The overnight session was uneventful for Treasuries, leading to almost perfectly unchanged opening levels in MBS.  Bond markets began moving quickly weaker following the 8:30am economic data.

At first glance, the weaker Housing Starts data should have had nothing to do with the selling pressure (weaker data is typically a net-positive for bonds).  That's probably a fair assessment, but less certainly than the headline suggests.  Reason being: the more even-keeled component of the data--Building Permits--actually improved nicely among single-family homes.  Multifamily permits dragged down the totals, but they're perennially more volatile.

All that having been said, the Consumer Price Index is likely the bigger deal at 8:30am.  That feels "weird" to consider as inflation data has been anything but a consideration when it comes to motivating weakness in bond markets for the past 4 years.  The accompanying selling pressure in stocks was a major clue though.

Old school market wisdom suggests bonds and inflation have a much stronger link than they've had in the age of QE.  Indeed, it would be odd to see increasing inflation, in and of itself, have as much of an effect on bond markets as it did this morning.  Thankfully, that's not exactly how it happened, so our sense of reality isn't overly violated.

Much more relevant than the actual effects of inflation are the effects that these sorts of inflation metrics will have on Fed policy!  Read that again if necessary, because that's what this morning's post-CPI sell-off is all about. 

Simply put, stronger inflation readings are one ingredient in accelerating the timeline for the removal of Fed accommodation.  That's why stocks got hit at the same time as bonds.  That's why we saw a surprising amount of movement from data that hasn't really mattered in four years.

Even then, bonds remain contained in the same range we've been evaluating almost all month (marked by 2.57-2.66 10yr yields).