|
|||||||||
Copyright © 2004 by servicing.com |
October
26th, 2004 |
||||||||
Servicing Values Hold Steady Accompanying a general flattening of the yield curve, mortgage rates hovered around the same level to slightly up in the month of September. The value of mortgage servicing rights did not materially change this period as prepayment speeds and earnings rates moved only slightly from the previous month. There was strong sales activity in the bulk MSR market in September. Several notable portfolio offerings were put out for bid, ranging in size from one to ten billion dollars, however pricing was relatively flat from last month. Also of note was the announcement by several mid-sized servicers of their intent to sell all of their servicing sometime this year. This unexpected increase in supply should not materially affect the overall market, as the appetite is still very strong for large bulk MSR portfolios. However, smaller bulk portfolios are still struggling with a general lack of demand and the changing dynamics of the marketplace. As of the end of June, MIAC completed its annual review of the characteristics of the Generic Servicing Asset (GSA) portfolio that is used to construct the MSR Product Type Price Indexes. MIAC has created some brand new asset classes and removed some older, smaller segments. In addition, some material changes to the characteristics of existing asset classes have been implemented. As a result, the change in the MSR Product Type Indexes from May to June may not be meaningful due to the distortion of the individual buckets. These MIAC Indexes represents the MSR price behavior of the entire 30-Year Conventional Agency MSR and 30-Year Jumbo MSR market and particular components within the marketplace have increased or declined to a greater or lesser extent. These Indexes DO NOT represent multiples indicative of the value of new production, current coupon servicing rights. |
|||||||||
|
|||||||||
 |
|||||||||
| Once again, the Federal Open Market Committee raised the target Fed Funds rate by another 25 bps on September 21st to 1.75%, but claimed that their stance on monetary policy continues to be accommodative. Expectations for inflation have eased even though energy prices themselves continue to increase. Labor market conditions and output growth have modestly improved, and together with an increase in productivity supports an expectation for further gains in economic activity. The 1 Month Libor rate moved up another 17 bps through the month of August to finish at 1.84%. The 5-Yr CMS rose 5 bps, but the 10-Yr CMS rate dropped by 1 bps. FNMA and GNMA 30-Yr Current Coupon Yields increased by 6-7 bps each. |
|||||||||
 |
|||||||||
Libor-swap spreads have tightened again in September, and the Swap curve has flattened further. After widening about 7 bps in the month of August, both FNMA and GNMA Mortgage/Swap spreads have widened further approximately 6-7 bps as of September-end. |
|||||||||
MSR OASs Tighten, and Widen! MIAC Analytics makes available Daily GSA OASs through the Daily GSA Pricing Service. Each day, MIAC Analytics will use the previous day’s current volatility of 1-Year into 10-Year swaption volatility as a proxy for mortgage volatility. The graph below illustrates the OASs computed in MIAC Analytics from the actual GSA prices for nine of the largest GSAs: Conventional 15-Year 5.0% Issue Year 2002, Conventional 15-Year 5.0% Issue Year 2003, Conventional 15-Year 5.5% Issue Year 2002, Conventional 30-Year 5.0% Issue Year 2003, Conventional 30-Year 5.5% Issue Year 2003, Conventional 30-year 6.0% Issue Year 2002, Conventional 30-year 6.0% Issue Year 2003, GNMA 30-Year 5.5% Issue Year 2003, and GNMA 30-Year 6.0% Issue Year 2002. The graph displays the monthly OASs from 10/1/2003 to 7/1/2004.
After peaking back in the summer of 2003, the GSA OASs followed
a tightening trend through the month of June 2004, but since then
have appeared to stay within a certain range. The OASs for the Conventional
15-Year assets have tightened whereas the OASs for the Conventional
30-Year assets have widened. |
|||||||||
|
|||||||||
Servicing.com
has provided daily IO/PO pricing on a portfolio of Trust IOs for
several years. As a result, we can provide a comparison of historical
Trust IO OASs with the same interest rate model, volatility inputs
and OAS cash flow model as the historical GSA OASs. One can see
from the table above that Trust IO OASs don’t appear to be
well correlated to OAS behavior of the GSAs. This behavior does
not support the use of Trust IO OASs as benchmark OASs for the MSR
marketplace. |
|||||||||
In
the table above, the OAS durations or measurements of the anticipated
MSR price sensitivity
|
|||||||||
OAS Convexities: MSR convexities have continued moving further into their typical negative territory for most Generic Servicing Assets, as rates persistently climb. When convexity is positive, MSR prices will increase more for an upward rate shift than they will decrease for a downward rate shift. When convexity is negative, MSR prices will decrease more for a downward rate shift than they will increase for an upward rate shift. |
|||||||||
|
|
|||||||||
| BMA
Prepay Speeds Tap the Brakes The following tables reflect BMA historical dealer median prepayment forecasts. After a noticeable pick-up in BMA dealer consensus prepayment speeds observed in the month of August, speeds have generally slowed down a little bit through the month of September, and are still well below their historical highs. The speeds for the 2004 Issue Year asset classes hit their peak levels in August, but have mostly declined since then. On 7/15/2003 and 5/15/2004 the Bond Market Association released a revised listing of mortgage assets and their prepayment speeds to reflect the current mortgage market at the time. These new asset classes do not have historical speeds prior to that date. |
|||||||||
Red indicates the highest forecasts on record and yellow the second highest. |
|||||||||
|
|||||||||
More detailed BMA prepayment information can be obtained at www.servicing.com by viewing the Daily MIMs (Mortgage Industry Medians) data product. |
|||||||||
|
|||||||||
