Compound’s model of Exeter Select Automobile Receivables Trust 2025-1 projects a lifetime cumulative net loss of 12.2% of the cutoff pool in its base case of a 12% CDR, a 1.20% ABS prepayment speed and 60% loss severity. On that projection the junior Class E can absorb 1.39x the base-case loss before it takes its first dollar of loss, which trips the request’s 1.5x flag. Class D, one step up, covers 1.51x. No trigger is breached, the reserve is never drawn, and every class repays at least 15 months before its legal final date.
The trust is performing in line with the issuer’s history: 5.35% cumulative net loss at month 18 against a 5.48% average for the issuer’s vintage pools at the same age. That history is the main caveat. 2025-1 was the first deal on the Exeter Select shelf, so its prospectus has no prior trusts to compare against. The loss curves come from 20 quarterly Exeter origination vintages, Q1 2020 to Q4 2024, that the prospectus presents as comparable, which is a different benchmark from the prior-trust curves the request asked for. The coverage figures are the model’s results under its assumptions, not a rating view on the notes.
You are on an asset-backed finance team at a credit fund looking at the mezzanine classes of a consumer auto deal. You need the pool as it stands today, how its losses compare with the issuer’s history, a defensible base case, and how much loss each class can take. We asked Compound to build that model on a real securitization from public filings. It ran on November 19, 2026 and returned one Excel workbook of nine sheets and a short written answer.
What the workbook shows:
- The trust. $336.8 million of auto loans at cutoff, financed with $327.52 million of notes in seven classes, a 2.75% initial overcollateralization building to a 6.80% target, and a reserve of 1.00% of the cutoff pool.
- The pool today. $188.6 million across 9,128 loans, a 56.0% pool factor, a 17.84% weighted average APR, a 671 weighted average credit score and 90% used vehicles.
- Performance to date. 5.35% cumulative net loss, 4.83% of the pool more than 60 days delinquent, overcollateralization at its 6.80% target and the reserve fully funded.
- The base case. 12.2% lifetime cumulative net loss, between two read-throughs from the vintage curves of 10.2% and 13.8%.
- The classes. Loss coverage of 2.83x for Class B, 2.10x for C, 1.51x for D and 1.39x for E. Class A-3 takes no loss even at a 100% CDR. Class E’s coverage runs from 1.16x to 1.72x across the sensitivity grid.
The prospectus, its static pool annex, eighteen Form 10-D servicer certificates (our count) and the Form ABS-EE asset data file all come from SEC EDGAR, and Compound used its stored EDGAR filing index to screen the trusts. The Sources sheet notes that no web search was needed. Every other input is marked “Analyst assumption” with a reason.
“I invest in asset-backed finance and want a collateral and structure model for a consumer loan pool, built on a real securitization. From SEC EDGAR, pick one recent U.S. auto loan securitization trust that files Form ABS-EE asset-level data, and take the pool's balances, interest rates, remaining terms, credit scores, delinquencies and losses from the latest filings, and the class sizes, coupons, overcollateralization target, reserve account and triggers from its prospectus. Build pool stratifications by credit score, rate, term and vehicle age. Build the issuer's static pool cumulative net loss curves from the prospectus for its prior trusts and place this trust's performance to date on them. Set default, prepayment and loss severity assumptions with a note on each, then project the pool's monthly cash flows and run them through the waterfall: servicing fee, interest and principal by class, overcollateralization build to target, and the reserve account. Report the weighted average life and the losses each class can absorb before its first dollar of loss, as a multiple of the base-case cumulative net loss. Add a sensitivity table of the most junior class's loss coverage to default and prepayment rates. Flag cumulative net losses running above the issuer's prior trusts at the same age, a trigger the base case breaches, and a junior class with loss coverage below 1.5x. Use the stored public data wherever it has the figure, and web search only for what it lacks. Put every input on an Assumptions sheet with one column that says where it came from: a cited source, or "analyst assumption" for a figure I would set myself, with the reason for the value you chose. Every output must be a live formula that traces back to the Assumptions sheet. Add a Sources sheet listing each data source, what it gave and its date. Deliver one Excel workbook and a short summary of the result, the flags, and what I would need from the issuer to firm it up. Cite every sourced figure.”
