 |
NEW YORK (Standard & Poor's) Aug. 7, 2007--Standard & Poor's Ratings Services
today placed its ratings on 207 classes of U.S. residential mortgage-backed
securities (RMBS) backed by U.S. first-lien Alternative A (Alt-A) mortgage
collateral issued from the beginning of October 2005 through the end of
December 2006 on CreditWatch with negative implications.
The complete ratings list is included in "U.S. Alt-A RMBS Classes Affected By
Aug. 7, 2007, CreditWatch Actions,” available on RatingsDirect, the real-time
Web-based source for Standard & Poor's credit ratings, research, and risk
analysis. The list can also be found on Standard & Poor's Web site at
www.standardandpoors.com. Select Products and Services and then Ratings.
Choose Standard & Poor's Views On The Subprime Mortgage Market and scroll down
to Structured Finance.
These CreditWatch actions affect 207 classes that had an original total
balance of approximately $913.9. million, which represents 0.20% of the
approximately $455.4 billion in U.S. RMBS backed by first-lien Alt-A
collateral, rated by Standard & Poor's from the beginning of October 2005
through the end of December 2006. This is in addition to the ratings on 30
classes from 27 transactions from this period that were previously placed on
CreditWatch, beginning in March 2007. During the same period, the total
balance of U.S. RMBS securities backed by all types of residential mortgage
collateral issued in the non-agency market was more than $1.2 trillion.
We are also conducting a review of CDO transaction ratings in which the
underlying portfolios contain any of the securities affected by these rating
actions. A separate press release will be published later today.
Alt-A loans are first-lien residential mortgages that generally conform to
traditional "prime" credit guidelines; however, the loan-to-value (LTV) ratio,
loan documentation, occupancy status, property type, or other factors cause
these loans not to qualify under standard underwriting programs for prime
jumbo and prime quality conforming loans. This review included payment option
ARMs (adjustable-rate mortgages), which have, as a subgroup of Alt-A,
exhibited lower delinquency rates and have greater credit protection than the
average for the entire Alt-A sector.
These actions reflect a rising level of delinquencies among the Alt-A
collateral supporting these transactions, as well as our expectation that
losses on the collateral will exceed historical precedent and may exceed our
original expectations. The weak collateral performance is due to a combination
of factors that include, but are not limited to, high combined LTVs (CLTVs);
home price declines; an environment of looser underwriting standards; risk
layering (the combination of several risk elements for one single borrower);
and speculative borrowing behavior. The increase in the number of loans with
the combination of these risk characteristics, coupled with a reversal in home
price appreciation and higher interest rates, has led to deteriorating
performance among the transactions that closed at the end of 2005 through
2006.
Realized cumulative losses within U.S. RMBS transactions backed by Alt-A
collateral issued during this period have been relatively low to date, at
approximately 4 basis points (bps; 0.04%). The Alt-A transactions are
evidencing lower levels of early payment defaults than transactions from the
same period that are backed by first-lien subprime and closed-end second-lien
collateral. However, the collateral underlying the Alt-A transactions has been
experiencing high levels of severe delinquencies (90-plus days, foreclosures,
REOs) that have not abated. Our assumptions had anticipated that the
delinquency and default patterns of these transactions would revert to trends
closer to historic norms. However, these transactions have now reached a
sufficient level of seasoning for us to conclude that, based on the factors
above, they will evidence delinquency and default loss trends that are
indicative of poor future performance and that these trends will continue to
exceed historic precedent and our original ratings assumptions.
