There is considerable evidence that people are more likely to arrive at the conclusions that they desire. Neural network models, for instance, suggest that brushing aside a contradiction seems to activate the reward mechanisms of the brain.

As a result, investors are motivated to agree more enthusiastically with information that suggests they might make money on their investments, but to disagree with data that suggest they might be on the losing end.

That’s why so many investors are willing to bet on underperforming/losing stocks, cutting them more slack and subsequently putting more money into them, suggests a paper by Pennsylvania State University professors Jere-my Ko and Oliver Hansch, entitled Persistence of Beliefs in an Investment Experiment.

Psychologists attribute this tendency to cognitive dissonance — a reluctance to admit that past decisions may have been incorrect. Investors distort or ignore pertinent information to relieve the discomfort caused by the conflicting views in their heads.

The authors studied a group of Penn State students who were given the equivalent of $10 a week and asked to allocate it, given three investment choices each week.

Subjects could allocate their money in any way among the three choices but profits from one week were not rolled over into the next.

The first choice asked participants to make a bet on which of two similarly valued stocks would go up or down that week. If they picked the correct stock, they doubled their money.

The students also had the option of a blind bet, making a double-or-nothing wager with a 50% chance of winning or losing, based upon a Powerball lottery draw the subsequent week.

If neither of these options appealed, the students could also choose to take the $10 in cash.

Every time the participants placed a bet, they were asked if they had observed any new data concerning their stock bet during that week. If they had, they were asked to rate the information as positive, negative or neutral.

By measuring how many people continued with their bet from the previous week vs how many went with a different option, the researchers were able to estimate how likely investors were to change their thinking and the degree to which discovering new information influenced their betting decisions.

Over the six-month experiment, 45% of the participants opted for the stock bet, 35% took the cash and 20% chose the chance bet. What was most surprising, the paper notes, was that 67% of the participants made the same choice week after week.

In addition, the participants appeared to pay more attention to new information and reported news more frequently when faced with a losing bet than when faced with a winning one. The participants were also more willing to take positive information about their stock pick at face value vs any bad news, which was more actively scrutinized.

The researchers had assumed that participants making the stock bet would stay with a winning stock and change from a losing one fairly frequently. At the very least, the researchers had expected to see a 50% persistence in the stock bet over the life of the experiment.

The results, however, suggest that the participants made the irrational choice 17% of the time, backing the stock that did not climb from the previous week’s levels, regardless of any indication that it might turn around.

From this, the paper’s authors concluded that people are more likely to bet again on a prior pick, even when the realized returns are negative. Investors are also more likely to increase the size of their overall bet, given the opportunity. IE