September

Probability weighting and stock trading behaviours

 By Alex S.L. Tse, CERF Research Associate 

Humans are far from being a perfect machine of decision making especially in the face of uncertainty. One prevalent phenomenon is that individuals tend to overweight probabilities associated with extreme events. Examples include lottery punters’ optimism towards winning a jackpot and air passengers’ anxiety towards plane crash. In the context of finance, what are the implications of such psychological bias on investment decisions?

CCFin research associate Alex Tse and his collaborators Vicky Henderson and David Hobson investigated the effect of probability weighting on stock trading behaviours through a theoretical model of asset sale. They found that agents with probability weighting will adopt trading strategies in form of stop-loss but not gain-exit: on the one hand, probability overweighting of the worst scenario encourages investors to offload a losing stock. On the other hand, probability magnification of the best outcome encourages investors to maintain participation on the rally. This provides a potential justification of the popular usage of stop-loss orders among retail investors.

Probability weighting is also useful to explain the “price disposition effect”, a well-documented financial anomaly where investors are selling winning stocks much more often than losing stocks. Existing models typically generate a very extreme disposition effect. With inclusion of probability weighting, however, investors are now more incentivised to hold a winning stock relative to a losing stock as they find a lottery-like payoff with positive skewness attractive. This enables the model to deliver a level of disposition effect much closer to what empirical literature suggests.