That is the request as we sent it, without its closing line, which asks that no name or email be put in the workbook.
Choosing the trust and reading the prospectus
Compound screened the auto loan trusts that file Form ABS-EE by the size of their asset data files and chose Exeter Select 2025-1, whose full September 2026 file could be carried loan by loan in the workbook. Every class, coupon, enhancement level and trigger is cited to its prospectus, filed on April 21, 2025.
The trust closed on April 23, 2025, with a cutoff pool of $336.8 million. Its Form ABS-EE for the period ended September 30, 2026 was filed on October 29, 2026, and the Form 10-D for the same collection period gives the note balances after the October 15, 2026 distribution. Classes A-1 and A-2 have paid off. $175.8 million of notes remain, with Class A-3 at $73.1 million and the subordinate classes still at their original balances.
| B | C | D | E | F | |
|---|---|---|---|---|---|
| 19 | Class | Original balance ($) | Coupon | Legal final date | Balance after last distribution ($) |
| 20 | Class A-1 | 30,000,000 | 4.607% | 5/15/2026 | - |
| 21 | Class A-2 | 105,300,000 | 4.830% | 10/16/2028 | - |
| 22 | Class A-3 | 89,500,000 | 4.690% | 4/15/2030 | 73,091,924 |
| 23 | Class B | 23,410,000 | 4.870% | 8/15/2031 | 23,410,000 |
| 24 | Class C | 37,550,000 | 5.400% | 8/15/2031 | 37,550,000 |
| 25 | Class D | 35,030,000 | 6.140% | 9/15/2031 | 35,030,000 |
| 26 | Class E | 6,730,000 | 8.440% | 12/15/2032 | 6,730,000 |
| 27 | Total notes | 327,520,000 | 175,811,924 |
The coupons step up from 4.69% on A-3 to 8.44% on Class E. Class E is the thinnest tranche at $6.73 million, and the answer lists “the final terms of the Class E notes, which weren’t publicly offered” among the items to confirm with the issuer.
| B | C | D | |
|---|---|---|---|
| 5 | Input | Value | Unit |
| 30 | Target OC: % of current pool balance | 6.80% | % |
| 31 | Target OC floor: % of cutoff pool balance | 1.50% | % |
| 32 | Specified reserve: % of cutoff pool balance | 1.00% | % |
| 35 | OC amount at model start (reported) | 12,827,479 | $ |
| 36 | Servicing fee: annual % of beginning pool balance | 2.00% | % |
| 38 | Clean-up call threshold: % of cutoff pool balance | 5.00% | % |
| 39 | Delinquency trigger: 61+ day delinquency rate | 40.00% | % |
| 40 | Initial OC at closing: % of cutoff pool balance | 2.75% | % |
The structure is a typical sequential-pay auto deal. Overcollateralization started at 2.75% of the cutoff pool and builds from excess spread to a target of 6.80% of the current pool, with a floor of 1.50% of the cutoff pool. The reserve is fixed at 1.00% of the cutoff pool, $3.4 million. The servicing fee is 2.00% a year. The trigger the model tests each month is the prospectus’s asset representations review trigger, a 40% rate of receivables more than 60 days delinquent. It also tests for a missed senior interest payment, an event of default, and for any class outstanding past its legal final date.
The ABS-EE loan tape and the pool strats
The Loan Tape tab holds all 9,128 active receivables from the ABS-EE file, and it ties to the 10-D pool balance of $188.6 million to the cent. Every strat on the Strats tab is a live formula over that tape, with the bucket edges set on the Assumptions sheet.
| C | D | E | F | G | H | I | J | K | L | |
|---|---|---|---|---|---|---|---|---|---|---|
| 5 | State | Origination (mm/yyyy) | Current balance ($) | Current APR | Remaining term (mo) | Original term (mo) | Credit score | Vehicle model year | New (1) / Used (2) | Days delinquent |
| 6 | FL | 01/2025 | 8,917.64 | 27.11% | 42 | 61 | 650 | 2013 | 2 | 94 |
| 7 | IN | 01/2025 | 17,543.58 | 22.00% | 55 | 74 | 710 | 2020 | 2 | 0 |
| 8 | TN | 10/2024 | 19,772.88 | 17.52% | 58 | 79 | 653 | 2021 | 2 | 0 |
| 9 | MI | 02/2025 | 28,287.95 | 16.15% | 60 | 80 | 655 | 2020 | 2 | 20 |
The pool is near-prime and seasoned. The weighted average credit score is 671, the weighted average remaining term is 56.6 months against an original 76.7, and the average balance is $20,666. The collateral is mostly older used cars: the weighted average vehicle age is 6.0 years, and 15.8% of the balance is on cars 10 years old or more.
