Credit Acceptance Auto Loan Trust 2019-3November 7, 2019
Preliminary Ratings
Class Preliminary rating Type Interest ratePreliminary amount (mil.
$)Legal final maturitydate
A AAA (sf) Senior Fixed 237.5 Nov. 15, 2028
B AA (sf) Subordinate Fixed 67.9 Jan. 16, 2029
C A (sf) Subordinate Fixed 46.3 March 15, 2029
Note: This presale report is based on information as of Nov. 7, 2019. The ratings shown are preliminary. Subsequent information may result inthe assignment of final ratings that differ from the preliminary ratings. Accordingly, the preliminary ratings should not be construed asevidence of final ratings. This report does not constitute a recommendation to buy, hold, or sell securities.
Profile
Expected closing date Nov. 21, 2019.
Collateral Nonrecourse loans to dealers secured by subprime automobile installment salescontracts (dealer advances) and subprime automobile installment sales contracts(purchased loans).
Originator and servicer Credit Acceptance Corp. (BB/Stable/--), a Michigan corporation.
Seller Credit Acceptance Funding LLC 2019-3, a Delaware limited liability company whosesole member is Credit Acceptance Corp.
Issuer Credit Acceptance Auto Loan Trust 2019-3.
Indenture trustee, trust collateralagent, and backup servicer
Wells Fargo Bank N.A. (A+/Stable/A-1).
Owner trustee U.S. Bank Trust N.A.
Initial purchasers Wells Fargo Securities LLC, BMO Capital Markets Corp., Credit Suisse Securities (USA)LLC, Fifth Third Securities Inc., and Citizens Capital Markets Inc.
Credit Enhancement Summary
CAALT 2019-3 CAALT 2019-1
Preliminary ratings and amounts (mil. $)
Class A - 'AAA (sf)' 237.50 247.20
Class B - 'AA (sf)' 67.90 83.20
Presale:
Credit Acceptance Auto Loan Trust 2019-3November 7, 2019
(i)Performing installments sales contracts (i.e., nondefaulted principal balance of contracts securing the NBV) are $550.2 million for series2019-3 and $672.7 million for series 2019-1. (ii)The reserve account amount is 2% of the rated note balance, nondeclining. CAALT--CreditAcceptance Auto Loan Trust. NBV--Net book value. N/A--Not applicable.
Rationale
The preliminary ratings assigned to Credit Acceptance Auto Loan Trust 2019-3's (CAALT 2019-3's)asset-backed notes reflect:
- The availability of approximately 58.9%, 51.1%, and 45.6% credit support for the class A, B, andC notes, respectively, which is based on stressed cash flow scenarios, including excess spread,in respect of the installment sales contracts. These provide coverage of more than 2.50x, 2.25x,and 1.75x our 19.75%-20.25% expected cumulative net loss range, as a percentage of theperforming retail installment sales contracts, for the class A, B, and C notes, respectively. Thecredit support levels are commensurate with the assigned preliminary 'AAA (sf)', 'AA (sf)', and 'A
(sf)' ratings on the class A, B, and C notes, respectively (see the S&P Global Ratings' Auto LoanExpected Loss and Cash Flow Modeling sections for more information).
- Our expectation that under a moderate ('BBB') stress scenario, we do not expect our ratings onthe notes to decline from our preliminary ratings, all else being equal. Our ratings stabilitycriteria describe the outer bound of credit deterioration over three years as three ratingcategories in the case of 'AAA', 'AA', and 'A' rated securities (see "Methodology: Credit StabilityCriteria," published May 3, 2010).
- The credit enhancement in the form of subordination, overcollateralization (O/C), a reserveaccount, and excess spread (see the Credit Enhancement Summary table above).
- The timely interest and principal payments made under the stressed cash flow modelingscenarios, which we believe are consistent with the assigned preliminary ratings.
- Credit Acceptance Corp.'s (CAC's) extensive securitization performance history dating back to2003 and performance history on dealer advances dating back to 1992.
- CAC's long corporate track record and history of consistent profitability.
- The transaction's payment and legal structures, including a full turbo of the notes following a24-month revolving period; amortization triggers tied to servicer and pool performance; anddealer concentration limits, which, if violated, end the revolving period early and cause allcollections to be paid full turbo to the notes sequentially.
Changes From The Series 2019-1 Transaction
The credit enhancement, collateral composition, and structural changes from the last ratedseries, 2019-1, are detailed below. In our view, these changes are not material and we did notadjust our expected cumulative net loss or ratings-specific stressed loss level based on thechanges.
The credit enhancement changes from the series 2019-1 transaction, as a percentage of thedealer and purchased loan advances' initial net book value (NBV), include:
- Total hard credit enhancement for the class A and B notes decreased due to a decrease in theirsubordination, as the class A notes comprise a larger percentage of the NBV at the expense ofthe Class B and C notes. The enhancement remains sufficient for the voted preliminary ratings;
- Total class A and B credit enhancement decreased to 47.58% and 32.13% from 52.47% and35.93%, respectively;
- The class A and B subordination decreased to 25.98% and 10.53% from 30.87% and 14.33%,respectively.
