Mean Reversion and Timeframe

  • Posted by
  • on November 12th, 2010

I’m a huge fan of James Montier, but he seems to be taking this mean reversion thing personally.  At a conference last week, he repeated his post from August that “the death of mean reversion is greatly exaggerated”.  As far as I can tell, the only widely published “death of mean reversion” commentary has come from PIMCO in one of its “New Normal” pieces.  Of course mean reversion isn’t dead, but those who force reversion strategies eventually die.

As I mentioned recently, I think certain market characteristics are undisputed in their reversionary nature…sentiment, asset class popularity, valuation, profit margins.  There are inherent tendencies for these to ebb and flow based on the the business cycle, product cycle, or herd behavior in any segment of the market.  The problem lies in matching the duration of these tendencies to the duration of our evaluation period…far tougher than we imagine.  Mean reversion strategies are useless and even dangerous if we fade price simply because something is too high or too low.

I don’t post a ton of charts, because without seeing the lesser and longer timeframe a single chart lacks the context needed for a thorough decision.  There are nearly always conflicts across timeframes, consider the current equity landscape:

Hours to Days:             Trending Down

Days to Weeks:              Trending Up

Weeks to Months:        Trending Up

Months to Quarters:   Balanced

Quarters to Years:       Trending Up

Years to Cycles: Reverting from Down

This is why it’s important to tune out noise…even if someone’s opinion is right, it may be right in a timeframe that’s not even close to where you tread.  Personally, I operate in weeks to months…that means I need to keep myself in gear with developments in the “days to weeks” frame and the “months to quarters” frame.  Shorter or longer than that is noise to my plan…I can’t subject myself to every intraday ripple, nor do I need to impose a macro view onto my strategies.

That still leaves a wide swath of analysis to cover, so I don’t have time to reach beyond my domain.  I accept input from a few(but only a few) of those more focused in scope.  I follow them for the same reason I watch or read a biography…I love to learn how experts become experts.  The value they provide for me is process refinement, and it’s up to me to take those lessons and apply them to the appropriate timeframe for me.

As for my favorite measures, they have informed me of no major changes yet.  My gut told me October 13 and October 25 could be significant, but my gut was wrong…that is why I measure.  November 5 felt significant, but to this point we’ve only seen a pause.  My gut was a key part of the development of my tools, but on any given day it can lead me astray so I listen but always measure.  My gut often knows before my brain does, but it doesn’t know the how, when, and how much that are the real keys to profits.

So we’re trending down within an uptrend that has reached the upper levels of a balance area inside of a major reversion from a downtrend.  How do I chart that?  I can’t…I assess each frame independently, then put the pieces together in a way that lines up enough factors in my favor to place a bet.  Trends can exist between mean reversion phases, and mean reversion can exist between trend phases.  What’s important is not proclamations, but developing and executing a plan that takes a few time-tested factors and fits them into our personal timeframe.

The information in this blog post represents my own opinions and does not contain a recommendation for any particular security or investment. I or my affiliates may hold positions or other interests in securities mentioned in the Blog, please see my Disclaimer page for my full disclaimer.

blog comments powered by Disqus
Derek Hernquest Blog