Expanded Pool Disclosure

Well into the 1980s, analysis of prepayment risk and value focused on the pass-through characteristics that related to refinancing incentive (or affordability restraints on turnover), seasoning, and burnout: WAC, WAM, estimated age, and factor or percent of principal balance still outstanding. However, as the market continued to grow and mature, MBS broker-dealers and investors began to focus on regional differences in prepayment behavior. For example, California loans, dominated by those originated in higher-cost coastal cities, tended to refinance more aggressively in rallies, whereas New York loans tended to refinance sluggishly owing to the 1% (and higher) tax imposed on mortgages. (For similar reasons, Florida and Texas are also important slow-pay states.) At the outset, geographic location was proxied using the address of the original seller/servicer,19 but by 1990, the agencies were reporting current loan concentrations by the state in which the property was located. Geographic location also can be a factor when economic activity in particular states or regions is significantly above or below national trends (e.g., southern California in the early 1990s).

How strong an impact location can have is shown in Exhibit 3-16, which tracks the prepayments of 2003 Fannie Mae 5.5s with 100% concentrations of loans on California, Florida, New York, or Texas properties.

Loan size became a valued indicator of prepayment risk in the late 1990s. The market had long understood that to achieve equivalent payment savings, borrowers with smaller loans needed much steeper rate incentives than other borrowers. (A more precise measure would be payment savings as a percentage of household income, but this is not available.) Many market participants also

19. Clearly a weak proxy because consolidation led to increasingly regional and national mortgage banking companies.


Prepayments of 2003 Vintage Fannie Mae 5.5s by State

50 45 40 35 30

20 15 10

Prepayments of 2003 Vintage Fannie Mae 5.5s by State

Source: eMBS Inc. Reproduced with permission.

Source: eMBS Inc. Reproduced with permission.

had observed that lower-priced properties tended to turn over more readily in rising-interest-rate environments because they were more affordable. The variations in prepayment speeds across Fannie Mae 5.5s with varying average loan sizes are shown in Exhibit 3-17. That smaller loans prepaid more slowly at rally peaks is evident; the effect of loan size is evident throughout the low mortgage rate environment of 2003-2004.

In 2003, the agencies agreed to expand disclosure of underlying loan characteristics to include six additional characteristics (reported as weighted averages and quartile distributions and updated monthly):

  • Loan-to-value (LTV) ratio at origination (OLTV ratio)
  • Loan purpose: refinancing or purchase
  • Servicer (seller name was already disclosed; this captured transfers of servicing)
  • Credit score (typically FICO)
  • Property type: one unit or two to four units
  • Occupancy: owner-occupied, investment property, second home

This disclosure was welcomed by the broker-dealer and investor communities. Loan-level analysis on private issue CMOs and residential ABS had long


Prepayments of 2003 Vintage Fannie Mae 5.5s by Original Loan Size

DC 25 CL

Borrowing Basics

Borrowing Basics

Some small business persons cannot understand why a lending institution refused to lend them money. Others have no trouble getting funds, but they are surprised to find strings attached to their loans.

Get My Free Ebook

Post a comment