Auto Loan ABS
Auto Loan ABS
Auto Loan ABS
RATING METHODOLOGY
Executive Summary
Table of Contents:
EXECUTIVE SUMMARY
MAIN RISKS OF TYPICAL AUTO LOAN
TRANSACTION
ESTIMATING THE POOLS EXPECTED LOSS
METHODS FOR ASSESSING THE
VARIABILITY OF LOAN LOSSES
FACTORS THAT AFFECT THE POTENTIAL
VARIABILITY OF A POOLS LOSSES
THE EXPECTED LOSS APPROACH: A
PROBABILISTIC APPROACH WITH A
LOGNORMAL DISTRIBUTION
MODELING THE BOND STRUCTURE
USING THE MODEL OUTCOME AS INPUT
IN THE RATING COMMITTEE PROCESS
LEGAL CONSIDERATIONS
MONITORING
APPENDICES
MOODYS RELATED RESEARCH
1
2
2
5
6
8
8
10
11
12
13
25
The report combines, updates, and replaces 2 previous rating methodologies used to rate auto
loans ABS transactions.*
Analyst Contacts:
NEW YORK
This methodology covers our global rating approach to rating securities backed by pools of
auto loans to individuals or products with similar characteristics. 1 We discuss the main risk
drivers for typical auto loan transaction, including portfolio credit quality, transaction
structure, counterparty risk and operational risk, legal risk and sovereign risk. The
methodology also provides a step-by-step description of the rating process for securities
backed by auto loans. First, we estimate the likely loss (expected loss) of the auto loan pool
and the variability of loss to derive a distribution of the losses on the pool. Because they are
granular pools, we assume that the loss distribution for auto loans pools is lognormal.
Second, we use that loan loss probability distribution to derive the expected loss for the
security, using a model of the transactions cash flow structure. Third, we then compare the
securitys expected loss to our benchmarks for each rating level to determine the rating for
the security. Fourth, we determine the actual rating considering the foregoing analysis
together with other quantitative analyses and subjective assessments of factors, including
operational risk, counterparty risk and the legal structure of the transaction.
+1.212.553.1653
Mack Caldwell
+1.212.553.4106
Senior Vice President - Manager
mcginnis.caldwell@moodys.com
>>contacts continued on the last page
Contributors:
Sophie Berthelon
Senior Vice President
Corey Henry
Vice President Senior Analyst
Steven Bild
Analyst
* Andrew Silver also contributed to this report as
a research consultant
Pools backed by auto finance products such as hire purchase contracts will generally be considered as sharing similar characteristics. Non-operating, auto finance leases in
Australia fall under this category as well. The methodology will also be applicable to pools with a proportion of corporate obligors but which remain very granular.
This Credit Rating Methodology is being implemented on a global basis, except in jurisdictions where certain regulatory requirements must be fulfilled prior to
implementation in those jurisdictions.
ASSET-BACKED SECURITIES
the risk profiles of auto loan obligors as illustrated when available by their credit scores
the underlying type of vehicle (e.g. new or used) and specific loan characteristics (term,
amortization profile, interest rate), which influence borrower performance and the level of
recovery upon borrower default
current and forecast macroeconomic environments, which affect consumer behavior as well as the
health of the automobile industry in the relevant country
historical performance of pools with similar characteristics, usually provided by the originator
Transaction Structure: Specific features such as cash flow allocations, forms of credit enhancement
and cash-trapping mechanisms have an impact on the expected loss for each tranche of securities. For
transactions with a revolving or pre-funding period, the ability to replenish the portfolio with new
loans will add some uncertainty to the portfolio composition. When modeling the transaction, we aim
to capture the main structural features described in the transaction documentation.
Counterparty Risk/Operational Risk: Our assessment focuses on various counterparties in the
transaction and identifies main areas of counterparty risk (servicing, cash management, swaps) and any
associated structural mitigants, such as counterparty replacement triggers.
Legal Aspects: This analysis ensures that assumptions we use with regards to the quality of assets and
transaction structure are the same as the assets and structure the transaction documentation describes.
We also assess risks with regard to the assignment of the assets to the special purpose vehicle (SPV),
bankruptcy remoteness of the SPV or other jurisdiction-specific issues.
Sovereign Risk: The jurisdiction of assets, originator or transaction SPV is subject to country-specific
systemic risks. We identify the risks and factor them into our analysis. Examples include modifying
appropriate assumptions, using rating caps or defining minimum credit enhancement levels required
to achieve a particular rating.
ASSET-BACKED SECURITIES
gross default with recoveries separately. In the latter case we analyze the two components separately
and derive a cumulative default projection and a recovery assumption 3 together with recovery timing.
In certain transactions the legal structure will also drive the recovery assumption: for example, in most Japanese transactions, recovery cash flows dont benefit senior
noteholders, therefore we assume a recovery rate of zero.
The cumulative-loss-to-liquidation rate at a point in time is the cumulative losses to date of the pool, divided by the difference between the original pool balance and the
current pool balance (e.g., the cumulative liquidations to date). In contrast, the traditional cumulative loss rate is the cumulative losses to date divided by the original
pool balance of the loans.
ASSET-BACKED SECURITIES
pool of loans being securitized, either because of recent changes in the originators origination,
underwriting and servicing policies and strategies, or because of our expectation that the future
economic environment will be materially different from the one from which the historical performance
data came. However, there could be cases in which we would not be able to assign a rating due to
insufficiency of reliable historical data.
