Stable Distributions: Why They Matter for Trading

Good markets, bad ones: the one thing we can be certain of is that "this, too, shall pass."  Trends and ranges don't last forever.  Periods of flat and narrow activity are replaced by volatile directional moves and vice versa.  When I wrote my most recent book, I sought to explain a puzzling finding in my observations of traders:  Those who were most obsessed with standardizing their trading processes and trading with discipline were the ones showing the worst returns!  That flew in the face of common wisdom.  When I looked at the traders performing the best, what I saw were very nimble participants who could change their views and approaches to markets quickly as conditions changed.  When I codified my observations in the book into an ABCD scheme, you can see which came first:

*  Adapt to changing markets
*  Build on your strengths
*  Cultivate creativity
*  Develop best practices and processes

The challenging issue, of course, is *when* to adapt and how quickly situations will change.  Is today trading like yesterday?  Can we expect wholly new patterns to emerge or a continuation of the ones we've been seeing?  Do I stick with what has worked in my recent trading or take each day as it comes?

A concept that I find very helpful in this regard is the notion of stable distributions.  A stable distribution is one in which the means and standard deviations from one portion of a time series do not significantly differ from those of another portion.  If I flip a fair coin 1000 times and then flip it another 1000 times, the distribution of outcomes will be very similar.  That's a stable distribution.

An interesting application of standard technical measures is using them to gauge the stability of the market's behavior.  For instance, I can look at the distribution of upticks and downticks over a past period and compare that with the market's most recent distribution.  That is very helpful in identifying whether new participants entering and leaving the market are making a meaningful difference in price behavior.  If the distribution of upticks and downticks is remaining stable, we can have greater confidence that the near-term future will look like the most recent past.

When the number of stocks registering fresh 52-week highs recently soared to levels not seen in many, many months, the distribution was suddenly not stable.  Assuming that the future would be like the past would have been hazardous to one's wealth.

Most recently, I've received quite a few questions about when this bull move will come to an end.  One thing I watch from day to day is the number of NYSE stocks moving into, above, and below their respective Ichimoku Clouds.  (Raw data from Stock Charts).  I've collected those data for over a year, so I have a reasonable sense for when distributions are relatively stable and unstable.  The clouds are like Bollinger Bands.  If we don't have many stocks moving above their upper extremes, below their lower extremes, or moving from current extremes back into their clouds, the odds are pretty good that nothing major has changed for the market.  If a market is going to reverse course, we generally see advance notice from the strongest and weakest sectors, which make their moves first.  That shows up in the cloud data.

On Friday, we had a total of 156 NYSE stocks change their cloud status.  That is in the lowest quartile of readings over the past year.  The market had been strong on Wednesday and Thursday and it would have been easy to want to sell an "overbought" market on Friday, but in fact stocks were not changing their distributions.  That is very helpful information for traders deciding to trade vs. fade most recent market moves.

The wise trader approaches market action as a Bayesian, constantly looking at the most recent activity and determining whether it falls into the pattern of the recent past or meaningfully diverges.  The market situation will always change--eventually.  A key to successful trading is figuring out whether we are in a stable trading period or one of transition.

Further Reading:  Creativity and Innovation in Trading
.