Vishny built a parsimonious model of investors. In their model they tried to show how investors form expectations of future expectations. Investors in the real markets can make systematic errors when evaluating public information. (Barberis et al., 1997, p.27) Investors are prone to a conservatism bias -- the underweighting of new evidence when updating probabilities. They are also prone to a particular manifestation of the representative heuristic, which is the tendency to expect even short sequences of realizations of a random variable to reflect the properties of the parent population from which the realizations are drawn (Glaser et al., 2003, p. 7).