Story Time
with Tradex: Understanding Mortgage Cohort “Stories”
Not all mortgages are made the same.
Some mortgages are small, and some are large. Others are young, and others are
old. Some are rich and some are poor. In this brief write-up, we attempt to clarify
common mortgage collateral stories and the anticipated effects on prepayments
and valuation.
While most fixed income products have
known cash flows, mortgage backed securities do not because they are dependent
upon homeowner behavior that can be difficult to predict. Each month, a
homeowner faces a choice of paying, refinancing, or defaulting on the mortgage.
While these decisions should be relatively easy to predict based on rate
incentive, empirical evidence suggests that borrowers do not always act rationally
to maximize economic utility, making prepayment analysis both a behavioral and
economic study. Since thousands of loans are pooled into a single pass-through,
the law of large numbers should in theory reduce noise from irrational
homeowner actions, or lack thereof. However, certain ‘cohorts’ of homeowners
may collectively act in an irrational manner. The stratification of homeowner
characteristics into a single security often facilitates prepayment analysis.
Collateral Story Proliferation
Following the financial crisis, the
FHFA developed the HARP program[1]
to help American homeowners refinance into mortgages with more affordable monthly
payments. This program eventually led to an exponential growth of collateral
stories, and increased the knowledge necessary to accurately value and
anticipate these cash flows. Examples of story collateral include MHA Program
(HARP), loan balance, seasoned, geographic exposure, mortgage purpose
(relocation, investor, etc.), FICO, Third Party Origination, and many others. These
collateral types will often exhibit prepayment behavior that vary from their
generic counterparts due to differing degrees of refi sensitivity.
HARP Program
In March 2009, the FHFA
introduced the Home Affordable Refinance Program to, “provide access to
low-cost refinancing for responsible homeowners suffering from falling home
prices1.” The program allowed responsible homeowners to refinance
their loans despite having loan to value (LTV) ratios above 80%. Simply stated,
as home prices fell, LTV ratios increased dramatically and homeowners could not
take advantage of falling mortgage rates due to low or negative home equity. The
HARP I and II programs served as a remedy to this issue and led to a meaningful increase in prepayment speeds
on HARP eligible collateral. Many of these HARP loans were subsequently
securitized into a new pool and designated as “MHA 80”, “MHA 90”, etc. pools.
Investors viewed these pools as call protected[2]
since a borrower who refinanced under the program would be ineligible to access
the HARP program again. Initially, these pools carried lower risk premium.
However, as home prices rebounded and LTVs fell, the pools began to prepay
faster as mortgage rates remained low.
Loan Balance
While the average loan size in a
new mortgage is approximately $268k[3],
some pools are composed of loans whose original size is below 85k, while others
are greater than 417k. The first case, “LLBs”, are Low Loan Balance pools. Investors
often view LLBs as call protected because the total dollar savings of
refinancing is less on small loans. These homeowners are thus less likely to
refinance and, in theory, LLB pools should have more stable and predictable
cash flow. The result of lower expected prepayments and greater degree of cash
flow predictability leads to higher relative valuation and smaller risk premium.
This collateral type is often highly desirable when interest rates decline. On
the other end of the spectrum, Jumbo loans will have higher dollar savings than
smaller loan balance collateral given the same rate incentive. Thus, bonds
backed by Jumbo loans often have a larger risk premium as those borrowers have
greater incentives to refinance.
Seasoned (burnout)
Collateral seasoning refers to
the age of the underlying loans in a mortgage pool. Most new loans have
mortgage rates that are at or around prevailing market rates and those
borrowers have no rate incentive to refi. Post origination, mortgage rates will
change as bank borrowing costs increase or decrease. Over time, a mortgage pool
may experience periods in which homeowners have significant incentives to
refinance. Many borrowers will refinance and will exit the pool. What remains
is mortgage loans that have above-market coupons that are deep in the money.
For some reason, the homeowners remaining in the pool have not refinanced their
loans despite the financial incentive to do so. We often refer to this
phenomenon as burnout, and this seasoned collateral tends to have a muted
response to increases in refi incentive. Furthermore, seasoned collateral often
has a steady prepayment profile and investors often place a low risk premium on
this collateral type.
Geographic Exposure
The colloquialism, “all politics
is local” comes to mind when considering the geographic distribution of
underlying mortgage loans. While economic activity is often summarized at the
US level, this economic activity is an amalgamation of local economies at the
MSA (Metropolitan Statistical Area) and state level. Some of these economies
may be expanding, while others may be shrinking. This activity directly affects
borrowers’ income and home value, ultimately affecting prepayment behavior.
Geographic stories, or “GEOs” for short, typically contain 100% borrowers from
a certain state, such as California, NY, TX, etc. California borrowers tend to be
more sophisticated and own larger homes. Thus, they are highly sensitive to
rate incentives. CA bonds typically carry a higher risk premium than say, NY,
TX, or PR (Puerto Rico) bonds. When investing in GEO stories, it is prudent to
be aware of homeowner and mortgage attributes as it directly influences cash
flow. For example, recent declines in oil prices has negatively impacted shale
states’ economies. This may translate into lower HPA, less turnover and muted
refinancing incentive.
Conclusion
In the wake of the financial
crisis, the number and complexity of collateral stories has significantly
increased. As a result, investment managers must be cognizant of the various
collateral types and have an understanding of how they affect prepayment behavior
and MBS valuation. While this editorial provides some clarification of collateral
stories, we encourage our readers to research mortgage collateral stories further
and to have a dialogue with the investment team at Tradex.
Tradex Global Advisory Services, LLC
investorrelations@thetradexgroup.com
203-863-1500
@Tradex_Global
[1]
Collectively HARP I and HARP II. https://www.fanniemae.com/singlefamily/making-home-affordable
[2]
Call protection refers to the call option held by the homeowner. If prevailing
mortgage rates are lower than a homeowner’s mortgage rate, the homeowner may exercise
the call option and refinance into cheaper loans.
[3]
MBA US Conventional Refinance Average Loan Size Index
No comments:
Post a Comment