We select comparable originators based on similar pool characteristics and origination, underwriting,
collection and charge-off policies. To incorporate data from other originators, we adjust our analysis
for any differences in definitions of defaults and loan losses. However, originators tend to be
idiosyncratic to some extent; thus, the applicability of other originators data performance is not
perfect, which adds uncertainty to the analysis.
loan type (interest rate type, fully amortizing vs. balloon portions)
the obligors data (individuals vs. corporate, geographical and obligor concentrations, FICO or
internal credit score, down payment, debt-to-income and payment-to-income ratios)
To use the stratified data, we construct a new static pool loss analysis, weighting the disaggregated
performance data of each sub-pool by the proportions of loans with each characteristic in the
securitized pool. We then project the expected loss for the pool by using the extrapolation method
referred to above. Alternatively, for example when we want to get a loss projection by loan type, we
may extrapolate expected loss for each sub-pool first and then derive the pool expected loss from the
weighted average of the extrapolated loss for each sub-pool, using the weights of the sub-pool in the
securitized pool or concentration limits as per the legal documentation for revolving or pre-funding
transactions.
ASSET-BACKED SECURITIES
The need for adjustment arises principally from the need to account for 1) the amount of amortization
versus the losses that have already occurred, 2) the typical exclusion of delinquent loans from a
securitization and 3) the effects of lags in recoveries on defaulted loans. For relatively unseasoned
securitization pools, each of the effects will be relatively small so that the net effect is usually negligible.
For more seasoned securitization pools, the adjustment either increases or decreases our expected loss
projection. The degree of effect depends ultimately on the interplay of the various underlying factors
such as the timing of default, recoveries, prepayments and delinquencies.
Adjusting for Changes in Servicing Practices
Changes in servicing practices affect the delinquency, loss and recovery performance of the pool of
loans. Those changes often affect performance with a lag, with the effects not appearing in the data at
the time of analysis. Consequently, we incorporate our assessment of recent trends in the servicers
practices into our analysis, based largely on an operations review meeting with the servicer. We also
make qualitative adjustments to our expected loss, default or recovery projections based on that
analysis even if the effects have not appeared in the performance data.
Adjusting for Potential Changes in the Macroeconomic Environment
The historical data that we analyze is, in part, a product of their macroeconomic environment. Therefore,
if we expect future macroeconomic conditions to be materially different than historical conditions, we
will adjust our projection of the expected loss accordingly. We do so by looking at our macroeconomic
board projections whenever available. When not available, we look at alternate sources. We focus on
macroeconomic variables which we consider as important drivers of performance for auto loans pools: i)
the countrys GDP growth rate, ii) unemployment rate and when available iii) used car values. For certain
regions with more volatile macroeconomic environments, such as Latin America, adjustments to historical
observation could be significant.
ASSET-BACKED SECURITIES
enhancement, aiming at capturing the tail risk of the distribution). 6 That Aaa level of credit
enhancement derives 1) from credit enhancement levels of the existing comparable transactions and 2)
adjustments made judgmentally between the given pool and the comparable transactions to account
for differences in the factors affecting variability. We use that rating committee assessment to infer the
standard deviation of the loss distribution as described later. For a given loss estimate, the higher the
Aaa level, the higher is the implicit standard deviation of the loss distribution. Factors driving the Aaa
level of credit enhancement are described in the next section.
Servicing Stability
In assessing the pool loss variability, we examine the stability of the servicer, from both financial
strength and operational perspectives, to determine the likelihood that the servicer will apply
6
For transactions in countries where the rating of the senior class is subject to the Local Currency Country Risk Ceiling, the volatility proxy level represents the level of
credit enhancement that the rating committee would deem to be consistent for a security backed by the given asset pool to receive the maximum achievable rating. See
Appendix 5 for details on the incorporation of country risk in the calibration of the loss distribution.
ASSET-BACKED SECURITIES
consistent servicing practices and policies. The ability of the servicer to collect on the loans, mitigate
losses, and maximize recoveries has a direct impact on the loss performance of a pool.
Another factor in assessing servicing stability is the servicers operational structure. It affects the degree
to which a dislocation would impact the pool's loss performance, including dislocation arising from a
servicing transfer owing to servicer financial stress or a natural disaster. For example, historical
experience tells us that performance deterioration can be greater when it occurs with a decentralized
operation.
those relating to the obligors' creditworthiness (e.g., FICO score or internal credit score), their
capacity to repay (e.g., payment-to-income ratio)
key loan characteristics (e.g., loan-to-value ratio, original loan term, whether the underlying
vehicle is new or used, whether the loans are fully amortizing or balloon loans). Balloon loans
present some specific risks that are further explained in Appendix 2.
The availability of such information for the comparable static pools and for the securitized pool helps
reduce the potential variability around the loss estimate for the securitized pool.
Prefunding and revolving periods both allow for additional receivables to be added to the trust after the closing date: in a "prefunded" transaction, some of the proceeds
from the closing of the transaction are set aside in a prefunding account to be used to purchase additional receivables during the prefunding period; in a "revolving" deal,
principal collections from the loans can be used to purchase additional receivables during the revolving period.
ASSET-BACKED SECURITIES
Probability density function (PDF) of Log-normal Distribution With Expected Losses of 2.00% and
Standard Deviation of 1.04%
10.0%
9.0%
8.0%
Probability
7.0%
6.0%
5.0%
4.0%
3.0%
2.0%
1.0%
0.0%
Loss Rate
ASSET-BACKED SECURITIES
analyzing the potential losses for the different classes of notes, perhaps supplemented by separate
modeling of one or more special features.
In other regions, such as EMEA, Australia or Latin America there are fewer repeat issuers, and
structures tend to be more varied and complicated. As a result, we use a more comprehensive tool that
can accommodate in a single cash flow model many of the specific structural elements and risks that
can lead to material differences in the rating analysis. 8 The key input parameters to that type of model
are detailed in Appendix 3 and typically include:
the yield earned on the assets for each period taking into account any stresses that may cause this
yield to decrease
an assumption about the timing of loan losses or defaults throughout the life of the transaction,
transaction fees, the interest rates on the bonds, including any interest rate swaps,
how the transaction allocates cash flows and losses among the various parties in the transaction,
including different classes or tranches of notes.