We are affirming the ratings on all U.S. RMBS transactions backed by Alt-A
collateral issued between October 2005 and December 2006 that are not
currently on CreditWatch. All of the transactions with affirmed ratings have
passed our stress tests. In addition, the percentage of loans with CLTVs
greater than 95% was less than 12% for the affirmed transactions, compared
with more than 17% for the pools backing the issues with ratings on
CreditWatch. As a general matter, the strong performing Alt-A collateral
within each transaction that does not contain risk layering has been
performing within historical norms and within our original expectations. This
type of collateral closely resembles collateral found in prime jumbo
transactions where borrowers had FICO scores in excess of 720 and CLTVs in the
low-to-mid 70% range; and to date, performance in the prime jumbo sector is
also within the bounds of historical norms. Through June 2007, the total
cumulative loss experienced was approximately 0.04%, or 4 bps.
ADJUSTMENTS TO SURVEILLANCE ASSUMPTIONS
Our review of outstanding U.S. RMBS backed by first-lien Alt-A collateral
employs different assumptions than those we used to review U.S. RMBS backed by
first-lien subprime loan collateral. Roll rate assumptions, or the number of
currently delinquent loans and loans in foreclosure that will ultimately
result in a loss, are lower than those we assumed in our review of U.S. RMBS
backed by first-lien subprime collateral. The time it takes to incur losses is
assumed to be longer for Alt-A loans than for subprime loans. Recovery rates
are assumed to be higher than for subprime loans because of lower LTVs and
lower foreclosure expenses as a percent of generally higher loan balances.
Our revised surveillance assumptions focus on three areas: current losses to
date, losses assumed on currently delinquent loans, and future losses for
borrowers who are current with their loan payments. Losses incurred to date
have already reduced current credit protection. The assumptions for how
delinquent loans will progress through the delinquency pipeline toward
liquidation are as follows: 50% of 30-day delinquencies, 50% of 60-day
delinquencies, 80% of 90-day delinquencies, 80% of foreclosures, and 100% of
REOs. Assumptions regarding the timing of the losses projected on loans
currently in REO will be recognized evenly over the next eight months. Losses
on loans currently in foreclosure will be recognized as follows: 5% evenly
over the following one to eight months; 70% evenly in months nine through 15;
and the remaining 25% evenly from months 16 through 21. Losses on 64% of the
loans that are currently 90-days delinquent will be recognized as follows: 5%
evenly in months five through 13; 70% evenly over months 14 through 20; and
25% evenly over months 21 through 24. Loans that are delinquent 60 and 30 days
follow a similar pattern, shifting the 90-day assumptions out one and two
months, respectively. We further stress future losses by taking losses at peak
in month 15 and maintaining losses at this level for an additional 12 months.
After this extended peak loss period, we expect losses to begin to revert to
more normalized levels, reflecting our expectation that most high-risk loans
will have already moved through the stress in the default curve.
For Alt-A transactions, we will apply a 25% loss severity in our surveillance
assumptions. This is a higher level of stress than recently observed
severities on Alt-A loan collateral. This is also lower than the loss severity
of 40% used in our subprime surveillance assumptions. On average, Alt-A loans
are much larger in size, include high levels of mortgage insurance, and have
considerably lower LTV ratios and weighted average coupons, all of which
should lead to lower loss severities compared with subprime loans.
We expect that the majority of the rating actions or affirmations on classes
that have been placed on CreditWatch negative will be resolved within the next
several weeks. For each transaction, we will perform a cash flow analysis that
stresses prepayment speeds (CPRs) together with its transaction-specific loss
projections. For each U.S. RMBS transaction backed by Alt-A collateral that
solely relies on credit support from subordination and fails this stress
within the next 36 months, consistent with other first-lien mortgage products,
rating actions will typically be taken in accordance with the following
guidelines:
-- To 'CCC' on any class that does not pass our stress test scenario (a
class is expected to experience a principal write-down or, with respect to the
senior classes, a principal shortfall) within 12 months, regardless of its
current rating;
-- To 'B' on any class that does not pass our stress test scenario within
13 to 24 months;
-- To 'BB' on any class that does not pass our stress test scenario
within 25 to 30 months; and
-- To 'BBB' on any class that does not pass our stress test scenario
within 31 to 36 months.
