I devoted years of my life championing the cause of educating mortgage originators on the realities of MBS price movement.  The most compelling call to action came during the meltdown where MBS told a story that wasn't being told by Treasuries. 

Everything changed after the Fed stepped in with a giant syringe full of bond volatility's favorite sedative: QE.  The final straw was QE3 which specifically targeted MBS buying in September 2012.  At the time, and ever since, I declared that to be proof positive that the Fed "gets it" with respect to mortgage performance vs benchmark rates and that it was proof positive that they wouldn't let spreads slide away into oblivion ever again.

The end of 2012 was the last time I would need to go into much detail to explain MBS vs Treasuries for years.  Sure, there were a few spats of bigger movement where we talked about how MBS and mortgage rates can lag an initial spike in Treasuries, but even then, mortgage rates were moving in the same direction as Treasury yields and  the latter could be used exclusively to track trends and identify both intraday and bigger-picture inflection points.  All of the above led me to constantly explain stuff like this: Why so much focus on 10yr if this is an MBS Site?

The utterly massive market response to coronavirus quickly reintroduced a focus on MBS prices.  In fact, this was already the case as early as August 2019 as Treasury yields moved lower far more quickly than mortgage rates.  MBS were obviously having a terrible time adjusting to the precipitous move in the bond market and were also plagued by heavy issuance (higher supply putting downward pressure on prices).  Once again, it was easy to lean on MBS underperformance as a way to explain why mortgage rates weren't lower than the overall bond market would suggest.

Now here we are with prices on 2.5 coupons trading over 104!  That's a milestone few saw coming--at least not this soon.  It's an absolutely staggering price--a record high and then some.  Clearly, then, if MBS prices are higher than they were in 2012, rates should be lower than they were in 2012.  BUT NO!  At least not in a consistent way and definitely not for every lender.   The reasons for this were explained in great detail in this post yesterday.

We're going to need clarity on how the mortgage servicing market and the agencies will access Fed/Treasury capital to maintain capital buffers and payment consistency to bondholders.  That was going to be part of the relief bill, but now it's supposed to come in a separate heads-up from the Fed after the bill passes.  Either way, we should expect lender pricing to defy MBS logic in many cases until the mortgage market gets a better sense of how it should value risk going forward.   

I would like to say that 30yr Fixed UMBS would be the least affected here, and that's true for many lenders, but at this very moment, some lenders are looking at much more risk than others even in that purest of vanilla spaces due to the way they fund loans.  Some c30 servicers will be on the hook for both principal and interest payments through to bondholders plus escrows to the respective entities (we're talking about a forbearance situation here)--full PITI no matter what the borrower is paying.  Other servicers are only on the hook for ITI. 

This is one of the reasons some c30 lenders suck right now while others are near record lows.  It's also a big reason that FHA/VA have been hit so hard.  They're in that full PITI camp.  They're also in a demographic that's less able to soldier through COVID-related financial stress (more 1st time buyers, fewer assets on average, lower FICO on average, much higher LTVs on average).  

And it's ALSO the reason that anything other than Fannie/Freddie/Ginnie is doing even worse.  Although the relief bill doesn't obligate those lenders to offer forbearances, it also doesn't offer any protections for them when their borrowers are unable to pay.