| B | C | D | E | F | G | H | J | |
|---|---|---|---|---|---|---|---|---|
| 19 | Bucket | Receivables | Principal balance ($) | % of balance | WA APR | WA credit score | WA remaining term (mo) | 60+ days delinquent (% of bucket) |
| 20 | Not available | 28 | 523,989 | 0.3% | 19.37% | 0 | 53.2 | 9.72% |
| 21 | Below 620 | 354 | 7,437,396 | 3.9% | 17.31% | 576 | 56.5 | 2.34% |
| 22 | 620-639 | 233 | 5,073,201 | 2.7% | 16.84% | 630 | 56.7 | 3.22% |
| 23 | 640-659 | 3,373 | 71,358,423 | 37.8% | 18.24% | 649 | 56.7 | 5.64% |
| 24 | 660-679 | 2,074 | 42,540,926 | 22.6% | 17.91% | 669 | 56.4 | 4.33% |
| 25 | 680-699 | 1,477 | 30,486,820 | 16.2% | 17.37% | 689 | 56.5 | 4.63% |
| 26 | 700-719 | 691 | 13,575,629 | 7.2% | 18.08% | 708 | 56.3 | 4.00% |
| 27 | 720-739 | 388 | 7,765,371 | 4.1% | 17.76% | 728 | 56.6 | 5.64% |
| 28 | 740 and above | 510 | 9,877,650 | 5.2% | 16.70% | 775 | 57.2 | 4.72% |
| 29 | Total | 9,128 | 188,639,404 | 100.0% | 17.84% | 671 | 56.6 | 4.83% |
The credit score strat is concentrated: 37.8% of the balance sits between 640 and 659, and 3.9% is below 620. It also separates risk poorly. Loans below 620 are 2.34% delinquent, under the pool’s 4.83%, and the 720-739 bucket is as delinquent as the 640-659 bucket at 5.64%.
The rate strat tells a cleaner story, because the APR is the issuer’s own risk-based price. The share of balance more than 60 days delinquent climbs from 0.62% below a 12% APR to 8.91% at 24% and above.
Share of balance more than 60 days delinquent, by APR
Active receivables at September 30, 2026. The pool average is 4.83%
Compound's Strats tab, rows 34 to 41: Form ABS-EE asset data for the September 2026 reporting period.
The term strat has one bucket to ask about. Loans with 61 to 66 months remaining are 14.5% of the balance and 11.12% delinquent, more than twice the pool rate (our arithmetic). With the loans 20.2 months seasoned on average, a long remaining term can point to extended contracts, and the answer puts extensions at 3.7% of the balance in September. The extension history on the issuer list would settle it.
Performance to date on the static pool curves
Over eighteen monthly reports, cumulative net loss rose to 5.35% and the 60+ day delinquency rate to 4.83%. Placed on the issuer’s vintage curves, the trust sits 0.13 points under the vintage average at month 18, with 5 of the 15 vintages reporting at that age lower. The flag does not trip.