The collateral composition changes from the series 2019-1 transaction pertaining to the dealeradvances include:
- The dollar amount of dealer advances securitized decreased to $307.8 million from $353.1million. This is stable at approximately 70% as a percentage of the NBV of the pool;
- The number of dealers decreased to 754 from 972, partially as a result of the dealer loanportion of the pool having longer seasoning and therefore, there is more time for dealeradvances to reach the 100-loan threshold and close;
- The number of dealer advances decreased to 966 from 1,176;
- The top dealer concentration increased to 1.11% from 1.03% but still remains below the 1.50%maximum dealer concentration of the NBV;
- The top 10 dealer concentrations increased to 9.11% from 8.07%;
- The top state concentration remained in Michigan and increased to 13.59% from 12.68%; and
- The top 10 state concentrations increased to 65.10% from 63.12%.
The collateral composition changes from the series 2019-1 transaction pertaining to thepurchased loans' aggregate purchase price include:
- The purchased loan advances securitized increased slightly to 29.98% of total advances for2019-3 from 29.81% for 2019-1;
- The number of purchased loans decreased to 13,670 from 14,411;
- The top purchased loan advance state concentration decreased to 8.03% (Ohio) from 9.46%(Ohio); and
- The top 10 purchased loan advance state concentrations remained essentially flat at 55.15%versus 55.10%.
The method used to calculate the weighted average spread rate used to test the amortizationperiod trigger event was changed starting with the series 2019-1, and the trigger level increased to25.0% from 16.5%. We view this change as neutral to our analysis because the change in thecalculation methodology is not material since it reflects a change from an absolute differencecalculation to a ratio comparison. For transactions prior to 2019-1, the trigger is calculated as theabsolute difference between the pool's weighted average forecast collection rate and theweighted average advance rate, and the trigger level is set at 16.5%. Beginning with the 2019-1transaction, the trigger is calculated as one minus the ratio of the weighted average originaladvance rate to the pool's weighted average forecast collection rate, and the trigger level is set at25.0%.
Table 1 shows the absolute difference between the pool's weighted average forecast collectionrate and weighted average advance rate by series, as well as the calculation of the weightedaverage spread rate trigger level. The absolute difference has remained constant at 19.8% but islower than series 2018-2's 20.3%. This spread provides the deal with a margin of safety if thecollateral pool experiences a lower collection rate than the forecast.
Table 1
Spread Between WA Forecast Collection Rate And WA Advance Rate At Origination(i)
(i)These figures are percentages of the initial origination balances and not as of their respective cut-off dates. (ii)Reflects calculation of theweighted average spread rate as defined for transactions prior to series 2019-1. (iii)The methodology for calculating the weighted averagespread changed beginning with series 2019-1. (iv)Reflects calculation of the weighted average spread rate as defined for the series 2019-1transaction on. WA--Weighted average. CAALT--Credit Acceptance Auto Loan Trust. N/A – not applicable
Transaction Overview
CAALT 2019-3 is CAC's third securitization in 2019, its 24th stand-alone securitization since 2008,and its 28th S&P Global Ratings credit-rated term securitization. Series 2019-2 was a private,unrated, transaction pursuant to which approximately $625.1 million of loans were contributed toa wholly-owned special purpose entity that pledged the loans to an institutional lender under aloan and security agreement. The first series 2019-3 distribution will be made on Dec. 16, 2019,and subsequent distributions will be paid on the 15th day of each month or the next business day.The class A, B, and C notes total $351.7 million. Each class will be paid a fixed-interest rate andreceive principal sequentially as described in the Payment Structure section below.
In evaluating the credit quality of the securitized assets, we applied our "General MethodologyAnd Assumptions For Rating U.S. Auto Loan Securitizations" criteria, published Jan. 11, 2011.Because the installment sales contracts backing the dealer advance portion of the collateral arepledged by the dealers to CAC (as opposed to being sold), we have applied dealer defaultassumptions to reflect the risk (regardless of how remote) of an automatic stay of these contracts'proceeds in the event of a dealer bankruptcy. (See the S&P Global Ratings' Auto Loan ExpectedLoss section below for further discussion of our auto loan and dealer default assumptions.)
Transaction Structure
The series 2019-3 transaction incorporates the following structural features:
- A 24-month revolving period, during which CAC will add collateral, as needed, to maintain therequisite enhancement levels.
- A revolving period that may terminate before it is scheduled to end if certain trigger eventsoccur, at which point, early amortization would begin.
- A full turbo sequential payment structure across the rated notes at the start of theamortization period, which is scheduled to begin at the close of business on the November2021 distribution date; however, it is subject to early amortization if a trigger event occurs.
- An initial O/C of 20% of the NBV that should build quickly during the amortization periodbecause of the full turbo structure.
- A reserve account that will be funded with an initial deposit of 2.00% of the initial note balance.The reserve account is nondeclining throughout the deal's life.
- Approximately 70% of the initial NBV consists of dealer advances that were originated underCAC's portfolio program. The remaining 30% consists of the purchase price of the purchasedloans (see the Business Model section below for more information on these programs).
