It's going to be a while until we can truly sort out whether the current bout of bond market weakness is the same sort of correction seen in early 2015 (i.e. almost done) or if it will be more like mid-2013 (i.e. just getting warmed up).  Of course there's always an outside chance that a sell-off could be even bigger--as many have feared for years--but I'll believe in that possibility after I see it happen (rising rates require rising growth, inflation, and government borrowing, so you'll forgive me if I'm skeptical about 2 of those 3 components).

If a sell-off were to truly get out of control, it would reach a point where it was disproportionately bad for lower MBS coupons.  This is due to the concept of negative convexity, which is a confusing way to say that investors holding lower coupon MBS will freak out if they see broader rates moving higher for fear that they'll be stuck holding below-market rates-of-return tied to loans that no one wants to refi.

Want an analogy?  You got it.  Imagine the overall bond market complex is on a cruise ship that stops off at various ports and islands.  Imagine that you and your buddies are MBS coupons on this cruise ship.  At one of the most distant and tropical stops on the trip, you come to an island where passengers are allowed to disembark and explore the island.  You and your MBS buddies are hanging out, having a good time when suddenly you see the cruise ship start to pull away from the island.  You run to the edge of the water and wave your arms trying to get someone's attention.  A few of your more desperate buddies swim for it (there's a ladder on the side, I guess), and general panic and pandemonium ensue.

Bottom line, MBS are happy to hang out on remote islands of low rates until it looks like the cruise ship is leaving.  At that point, they'll lose their composure and sell-off disproportionately in order to get back to the ship. 

But we're not there yet.  The ship has fired up its engines (i.e. rates have begun rising), but it hasn't weighed anchor yet (i.e. broader rate levels are not yet super scary if you're a Fannie 3.0 MBS).  So, between now and the time that broader rate levels become super scary for Fannie 3.0 MBS, the initial thrust of rising rates affords MBS an opportunity to outperform.

Here's what it looks like in terms of MBS prices vs 10yr yields.  In this chart, I flipped MBS prices upside down so they move the same direction as Treasury yields during rallies and sell-offs.  In this chart, lower = stronger.  Notice MBS staying strong relative to Treasuries starting in July.

Another way to observe the outperformance is via actual yield spread.  This involves a somewhat subjective calculation of MBS yield (because true yield depends on the unknown variable of borrower behavior) and plots it against 10yr yields.  The following lines represent the distance between the 2 yields.  The lower the line, the closer MBS are getting to Treasuries (i.e. lower = MBS are improving vs Treasuries, aka "spreads are tightening"). 

While we aren't back to late 2012 levels, we really shouldn't expect to be, given that late 2012 was driven purely by the QE3 announcement.  All things considered, 70bps is an incredibly strong/tight spread for MBS vs Treasuries, and a definite silver lining to the broader sell-off.  The crew of MBS coupons on the tropical island are very confident in their ability to swim back to the ship if it starts moving, but until then, it's fun in the sun (relatively).

Keep in mind, this doesn't mean rates are lower or that they'll get lower necessarily.  The only point of all this is to say "hey everyone, yes, things are bad for bonds, but at least MBS aren't having as tough a time as Treasuries."