In any case, the model calculates the bond loss for each portfolio loss scenario of the lognormal curve.
The model then weights each bond loss by the frequency implied by the probability distribution. We
then sum the weighted losses to calculate the bonds expected loss.
The report Moodys Approach to Rating Securities Backed by Brazilian Consumer Assets, April 2013, (SF294535) details the rating considerations specific to Brazilian
securitizations backed by consumer assets and issued through Fundo de Investimento em Direitos Creditorios (FIDC).
ASSET-BACKED SECURITIES
rate loans tends to lower the weighted average rate of the remaining loans; we use that calculated lower
interest rate in the cash flow modeling.
We model the effects of the other two factors in two different ways, depending on whether we use a
single comprehensive cash flow model or a simpler, generic model. In a comprehensive model, the
effects of the last two factors are incorporated within the modeling, through the assumed prepayment
rate, the default or loss timing curve and the modeling of the cash flow allocations among the
participants, respectively.
In contrast, when we use a generic model, we typically use a separate model to determine the amount
of excess spread protection that could be lost in stress scenarios for prepayments and excess spread
leakage. Our stressed prepayment assumptions are typically tied to the credit quality of obligors in
the pool. We then subtract the lost protection from the amount that would be available in an
expected scenario, which gives us the net amount of protection that we assume will be available from
excess spread. We add that net amount to the other forms of credit protection (e.g., subordination,
over-collateralization, reserve fund, etc.) to obtain the total amount of credit protection that we
include in our generic model. Appendix 4 provides an illustration of the calculation of the credit for
spread for typical US auto loan transactions.
The final hard credit enhancement (e.g. subordination and reserve fund) under the senior tranche will
in general be different to the Aaa level or portfolio credit enhancement used to determine the
variability of the probability distribution. This is due to i) additional credit protection available from
excess spread and ii) the impact on the credit risk of the tranches of various structural features as
modeled in the cash flow model. However, the Aaa level and the final hard credit enhancement under
the senior tranche can be similar for very straight-forward sequential structures with limited excess
spread.
10
the risk of disruption in the transactions cash flows that could result from the non-performance of
a third party or from a natural disaster (operational risk)
counterparty risk
ASSET-BACKED SECURITIES
Legal Considerations
Our analysis focuses on the legal risks posed by the potential bankruptcy of the transaction originator,
securitization vehicle, servicer, collections account bank and other relevant party. We also pay
attention to the consumer protection laws and regulations applicable to the auto loan contracts, the
obligors and the originators. We review legal opinions to obtain external comfort in relation to the key
legal risks identified in a transaction.
whether the originator has actually sold the receivables, known as true sale
whether a court would consolidate the owner of the assets, the securitization trust, with the
sponsor in the event of the sponsors bankruptcy, known as substantive consolidation, and
whether the securitization trustee can enforce its ownership or security interest in the collateral
once the originator has filed for bankruptcy protection (perfection).
Our legal analysis of all these risks will depend on jurisdiction and applicable securitization laws.
The bankruptcy of the originator can also pose other risks that could reduce the cash-flow available to
repay the notes, such as set-off risk, and when the originator is also the servicer, cash commingling
risk.
Set-off Risk
The main risk posed by a potential bankruptcy of the originator is that loan obligors to whom the
originator owes money might be able to set off those amounts against the loan balance. The typical
situation in which this risk arises is when the originator is a bank and the loan obligors have deposits at
that bank. The amount of the set off represents a reduction in the principal amount of the loan pool
and, effectively, a loan loss.
To analyze this risk, we assess jurisdiction-specific laws and regulations governing the right to set off
deposits in the event of bankruptcy. In jurisdictions that allow set-off, and for transactions without
structural protections to fully mitigate set-off risk, we will conservatively estimate the potential set-off
exposure in modeling the likelihood of a default by the originator and the extent to which the
originator owes money to loan obligors.
11
The likelihood of a servicer default, measured by the servicers credit strength, and affected by any
transaction document provisions requiring the trustee to transfer servicing to a backup if the
servicers rating falls below a specified rating
ASSET-BACKED SECURITIES
The potential amount of the transactions cash that the servicer holds at the time of bankruptcy,
which reflects
the frequency with which the transaction documents require the trustee to sweep cash from
the servicers collection account to the trusts account
The potential for cash to continue to flow to the servicer after bankruptcy and become part of the
servicers bankruptcy estate, resulting from any processes in the documentation that redirect
collections to another account in the event of servicer bankruptcy or pre-bankruptcy event.
We analyze the extent to which the transactions credit enhancement and liquidity protections are
sufficient to prevent shortfalls to investors. When structural mitigants are not sufficient, we model the
rating impact by incorporating our assessment of the commingling exposure amount in the cash flow
model using the credit rating of the servicer 9 as a proxy for the servicers probability of default.
Collections are also at risk if the collections account bank holding the securitization trusts cash
becomes insolvent. Our analysis of this risk focuses on the minimum required rating for the collection
account bank specified in the transaction documentation and the actions that result if the banks rating
falls below that specified rating.
Monitoring
Our approach to monitoring the ratings of outstanding auto loan ABS is similar to the approach we
use to assign the initial ratings. It is a probability-based expected-loss approach, which involves
weighting the bond losses incurred in potential future scenarios by their probabilities of occurring and
summing those weighted losses. The approach relies on our estimates of the expected value and
variance of the pool losses to derive the lognormal probability distribution of the pool losses.
When monitoring the performance of outstanding ABS, we track the performance of the underlying
collateral, developments regarding the originator, servicer and other participants in the transaction, the
amount and form of credit enhancement and factors that affect the integrity of the legal structure. The
starting point is typically the monitoring of the collateral performance relative to our initial expectations.