- The installment sales contracts securing the dealer advances are pledged by the dealers, asopposed to being sold, to CAC. Therefore, our analysis also reflects our view of the risk, howeverremote, that cash flows from certain installment sales contracts could be delayed by anautomatic stay if a dealer whose advances are included in the securitized pool becomesbankrupt. We assumed a certain number of bankrupt dealers at each rating category (see theS&P Global Ratings' Auto Loan Expected Loss section below).
As noted above, the full turbo structure causes O/C as a percentage of the NBV to grow quicklyonce a securitized pool begins to amortize. Table 2 shows the O/C levels of those deals for whichdata are available at various times during the first six months of amortization, as a percentage ofthe initial NBV. With the exception of series 2016-3, 2017-1, and 2017-2, all of the deals listed intable 2 have fully paid down. The more recent transactions, series 2017-3 through 2019-1, are stillin their revolving periods.
Table 2
O/C Levels
CAALT
2017-2 2017-1 2016-3 2016-2 2015-2 2015-1
Status Amortizing Amortizing Amortizing Paid-off Paid-off Paid-off
O/C level during revolving period (%) 20 20 20 20 20 20
O/C level during amortization period (%)
Month one 23 24 23 24 24 24
Month three 28 32 29 31 31 33
Month five N/A 39 35 39 37 41
Month six N/A 43 39 43 40 45
O/C--Overcollateralization. CAALT--Credit Acceptance Auto Loan Trust. N/A--Not applicable.
As a corollary, the notes' swift paydown is also presented in the note factor analysis in chart 2.
Some of the performance triggers that will cause the revolving period to end and the amortizationperiod to begin include:
- The adjusted collateral amount, namely the dealer advances and purchased loan collateral,and any amounts on deposit in the principal collections account cannot be less than theamount required to ensure at least 20% O/C for two or more business days.
- The component of the adjusted collateral amount that consists of a deposit in the principalcollections account cannot exceed 5% of the adjusted collateral amount for two or morebusiness days.
- Cumulative collections for any three consecutive periods must equal at least 90% of thecumulative forecast collections.
- The weighted average spread rate must be at least 25.0%. It is calculated as the ratio of a) theweighted average percentage that is not originally advanced (i.e., one minus the weightedaverage original advance rate) to b) the pool's weighted average forecast collection rate. Asstated above, the series 2019-3 pool's percentage-based spread is 30.4%.
In rating this transaction, we will review the legal matters that we believe are relevant to ouranalysis, as outlined in our criteria.
Incorporated in 1972, CAC targets auto dealerships throughout the U.S. that write installmentsales contracts for the subprime and deep subprime auto market. The company has two lendingprograms: the portfolio and the loan purchase programs.
Portfolio program/dealer advances
The portfolio program is the older of the two programs and the company's central focus. Underthis program, CAC makes loans directly to dealers who are members of its dealer network. Thesedealer advances, which are securitized, are made in connection with the dealers' origination ofinstallment sales contracts that obligors enter into to purchase vehicles on the dealers' lots. Theinstallment sales contracts secure the dealer advances and are pledged by the dealers to CAC.The company services those contracts and transfers its secured rights in, and the cash flows from,the contracts to the securitization trust via the seller. The installment sales contract cash flowsare used to repay the dealer advances. The portfolio program currently accounts for approximately70% (as a percentage of NBV) of the series 2019-3 initial pool.
Generally, in the portfolio program, each payment received under the installment sales contractsis applied in the following priority:
- Pay CAC a finance charge that is 20% of the monthly collections for servicing the contracts.
- Repay the advance CAC made to the dealer.
- When the advance is paid in full, pay any remaining amounts to the dealer.
Loan purchase program
In March 2005, CAC introduced the newest version of its loan purchase program, whereby itpurchases contracts directly from dealers. This program provides incremental volume andaccommodates dealers that do not want to participate in the portfolio program, subject toeligibility requirements. Each monthly payment under the installment sales contract belongs toCAC to cover the contract's purchase price that it paid to the dealer and to pay its finance charge.
This program has grown as a percentage of the company's business over the last several quarters.Increasing penetration among large predominantly franchised dealerships, which tend to preferthe purchase loan program, has contributed to an increase in the average amount paid relative tolarger, newer, more expensive vehicles purchased over longer terms of up to 72 months.
Treatment of finance charges
CAC's finance charge is one of the assets transferred to the securitization trust and belongs to thetrust. It is used to cover the servicing and trustee fees, with the balance being available to thenoteholders. The remaining monthly collections are available for interest, principal, and turboedprincipal payments to the noteholders after the revolving period.
Collection forecasts
CAC's forecast of its expected principal and interest collections on the installment sales contracts
originated in relation to advances made under its portfolio program or purchased under itspurchased loan program, as well as its ability to actualize that forecast, is central to the business.Using a proprietary statistical model, the company makes a collection forecast for each loan in itsportfolio and updates this forecast monthly as it gradually actualizes collections.