| B | C | H | K | L | R | AA | AH | AI | AN | |
|---|---|---|---|---|---|---|---|---|---|---|
| 5 | Deal month | Collection period end | Ending pool balance | Cumulative net losses | CNL ratio (reported) | Delinquency rate (60+, reported) | OC amount | Annualized CDR | Period loss severity | Interest collected / contractual interest |
| 9 | 4 | 7/31/2025 | 308,167,438 | 1,572,994 | 0.47% | 2.49% | 19,675,168 | 5.47% | 76.7% | 95.2% |
| 12 | 7 | 10/31/2025 | 280,005,807 | 5,242,549 | 1.56% | 2.98% | 19,040,395 | 6.40% | 54.1% | 96.3% |
| 15 | 10 | 1/31/2026 | 253,708,828 | 9,104,274 | 2.70% | 4.06% | 17,252,200 | 9.04% | 60.6% | 96.9% |
| 18 | 13 | 4/30/2026 | 227,550,071 | 12,297,182 | 3.65% | 3.83% | 15,473,405 | 10.57% | 53.7% | 94.4% |
| 21 | 16 | 7/31/2026 | 203,110,817 | 15,979,114 | 4.74% | 4.37% | 13,811,536 | 10.04% | 54.9% | 99.2% |
| 22 | 17 | 8/31/2026 | 195,819,554 | 16,959,303 | 5.04% | 4.61% | 13,315,730 | 11.19% | 49.0% | 93.5% |
| 23 | 18 | 9/30/2026 | 188,639,404 | 18,013,454 | 5.35% | 4.83% | 12,827,479 | 10.41% | 59.1% | 92.4% |
The annualized default rate has run between 10.04% and 12.30% since month 13, and the trust’s trailing six-month CDR is 10.93%. Monthly severity has moved between 49.0% and 59.5% over the same months, with a cumulative severity of 59.18%. Delinquencies are the line moving the wrong way: the 60+ day rate rose from 3.83% at month 13 to 4.83% at month 18. Overcollateralization built from closing to $19.7 million by month 4, then stepped down with the pool while holding its 6.80% target.
The static pool annex covers 20 quarterly Exeter origination vintages with data through December 31, 2024. Compound built cumulative net and gross loss, ABS speed, pool factor and implied default rate curves for each, then set the trust beside them by months seasoned.
Cumulative net loss by months seasoned: the trust against the issuer's vintage pools
% of original balance. Vintage pools Q1 2020 to Q4 2024 from the prospectus static pool annex; the trust from its Form 10-D reports
Compound's Static Pool tab, rows 30 to 60, columns W to Z: Exeter Select Automobile Receivables Trust 2025-1 prospectus, Annex A, and Forms 10-D for April 2025 to September 2026.
| B | W | X | Y | Z | AA | AB | |
|---|---|---|---|---|---|---|---|
| 29 | Months seasoned | Vintage average | EXSA 2025-1 (trust) | Vintage minimum | Vintage maximum | Vintages reporting | Trust less average |
| 33 | 3 | 0.03% | 0.13% | 0.01% | 0.09% | 20 | 0.10% |
| 36 | 6 | 1.23% | 1.30% | 0.63% | 1.82% | 19 | 0.07% |
| 39 | 9 | 2.47% | 2.33% | 1.29% | 3.60% | 18 | -0.14% |
| 42 | 12 | 3.44% | 3.30% | 1.74% | 5.31% | 17 | -0.14% |
| 45 | 15 | 4.51% | 4.44% | 2.09% | 7.06% | 16 | -0.07% |
| 48 | 18 | 5.48% | 5.35% | 2.42% | 8.38% | 15 | -0.13% |
The trust ran slightly above the average for its first six months and has been under it from month 7 on. The range is wide. At month 18 the vintages run from 2.42% to 8.38%, and the 2020 and early 2021 pools sit at the bottom while the 2022 pools sit at the top. A trust at the average is in the middle of a spread of almost six points (our arithmetic).
Two features of the benchmark matter. The vintages are Exeter’s origination pools, none of them a Select trust, and their weighted average FICO scores ran from 632 to 658 against the trust’s current 671. And the average at month 18 mixes vintages with very different credit cycles. The flag compares the trust with that average, so a trust tracking the 2022 pools would trip it and one tracking the 2020 pools would not.
Default, prepayment and severity assumptions
The base case is a 12% CDR, a 1.20% ABS voluntary prepayment speed, 60% loss severity with a one-month recovery lag, and 94% of contractual interest collected. Each is set between the trust’s own history and the vintage pools, and each reason is on the Assumptions sheet.
