A $1.4 Billion Ticket to Better Carry

Well, we finally got a little relief in rates this week.

The 10-year Treasury fell 9 basis points, and the 30-year dropped about 10 basis points to 533.

But I wouldn't get too excited just yet.

September payrolls were only 29,000, while headline CPI accelerated to 3.7% year-over-year.

The Fed's kind of caught between a weakening labor market and renewed inflation pressure.

And the long end isn't necessarily going to cooperate with any effort to ease policy.

True indeed.

The very modest relief we've seen certainly isn't going to cause the housing market to respond in any meaningful fashion.

Mortgage rates fell a magnificent eight basis point to 732%.

Yet, applications still decline more than 5% for the week.

Clearly, it's not a fair aligning given the time lag between the two.

But still, Stephen, purchase applications were down 18% from a year ago, and refi activities remain almost 60% below last year's pace.

So, 8 basis point improvement clearly isn't enough to get borrowers moving again.

Exactly, and the housing turnover data tell a similar story.

Existing home sales are still running below $4 million annually, pending sales fell more than 7% in September, and housing starts declined as well.

We're seeing more inventory and softer home price appreciation, but I'd say the adjustment is happening primarily through fewer transactions rather than a major correction in prices.

and this is

super important for mbs investors because the the broad housing numbers are telling us to expect slower speeds so turnover stories remain the investor focus for the majority of the outstanding universe and production coupons especially generic ones naturally come with risks on both sides of equation significant extension should rates move higher, but also the potential for a modest rally to bring selected borrowers back into refi range in style.

So here's the uncomfortable question.

After this past week's rally, how much valuation cushion do we really have left, Steven?

I'd say less than last week, I guess.

That's nice.

In fact, if we're talking about overall MBS basis, the 30-year current coupon is around 125 basis points, which is six basis points tight to our model fair value, which is coming out to about 131 right now.

And the 15-year basis is about four basis points rich to its model.

So while the absolute spreads are near the wides of the year, after accounting for the broader macro variables, MBS' current coupon is on the tighter side.