Table 3 shows the company's original collection forecast on its installment sales contract portfolioby annual vintage, beginning in 2010. The initial forecasts are also compared with updatedforecasts as of June 30 and Sept. 30, 2019. In addition, the table shows the variance between theinitial and most recent forecasts, together with the percentage of the latter that has been realizedto date. The forecasts' accuracy increases as the loans season.
Furthermore, there is a positive variance between 2010 and 2013, and between 2017 and 2019through September, with 2014 currently flat to forecast. The company benefitted from lessintense competition in the wake of the Great Recession and was able to reduce its advance ratesin 2008 and 2009. The 2010 vintage also showed strength from a collections point of view withpositive variance of 4.1% between initial and latest forecasts. By contrast, the 2015 and 2016vintages currently show negative variance related primarily to heightened competition. Adjustinginitial collection forecasts downward in response to the observed decline in forecast collectionrates experienced in 2016, may well have adversely affected unit volumes, given the resultinglower advance rates. The 2015 vintage shows the widest negative variance, with the currentforecast adjusted down to 65.5% (in-line with the June 30 updated forecast) from the initialforecast of 67.7%.
Lastly, table 3 shows that the updated forecasts of the 2010-2015 vintages are now close to fullactualization, rendering these forecasts as essentially reality. Taken as a whole, the tableindicates the accuracy with which the company can forecast the collections that it expects tomake for the installment sales contracts it originates. This is an important consideration becausea forecast accuracy of less than 90% for three consecutive months will cause the securitization tobegin amortizing.
Table 3
Consumer Loan Forecast Collection Rates
Initial forecast(%)
As of Sept 30,2019 (%)
As of June 30,2019 (%)
Variance between initial andlatest forecast (%)
% of forecastrealized
2010 73.6 77.7 77.7 4.1 99.8
2011 72.5 74.8 74.8 2.3 99.5
2012 71.4 73.9 73.9 2.5 99.2
2013 72.0 73.5 73.5 1.5 98.7
2014 71.8 71.8 71.8 0.0 97.6
2015 67.7 65.5 65.5 (2.2) 93.2
2016 65.4 64.1 64.2 (1.3) 81.9
2017 64.0 65.0 65.1 1.0 64.1
2018 63.6 65.4 65.5 1.8 38.5
2019 64.1 64.8 64.7 0.7 11.2
Although underlying contract defaults can be high, CAC's losses are almost negligible because,from 2010 through September 2019, it advanced approximately 42%-47% of the contracts' fullprincipal and interest gross amount under the portfolio program and approximately 45%-52%under the loan purchase program. It advanced 43.1% and 45.7% for the portfolio and loan
purchase programs, respectively, for the nine months ended Sept. 30, 2019. The company setseach advance rate based on an acceptable level of profitability relative to its forecast collectionrate on the installment sales contracts associated with that advance.
Both table 4 and chart 3 show the updated forecast/actual collection rates on all of theinstallment sales contracts by annual vintage, as well as those vintages' associated advance ratesand the spread between the two. The spread provides a cushion if collection rates are lower thanthose that CAC originally forecast. The spread is also an important risk management metric forboth the company's business as a whole and its securitizations. From 2010 through September2019, forecast/actual collection rates averaged approximately 70%, while advance rates averagedapproximately 45%, resulting in an average spread of approximately 25%. Chart 3 illustrates thepercentage of forecast collections actually realized.
The NBV of an identified pool of advances is securitized in these transactions, along with thecompany's rights in the installment sales contracts associated with those advances. Chart 4shows the losses on the advances by annual vintage over the past 11 years. The highest lossoccurred in 2015 and is currently forecast to be 1.35%.
To help mitigate its risk under the portfolio program, CAC cross-collateralizes the loan repaymentsmade to dealers by aggregating them into pools of either 50 or 100 loans. The choice of closing outa pool at 50 loans has been available to dealers in CAC's network since Aug. 5 of this year. Eachpool represents the sum of advances made to a dealer and must be repaid from the aggregatecollections on the installment sales contracts backing each pool. As stated above, CAC receives20% of all collections on the installment sales contracts as its finance charge. The 80% balance ofmonthly collections is first used to pay the advances in each pool. Once all of the advances in adealer's pool have been repaid, 80% of any subsequent collections are remitted to the dealer asback-end profit. This helps align the interests of CAC with those of each dealer. In addition, notonly are a dealer's pool of loans cross-collateralized, all of its pools are themselvescross-collateralized if a dealer has more than one pool.
Additionally, a dealer cannot account for more than 1.50% of the securitized dealer advances, normay the aggregate of the 20 largest dealer concentrations account for more than 20.00% of thesecuritized dealer advances to mitigate dealer bankruptcy risk. The highest dealer concentrationin the series 2019-3 pool at cut-off is approximately 1.11%, and the top 20 dealers account forapproximately 15.86% of the securitized dealer advances.
The company verifies numerous aspects of a loan application including the income andemployment status of each of its obligors.
Similar to other finance companies, CAC has received notices, subpoenas, and civil investigative
demands over the last few years from various regulatory agencies, such as the U.S Department ofJustice, Federal Trade Commission, Consumer Financial Protection Bureau (CFPB), and variousstate attorneys general requesting information and documents related to its origination,underwriting, servicing, and securitization of auto loans, as well as information on the company'spolicies and procedures around allowing car dealers to use GPS starter-interrupt devices onvehicles.