The key metric that we track is the then-current cumulative net loss rate or cumulative default and
recoveries for the transaction. We combine that loss rate with the issuer's historical loss experience to
update our estimate of the ultimate lifetime net loss rate on the pool of loans. We take into account
any material change in the macroeconomic environment that could impact future performance. We
then use that updated estimate to assess whether the current ratings assigned to the transaction are still
appropriate based on the credit protection available to investors. Our evaluation of the credit
protection takes into account both the current levels of credit enhancement as well as how the
transactions structural features, such as the cash allocation mechanics among the various classes of
investors, is likely to affect the credit enhancement and the extent to which the transaction allows the
release of credit enhancement.
Our monitoring analysis also includes an assessment of the stability of the originator, servicer, swap
counterparties and credit support providers. Should these entities become unable to fulfill their
obligations to the transaction, the risk is greater that cash flows to investors will decline. Thus, changes
in the financial stability of an entity that has a weight in the rating of the securities can result in a
rating action on the securities.
9
12
In absence of a credit rating or in situations where the transaction structure aims at provisioning for the commingling risk independently of the credit quality of the
servicer, we will typically model the commingling risk without weighting it by the credit quality of the servicer.
ASSET-BACKED SECURITIES
The approach is based on the calculation of the growth rate of the average cumulative defaults
observed during previous periods. If we consider the percentage increase in average cumulative defaults
from period to period after origination (using a comparable amount of data points), we get an
estimation of the possible future growth rates of cumulative defaults for each period.
We obtain extrapolated default data for the future by multiplying the last historical data point of a
specific vintage by one plus the growth rate of the average cumulative defaults of the specific period
(and so on with the subsequent growth rates and the resulting extrapolated data).
When the observation period covered by historical data is shorter than the average maturity of loans, we
may extend the observed default curves in order to capture the impact of potential defaults after the
observation period and build a full default timing curve. In order to simulate these unobserved defaults,
one approach is to extrapolate the default rate of the longest observed period to the weighted average
maturity of the pool for each vintage curve, at a rate equal to the last actually observed growth rate.
TABLE 1
10
11
12
13
14
15
16
1.28%
Q1
2001
6,734,496
0.01%
0.07%
0.14%
0.20%
0.24%
0.31%
0.56%
0.71%
0.76%
0.88%
1.10%
1.13%
1.17%
1.22%
1.27%
Q2
2001
17,798,000
0.00%
0.02%
0.10%
0.29%
0.53%
0.70%
0.85%
0.95%
1.01%
1.17%
1.27%
1.46%
1.49%
1.51%
1.58%
1.59%
Q3
2001
13,456,298
0.00%
0.03%
0.04%
0.23%
0.34%
0.42%
0.50%
0.66%
0.79%
0.88%
1.12%
1.18%
1.24%
1.31%
1.37%
1.38%
Q4
2001
12,884,480
0.03%
0.07%
0.07%
0.12%
0.24%
0.44%
0.64%
0.80%
0.91%
1.11%
1.27%
1.33%
1.36%
1.41%
1.47%
1.48%
Q1
2002
19,509,488
0.02%
0.06%
0.11%
0.18%
0.30%
0.44%
0.52%
0.61%
0.89%
1.05%
1.19%
1.25%
1.29%
1.34%
1.39%
1.41%
Q2
2002
21,876,657
0.00%
0.03%
0.14%
0.31%
0.41%
0.51%
0.70%
0.80%
0.90%
0.97%
1.32%
1.41%
1.45%
1.51%
1.57%
1.58%
2.01%
Q3
2002
28,659,946
0.00%
0.04%
0.21%
0.33%
0.50%
0.68%
0.94%
1.04%
1.23%
1.40%
1.68%
1.79%
1.85%
1.92%
2.00%
Q4
2002
22,374,331
0.01%
0.05%
0.17%
0.43%
0.56%
0.75%
0.99%
1.10%
1.12%
1.29%
1.54%
1.65%
1.70%
1.76%
1.84%
1.85%
Q1
2003
28,772,302
0.00%
0.04%
0.16%
0.46%
0.60%
0.71%
0.90%
1.08%
1.23%
1.42%
1.70%
1.81%
1.87%
1.94%
2.02%
2.04%
1.95%
Q2
2003
28,093,680
0.00%
0.10%
0.23%
0.41%
0.61%
0.73%
0.88%
1.03%
1.18%
1.36%
1.63%
1.74%
1.79%
1.85%
1.94%
Q3
2003
30,675,247
0.01%
0.04%
0.18%
0.37%
0.49%
0.62%
0.82%
0.96%
1.09%
1.26%
1.51%
1.61%
1.66%
1.72%
1.79%
1.81%
Q4
2003
32,602,184
0.02%
0.06%
0.21%
0.39%
0.58%
0.76%
1.00%
1.17%
1.34%
1.54%
1.84%
1.97%
2.03%
2.10%
2.20%
2.21%
Q1
2004
4,187,826
0.03%
0.08%
0.15%
0.41%
0.60%
0.78%
1.02%
1.20%
1.37%
1.58%
1.89%
2.02%
2.08%
2.16%
2.25%
2.27%
Q2
2004
57,008,449
0.00%
0.02%
0.20%
0.43%
0.63%
0.82%
1.08%
1.27%
1.45%
1.66%
2.00%
2.13%
2.20%
2.28%
2.38%
2.39%
Q3
2004
62,510,583
0.03%
0.06%
0.18%
0.39%
0.56%
0.73%
0.96%
1.13%
1.29%
1.48%
1.78%
1.90%
1.96%
2.03%
2.12%
2.13%
Q4
2004
69,544,482
0.01%
0.05%
0.14%
0.31%
0.45%
0.59%
0.77%
0.91%
1.04%
1.19%
1.43%
1.52%
1.57%
1.63%
1.70%
1.71%
0.01%
0.05%
0.15%
0.33%
0.48%
0.63%
0.82%
0.97%
1.10%
1.27%
1.52%
1.62%
1.68%
1.74%
1.81%
1.83%
381%
197%
116%
45%
31%
31%
17%
14%
15%
20%
7%
3%
4%
4%
1%
13
For more details, see our report Historical Default Data Analysis for ABS Transactions in EMEA, November 2005
ASSET-BACKED SECURITIES
The starting point in projecting losses based on the static pool cumulative loss data is creating a loss
timing curve for the originator. The loss timing curve provides the percentage of the overall lifetime
losses likely to be incurred by the receivables at various intervals of the pool's life. The loss timing
curve can then be used to extrapolate the cumulative losses on a static pool of receivables from its
current level to the expected level at maturity.
