New York, October 09, 2019 -- Moody's Investors Service ("Moody's") has assigned provisional ratings
to eighteen classes of notes issued by Mill City Mortgage Loan Trust ("MCMLT")
The certificates are backed by one pool of 2,356 seasoned performing
and modified re-performing loans which includes a small portion
of junior liens mortgage loans (13.66%) and loans that are
currently 60+ days MBA delinquent (5.14%). The
collateral pool has a non-zero updated weighted average FICO score
of 642 and a weighted average current LTV of 93.91% (including
the deferred balance for calculation).
Similar to MCMLT 2019-1, this pool does not have any HELOC
loans. In addition, approximately 12.33% of
the loans are originated on or after January 1, 2010 ("newly originated
loans"). 78.62% of the loans in the collateral pool
were also previously modified and the remaining loans have never been
Fay Servicing LLC ("Fay") and Shellpoint Mortgage Servicing ("Shellpoint"),
are the servicers for the loans in the pool. The servicers will
not advance any principal or interest on the delinquent loans.
However, the servicers will be required to advance costs and expenses
incurred in connection with a default, delinquency or other event
in the performance of its servicing obligations.
Goldman Sachs Mortgage Company (GSMC) is acquiring the collateral from
a CarVal Investors limited fund. GSMC as a sponsor and loan seller
is structuring the transaction. This arrangement is weaker than
the alignment of interests in the prior MCMLT transactions where the sponsor,
the depositor and the R&W provider are affiliates of CarVal Investors.
A mitigating factor is that GSMC will retain at least 5% of the
notes to satisfy U.S. risk retention requirements.
Similar to other MCMLT transactions, a non-rated limited
fund from Carval Investors will be the R&W provider.
The complete rating actions are as follows:
Issuer: Mill City Mortgage Loan Trust 2019-GS1
Cl. A1, Assigned (P)Aaa (sf)
Cl. A1A, Assigned (P)Aaa (sf)
Cl. A1B1, Assigned (P)Aaa (sf)
Cl. A1B2, Assigned (P)Aaa (sf)
Cl. A2, Assigned (P)Aa3 (sf)
Cl. A3, Assigned (P)A1 (sf)
Cl. A4, Assigned (P)A3 (sf)
Cl. B1, Assigned (P)Ba3 (sf)
Cl. B1A, Assigned (P)Ba1 (sf)
Cl. B1B, Assigned (P)Ba3 (sf)
Cl. B2, Assigned (P)Caa1 (sf)
Cl. B2A, Assigned (P)B1 (sf)
Cl. B2B, Assigned (P)Caa1 (sf)
Cl. M1, Assigned (P)Aa3 (sf)
Cl. M2, Assigned (P)A3 (sf)
Cl. M3, Assigned (P)Baa3 (sf)
Cl. M3A, Assigned (P)Baa3 (sf)
Cl. M3B, Assigned (P)Baa3 (sf)
Summary Credit Analysis and Rating Rationale
Moody's expected losses on MCMLT 2019-GS1's collateral pool average
17.25% in our base case scenario. Our loss estimates
take into account the historical performance of loans that have similar
collateral characteristics as the loans in the pool.
For the non-modified portion of this pool, we analyzed data
on delinquency rates for always current (including self-cured)
loans. Similarly, for the modified portion of this pool,
we analyzed data on delinquency rates for modified loans. Our final
loss estimates also incorporates adjustments for the strength of the third
party due diligence, the servicing arrangement and the representations
and warranties (R&W) framework of the transaction.
The methodologies used in these ratings were "Moody's Approach to Rating
Securitizations Backed by Non-Performing and Re-Performing
Loans" published in February 2019, and "US RMBS Surveillance Methodology"
published in February 2019. Please see the Rating Methodologies
page on www.moodys.com for a copy of these methodologies.
MCMLT 2019-GS1 is a securitization of 2,356 loans and is
primarily comprised of seasoned performing and modified re-performing
mortgage loans. Approximately 78.62% of the loans
in the collateral pool have been previously modified (including the deferred
balance for calculation).
We based our expected losses on our estimates of 1) the default rate on
the remaining balance of the loans and 2) the principal recovery rate
on the defaulted balances. The two factors that most strongly influence
a re-performing mortgage loan's likelihood of re-default
are the length of time that the loan has performed since a loan modification,
and the amount of the reduction in the monthly mortgage payment as a result
of the modification. The longer a borrower has been current on
a re-performing loan, the less likely the borrower is to
re-default. Approximately 57.67% of the borrowers
have been current (OTS method of delinquency) on their payments for at
least the past 24 months.
We estimated expected losses for the pool using two approaches --
(1) pool-level approach, and (2) re-performing loan
In the pool-level approach, we estimate losses on the pool
by using a approach similar to our surveillance approach wherein we apply
assumptions on expected future delinquencies, default rates,
loss severities and prepayments as observed from our surveillance of similar
collateral. We project future annual delinquencies for eight years
by applying an initial annual default rate and delinquency burnout factors.
