Bond markets were at their best levels since Monday heading into today's 1pm 30yr Bond auction (MBS prices don't reflect that only because of the roll).  In so doing, they were rushing to price in a certain amount of positivity (or rather, a fear of being too negative) expected to be validated by the auction results.  They overshot the mark as the auction stats ended up being  slightly weaker than average/  That was still a relatively impressive result given the that bond markets are more prone to weakening ahead of auctions.

In other words, as opposed to betting that the auction would be challenging and breathing a sigh of relief after it was done, traders instead bet on the auction being strong and pulled back a bit after seeing they overshot the mark.

But why take the opposite stance this time around?  It has to do with the big shift in bond markets in February where more bets were made that rates would rise--aka "shorts" or short positions.  When there are a significant amount of shorts in the market and when something comes along like a much weaker than expected Retail Sales report that guarantees a certain amount of bond market positivity, the shorts are forced to cover.  Since a short equates to "bond selling," covering the short means bond buying.  Bond buying in turn raises prices, forcing more shorts to cover.

Shorts further had to protect against the risk that the auction would move against them as well (hence "fear of being too negative"), so trading levels stayed solid right up until we saw the slightly weaker auction results.  That left traders free to redistribute their positions, which eventually left us right about where we were before the short covering rally started.  MBS outperformed Treasuries in that endeavor and ended near the highs of the day.