Expectation Trap Pingyang Gao June 26, 2015
Old game and new event Market prices aggregate partial information. (Plott and Sunder 1988). Does market do so when market interventions are based on prices? How can we use the information embedded in prices?
Theoretical Prediction M(dp)*dp=0 If dp>0, M(dp)=0; If dp<=0, M(dp)=1 Investors are always weakly better off by the proposed new regulations.
Our game At the end of each period, if the average trading price for the period is less than or equal to 225 points (Y-dividend), the original realized state will be replaced by state Y; otherwise, the original state will remain. The current state is Y. If it’s changed to Z and prices reflect this change, then Z will be retained; If It’s changed to X and prices reflect this change, then X will be reversed to Y.
Predictions When the drawn state is X, average price is equal to or slightly lower than 225, but never higher than 225. When the drawn state is Y, average price is around 225. When the drawn state is Z, average price is above 225 and near 325.
More Predictions Table 1: Price Predictions XYZ Unconditional RE<= Ignoring Rule Bayesian Update
Data Review PeriodAVPSTDNMAXMin Statet (225)t (105)t (325) X Y X Y X Z Z X Y Z
Three Hypotheses
Average Prices
Efficiency Lost
Statistical Significance PeriodDrawn Statest(AVP=225)t(AVP=105)t(AVP=325) 1X Y2.53 3X Y0.43 5X Z Z X Y Z
Preliminary Conclusions Market is able to aggregate information when the impact of intervention is positive. Market is able to aggregate information when the impact of intervention is positive Aggressive use of information embedded in prices may destroy the usefulness of the information.