Based on the loan characteristics of the pool and the demonstrated pay
histories, we expect an annual delinquency rate of 12.88%
on the first lien portion of the collateral pool for year one.
We then calculated future delinquencies on the pool using our default
burnout and voluntary conditional prepayment rate (CPR) assumptions.
Our assumptions also factor in the high delinquency rates expected in
the early stages of the transaction due to payment shock expected for
step-rate loans. The delinquency burnout factors reflect
our future expectations of the economy and the U.S. housing
market. We then aggregated the delinquencies and converted them
to losses by applying pool-specific lifetime default frequency
and loss severity assumptions. Our loss severity assumptions are
based off observed severities on liquidated seasoned loans and reflect
the lack of principal and interest advancing on the loans.
We also conducted a loan level analysis on MCMLT 2019-GS1's collateral
pool. We applied loan-level baseline lifetime propensity
to default assumptions based on the historical performance of loans with
similar collateral characteristics and payment histories. We then
adjusted this base default propensity up for (1) adjustable-rate
loans, (2) loans that have the risk of coupon step-ups and
(3) loans with high updated loan to value ratios (LTVs). We applied
a higher baseline lifetime default propensity for interest-only
loans, using the same adjustments. To calculate the final
expected loss for the pool, we applied a loan-level loss
severity assumption based on the loans' updated estimated LTVs.
We further adjusted the loss severity assumption upwards for loans in
states that give super-priority status to homeowner association
(HOA) liens, to account for potential risk of HOA liens trumping
a mortgage. The deferred balance in this transaction is approximately
$45.45 million, representing approximately 11.77%
of the total unpaid principal balance.
Three loans in the pool currently feature an active HAMP Principal Reduction
Alternative (HAMP-PRA). Under HAMP-PRA, the
principal of the borrower's mortgage may be reduced by a predetermined
amount called the PRA forbearance amount if the borrower satisfies certain
conditions during a trial period. If the borrower continues to
make timely payments on the loan for three years, the entire PRA
forbearance amount is forgiven. Also, if the loan is in good
standing and the borrower voluntary pays off the loan, the entire
forbearance amount is forgiven.
For non-PRA forborne amounts, the deferred balance is the
full obligation of the borrower and must be paid in full upon (i) sale
of property (ii) voluntary payoff or (iii) final scheduled payment date.
Upon sale of the property, the servicer therefore could potentially
recover some of the deferred amount. For loans that default in
future or get modified after the closing date, the servicer may
opt for partial or full principal forgiveness to the extent permitted
under the servicing agreement.
Based on performance data and information from servicers, we assume
that 100% of the remaining PRA amount would be forgiven and not
recovered. For non-PRA deferred balance, we applied
a slightly higher default rate for these loans than what we assumed for
the overall pool given that these borrowers have experienced past credit
events that required loan modification, as opposed to borrowers
who have been current and have never been modified. Also,
for non-PRA loans, based on performance data from an RPL
servicer, we assumed approximately 95% severity as servicers
may recover a portion of the deferred balance. For this pool,
non-PRA deferred balance account for 100.00% of the
deferred balance. The final expected loss for the collateral pool
reflects the due diligence scope and findings of the independent third
party review (TPR) firms as well as our assessment of MCMLT 2019-GS1's
representations & warranties (R&Ws) framework.
Similar to other MCMLT transactions, MCMLT 2019-GS1 has a
simple sequential priority of payments structure without any cash flow
triggers. The transaction allocates excess cash flow (net of realized
losses and certain unreimbursed amounts) sequentially to the senior notes
A1A, A1B1 and A1B2 and then to the subordinated tranches,
class M1 through class B6B up to a target payment amount. This
arrangement is weaker than previous MCMLT transactions (except for MCMLT
2019-1) where all the remaining excess cash flow (net of realized
losses, certain unreimbursed amount and payment to senior notes)
is distributed to the subordinated tranches. In this transaction,
after the excess cash flow is used to pay the subordinated tranches the
target payment amount, the remaining excess cash flow will be distributed
to the residual certificates. We took into consideration the change
in the monthly excess cashflow waterfall in our modeling of the transaction's
The transaction will benefit from overcollateralization but it will experience
spread compression due to deleveraging, prepayment and modification.
We took this into account in our analysis. The servicer will not
advance any principal or interest on delinquent loans. However,
the servicer will be required to advance costs and expenses incurred in
connection with a default, delinquency or other event in the performance
of its servicing obligations.
Credit enhancement in this transaction is comprised of subordination provided
by mezzanine and junior tranches and the buildup of overcollateralization
from available excess interest. The principal payment received
from excess interest collections will limit the faster pay down on the
senior notes as a smaller percentage of the excess cash flow would be
allocated to the seniors due to structural features mentioned previously.
