This commentary will begin it's life as the "day ahead" entry for July 19th, 2017, but it may as well be a "knowledge base" entry for the validity of technical analysis.  There are already several knowledge base articles designed to help people understand technical analysis on MBS Live.  These include:

Basic Concepts of Technical Analysis (And Some Jargon Definitions)
Technical Analysis 101, Trends and Trend-Lines
Max Support/Resistance
Pivot Points (inflection points, floors, ceilings, etc.)

Today's addition has to do with how much faith we should put in any given technical study.  Are some studies better than others?  When multiple studies are saying the same thing, should we put more stock in the conclusion?  If the technicals can be reconciled with fundamental developments, should we take them more seriously?  Despite all that, can technicals still give the WRONG signal at the WORST possible time for mortgage borrowers/originators playing the lock/float game?

The answer to all of the above is YES!  

Naturally, the more agreement we see between different technical studies, the stronger the conclusion is.  If that conclusion can be argued in terms of fundamental developments (econ data, news, central bank policy, tradeflow considerations), it's that much stronger.  

But keep in mind the fatal flaw of technical analysis: because any technical study is based on the hard numbers associated with the movement of a particular security, technical studies generally run the risk of telling the same story.  In other words, if 7 different journalists with drastically different writing styles are all covering the eruption of a volcano, the audience will likely conclude that a big mountain blew up regardless of the journalist they like best.

Let's look at an example comparing 2017 and 2015.  In the 2017 example, we have 10yr yields breaking well below the middle bollinger band for the first time in several weeks, and for the first time since spiking somewhat abruptly above that middle line (and pushing the upper line for multiple days).  All that last sentence can be seen in the dotted-red box and white boxes below.  FYI, the purple wavy lines are the Bollinger Bands and the chart is of 10yr Treasury yields.

The other boxes on the charts simply show similar developments between the two time-frames.  The lower portions are fast and slow stochastics (momentum indicators).  The green dotted box shows how both topped out at the end of the last sell-off (May 2017 and March 2015) before entering a relatively flat, stable trend in the teal boxes.  During that time, momentum indicators (grey boxes) ebbed higher and ultimately topped out as yields broke out as mentioned above (red boxes).

The relevant part for today and for the current week is in the white box.  Yields are breaking below the middle bollinger band.  Longer-term momentum (bottom section of chart/slow stochastics) is shifting positively.  Short-term momentum has already shifted all the way back down to the lower blue "overbought" line.

 All the ingredients are there for a positive technical conclusion.  But 2015 shows us that positive technical conclusions don't always matter.  In fact, oftentimes, the first break of the middle bollinger band results in a bounce followed by the true test at the levels where yields first peeled away from the outer bollinger band.  In 2015, that meant yields needed to hold below roughly 2.30 in order to form an "M-top" (a pattern that students of bollinger bands look for).  The M-top would later happen in late June 2015.  

With all of that in mind, the best conclusion in the case of weakness in the coming days would be that 10yr yields need to hold under 2.40% in order for the best long-term technical signal.  Of course, there's every possibility that we simply continue rallying here, but that could depend entirely on Mario Draghi's tone tomorrow morning.