| B | C | D | |
|---|---|---|---|
| 5 | Input | Value | Unit |
| 72 | Default rate: annual CDR on the performing balance | 12.00% | % per year |
| 73 | Voluntary prepayment speed (ABS) | 1.20% | % ABS |
| 74 | Loss severity: % of defaulted balance not recovered | 60.00% | % |
| 75 | Recovery lag | 1 | months |
| 80 | Interest collection rate: % of contractual interest on the performing balance | 94.0% | % |
| 82 | Trust trailing 6-month CDR | 10.93% | % |
| 83 | Trust trailing 12-month CDR | 9.84% | % |
| 84 | Comparable vintages: average CDR at months seasoned 19-60 | 12.98% | % |
| 85 | Trust cumulative loss severity (net losses / liquidations) | 59.18% | % |
| 86 | Trust trailing 6-month loss severity | 55.44% | % |
| 87 | Comparable 2022-2023 vintages: CNL / CGL at latest month (average) | 64.66% | % |
| 88 | Trust trailing 6-month cash principal rate (% of beginning balance) | 2.722% | % per month |
| 91 | Implied voluntary ABS at current pool age | 1.21% | % ABS |
| 92 | Comparable vintages: average total ABS speed at months 19-60 (includes defaults) | 1.28% | % ABS |
| 93 | Trust trailing 6-month interest and fees collected / contractual interest | 94.27% | % |
The default rate sits between two reference points. The note on the CDR reads: “trust’s trailing 6-month CDR is 10.9% (row 82) and 60+ day delinquencies have risen from 3.8% to 4.8% over six months; comparable vintages averaged about 13.0% CDR at months 19-60 (row 84). 12.0% sits between them”. The trailing twelve-month CDR is lower again, at 9.84%.
For prepayments, Compound backed the voluntary speed out of the trust’s own cash principal. The trailing six-month cash principal rate is 2.722% of the beginning balance a month. Subtracting the model’s scheduled amortization leaves an implied voluntary ABS of 1.21% at the pool’s age. The vintages’ 1.28% is higher, and the note gives the reason not to use it: “vintage ABS speeds include defaults, so they overstate voluntary prepayments.”
Severity is set at 60%, against the trust’s 59.18% cumulative and 55.44% trailing six-month figures and 64.66% for the 2022 and 2023 vintages. The note says 60% “leans on the trust’s own history with a margin.” The interest collection rate of 94% comes from the trust’s trailing six months, when it collected 94.27% of contractual interest.
The result is a lifetime cumulative net loss of 12.2% of the cutoff pool, including the 5.35% already taken. Two simple read-throughs from the vintage curves bracket it.
| B | C | |
|---|---|---|
| 367 | Average implied CDR at months seasoned 19-60 | 13.0% |
| 368 | Average total ABS speed at months seasoned 19-60 (includes defaults) | 1.28% |
| 369 | 2022-2023 vintages: average CNL / CGL at latest month | 64.7% |
| 370 | Trust CNL at latest month | 5.35% |
| 371 | Trust months seasoned | 18 |
| 372 | Vintage average CNL at the same month | 5.48% |
| 373 | Trust less vintage average | -0.13% |
| 374 | Vintages below the trust at the same month | 5 |
| 375 | Vintages reporting at that month | 15 |
| 376 | Flag: trust CNL above the vintage average at the same month | No |
| 377 | Static-pool projection of trust lifetime CNL: multiplicative | 13.76% |
| 378 | Static-pool projection of trust lifetime CNL: additive | 10.19% |
| 379 | Model base-case lifetime CNL | 12.17% |
The multiplicative projection, 13.76%, scales the trust’s 5.35% by how much the older vintages’ losses grew after month 18. The additive one, 10.19%, adds the points they added. Both lean on the only five vintages with 48 months or more of history, Q1 2020 to Q1 2021, which are also the lowest-loss pools in the set. The base case at 12.17% sits between them.
The monthly cash flows and the waterfall
Four rep lines, bucketed by remaining term, project the pool for 72 months. Each month’s collections pay the 2.00% servicing fee and trust fees, interest by class, then principal sequentially, with all excess spread turbo-paying notes until overcollateralization reaches target and the rest released to the certificateholders.