To the extent that the overcollateralization amount is zero or insufficient
monthly excess cash flow, realized losses will be allocated to the
notes in a reverse sequential order starting with the lowest subordinate
bond. The Class A1A, Class A1B1, Class A1B2,
Class A1, Class A2, Class A3, Class A4, Class
M1, Class M2, Class M3A, Class M3B, Class M3,
Class B1A, Class B1B, Class B1, Class B2A, Class
B2B and Class B2 notes carry a fixed-rate coupon subject to the
collateral adjusted net weighted average coupon (WAC) and applicable available
funds cap. The Class B3A, Class B3B, Class B3,
Class B4A, Class B4B, Class B4, Class B5A, Class
B5B, Class B5, Class B6A, Class B6B and Class B6 are
variable rate notes where the coupon is equal to the lesser of adjusted
net WAC-0.25% and applicable available funds cap.
To assess the final rating on the notes, we ran 96 different loss
and prepayment scenarios through our cash flow model. The scenarios
encompass six loss levels, four loss timing curves, and four
prepayment curves. The structure allows for timely payment of interest
and ultimate payment of principal with respect to the notes by the legal
Third Party Review
Three third party review (TPR) firms conducted due diligence on all but
few loans in MCMLT 2019-GS1's collateral pool. The TPR firms
reviewed compliance, data integrity and key documents, to
verify that loans were originated in accordance with federal, state
and local anti-predatory laws. The TPR firms also conducted
audits of designated data fields to ensure the accuracy of the collateral
tape. An independent firm also reviewed the title and tax reports
for all the loans in the pool.
Based on our analysis of the third-party review reports,
we determined that a portion of the loans with some cited violations are
at enhanced risk of having violated TILA through an under-disclosure
of the finance charges. In addition, the diligence providers
were unable to determine if four loans were originated in accordance with
ATR rules. We have made adjustments consistent with our approach
to account for the rating impact of ATR rules that could cause future
losses to the trust. We incorporated an additional hit to the loss
severities for these loans to account for this risk. The title
review includes confirming the recordation status of the mortgage and
the intervening chain of assignments, the status of real estate
taxes and validating the lien position of the underlying mortgage loan.
Once securitized, delinquent taxes will be advanced on behalf of
the borrower and added to the borrower's account. The servicer
will be reimbursed for delinquent taxes from the top of the waterfall,
as a servicing advance. The representation provider has deposited
collateral of $650,000 in the Assignment Reserve Account
(ARA) to ensure one or more third parties monitored by the Depositor completes
all assignment and endorsement chains and record an intervening assignment
of mortgage as necessary. The amount deposited in the ARA at the
closing date is lower than the previous Mill City transaction, MCMLT
2019-GS1. We have considered the ARA deposit and factors
such as: (i) the high historical cure rate in the previous Mill
City transactions and(ii) the low delinquency rate of the previous Mill
City transactions. We did not make any additional adjustment for
Representations & Warranties
Our ratings also factor in MCMLT 2019-GS1's weak representations
and warranties (R&Ws) framework because they contain many knowledge
qualifiers and the regulatory compliance R&W does not cover monetary
damages that arise from TILA violations whose right of rescission has
expired. The breach discovery process for this transaction is also
weaker than previous Mill City securitizations (except MCMLT 2017-3,MCMLT
2018-3 and MCMLT 2018-4) and other rated RPL transactions.
Previously, with the exception of MCMLT 2017-3, MCMLT
2018-3 and MCMLT 2018-4, an independent party reviewed
R&W breaches for every loan that became 120 days delinquent.
For this transaction similar to MCMLT 2017-3, MCMLT 2018-3
and MCMLT 2018-4, an independent party reviews R&W breaches
for every loan that incurs a realized loss.
While the transaction provides for a Breach Reserve Account to cover for
any breaches of R&Ws, the size of the account is slightly smaller
for MCMLT 2019-GS1 ($775,000 relative to aggregate
collateral balance of $386.29 million) compared to MCMLT
2019-1 ($0.9 million relative to aggregate collateral
pool $440.2 million). An initial deposit of $775,000
will be remitted to the Breach Reserve Account on the closing date,
with an initial Breach Reserve Account target amount of $1.2
million. We did not make any adjustment for this as it was not
much different in terms of percentage of balance.
We believe there is a very low likelihood that the rated notes in MCMLT
2019-GS1 will incur any loss from extraordinary expenses or indemnification
payments owing to potential future lawsuits against key deal parties.
First, majority of the loans are seasoned with demonstrated payment
history, reducing the likelihood of a lawsuit on the basis that
the loans have underwriting defects. Second, historical performance
of loans aggregated by the sponsor to date has been within expectation,
with minimal losses on previously issued Mill City transactions.
