Who remembers asymptotes from high school?  Those are the curved lines that gradually approach a certain level--getting closer and closer, but never reaching it.  Sometimes in bond markets, instead of the normal triangular consolidations (i.e. linear trends of "higher lows" and "lower highs" converging), we instead have a static level on one side and an asymptotic line on the other side.  

A picture makes it easier to see, and it just so happens to apply to the current environment where 10yr yields have had a ceiling around 1.60 (with the brief exception of the Jackson Hole breakout) offset by a lower boundary that has constantly been approaching the upper boundary by smaller and smaller amounts.

As the chart shows, the trend strength (measured by the ADX technical overlay--the green line) is still very low in a historical context, but it has also bounced off the recent lows.  Taken together with the "higher lows," and the fact that some of our resilience could be a factor of month-end trading, there's reason to be at least a little bit cautious heading into the end of this week.

As a refresher for those who need it, month-end trading could be helping bond market resilience because certain traders are required to be holding a certain mix of bonds at the end of any given month.  Because bonds have a fixed series of payments, the total life span (or "duration" to use the industry jargon) of the bond holdings in a portfolio will always be getting shorter unless it's replenished by new bond buying.  

The indices that dictate the required duration for bond portfolios are updated at the end of the month and traders are buying bonds as needed to meet those duration requirements.  

Here's an analogy.  The trader is like a chef who has to keep the kitchen stocked.  As the month progresses, food is eaten by clients.  Near the end of the month, the restaurant does an analysis on what was sold and factors in what's expected to be sold in the coming month in order to estimate what the chef needs to buy in order to keep food coming out of the kitchen as normal.  

The chef is also required to show restaurant management that all of the required ingredients have been bought in an end-of-month report.  The need to create such a report serves as motivation for the chef to be buying with a greater sense of urgency near the end of the month.

Granted, this analogy can go both ways.  The chef could have too much food on hand and need to sell some, but because food is always being eaten by patrons, any such oversupply would only result from the chef having bought too much throughout the course of the month.  

As such, month-end bond BUYING is more common than month-end bond SELLING.  The caveat is that this only applies to money managers who are holding portfolios based on indices.  They're only one corner of the market.  Other traders could certainly have an agenda to sell bonds for other reasons.  Because of this, month-end bond buying among money managers is never a guarantee of gains--just something we know is adding a slight measure of support for bonds around the end of the month--all other things being equal.