Written by Darren Stumberger, as originally published in The Reverse Review.

Floater spreads have leaked wider into quiet August flows along with IOs (interest-only securities). IOs have been off a half of a point over the past several weeks. Seasoned fixed rates have stabilized and tightened after widening 40-50 bps in late June. Three-year and five-year-plus average life papers are trading in the very high 70s. The four-year part of the curve has been more challenged with spreads 5-10 bps wider.

With the upcoming program changes, the market will continue to be dominated by floating-rate products (call it an 80/20 split) and the cash flows will be incrementally longer because LTVs have been lowered. Even though the curve is steeper, which shortens average life, the drop in LTVs will outweigh the effect on cash flows. Moreover, the modeling of draw behavior will become more nuanced as borrowers will be capped in one way or another in drawing on their available proceeds. It’s possible we’ll see “pent-up” draw demand during the initial 12 months and then a spike upon expiration, but that’s pure speculation and time will tell.

Operationally, it seems like a tall order for servicers to monitor draw requests against the absolute prohibition to draw in the face of extenuating circumstances or an increased need for additional proceeds. Again, we’ll see how that plays out.

Most importantly, the program is being strengthened with robust full-document underwriting, the rejection of borrowers with a propensity to go delinquent on their taxes and insurance, and the elimination of non-borrowing spouses and broker steering. All of these changes keep the existing universe of bonds safe from any type of refinance/prepay scenario, and I would expect the new production post-October 1 to be even less callable than prior vintages (although that’s not saying a lot).

I also expect market spread levels to continue to drift wider as the Fed tapers uncertainty and time is needed for investors to digest the new program parameters and speed history. Given most, if not all, floating HMBS is structured in CMO form, execution will be driven by IO investors, and given their sophistication level, it’s likely that a track program will be needed before they jump into something new. I think prepays will actually be more muted in the new program, with spreads eventually tightening; it will just take some time for this to play out.