Thursday, November 13, 2008

Toward a Cognitive Theory of Trader Performance

We tend to think of trading ideas and positions in terms of their objective characteristics: the market traded, the location of the trade, its direction, the size of the position, the target, etc. This can have the result of making any one trade look similar to others when a trader behaves in a disciplined fashion. It becomes confusing to the trader why some seemingly good ideas work out and others don't.

Suppose, however, we think of trading in subjective terms: how ideas present themselves to the trader. For example, some ideas are experienced by the trader primarily as subjective hunches and felt tendencies; other ideas may be expressed as explicit, planned ideas. Still other ideas can manifest themselves as images: pictures of what the trader thinks will happen. Yet other ideas take the form of analogies and metaphors, sometimes drawn from other markets, other times from entirely other spheres of experience.

Moreover, traders in the same market, looking at the same data (technical, fundamental, etc.) may process this information quite differently. Just as we have unique combinations of personality traits, we also possess distinctive cognitive styles.

Where fear, greed, overconfidence, stress, risk, and uncertainty may affect performance is by nudging traders out of their normal, effective cognitive styles. Under duress, a conceptual trader may act impulsively on a hunch. An intuitive trader might overthink markets. The trader who reasons in images or metaphors may become too detail oriented, analyzing where they should be synthesizing.

This suggests limitations to the usual coaching of traders, which tend to focus on strategies and feelings, not the cognitive path by which ideas are generated and expressed as trades. Traders may enter slumps, not because of market dislocations or psychological stress alone, but because of subtle and unrecognized shifts in how they process information.

A cognitive framework for thinking about trading performance strikes me as uniquely promising and fits well with my observations of hedge fund portfolio managers. In my next post on this topic, I'll offer a specific example and elaborate some of the implications of this view for self-coaching.
.

6 comments:

SSK said...

Great, very insightful and well put. Keep up the good work. Best, Steve (SSK)

Finance Fanatic said...

I like your blog, very insightful, We should trade backlinks, my site is www.crashmarketstocks.com

Firebird said...

So true, what you mention is like the aggressive chess player that tries to play defense against a superior rival, or a ball team that changes its play in the play-offs: it usually doesn't work because it does not come naturally.

Best trading,

Jorge

katallaxia said...

!!!

ken long said...

i have been thinking about a continuum of environmental conditions ranging from Simple<>Complicated<>Complex<>Chaotic<>Random

in each condition a different set of cognitive skills and response strategies are required

i elaborate on these ideas at http://kansasreflections.wordpress.com/2008/10/22/reflections-on-ray-isons-work-in-systems-thinking-in-action-research/

its clear that if things are Simple, then too much thinking and processing are bad for returns because simple fast action will win. But the systems that we use to invest in "normal" complex conditions will fail when volatility increases beyond usual limits (LTCM) and what is needed may be metaphorical, situational trading or no trading at all until some form of normal is required.

I am beginning to think that no single response is appropriate for all of these different market states; what's needed is a market classification scheme that lets us know when we have an edge based on our particular edge.

i think these ideas are in harmony with your thoughts on psychology and cognitive responses

Brett Steenbarger, Ph.D. said...

Thanks for the comments; your ideas, Ken, might help to explain why traders see the markets so well at some times and not well at all at other times--

Brett