We frequently employ the delta loss curve method to construct the loss curve. This method uses the
incremental (delta) losses experienced by the vintages during each period. The first step is to calculate
average incremental losses across vintages for each period (average delta loss). Next, the cumulative
average delta loss is calculated for each period by adding the incremental delta losses up through that
period (Cumulative Delta Loss). If the static pool performance history does not include pools that
are fully paid down, there are more losses to be incurred in these static pools over their remaining lives.
Therefore, the next task is to determine the "anchor" or terminal value of the cumulative delta loss
curve. There are various methods for forecasting the anchor value: one such method is to analyze the
trend line of six-month deltas to determine the projected six-month deltas over the remaining life.
Those projections are added to the life-to-date losses to determine the anchor or terminal loss.
The loss curve is created by calculating the percentage of the total cumulative delta loss incurred
through each period after origination. The loss timing curve can then be used to project the
cumulative loss for each of the vintages with incomplete history by dividing the life-to-date loss for any
vintage by the corresponding value of the loss timing curve.
TABLE 2
14
ASSET-BACKED SECURITIES
Down payment
car value
amortisation
Balloon payment
time
Source: Moodys Investors Service
Typically, the borrower has the obligation to repay the entire loan, including the balloon payment.
Nonetheless, certain balloon products will also offer the borrower further options such as returning the
vehicle to the dealership before making the balloon payment at maturity and receiving repayment on
the final balloon installment typically from the dealership. Coverage for the final balloon installment
through the dealership under the loan typically refers to the so-called dealer buy-back.
Risk analysis
Among the things we focus on when assessing balloon loan aspects in auto loan ABS is the additional
risk that this product may present to the transaction. First of all, we evaluate the details of the product
type. In particular, we review the borrowers contractual obligation to repay the balloon payment in all
circumstances even where the dealer buy back is not available. In addition, historical data is analyzed
slightly differently for balloon loans compared to normal amortizing loans due to the back ended
amortization and hence large cash flows also due at contract maturity.
Even if the borrower must contractually repay the balloon loan portion, a transaction could be exposed
to further losses compared to normal amortizing products, due to defaulting balloon loans in a stressed
manufacturer default scenario. This represents one of the main risk components of balloon loans as
11
15
The future vehicle value estimate also considers a contractual fixed mileage on the car and the assumption of regular wear and tear.
ASSET-BACKED SECURITIES
following the manufacturer group default, the dealership could suffer financial distress and would not
be able to provide the dealer buy back. As a consequence, the borrowers will find themselves having to
pay a large sum, which they expected to simply receive as part of the dealer buy back option and may
not be able to cover the final balloon installment due at contract maturity. Furthermore, if the
borrower was planning to sell the car on the secondhand car market at contract maturity in order to
pay the final balloon installment, secondhand car values for the respective brands would be negatively
affected due to the stressed scenario.
Hence, our analysis will reflect this increased risk in its quantitative analysis. We evaluate the risks by
considering the following key components, namely (1) percentage of balloon loans in the total
portfolio and balloon loan principal due at contract maturity of these loans, (2) the car
make/manufacturer concentration in portfolio and relevant manufacturer ratings, (3) any dealership
concentration and dealer multi-/mono- brand aspects, (4) for revolving period transactions, eligibility
criteria aspects. Based on this analysis, we stress key model inputs depending on the degree of risk
present.
16
ASSET-BACKED SECURITIES
Appendix 3 Use of Cash Flow Model for Certain Auto Loans ABS Transactions
In certain markets such as EMEA or Australia where structures tend to be more varied and
complicated, we use a comprehensive cash flow model, such as ABS ROMTM, which aims at capturing
the structural aspects of the transaction and risks that can lead to material differences in the rating
analysis. The model incorporates the lognormal loss distribution and assumptions with regards to
assets, liabilities and other transaction-specific factors. A simplified version of ABS ROMTM is
available on moodys.com.
The model basically produces a series of loss scenarios. In each loss scenario, the corresponding loss for
each class of notes is calculated given the incoming cash flows from the assets and the outgoing
payments to third parties and noteholders.
The expected loss (EL) for each tranche is the sum product of: (i) the probability of occurrence of each
loss scenario; and (ii) the loss expected in each default scenario for each tranche.
The EL of each tranche is associated with a particular time horizon in order to compare the EL to our
benchmark for that time horizon as per our Idealised Expected Loss table. The relevant time horizon is
the weighted-average life of the tranche, which is calculated based on the timing of payment of
principal to the tranche under each default scenario.
We also run sensitivities to a variety of key asset inputs and structural features in order to test the
sensitivity of the notes ratings.
In this appendix, we explain how we derive for an auto loan ABS our assets and liabilities assumptions
necessary to run such model.