Earlier this year, the attorney general of Mississippi filed a complaint in Chancery Court inMississippi alleging that the company engaged in unfair and deceptive practices relating to itsorigination and collections of auto loans. More or less contemporaneously, CAC received a civilinvestigative demand from the CFPB requesting information relating to Credit Acceptance'sorigination and collection of consumer loans, third-party providers, and credit reporting. CAC alsoreceived a subpoena from the Office of the New York State Attorney General relating to CreditAcceptance's origination and collection policies and procedures in New York.
The company has indicated that it cannot predict the scope, duration, or outcome of these mattersand cannot estimate a possible loss or range of reasonably possible loss.
Payment Structure
The distributions will be made from available funds according to the priority shown in table 5.
Table 5
Payment Waterfall
Priority Payment
1 The servicing fee (6% of monthly collections), the backup servicing fee (10% of monthly collections if thebackup servicer has become the servicer or $4,000 if the backup servicer has not become the servicer), andtrustee fees and expenses ($17,333). Our cash flow analysis incorporated a 10% servicing fee.
2 Class A, B, and C note interest, paid sequentially.
3 Restore the reserve account to its required level.
4 During the revolving period, purchase additional collateral in an amount sufficient to maintain the requiredovercollateralization.
5 During the amortizing period, use all available funds sequentially to reduce the class A note principal balance tozero, then to reduce the class B note principal balance to zero, and then to reduce the class C note principalbalance to zero.
6 Any unpaid amounts owed to the backup servicer, owner trustee, indenture trustee, and trust collateral agent.
7 Any remainder to the certificateholder.
If an event of default occurs under the indenture, the notes will become immediately due andpayable, whether automatically, as in the case of the seller's or issuer's voluntary or involuntarybankruptcy or by a majority vote by principal amount of the senior-most outstanding noteholders.
If an indenture event of default occurs and the notes have been accelerated, the indenture trusteemay exercise certain remedies, such as the liquidation or sale of the trust estate, albeit only incertain circumstances. First, if the event of default is due to a failure to pay principal on the notesat final maturity or the seller's or issuer's voluntary or involuntary bankruptcy, the trust estatemay be liquidated without further warning. Second, for any other indenture event of default, thetrust estate may be sold or liquidated if all of the noteholders consent, the sale or liquidationproceeds are sufficient to ensure that all of the noteholders are paid in full, or if the indenturetrustee determines that the trust estate will not continue to provide sufficient funds to pay
principal and interest on the notes as they would have become due had they not been accelerated.
If the notes are accelerated in the case of a failure to pay or bankruptcy event, distributions will bemade from available funds according to the priority shown in table 6. In the case of any otherindenture event of default, distributions will be made from available funds as described in table 5.However, the fees and expenses in item 1 will not be limited, and no amounts will be distributed tothe reserve account per item 3.
Table 6
Event Of Default Payment Waterfall
Priority Payment
1 The servicing fee; any indemnification amounts owed to the backup servicer ($17,000); trustee fees,indemnification amounts, and expenses ($17,333); and unpaid transition expenses due to anysuccessor servicer, paid pari passu.
2 Class A note interest.
3 Class A note principal until the class A note balance has been reduced to zero.
4 Class B note interest.
5 Class B note principal until the class B note balance has been reduced to zero.
6 Class C note interest.
7 Class C note principal until the class C note balance has been reduced to zero.
8 Any remainder to the certificateholder.
Pool Analysis
As of the Sept. 30, 2019, cut-off date, the series 2019-3 collateral pool contained approximately$439.62 million of advances secured by approximately $802.11 million principal installment salescontracts. The installment sales contracts' characteristics are shown in table 7. Typical of thecompany's securitizations, the pool at cut-off contains defaulted contracts that we did not giveany credit in our cash flow runs to (approximately $250.85 million of principal outstanding). Thecompany's collection forecast considers each loan's status, whether performing or defaulted.
We currently maintain ratings on eight outstanding CAALT transactions. Four of the eighttransactions are still in their revolving period, while four are currently in their amortization periodas shown in table 8 below. We recently raised our ratings on the class B and C notes of each ofseries 2016-3 and 2017-1 to 'AAA (sf)' from 'AA (sf)' and 'A (sf)', respectively. We also affirmed the'AAA (sf)' ratings on the series 2017-1 class A notes. These rating actions reflect our viewsregarding future collateral performance, as well as each transaction's structure and creditenhancement, among other factors. We will continue to monitor the performance of eachtransaction and take rating actions as we deem appropriate.
S&P Global Ratings' Auto Loan Expected Loss: 19.75%-20.25%
To derive the base-case loss for the CAALT 2019-3 transaction, we considered:
- Net loss static pool data relating to the company's paid-off and outstanding amortizingsecuritizations;
- Net loss static pool data for the proposed securitization;
- The extensive origination net loss static pool data provided by the company;
- The collateral characteristics of the pool of installment sales contracts securing the securitizedpool of advances;
- The pool's seasoning;
- Portfolio recovery data; and
- Our forward-looking view of the auto sector and the economy.