| B | D | AD | AE | AF | AG | AH | AO | AV | CT | CU | DG | DI | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 8 | Month | Distribution date | Pool beginning balance | Defaults | Scheduled principal | Prepayments | Interest collected | Net losses | Servicing fee due | OC principal (priority 19) | Release to certificateholders | Total notes ending | OC % of pool |
| 9 | 0 | 10/15/2026 | - | 175,811,924 | 6.80% | ||||||||
| 10 | 1 | 11/15/2026 | 188,639,404 | 1,998,869 | 2,150,925 | 2,921,959 | 2,608,330 | 1,199,321 | 314,399 | 6,584,554 | - | 169,227,371 | 6.80% |
| 11 | 2 | 12/15/2026 | 181,567,651 | 1,923,935 | 2,125,715 | 2,856,736 | 2,510,329 | 1,154,361 | 302,613 | 6,443,071 | 790,118 | 162,784,300 | 6.80% |
| 12 | 3 | 1/15/2027 | 174,661,266 | 1,850,753 | 2,100,256 | 2,792,059 | 2,414,622 | 1,110,452 | 291,102 | 6,284,540 | 769,525 | 156,499,760 | 6.80% |
| 20 | 11 | 9/15/2027 | 125,123,597 | 1,325,840 | 1,887,412 | 2,293,563 | 1,728,203 | 795,504 | 208,539 | 5,132,351 | 571,690 | 111,482,842 | 6.80% |
| 33 | 24 | 10/15/2028 | 64,449,466 | 682,922 | 1,504,919 | 1,549,247 | 887,822 | 409,753 | 107,416 | 3,737,089 | 75,047 | 55,660,539 | 8.32% |
| 41 | 32 | 6/15/2029 | 37,978,002 | 402,424 | 1,245,439 | 1,127,165 | 521,515 | 241,454 | 63,297 | 2,775,028 | 41,602 | 30,151,135 | 14.35% |
| 45 | 36 | 10/15/2029 | 27,558,014 | 292,011 | 1,108,406 | 925,379 | 377,473 | 175,207 | 45,930 | 2,325,796 | 32,565 | 20,180,380 | 20.02% |
In month 1, $2.0 million of the pool defaults and $1.2 million becomes net loss, with recoveries arriving a month later. After the $314,399 servicing fee and class interest, $6.58 million is left, and all of it goes to principal to hold overcollateralization at 6.80%. From month 2, a release to the certificateholders starts at $790,118 and falls with the pool.
The floor changes the picture in the deal’s third year. From month 21, 6.80% of the pool is less than the 1.50% floor on the cutoff pool, about $5.05 million (our arithmetic), so the target stops falling. Overcollateralization as a share of the pool then rises, to 8.32% at month 24 and 20.02% at month 36, and the releases shrink to a trickle. The back end of the deal is where the subordinate classes get their protection.
| B | C | |
|---|---|---|
| 97 | Lifetime cumulative net loss (% of cutoff pool) | 12.17% |
| 98 | Projected net losses from model start ($) | 22,968,151 |
| 99 | Maximum projected 60+ delinquency rate | 5.41% |
| 100 | Months the delinquency trigger is breached | 0 |
| 101 | Months with a senior-class interest shortfall | 0 |
| 102 | Class-months outstanding past legal final | 0 |
| 103 | Reserve account draws ($) | - |
| 104 | Release to certificateholders ($) | 20,915,362 |
| 105 | Clean-up call eligible: month | 41 |
| 106 | Clean-up call eligible: distribution date | 3/15/2030 |
The base case projects $22,968,151 of further net losses and $20,915,362 released to the certificateholders. The 60+ day delinquency rate peaks at 5.41%, far from the 40% review trigger. No month has a senior interest shortfall, no class runs past its legal final date and the reserve is never drawn. The pool reaches the 5% clean-up call threshold on March 15, 2030, and the model leaves the call unexercised so that loss coverage does not rely on it.
Weighted average life and loss coverage by class
At the base speed, Class E takes its first dollar of loss at a 22.0% CDR, a 16.88% lifetime cumulative net loss. That is 1.39x the base case. Class D breaks at 18.34%, 1.51x. Class C covers 2.10x, Class B 2.83x, and Class A-3 takes no loss at any CDR up to 100%.