Asset Modeling
17
Default Definition
Timing of Defaults
Recovery Rate
ASSET-BACKED SECURITIES
Asset Modeling
Recovery Timing
Portfolio Yield
We derive our assumption from the portfolio yield vector provided for the
portfolio based on the portfolios scheduled amortization. To take into
account the yield compression resulting from prepayments and defaults,
we examine the dispersion of interest rates within the pool and assume
that prepayments and defaults result from highest yielding assets. Hence,
we determine a haircut reflecting the potential yield reduction due to
prepayments and defaults and reduce the yield vector accordingly.
We may stress further the yield vector if the transaction is partially
unhedged or for possible renegotiations of loans terms and conditions.
Liabilities Modeling
12
18
Transaction Structure
The cash flow model allows us to reflect various structural aspects of the
transaction, such as the different classes of notes, priority of payments,
reserve fund, principal deficiency ledgers. Options such as principal-topay interest mechanism and interest earned on cash in the transaction
can also be captured by the cash flow model where relevant.
Triggers
The cash flow model allows us to capture various triggers (such as typical
triggers based on the amount of losses, draw of the reserve fund or
unpaid balance in the Principal Deficiency Ledger or PDL) in the
transaction. For example, many common triggers that would change the
waterfall from sequential to pro-rata or vice versa or result in the
accelerated amortization of the senior notes can be specified in the model
if they are based on performance tests (such as default levels or PDL
tests). In addition, the model allows us to capture any reserve fund build
up/amortization triggers and triggers that would result in interest deferral
on the notes.
Transaction Expenses
Swaps
Prepayments are unscheduled principal collections (i.e. partial or total repayment of the outstanding debt before the amounts become due). Dynamic prepayment data
is calculated as the ratio between prepayment amount received during each period and the outstanding portfolio as of the same date. The Constant Prepayment rate
(CPR) is often expressed as an annualized percentage.
ASSET-BACKED SECURITIES
The amount by which the average interest rate on the loans may decline over the life of the
security, which we refer to as WAC (weighted average coupon) deterioration;
2.
The speed with which loans prepay during the life of the security; and
3.
The amount of excess spread that leaks out of the transaction before it is needed to protect
investors.
In this appendix, we describe how we determine how much credit to give to excess spread as credit
enhancement for US auto loan ABS.
WAC Deterioration Stress
Our analysis of the changes over time in the WAC of auto loan pools showed that the WACs of
higher-credit-quality pools tend to increase over time, while the WACs of lower-credit-quality pools
tend to decrease over time. The prime pools that were studied experienced an average WAC increase
of 0.43% from the cutoff date to the point of reaching a 15% pool factor, while subprime pools
experienced a WAC decrease of 0.54%.
The analysis highlighted the expected changes in WAC based on the actual prepayments and defaults
that occurred on the pools. However, we are more concerned about how much excess spread will be
available when it is needed - that is, in high default scenarios - than in the average, or expected, case.
Consequently, we focus on the WAC of the pool in a high-loss, or "stressed," scenario by assuming
that 2-10% of the highest WAC loans in a pool prepay, thus reducing its WAC.
Prepayment Rate Stress
Our analysis of prepayment rates in auto loan pools has shown that the absolute prepayment speed (or
ABS) 13 of US auto loans typically to be around of 1.50% of the initial contracts per month.
We stress the prepayment speeds from this expected prepayment level, which results in a shorter
weighted average life (WAL) of the transaction and, thus, less excess spread over the life of the deal to
cover losses. Our stressed prepayment rate depends upon the type of the borrower (subprime
borrowers prepay at a higher rate than prime borrowers) and the types of incentives offered to the
borrower (borrowers offered interest rates below market rates are expected to prepay at lower rates).
Credit Enhancement (CE) Leakage Stress
Another stress incorporated into the excess spread analysis is to account for the possibility that excess
spread may not be available to support the rated bonds; instead it may have been released or "leaked"
to the unrated bonds or residual interest because of the timings of loan losses. For example, excess
spread and other cash (e.g., pro-rata principal allocations and reserve account releases) may be released
to junior interests in the early months of a transaction before losses have reached a sufficiently high
level to utilize that credit enhancement.
The risk is incorporated in the analysis of US auto loan transactions by applying the simplified cash
flow model, which assumes into its that a certain amount of excess spread leaks out of a transaction in
the first 12 months. From the 13th month forward, the model assumes that any remaining credit
enhancement will be fully utilized to cover losses in the breakeven loss scenarios. The amount of credit
enhancement leakage in the first year is also dependent on the weighted average remaining maturity
(WARM) of the asset pool. Pools with longer WARMs are assumed to have a greater CE leakage in the
13
19
Absolute Prepayment Speed (ABS) usually measures the sum of voluntary and involuntary (e.g. due to defaulted loans) prepayments as the rate of prepayment each
month which is related to the original number of receivables in a pool of receivables.
ASSET-BACKED SECURITIES
first 12 months than pools with shorter WARMs, reflecting the potential for more back-ended losses
in longer WARM pools.
Calculation of the Excess Spread Benefit: An Example
The implementation of the WAC deterioration, prepayment speed and CE leakage stresses within our
bond breakeven cash flow model is shown below for an auto loan ABS deal example. It includes the
following characteristics:
Pool Overview
Balance
# Contracts
Avg Bal
WAC
1,000,000,000
50,000
20,000
8.38%
Bond Coupon
Servicing Fees
4.00%
1.00%
A
B
OC
Size
Target Rating
94%
5%
1%
Aaa (sf)
A3 (sf)
unrated
Init (% orig)
Target (% os)
Floor (% orig)
Reserve Account
Overcollateralization
Total
0.50%
1.00%
2.00%
0.50%
1.50%
0.50%
The asset pool in the example has a WARM of 60 months and a WAC of 8.38%. The weighted
average bond coupon is assumed to be 4.00% and the servicing fee is assumed to be 1.00%.