Chart 5 shows the origination cumulative net loss static pools by annual vintage from 2002through 2019. With no seasoning credit, they suggest a weighted average net loss expectation ofapproximately 24.5%.
Chart 5
The losses on the series 2012-2 through 2016-2 paid-off deals are summarized in table 9.Because these deals are paid off, it is possible to isolate the dollar amount of losses on theunderlying installment contracts attributable to the amortization period. This dollar amount canbe measured against the installment contracts' principal balance as of the beginning of theamortization period plus the cumulative collateral added throughout the amortization period to
The losses on the amortizing deals are summarized in table 10. The series 2016-3, 2017-1, and2017-2 deals are the only deals currently in their amortization period. After 12, eight, and fourmonths of amortizing, respectively, these series are currently at a 24.39%, 46.01%, and 73.64%note factor, respectively. The current cumulative net losses sustained during the amortizationperiod to date are 8.80%, 5.91%, and 3.95% respectively, measured as a percentage of theinstallment contracts' principal balance as of the beginning of the amortization period plus thecumulative collateral added to any existing open pools throughout the amortization period to date.
We projected losses on series 2016-3, 2017-1, and 2017-2 using loss curves derived from theloss-timing curves of the paid-off series 2015-1, 2015-2, and 2016-2. On average, we project thepools to incur total cumulative net losses during their respective amortization periods of 9.71%,8.52%, and 10.51%, respectively. These percentages are measured against the installmentcontracts' principal balance as of the beginning of the amortization period plus the cumulativecollateral added through the amortization period.
Series 2017-3 through 2019-1 are still in their respective revolving periods, have not breached anyearly amortization triggers, and continue to maintain their required O/C and reserve accountamounts.
Chart 6 shows the cumulative recoveries on the company's portfolio as a whole. This indicates
cumulative recoveries of approximately 35%, though the average is closer to 25% over the firstthree years.
Chart 6
To derive the 20.00% expected loss, we gave some credit to seasoning but considered that thecompany will be funding additional collateral during the transaction's two-year revolving periodand that this additional collateral could be of a somewhat weaker credit profile. We believe thisrisk is limited due to the triggers regarding forecast collection accuracy and the minimum 25%spread requirement for the ratio of the collection and advance rates, both referenced in theBusiness Model section above. If either trigger is breached, the securitization will automaticallyamortize. Using a 2.50x multiple at 'AAA', 2.25x at 'AA', and 1.75x at 'A' equates to stress losses of50%, 45%, and 35%, respectively.
As previously discussed, we assume the bankruptcy of a certain number of dealers at each ratingcategory under both the largest dealer default amount method and the hybrid default amountmethod. We further assume that each bankrupt dealer represents the 1.5% per dealer maximumconcentration amount permitted in the transaction documents subject to a 95% severity. Ourrating-specific cumulative net loss assumption is equal to the greater of the two results, which, inthe case of the series 2019-3 transaction, is engendered by the hybrid default amount method.
Largest dealer default amount method
Under the largest dealer default amount method, at the 'AAA' level, the transaction must cover a95.0% loss severity on the collateral associated with the top 24 dealer concentrations,aggregating to 34.2% (i.e., 1.5% per dealer maximum concentration for each of these dealers andallowing for 5.0% recoveries). At the 'AA' level, the transaction must cover a 95.0% loss severity onthe collateral associated with the top 20 dealer concentrations, aggregating to 28.5% even thoughthe top 20 dealers cannot form more than 20.0% of the advances. At the 'A' level, the transaction
must cover a 95.0% loss severity on the collateral associated with the top eight dealerconcentrations, aggregating to 11.4%. These concentrations are, in turn, applied to the performingcontracts securing the advances made under CAC's portfolio program. These compriseapproximately 70.6% of total performing contracts. This means that the 'AAA', 'AA', and 'A'required credit enhancement under the largest dealer default amount method is 38.9%, 33.4%,and 18.3%, respectively.
Table 11
Largest Dealer Default Amount Method
Stress level 'AAA'
Dealer advance portion (70.6%)
Number of dealers 24
Maximum concentration per dealer (%) 1.50
Dealer concentration penalty (%) 36.00
95% severity on the dealer concentration penalty (%) 34.20
Purchased loan portion (29.4%)
Base-case loss (%) 20.00
Multiple (x) 2.50
Total purchased loan portion loss (%) 50.00
Weighted average loss
Dealer advance portion (70.2%) 24.14
Purchased loan portion (29.8%) 14.71
Combined loss level (%) 38.85
Hybrid method
Under the hybrid default amount method, at the 'AAA' level, the transaction must cover a 95.0%loss severity on the collateral associated with the top five dealer concentrations, aggregating to7.5% plus a 50.0% loss on the remaining collateral balance, according to our auto loan criteria(see table 12). At the 'AA' level, the transaction must cover a 95.0% loss severity on the collateralassociated with the top four dealer concentrations, aggregating to 6.0% plus a 45.0% loss on theremaining collateral balance. At the 'A' level, the transaction must cover a 95.0% loss severity onthe collateral associated with the top three dealer concentrations, aggregating to 4.5% plus a35.0% loss on the remaining collateral balance. This means that the 'AAA', 'AA', and 'A' requiredcredit enhancement under the hybrid method and for this transaction is approximately 52.4%,47.1%, and 36.9%, respectively.
