In addition to writing about mortgage rates, I also follow and analyze the underlying bond markets.  Early on in that endeavor, one of my favorite phrases was "play the range before the range plays you."  This was every bit as simple as it sounds.  Like mortgage rates, financial market metrics tend to get stuck in a range from time to time.  Frequent bounces around the same ceilings and floors let us know a range is taking shape.  From that point on, the highest probability bet is to assume the floor and ceiling will remain intact until one is clearly broken.

That might sound overly simplistic.  But the point isn't to call attention to the obvious (after all, of course a ceiling isn't a ceiling any more if it has been broken).  Rather, the point is to not get caught up in rapid movement from one side of the range to the other.  For instance, it's tempting to worry that rates will continue higher if they move quickly from the floor to the ceiling.  This notion of "playing the range" suggests that we don't worry about such moves higher until we have actual confirmation that the range is broken.  In other words, you'll only "get played" by the range one time (when the break actually happens) as opposed to getting faked out every time rates look like they're on the move.  

Fortunately, the current range is so exceedingly narrow that there's not very much at stake.  The top tier conventional 30yr fixed quote has been between 3.375% and 3.5% for more than a month.  The application of the "play the range" wisdom would quite simply suggest that locking makes more sense when we're at 3.375% and floating makes more sense when we're at the higher end of the range.  As always floaters must be aware that when the ceiling is finally broken, they'll be in a worse spot than if they'd panicked and locked after the first move higher.  If that sort of risk doesn't sound pleasant, you won't be foregoing too much benefit by locking, even if rates improve.  

As for today specifically, rates moved back up a bit after having made it to 1-month lows on Friday.  They're still nearer to lower end of "the range," so that's been a statistically smarter time to lock.  But again, the range is so narrow, that the consequences are far from dire no matter what you do.  If you choose to float, just be sure to set a line in the sand to act as your signal to lock if it's broken.

Today's Best-Execution Rates

  • 30YR FIXED - 3.375%
  • FHA/VA - 3.0 - 3.25%
  • 15 YEAR FIXED - 2.75%
  • 5 YEAR ARMS -  2.75 - 3.25% depending on the lender

Ongoing Lock/Float Considerations

  • In the biggest of pictures, "global growth concerns" remain the driving force behind the long-term trend toward lower rates
  • Amid that trend, periodic corrections toward higher rates can and will happen.  These can happen for no apparent reason, or they can be brought on by changes in expectations surrounding central bank policy at home and abroad, as well as geopolitical and systemic risks

  • Time horizon and risk tolerance are 2 variables to consider when it comes to locking.  If you have plenty of time and don't mind losing some ground, set a limit as to how much higher rates could go before you'd lock to avoid further losses, and then float in the hopes of never seeing that limit.
  • In the shorter-term, it's always good to look for lock opportunities after rates have been moving lower or sideways repeatedly, especially if they've since begun to move back up in any sort of consistent way. 
  • As always, please keep in mind that the rates discussed generally refer to what we've termed 'best-execution' (that is, the most frequently quoted, conforming, conventional 30yr fixed rate for top tier borrowers, based not only on the outright price, but also 'bang-for-the-buck.'  Generally speaking, our best-execution rate tends to connote no origination or discount points--though this can vary--and tends to predict Freddie Mac's weekly survey with high accuracy.  It's safe to assume that our best-ex rate is the more timely and accurate of the two due to Freddie's once-a-week polling method).