| B | C | D | E | F | G | I | K | L | M | N | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 20 | Class | Balance ($) | Coupon | Hard credit enhancement (% of pool) | WAL to maturity (yrs) | WAL to 5% call (yrs) | Legal final date | Breakeven CDR | Breakeven lifetime CNL | Loss coverage (x base CNL) | Remaining-loss coverage (x) |
| 21 | A-3 | 73,091,924 | 4.69% | 63.0% | 0.55 | 0.55 | 4/15/2030 | None to 100% | 38.96% | >3.2x | >4.9x |
| 22 | B | 23,410,000 | 4.87% | 50.6% | 1.32 | 1.32 | 8/15/2031 | 88.0% | 34.44% | 2.83x | 4.27x |
| 23 | C | 37,550,000 | 5.40% | 30.7% | 1.94 | 1.94 | 8/15/2031 | 46.4% | 25.50% | 2.10x | 2.95x |
| 24 | D | 35,030,000 | 6.14% | 12.2% | 2.93 | 2.92 | 9/15/2031 | 25.5% | 18.34% | 1.51x | 1.90x |
| 25 | E | 6,730,000 | 8.44% | 8.6% | 3.85 | 3.42 | 12/15/2032 | 22.0% | 16.88% | 1.39x | 1.69x |
The breakevens come from a 101-row data table on the Sensitivity tab that reruns the full cash flow at each CDR from 0% to 100% in 1% steps, holding the base prepayment speed. Each class’s breakeven is interpolated between the last row without a loss and the first row with one. At a 22% CDR Class E loses its first $44,912; at a 26% CDR Class D loses $935,775.
Breakeven lifetime cumulative net loss by class, against the base case
% of the cutoff pool, at a 1.20% ABS. 1.5x the base case is 18.25% (our arithmetic). Class A-3 has no loss at a 100% CDR; its bar is the loss at that CDR
Compound's Summary tab, column L, and Cash Flow tab, cell C97.
The weighted average lives run from 0.55 years on A-3 to 3.85 years on Class E to maturity, or 3.42 years to the 5% call. E’s last payment falls on November 15, 2030, against a legal final of December 15, 2032.
The coverage multiple has one convention to read carefully. Lifetime CNL includes the 5.35% the trust has already lost, which inflates both sides of the ratio by the same amount. The Summary also reports a remaining-loss coverage, future losses at breakeven over future losses in the base case. On that measure Class E covers 1.69x and Class D 1.90x. The request’s 1.5x test uses the lifetime figure, which is the stricter of the two for a seasoned deal.
Hard credit enhancement explains the gap between D and E. Class E is protected only by overcollateralization and the reserve, 8.6% of the current pool; Class D adds E’s $6.73 million beneath it, for 12.2%. Excess spread does the rest: at a 17.84% pool APR and coupons of 8.44% or less, the deal earns several points a year above its note interest and servicing fee, which is how a class with 8.6% of hard enhancement absorbs a 16.88% lifetime loss.
Class E’s coverage across default and prepayment rates
Across 8% to 16% CDR and 0.8% to 1.6% ABS, Class E’s coverage runs from 1.16x to 1.72x. It stays under 1.5x at a 12% CDR or higher at every prepayment speed tested, and reaches 1.50x or more at a 10% CDR at every speed from 1.0% ABS up.
| B | C | D | E | F | G | |
|---|---|---|---|---|---|---|
| 38 | CDR \ ABS | 0.80% | 1.00% | 1.20% | 1.40% | 1.60% |
| 39 | 8.0% | 1.66x | 1.67x | 1.68x | 1.70x | 1.72x |
| 40 | 10.0% | 1.49x | 1.50x | 1.52x | 1.54x | 1.56x |
| 41 | 12.0% | 1.36x | 1.37x | 1.39x | 1.41x | 1.44x |
| 42 | 14.0% | 1.25x | 1.26x | 1.28x | 1.30x | 1.33x |
| 43 | 16.0% | 1.16x | 1.18x | 1.19x | 1.22x | 1.24x |
The grid moves far more with the default rate than with prepayments. Each two points of CDR cost Class E between 0.09x and 0.17x of coverage across the grid (our arithmetic), while the full range of prepayment speeds moves it by less than 0.1x at any CDR.
