Bonds began the day in weaker territory and were approaching the highest yields of the past 2 months by mid-morning.  Part of the negativity could have been driven by the higher Labor Cost data (+4.3 vs +2.1 forecast) which can speak to higher worker compensation, and thus, inflation.  But the lion's share of weakness was simply due to the way the chips fell with respect to the new month's trading positions. 

More simply put, there is always a certain amount of trading momentum that just 'happens' at the beginning and end of any given month.  There's no way to know which way it will push markets, but it's easy to identify in hindsight.  Bottom line, that's just the way traders were trading this morning.  All we could do was hope for a weak ISM number.

Indeed, we got a weak ISM number (49.4 vs 52.0 forecast and 52.6 previously).  This was bad enough to instantly erase the losses that had occurred in the first few hours of the day.  The stronger trading levels forced some traders to cover their short positions in bonds (bets on rates moving higher).  So-called "short covering" is accomplished by buying bonds.  

Short-covering isn't really a positive development in the bigger picture.  It's sort of an artificial rally, because the gains aren't driven by traders with a desire to own bonds (again, it's a desire to get back to 'neutral' on bonds that had previously been sold).  This point was emphasized by the fact that yields never made it past yesterday's lows, despite a good amount of intraday improvement.  As such, we head into tomorrow's NFP data with bonds STILL almost perfectly in the center of their recent range, just waiting for NFP to give them a reason to test that higher range boundary again.

I'll be extra clear here: I don't think an exceptionally strong NFP is the sort of thing bonds can bounce back from at the moment.  In the event of a major sell-off tomorrow, the wild card would be next week's ECB Announcement (Thursday).