Cash Flow Modeling
We modeled the performing portion of the underlying installment sales contracts to simulate therated stress scenarios appropriate for the assigned preliminary ratings (see table 12).
(i)For 'AAA', 18.5% equals 20.0% of base-case cumulative net losses multiplied by 1 minus the 7.5% dealer concentration penalty. For 'AA',18.8% equals 20.0% of base-case cumulative net losses multiplied by 1 minus the 6.0% dealer concentration penalty. For 'A', 19.1% equals20% of base-case cumulative net losses multiplied by 1 minus the 4.5% dealer penalty. (ii)The servicing fee is 6% of monthly collections butwas run at the 10% rate that applies if the backup servicer is appointed to take over servicing. (iii)The maximum cumulative net losses on thepool that the transaction can withstand without triggering a payment default on the relevant note classes. ABS--Absolute prepayment speed.
The break-even results show that the class A, B, and C notes are credit-enhanced to the degreenecessary to withstand stressed net loss levels that are consistent with the assigned preliminaryratings.
In addition to running break-even cash flows, we ran a sensitivity analysis to see howhigher-than-expected losses could affect our preliminary ratings on the notes (see table 13 andchart 7).
Table 13
Scenario Analysis Summary
Stress level 'BBB'
Hybrid method
Dealer advance portion (70.6%)
Number of dealers 2
Maximum concentration per dealer (%) 1.50
Dealer concentration penalty (%) 3.00
95% severity on the dealer concentration penalty (%) 2.85
Base-case loss (%) 20.00
Base-case loss adjusted for dealer concentration penalty (%)(i) 19.40
Multiple (x) 1.40
Adjusted stress loss level (%) 27.16
Total dealer advance portion loss (%) 30.01
Purchased loan portion (29.4%)
Base-case loss (%) 20.00
Multiple (x) 1.40
Total purchased loan portion loss (%) 28.00
Weighted average loss (%)
Dealer advance portion (70.6%) 21.18
Purchased loan portion (29.4%) 8.24
Combined loss level (%) 29.42
Cumulative net loss timing (mos.) 12/24/36
Cumulative net loss (%) 61/95/100
ABS voluntary prepayments (%) 0.60
Recoveries (%) 30.00
Recovery lag (mos.) 4
Servicing fee (%) 10% of monthly collections
(i)19.4% equals the 20% base-case cumulative net losses multiplied by 1 minus the 3.0% dealer concentration penalty. ABS--Absoluteprepayment speed.
Scenario analysis: 29.42% cumulative net loss results
In our 'BBB' scenario, the transaction's credit enhancement is sufficient to cover a 100% lossseverity on the collateral associated with the top two dealer concentrations, aggregating to 2.85%(i.e., 1.50% per dealer maximum concentration for each of these two dealers at a 95% severity) aspart of a total loss of 29.42% on the collateral, according to the application of our auto loancriteria.
Under this stressed loss scenario, the transaction reaches its highest O/C level as a percentage ofinitial performing pool balance in month four of the amortization period, which is approximately40.58% of the initial performing pool balance, and reduces to approximately 22.44% of the initialperforming pool balance by the time the notes are fully paid off in month 28. Consistent with thefull turbo structure of the amortization period, excess cash flow is released once all of the noteshave paid off starting in month 28. The reserve account is never drawn and is released in month29.
Interest is paid on a timely basis for all classes, and the class A, B, and C notes are paid fullprincipal in months 15, 22, and 28 of the amortization period, respectively.
Our sensitivity analysis allows us to simulate a moderate, or 'BBB', loss scenario to determine thedegree to which the preliminary ratings are susceptible to a negative rating action. Our resultsshow that, all else being equal, we do not expect our ratings on the class A, B, and C notes todecline from our preliminary 'AAA (sf)', 'AA (sf)', and 'A (sf)' ratings, respectively.
Our ratings stability criteria describe the outer bounds of credit deterioration over three years asthree rating categories in the case of 'AAA', and 'AA', and 'A' rated securities. (For more
information, see "Methodology: Credit Stability Criteria," published May 3, 2010.)