Third, the transaction has reasonably well defined processes in
place to identify loans with defects on an ongoing basis. In this
transaction, an independent breach reviewer must review loans for
breaches of representations and warranties when a realized loss is incurred
on a loan, which reduces the likelihood that parties will be sued
for inaction. Furthermore, the issuer has performed nearly
73.4% due diligence by independent third parties with respect
to compliance and payment history and has disclosed the results of the
The transaction benefits from an adequate servicing arrangement.
Shellpoint will service 73.95% of the pool, Fay will
service 26.05% of the pool. Wells Fargo Bank,
N.A. is the Custodian of the transaction. MCMLT 2019-GS1's
Indenture Trustee is U.S. Bank National Association.
Factors that would lead to an upgrade or downgrade of the ratings:
Levels of credit protection that are higher than necessary to protect
investors against current expectations of loss could drive the ratings
up. Losses could decline from our original expectations as a result
of a lower number of obligor defaults or appreciation in the value of
the mortgaged property securing an obligor's promise of payment.
Transaction performance also depends greatly on the US macro economy and
housing market. Other reasons for better-than-expected
performance include changes to servicing practices that enhance collections
or refinancing opportunities that result in prepayments.
Levels of credit protection that are insufficient to protect investors
against current expectations of loss could drive the ratings down.
Losses could rise above our original expectations as a result of a higher
number of obligor defaults or deterioration in the value of the mortgaged
property securing an obligor's promise of payment. Transaction
performance also depends greatly on the US macro economy and housing market.
Other reasons for worse-than-expected performance include
poor servicing, error on the part of transaction parties,
inadequate transaction governance and fraud.
For further specification of Moody's key rating assumptions and sensitivity
analysis, see the sections Methodology Assumptions and Sensitivity
to Assumptions of the disclosure form.
Further information on the representations and warranties and enforcement
mechanisms available to investors are available on http://www.moodys.com/researchdocumentcontentpage.aspx?docid=PBS_1197795.
The analysis includes an assessment of collateral characteristics and
performance to determine the expected collateral loss or a range of expected
collateral losses or cash flows to the rated instruments. As a
second step, Moody's estimates expected collateral losses or cash
flows using a quantitative tool that takes into account credit enhancement,
loss allocation and other structural features, to derive the expected
loss for each rated instrument.
Moody's quantitative analysis entails an evaluation of scenarios
that stress factors contributing to sensitivity of ratings and take into
account the likelihood of severe collateral losses or impaired cash flows.
Moody's weights the impact on the rated instruments based on its
assumptions of the likelihood of the events in such scenarios occurring.
For ratings issued on a program, series, category/class of
debt or security this announcement provides certain regulatory disclosures
in relation to each rating of a subsequently issued bond or note of the
same series, category/class of debt, security or pursuant
to a program for which the ratings are derived exclusively from existing
ratings in accordance with Moody's rating practices. For ratings
issued on a support provider, this announcement provides certain
regulatory disclosures in relation to the credit rating action on the
support provider and in relation to each particular credit rating action
for securities that derive their credit ratings from the support provider's
credit rating. For provisional ratings, this announcement
provides certain regulatory disclosures in relation to the provisional
rating assigned, and in relation to a definitive rating that may
be assigned subsequent to the final issuance of the debt, in each
case where the transaction structure and terms have not changed prior
to the assignment of the definitive rating in a manner that would have
affected the rating. For further information please see the ratings
tab on the issuer/entity page for the respective issuer on www.moodys.com.
For any affected securities or rated entities receiving direct credit
support from the primary entity(ies) of this credit rating action,
and whose ratings may change as a result of this credit rating action,
the associated regulatory disclosures will be those of the guarantor entity.
Exceptions to this approach exist for the following disclosures,
if applicable to jurisdiction: Ancillary Services, Disclosure
to rated entity, Disclosure from rated entity.
Regulatory disclosures contained in this press release apply to the credit
rating and, if applicable, the related rating outlook or rating
Please see www.moodys.com for any updates on changes to
the lead rating analyst and to the Moody's legal entity that has issued
Please see the ratings tab on the issuer/entity page on www.moodys.com
for additional regulatory disclosures for each credit rating.
Asst Vice President - Analyst
Structured Finance Group
Moody's Investors Service, Inc.
250 Greenwich Street
New York, NY 10007
JOURNALISTS: 1 212 553 0376
Client Service: 1 212 553 1653
VP - Senior Credit Officer/Manager
Structured Finance Group
JOURNALISTS: 1 212 553 0376
Client Service: 1 212 553 1653
Moody's Investors Service, Inc.
250 Greenwich Street
New York, NY 10007
JOURNALISTS: 1 212 553 0376
Client Service: 1 212 553 1653