Tuesday sees the inception of both the Treasury Auction cycle and the MBS Settlement cycle for the week and month.  For MBS, this merely means that September coupons are done trading today and October coupons will take over between today's close and tomorrow's open.  That makes for the appearance of a drop in prices, but we're simply shifting our focus to the next box of MBS coming down the assembly line for an indication of "MBS Prices." 

In other words, "MBS Prices" that we discuss are only ever for the upcoming delivery month.  Other, lower prices for future months exist, and in this case October's prices will become our representative for MBS Prices after today.  October is already trading at a lower price (recent average gap between months has been 8-12 ticks or .25-.375).  If this is confusing, see last month's discussion, which had some more detail (if it's still confusing after that, let me know!).  Here's an approximation of "before and after" prices, had MBS rolled yesterday:

Before that, Treasuries kick off the week of auctions with 3yr Notes at 1pm.  These shorter term auctions had been almost totally inconsequential for most of the past 3 years but have seen a resurgence as taper talk enveloped the short end of the yield curve with the long.  In other words, not only are the longer term maturities like 10yr Treasuries being affected by the tidal shift in QE expectations, but the short end (2's and 3's) have been increasingly under pressure as the timeline for the Fed's first rate hike comes into focus in mid 2015. 

This can certainly change a lot between now and then, but the rule of thumb for short term notes and T-bills is that they haven't had much reason to move while their settlement dates remained inside the window of time wherein the Fed was expected to hold the Fed Funds Rate near zero.  This phenomenon was christened with a sledgehammer when the Fed introduced the calendar date verbiage into the FOMC Announcement in mid 2011.  It's since been changed to the Unemployment rate threshold of 6.5%, and heavily qualified as having relatively little to do with the Unemployment rate of 6.5%.  We're guessing the Fed would play that slightly differently in retrospect, but the inconsistency would cause greater damage than the confusion at this point.