Over-reaction Bias

Behaviourists have demonstrated that humans have an over-reaction to news. People are fast to react to bad news and slow to react to good. It takes years of bull markets to coax people into stocks, but one big fall and they are outta here. Stocks can sit at a big discount to book value for years, providing amazing opportunities to buy and yet no one takes them until the herd moves in, yet a short term earnings report with a few minor abnormal losses written up often has an immediate and calamitous effect on a stock price.

People put too much significance on chance events, thinking they spot a trend. (Especially a problem with people that use neural nets and high-tech genetic algorithms for trading, as you are virtually guaranteed to spot spurious trends by using them). Believing they see what the others do not, people make hasty decisions on superficial analysis.

Economist, Richard Thaler made an interesting demonstration of over-reaction bias with his students. He divided them into two groups, both of whom were to design a hypothetical portfolio of stocks and treasury bills. First the students were shown a market simulator of a trailing 25 year period. One group was given large amounts of short term information, showing the full extent of the market's volatility on a daily basis. The other group were only shown end-of-five-year results. The first group went on to design a portfolio of 40% stocks, 60% T-bills. The second group's portfolio averaged about 70% stocks, even though the overall 25 year results reflected the same returns on the same market. Thaler sums up the lesson by advising them, "My advice to you is to invest in equities and then don't open the mail."