MBS Day Ahead: Bounce Potential. Let's Hope it Holds

Have you ever wondered why technical analysts and other seemingly savvy market commentators tend to reference moving averages that are nice, round numbers--25, 50, 100, 200-days?  What about 75?  Why not 150?  Why not 137.3?  Indeed, why should we pay more attention to one moving average versus another simply because it is calculated based on a nice, tidy factor of 100.

I can appreciate that the concept of "100" is a big deal in America, relative to say, 88, but it's just the opposite in Japan (or so I've been told).  Whereas 100th birthdays are the big deal here (just ask Willard Scott and his lurid pantry of jam jars), it's the 88th birthday that's the big deal in Japan.  I don't bring Japan up for any particular market-related reason--only to illustrate the point that this "100" business is an arbitrary, and very American thing.

What if moving averages weren't arbitrary?  After all, wouldn't it make more sense to tie them to the actual number of days that exist in various chunks of market time?  For instance, if we know that a fair amount of trading momentum in bond markets is dictated by "month-end" and that new positions often flood in during the beginning of the month, how about we put a monthly moving average on a pedestal above  the next closest popular number?  

It turns out this chip is already in play, and it has been for quite some time.  For decades, technicians have been using the Bollinger Band study to analyze market behavior, and at its core is a 21-day moving average.  There's a very simple reason that 21 days makes all kinds of sense.  There are an average of 21 business days in any given month!  So the next time someone tells you that bonds are going to do one thing or another based on a 25 day (or 50, or 100) moving average, ask them why not the 21 day? Or the incredibly badass 63-day (one quarter of business days) or the extremely awesome 248 day moving average (typical number of full trading days each year).  

The truth is that moving averages don't predict the future, but they can certainly help us know when things may be changing.  Indeed that's why the 21-day is at the heart of the Bollinger Band study.  It's coming into play now as 10yr yields moved up to the important 1.53% pivot point yesterday.  They will be trying to bounce and hold below it today, and it just so happens that recent market movement has the 21-day moving average crossing right through 1.53 today.  I like to see these sorts of confluences.  It adds weight to the trading level in question, but again, not as a predictor of the future.  The true importance of these technical levels is as a trip wire to let us know if things are changing.  As long as we're holding under 1.53 today, the broader positive trend remains intact.