Prepay Commentary

September Prepayment Speeds

October 6, 2017


MBS prepayments speeds slowed as expected, a complete reversal of the prior month’s increases. Again, technical factors were the main driver behind prepayment differentials MoM, as the housing market and mortgage rates have been locked in a near stagnant range for several months. For thirty-year MBS, both FNMA and FHLMC decreased 10% overall, GNMA Is decreased 11%, and GNMA IIs decreased 8%. Both in-the-money and out-of-the-money coupons slowed by similar amounts, confirming the technical and calendar related nature of the influences driving change.

Fifteen-year and thirty-year mortgage rates continued to fall from their March highs from August through early September, with the recent bounce pushing mortgage rates back to July levels. Conventional fifteen-year and thirty-year mortgages hit their year-to-date lows of 2.91% and 3.67%, respectively, in early September, according to Even though mortgage rates have steadily declined since March, prepayment speeds do not seem to reflect this downward trend, and consequently are more a reflection of seasonal factors, including day count. September had three less closing days than August, which accounts for a prepayment decrease of around 13% for the month of September. Meanwhile, the MBA Refi Index continues to be range-bound, holding far below levels of 2016. The Index reached a year-to-date high the week ending September 8th at 1636.80; however, it quickly dropped back closer to its six-month average of around 1400.

October has one more business day than September, and mortgage rates stayed in fairly close proximity to prior months during the pertinent application period. Prepayments are therefore expected to increase slightly, falling somewhere between this month’s and last month’s speeds.

Hurricanes Harvey and Irma should influence prepayment trends in a significant way for the affected regions for the next several months, while exerting very minor impacts on overall speeds. The storms certainly deterred voluntary prepayments. Additionally, there is a forbearance period for all counties declared disaster areas. This generally means a moratorium on evictions and foreclosures. The first 90 days is automatic, with the servicers’ discretion determining whether this period needs to be extended for another nine months after that. Meanwhile, both FNMA and FHLMC have said they intend to waive the four-month automatic purchase out of pools for all loans under forbearance. Assuming that happens, the effect on FNMA and FHLMC loan populations in the affected areas should print slower CPRs for the near term, a little over a year, due to reduced voluntary prepayments and minimal foreclosure related buyouts. Subsequently, as insurance claims and FEMA and other settlements pay out, and also as some irreversible and unfortunate situations conclude, CPRs for the affected areas should pick up sharply. While loans in GNMA pools are similarly covered by the forbearance policies announced by the FHFA, servicer buyouts are likely to accelerate, increasing speeds for the population of affected loans in GNMA pools during the coming months.



FNMA Speeds by Vintage Year

FHLMC Speeds by Vintage Year

GNMA Speeds by Vintage Year

Non-Generic Collateral Comparison Tables

James Plunkett

Director of Investment Product Strategies

Vining Sparks

Amanda Noa

Vining Sparks

The information included herein has been obtained from sources deemed reliable, but it is not in any way guaranteed, and it, together with any opinions expressed, is subject to change at any time. Any and all details offered in this publication are preliminary and are therefore subject to change at any time. This has been prepared for general information purposes only and does not consider the specific investment objectives, financial situation and particular needs of any individual or institution. This information is, by its very nature, incomplete and specifically lacks information critical to making final investment decisions. Investors should seek financial advice as to the appropriateness of investing in any securities or investment strategies mentioned or recommended. The accuracy of the financial projections is dependent on the occurrence of future events which cannot be assured; therefore, the actual results achieved during the projection period may vary from the projections. The firm may have positions, long or short, in any or all securities mentioned. Member FINRA/SIPC.
Copyright © 2022
This is a publication of Vining-Sparks IBG, LLC
775 Ridge Lake Blvd., Memphis, TN 38120