The breakeven level of credit enhancement is calculated for the subordinate Class B bond, which is
supported by a non-declining 0.50% reserve account and overcollateralization which is built from
1.00% initially to a target of 2.00% of the outstanding pool, subject to a floor of 0.50% of the initial
pool.
We assume that the rating committee decided that the expected lifetime cumulative loss for the pool
backing the securities to be rated was 2% and the Aaa level 8%.
1.
2.
20
Stressed
TotABS
1.50%
2.25%
Avg Life
1.89
1.49
ASSET-BACKED SECURITIES
3.
CNL
2.00%
Recovery Rate
50.00%
Year
Timing of Losses
1
2
3
4
5
32.00%
31.00%
22.00%
15.00%
0.00%
Based on the structural features that trap a certain amount of excess spread, and on the remaining term
of the loan pool, 0.90% of excess spread is assumed to leak out of the transaction in the first 12
months.
Excess Spread Leakage
Month
1
2
3
4
5
6
7
8
9
10
11
12
0.00%
0.00%
0.00%
0.13%
0.32%
0.47%
0.60%
0.71%
0.79%
0.85%
0.88%
0.90%
In the first three months, excess spread is being trapped to build to the target OC level of 1%. In the
following nine months, excess spread is first used to cover losses, which are assumed to be low in the
first year of the deal, then released out of the transaction. After 12 months, losses are assumed to be
high and excess spread is fully utilized to cover losses.
Calculation of Credit for Excess Spread
Excess Spread Calculation
21
3.29%
1.49
4.90%
0.90%
-Expected CE Leakage
4.00%
ASSET-BACKED SECURITIES
We then add the amount of excess spread to the other forms of outside enhancement to determine the
bond ratings suggested by the modeling process.
Total Credit Enhancement (excl. subordination)
22
Reserve Fund
0.50%
4.00%
Overcollateralization
1.00%
Total
5.50%
ASSET-BACKED SECURITIES
country-specific factors, such as our expectation of the level of increased unemployment rates,
consumer leverage levels and economic development
the effects of possible adverse changes to the legal and institutional environment in the country
We will apply such minimum portfolio CE levels as long as we assume that those conditions will
prevail.
We may also apply a minimum expected loss multiple to ensure that extreme loss scenarios have an
adequate probability of occurrence in our analysis. We apply this multiple when we assign or update
the expected loss. We determine it as a multiple of the transactions expected loss to ensure that we
maintain a minimum level of difference between the expected loss and the portfolio CE. The method
for calculating the multiple allows the loss distribution used to simulate losses incurred by the
securitised portfolio to maintain a minimum coefficient of variation. Moreover, this method is
particularly important for transactions with high expected loss assumptions or where there is an
expectation of adverse performance, which the arrears performance of the collateral portfolio is not yet
14
23
In certain circumstances, depending on the drivers of the LCC we may consider alternative loss distribution assumptions or may not adjust our loss distribution
assumptions taking into consideration the LCC.
ASSET-BACKED SECURITIES
reflecting but is already qualitatively incorporated into the expected loss assumption. The multiples
differ based on the level of the expected loss assumed for the portfolio, but typically will range from
between 3x (for high expected loss assumptions) and 5x (for low expected loss assumptions).
[Link to Excel file with Minimum Portfolio CE]
Probabilities
0.00%
5.00%
10.00%
15.00%
20.00%
Cumulative Losses
24
ASSET-BACKED SECURITIES
25
ASSET-BACKED SECURITIES
Analyst Contacts:
2013 Moodys Investors Service, Inc. and/or its licensors and affiliates (collectively, MOODYS). All rights reserved.
MILAN
+39.02.9148.1100
Alex Cataldo
+39.02.91481103
Associate Managing Director
alex.cataldo@moodys.com
TOKYO
+81.3.5408.4000
Atsushi Karikomi
+81.3.5408.4185
Vice President - Senior Analyst
atsushi.karikomi@moodys.com
NEW YORK
+1.212.553.1653
Maria Muller
+1.212.553.4309
Senior Vice President - Manager
maria.muller@moodys.com
HONG KONG
+852.3758.1309
Jerome Cheng
+852.3758.1309
Vice President - Senior Credit Officer/Manager
jerome.cheng@moodys.com
SYDNEY
+61.2.9270.8169
Jennifer Wu
+61.2.9270.8169
Vice President - Senior Credit Officer/Manager
jennifer.wu@moodys.com
MOODY'S CLIENT SERVICES:
New York: +1.212.553.1653
Tokyo: +81.3.5408.4100
London: +44.20.7772.5454
Hong Kong: +852.3551.3077
Sydney: +612.9270.8100
Singapore: +65.6398.8308
ADDITIONAL CONTACTS:
Website: www.moodys.com
Contributors:
Sophie Berthelon
Senior Vice President
Corey Henry
Vice President Senior Analyst
Steven Bild
Analyst
* Andrew Silver also contributed to this
report as a research consultant
26
CREDIT RATINGS ISSUED BY MOODY'S INVESTORS SERVICE, INC. (MIS) AND ITS AFFILIATES ARE MOODYS
CURRENT OPINIONS OF THE RELATIVE FUTURE CREDIT RISK OF ENTITIES, CREDIT COMMITMENTS, OR DEBT OR
DEBT-LIKE SECURITIES, AND CREDIT RATINGS AND RESEARCH PUBLICATIONS PUBLISHED BY MOODYS
(MOODYS PUBLICATIONS) MAY INCLUDE MOODYS CURRENT OPINIONS OF THE RELATIVE FUTURE CREDIT
RISK OF ENTITIES, CREDIT COMMITMENTS, OR DEBT OR DEBT-LIKE SECURITIES. MOODYS DEFINES CREDIT RISK
AS THE RISK THAT AN ENTITY MAY NOT MEET ITS CONTRACTUAL, FINANCIAL OBLIGATIONS AS THEY COME DUE
AND ANY ESTIMATED FINANCIAL LOSS IN THE EVENT OF DEFAULT. CREDIT RATINGS DO NOT ADDRESS ANY
OTHER RISK, INCLUDING BUT NOT LIMITED TO: LIQUIDITY RISK, MARKET VALUE RISK, OR PRICE VOLATILITY.