Related Criteria
- Criteria | Structured Finance | Legal: U.S. Structured Finance Asset Isolation AndSpecial-Purpose Entity Criteria, May 15, 2019
- Criteria | Structured Finance | General: Incorporating Sovereign Risk In Rating StructuredFinance Securities: Methodology And Assumptions, Jan. 30, 2019
- Criteria | Structured Finance | General: Methodology: Criteria For Global Structured FinanceTransactions Subject To A Change In Payment Priorities Or Sale Of Collateral Upon ANonmonetary EOD, March 2, 2015
- Criteria - Structured Finance - General: Criteria Methodology Applied To Fees, Expenses, AndIndemnifications, July 12, 2012
- General Criteria: Global Investment Criteria For Temporary Investments In TransactionAccounts, May 31, 2012
- Criteria | Structured Finance | ABS: General Methodology And Assumptions For Rating U.S. AutoLoan Securitizations, Jan. 11, 2011
- Criteria | Structured Finance | General: Methodology For Servicer Risk Assessment, May 28,2009
Related Research
- Four Ratings Raised, One Affirmed On Two Credit Acceptance Auto Loan Trust Transactions,Oct. 3, 2019
- Global Structured Finance Scenario And Sensitivity Analysis 2016: The Effects Of The Top FiveMacroeconomic Factors, Dec. 16, 2016
In addition to the criteria specific to this type of security (listed above), the following criteriaarticles, which are generally applicable to all ratings, may have affected this rating action:"Counterparty Risk Framework: Methodology And Assumptions," March 8, 2019; "Post-DefaultRatings Methodology: When Does Standard & Poor's Raise A Rating From 'D' Or 'SD'?," March 23,2015; "Global Framework For Assessing Operational Risk In Structured Finance Transactions,"Oct. 9, 2014; "Methodology: Timeliness of Payments: Grace Periods, Guarantees, And Use of 'D'And 'SD' Ratings," Oct. 24, 2013; "Criteria For Assigning 'CCC+', 'CCC', 'CCC-', And 'CC' Ratings,"Oct. 1, 2012; "Methodology: Credit Stability Criteria," May 3, 2010; and "Use of CreditWatch AndOutlooks," Sept. 14, 2009.
S&P may receive compensation for its ratings and certain credit-related analyses, normally from issuers or underwriters of securities or from obligors.S&P reserves the right to disseminate its opinions and analyses. S&P's public ratings and analyses are made available on its Web sites,www.standardandpoors.com (free of charge), and www.ratingsdirect.com and www.globalcreditportal.com (subscription), and may be distributedthrough other means, including via S&P publications and third-party redistributors. Additional information about our ratings fees is available atwww.standardandpoors.com/usratingsfees.
S&P keeps certain activities of its business units separate from each other in order to preserve the independence and objectivity of their respectiveactivities. As a result, certain business units of S&P may have information that is not available to other S&P business units. S&P has establishedpolicies and procedures to maintain the confidentiality of certain non-public information received in connection with each analytical process.
Credit-related and other analyses, including ratings, and statements in the Content are statements of opinion as of the date they are expressed andnot statements of fact. S&P's opinions, analyses and rating acknowledgment decisions (described below) are not recommendations to purchase,hold, or sell any securities or to make any investment decisions, and do not address the suitability of any security. S&P assumes no obligation toupdate the Content following publication in any form or format. The Content should not be relied on and is not a substitute for the skill, judgment andexperience of the user, its management, employees, advisors and/or clients when making investment and other business decisions. S&P does not actas a fiduciary or an investment advisor except where registered as such. While S&P has obtained information from sources it believes to be reliable,S&P does not perform an audit and undertakes no duty of due diligence or independent verification of any information it receives. Rating-relatedpublications may be published for a variety of reasons that are not necessarily dependent on action by rating committees, including, but not limitedto, the publication of a periodic update on a credit rating and related analyses.
To the extent that regulatory authorities allow a rating agency to acknowledge in one jurisdiction a rating issued in another jurisdiction for certainregulatory purposes, S&P reserves the right to assign, withdraw or suspend such acknowledgment at any time and in its sole discretion. S&P Partiesdisclaim any duty whatsoever arising out of the assignment, withdrawal or suspension of an acknowledgment as well as any liability for any damagealleged to have been suffered on account thereof.
No content (including ratings, credit-related analyses and data, valuations, model, software or other application or output therefrom) or any partthereof (Content) may be modified, reverse engineered, reproduced or distributed in any form by any means, or stored in a database or retrievalsystem, without the prior written permission of Standard & Poor's Financial Services LLC or its affiliates (collectively, S&P). The Content shall not beused for any unlawful or unauthorized purposes. S&P and any third-party providers, as well as their directors, officers, shareholders, employees oragents (collectively S&P Parties) do not guarantee the accuracy, completeness, timeliness or availability of the Content. S&P Parties are notresponsible for any errors or omissions (negligent or otherwise), regardless of the cause, for the results obtained from the use of the Content, or forthe security or maintenance of any data input by the user. The Content is provided on an “as is” basis. S&P PARTIES DISCLAIM ANY AND ALL EXPRESSOR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE ORUSE, FREEDOM FROM BUGS, SOFTWARE ERRORS OR DEFECTS, THAT THE CONTENT'S FUNCTIONING WILL BE UNINTERRUPTED OR THAT THECONTENT WILL OPERATE WITH ANY SOFTWARE OR HARDWARE CONFIGURATION. In no event shall S&P Parties be liable to any party for any direct,indirect, incidental, exemplary, compensatory, punitive, special or consequential damages, costs, expenses, legal fees, or losses (including, withoutlimitation, lost income or lost profits and opportunity costs or losses caused by negligence) in connection with any use of the Content even if advisedof the possibility of such damages.
Standard & Poor’s | Research | November 7, 2019 26
