Monday, June 30, 2014

Trading and Cognitive Bias

Traders are generally sensitive to emotional factors that take them out of their zone and lead them to trade either too impulsively or too cautiously.  Less well recognized are biases in our thinking processes that lead us to erroneous conclusions and poor trades.  This is because not all cognitive biases are accompanied by emotional upheaval.  Indeed, many exist precisely to maintain an emotional status quo--even at the cost of distorting perception.

There are many cognitive biases that can impact trading decisions; more than we can typically be aware of at any given point in time.  One of the most common that I observe is in-group bias.  We allow our personal affiliations to overvalue and overweight the observations and decisions of our peers.  This leads to a kind of bandwagon effect, where we no longer make truly independent market observations and trading decisions.  For trading firms this is particularly pernicious, as--combined with overconfidence bias--it can lead to concentrated bets and correlated returns just as thought is most biased. 

Another common bias is the anchoring effect.  This occurs when we seize upon a particular piece of information--often the most recent, salient observation--and draw conclusions from that limited data point.  This frequently happens when traders don't have strong views on markets, but feel a need to place trades.  The desire for a trade rationale leads them to overvalue recent observations.

Of course, these biases can operate in concert.  For example, a group of traders can anchor off a piece of salient data and set off a bandwagon effect that impacts other traders, leading to biased group-think.

Some traders seek to minimize cognitive bias by taking a quantitative approach to their trading.  Alas, this can have the effect of substituting even more subtle distortions for the better known ones.  Strategies that look too good to be true often are, as the search for patterns can find seemingly good results in random ways.  This most notably occurs when backtests of strategies are overfit:  we keep searching for "significant" results until we find them.  Such strategies typically boast phenomenal Sharpe ratios--lots of gain for little pain--but, as Marcos Lopez de Prado has recently observed, such ratios are deceptive if they are not deflated for selection bias.

Traders typically think of discipline in emotional terms:  taming one's fear, greed, and frustration in order to stick to trading plans.  These plans, however, are only valuable insofar as they reflect a different, cognitive discipline.  Rules that guide our trading are only as good as the rules that guide our identification of trading opportunities.  More on that topic to come.

Further Reading:  Changing Self-Talk by Talking to Yourself