CREDIT RATINGS AND MOODYS OPINIONS INCLUDED IN MOODYS PUBLICATIONS ARE NOT STATEMENTS OF
CURRENT OR HISTORICAL FACT. CREDIT RATINGS AND MOODYS PUBLICATIONS DO NOT CONSTITUTE OR
PROVIDE INVESTMENT OR FINANCIAL ADVICE, AND CREDIT RATINGS AND MOODYS PUBLICATIONS ARE NOT
AND DO NOT PROVIDE RECOMMENDATIONS TO PURCHASE, SELL, OR HOLD PARTICULAR SECURITIES. NEITHER
CREDIT RATINGS NOR MOODYS PUBLICATIONS COMMENT ON THE SUITABILITY OF AN INVESTMENT FOR ANY
PARTICULAR INVESTOR. MOODYS ISSUES ITS CREDIT RATINGS AND PUBLISHES MOODYS PUBLICATIONS WITH
THE EXPECTATION AND UNDERSTANDING THAT EACH INVESTOR WILL MAKE ITS OWN STUDY AND
EVALUATION OF EACH SECURITY THAT IS UNDER CONSIDERATION FOR PURCHASE, HOLDING, OR SALE.
ALL INFORMATION CONTAINED HEREIN IS PROTECTED BY LAW, INCLUDING BUT NOT LIMITED TO, COPYRIGHT LAW,
AND NONE OF SUCH INFORMATION MAY BE COPIED OR OTHERWISE REPRODUCED, REPACKAGED, FURTHER
TRANSMITTED, TRANSFERRED, DISSEMINATED, REDISTRIBUTED OR RESOLD, OR STORED FOR SUBSEQUENT USE FOR
ANY SUCH PURPOSE, IN WHOLE OR IN PART, IN ANY FORM OR MANNER OR BY ANY MEANS WHATSOEVER, BY ANY
PERSON WITHOUT MOODYS PRIOR WRITTEN CONSENT.
All information contained herein is obtained by MOODYS from sources believed by it to be accurate and reliable. Because of
the possibility of human or mechanical error as well as other factors, however, all information contained herein is provided
AS IS without warranty of any kind. MOODY'S adopts all necessary measures so that the information it uses in assigning a
credit rating is of sufficient quality and from sources MOODY'S considers to be reliable including, when appropriate,
independent third-party sources. However, MOODYS is not an auditor and cannot in every instance independently verify or
validate information received in the rating process. Under no circumstances shall MOODYS have any liability to any person
or entity for (a) any loss or damage in whole or in part caused by, resulting from, or relating to, any error (negligent or
otherwise) or other circumstance or contingency within or outside the control of MOODYS or any of its directors, officers,
employees or agents in connection with the procurement, collection, compilation, analysis, interpretation, communication,
publication or delivery of any such information, or (b) any direct, indirect, special, consequential, compensatory or incidental
damages whatsoever (including without limitation, lost profits), even if MOODYS is advised in advance of the possibility of
such damages, resulting from the use of or inability to use, any such information. The ratings, financial reporting analysis,
projections, and other observations, if any, constituting part of the information contained herein are, and must be construed
solely as, statements of opinion and not statements of fact or recommendations to purchase, sell or hold any securities. Each
user of the information contained herein must make its own study and evaluation of each security it may consider
purchasing, holding or selling.
NO WARRANTY, EXPRESS OR IMPLIED, AS TO THE ACCURACY, TIMELINESS, COMPLETENESS, MERCHANTABILITY OR
FITNESS FOR ANY PARTICULAR PURPOSE OF ANY SUCH RATING OR OTHER OPINION OR INFORMATION IS GIVEN OR
MADE BY MOODYS IN ANY FORM OR MANNER WHATSOEVER.
MIS, a wholly-owned credit rating agency subsidiary of Moodys Corporation (MCO), hereby discloses that most issuers of
debt securities (including corporate and municipal bonds, debentures, notes and commercial paper) and preferred stock
rated by MIS have, prior to assignment of any rating, agreed to pay to MIS for appraisal and rating services rendered by it fees
ranging from $1,500 to approximately $2,500,000. MCO and MIS also maintain policies and procedures to address the
independence of MISs ratings and rating processes. Information regarding certain affiliations that may exist between
directors of MCO and rated entities, and between entities who hold ratings from MIS and have also publicly reported to the
SEC an ownership interest in MCO of more than 5%, is posted annually at www.moodys.com under the heading
Shareholder Relations Corporate Governance Director and Shareholder Affiliation Policy.
For Australia only: Any publication into Australia of this document is pursuant to the Australian Financial Services License of
MOODYS affiliate, Moodys Investors Service Pty Limited ABN 61 003 399 657AFSL 336969 and/or Moodys Analytics
Australia Pty Ltd ABN 94 105 136 972 AFSL 383569 (as applicable). This document is intended to be provided only to
wholesale clients within the meaning of section 761G of the Corporations Act 2001. By continuing to access this document
from within Australia, you represent to MOODYS that you are, or are accessing the document as a representative of, a
wholesale client and that neither you nor the entity you represent will directly or indirectly disseminate this document or
its contents to retail clients within the meaning of section 761G of the Corporations Act 2001. MOODYS credit rating is an
opinion as to the creditworthiness of a debt obligation of the issuer, not on the equity securities of the issuer or any form of
security that is available to retail clients. It would be dangerous for retail clients to make any investment decision based on
MOODYS credit rating. If in doubt you should contact your financial or other professional adviser.