Thursday, July 31, 2014

Idea Generation, Originality, and Natural Selection

It's difficult to imagine achieving uniquely positive results in trading by looking at the same information in the same way as everyone else.  Distinctive returns require distinctiveness of information processing:  either seeing old things in new ways or seeing entirely new things.

Can you name specific old things that you process in new ways to generate trading ideas?

Can you identify new things that you process that others don't look at?

If you were applying for a trading job and I was the recruiter, those are questions you could expect to be asked.  If you were running your own fund, those are questions you could expect seasoned investors to ask.
So it's worth asking yourself:  When it comes to idea generation, what makes you unique?

I recently downloaded intraday information on the number of NYSE stocks making new highs for the day session vs. those making new lows.  The new highs and lows are for the current trading session only and the numbers are cumulated for every stock, every minute of the session.

Suppose I choose a time of the morning--say, 10 AM EST.  The stock market has been open for 30 minutes.  What does it mean if many stocks are making fresh daily highs or lows precisely at 10 AM?  For one thing, it means that many stocks have broken out of their opening ranges.

With that recognition, now we can ask new questions:

If many stocks break out of their opening ranges, is the day more likely to be a trend day?  Does opening range breakout breadth (the number of stocks breaking out of their opening ranges) predict trend better than opening range breakouts of a single broad market index?

What is the earliest time of day that we can detect broad opening range breakouts that is predictive of trend?  If we see few opening range breakouts by a certain time of the morning, can we validly conclude that we're likely to see a range bound session?

This is but one research project I'm working on.  There are several others, all just as out of the box, but all similarly built on old, familiar market concepts.

90% of these ideas will not bear significant fruit.  If I generate one new idea to research every week, however, that will leave me at the end of the year with five new, significant ideas.  Over the course of a career, that adds up.

Simonton has it right:  It's all about natural selection and sustaining the originality to generate plenty of mutations.

Further Reading:  Creativity and Greatness in Trading