We also placed our ratings on CreditWatch for those transactions in which our
loss assumptions, as stated above, materially exceeded the available
subordination and reserve funds, for the purpose of conducting a specific
review of how these transactions will perform under various prepayment and
delinquency stresses involving loan resets. Rating actions will be taken
according to the guidelines listed above.
In cases where the remaining loss protection on a more senior class is
materially eroded by projected losses and a subordinate class is downgraded,
we will adjust our rating lower to reflect the reduced relative protection of
that class.
ADJUSTMENTS TO RATING ASSUMPTIONS FOR NEW TRANSACTIONS
We are also implementing several revisions to our assumptions for rating new
transactions that close on or after Aug. 7, 2007, as described below. In July
2006, we adjusted our analysis of new transactions to assume increased default
frequency rates, and the changes described below are further refinements to
these assumptions.
Recent performance data indicates that loans with layered risk and high CLTV
purchase money loans are highly correlated with the increased rate of
delinquencies and defaults being experienced on the transactions with ratings
placed on CreditWatch negative. As a result, as highlighted below, we are
increasing credit enhancement for purchase money mortgage loans with high
CLTVs and reducing our reliance on FICO as a risk mitigant to significant
levels of layered risk.
In late 2005 and 2006, mortgage origination underwriting guidelines expanded
rapidly, which allowed the proliferation of layered risks within the Alt-A
market. This combination of multiple risk factors for a single loan is the
principal driving force behind the deteriorating performance of the 2006
vintage. Historically, the presence of high FICO scores within a loan has
proved an effective mitigant to increased risk factors elsewhere, such as
higher CLTVs. However, the increase in recent delinquencies across all FICO
bands indicates that a borrower's previous credit performance is less
predictive of stronger performance for loans with increased risk layering.
This emerging delinquency performance has prompted us to reduce our emphasis
on FICO score as an offset to layered risk. Recent delinquency data also
indicates a need to adjust default expectations for certain purchase loans.
These loans are underperforming our initial assumptions, particularly when
combined with high CLTVs. The performance related to purchase loans is
unprecedented in historical data. We will increase our default expectations
for the increased risk at high CLTVs, particularly those with CLTVs that
exceed 90%.
Our default expectation for purchase loans with high CLTV ratios will increase
by approximately 5%-50%, depending on the corresponding CLTV ratio.
IMPACT ON CURRENT RATINGS
The CreditWatch actions on the 207 different classes are spread across the
various ratings categories as follows: 54.11% are from the 'BBB' rating
category; 22.22% are from the 'BB' rating category; 14.97% are from the 'B'
rating category; 8.70% are from the 'A' rating category. Therefore, 91.30% of
the lowered ratings affected classes that were rated 'BBB+' or lower. (It
should be noted that we did not place any 'AAA' or 'AA' ratings on CreditWatch
negative.)
U.S. RMBS transactions backed by Alt-A collateral issued beginning January
2007 have not had adequate seasoning to establish a payment history that would
make the outcomes of the delinquency and loss tests described above
meaningful. However, since the same asset risks that are apparent in the
transactions issued in 2005 and 2006 may also be present in the January 2007
through July 2007 transactions, we will continue to monitor the 2007 vintage
securitizations and apply our revised surveillance assumptions to these
transactions as they season and as delinquency and loss data become available.
We will also review these transactions using our revised assumptions for new
ratings and may take rating actions, as deemed appropriate, going forward.
Standard & Poor's, a division of The McGraw-Hill Companies (NYSE:MHP), is the
world's foremost provider of independent credit ratings, indices, risk
evaluation, investment research and data. With approximately 6,300 employees
located in 20 countries and markets, Standard & Poor's is an essential part of
the world's financial infrastructure and has played a leading role for more
than 140 years in providing investors with the independent benchmarks they
need to feel more confident about their investment and financial decisions.
For more information, visit http://www.standardandpoors.com.
|