Faster prepayments help coverage in this grid, but they hurt the class’s breakeven. At a 0.80% ABS Class E breaks at a 17.96% lifetime loss; at 1.60% it breaks at 15.68%, because faster payoffs leave less excess spread to absorb losses. Coverage still rises with speed because the scenario’s own lifetime loss falls faster than the breakeven does.
| B | C | D | E | F | G | |
|---|---|---|---|---|---|---|
| 255 | ABS speed | 0.80% | 1.00% | 1.20% | 1.40% | 1.60% |
| 256 | Breakeven CDR | 20.95% | 21.42% | 21.98% | 22.77% | 23.75% |
| 257 | Breakeven lifetime CNL | 17.96% | 17.45% | 16.88% | 16.28% | 15.68% |
The breakeven CDR runs from 20.95% to 23.75%, all well above the trust’s trailing 10.93%. The flag says the margin between the base case and Class E’s first dollar of loss is thinner than the request’s threshold. In the base case itself, at a 12% CDR, Class E repays in full.
| B | C | D | |
|---|---|---|---|
| 29 | Test | Result | Detail |
| 30 | CNL running above the issuer's prior pools at the same age | No | Trust 5.35% vs vintage average 5.48% at month 18; 5 of 15 vintages were lower |
| 31 | Delinquency trigger (61+ days, 40%) breached in base case | No | Projected 60+ delinquency peaks at 5.4% against the 40% trigger |
| 32 | Reserve account drawn in base case | No | Draws of $0; reserve stays at the $3,367,892 specified amount until notes fall below it |
| 33 | Senior interest shortfall (event of default) in base case | No | 0 months with a shortfall on the most senior class outstanding |
| 34 | Class outstanding past its legal final date in base case | No | All classes repaid before legal final; smallest cushion 15 months |
| 35 | Most junior class (E) loss coverage below 1.5x | FLAG | Class E covers 1.39x base-case CNL; Class D covers 1.51x |
What the public filings cannot tell you
The prospectus, the servicer reports and the tape give the structure, the history and today’s pool. The recovery timing, the voluntary prepayment history and the issuer’s own trust performance come only from Exeter, and each replaces a specific assumption in the model. Compound’s answer lists what to ask for:
- Performance of Exeter’s own Select trusts, and vintage curves updated past December 2024. This replaces the origination vintages as the benchmark for the flag and the read-throughs.
- Recovery data: time from charge-off to recovery, repossession and auction proceeds, and severity by vintage. This replaces the 60% severity and the one-month lag.
- Whether the static pool prepayment speeds include defaults, plus a history of voluntary prepayments only. This replaces the 1.20% ABS backed out of cash principal.
- Extension, deferral and modification history, and delinquency roll rates. Extensions were 3.7% of the balance in September, and roll rates would show whether the rise to 4.83% delinquency feeds the CDR.
- The servicer’s own loss forecast, the transaction documents to confirm the fee caps and reserve release mechanics, and the final terms of the Class E notes.
Our view
The model gives a clear read on where the subordinate classes stand: losses tracking the issuer’s average, a base case inside the vintage read-throughs, and a Class E whose cushion is real but under 1.5x. We would keep the structure and spend the diligence on the default rate, which moves coverage most.
Under the model’s assumptions the senior and mezzanine classes are well protected. Class C covers 2.10x and B 2.83x, nothing in the base case touches the reserve, and the delinquency trigger sits an order of magnitude away. The trust’s 5.35% at month 18 is in the middle of a wide vintage range. These are the model’s results under its assumptions, not a rating view on the notes or a view on Exeter.
The other side is that Class E’s 1.39x depends on a 12% CDR that sits above the trust’s own trailing rate and below the vintages’, with delinquencies rising. At a 10% CDR Class E clears 1.5x at most speeds; at 14% it covers 1.28x at the base speed. The benchmark is also softer than the request asked for, since no prior Select trust exists, and the two read-throughs rest on the five lowest-loss vintages.
Three findings would change our view: Select trust performance running above the origination vintages, recovery data pointing to severity well above 60%, and delinquency roll rates that lift the trailing CDR toward the vintages’ 13%. From here the work is a sensitivity analysis on severity alongside the CDR grid, and monthly portfolio monitoring of each new 10-D against the base case.
Exeter auto ABS collateral and structure model (Excel)Compound’s workbook from this run: a Summary with the flags, an Assumptions sheet with a source or reason beside every input, the monthly Cash Flow and waterfall, breakeven and Sensitivity tables, the Static Pool curves, Strats, eighteen months of trust Performance, the 9,128-loan Loan Tape and Sources.Download · FreeFAQ
How do you calculate breakeven loss coverage for an ABS class?
Rerun the full cash flow at rising default rates, holding prepayments and severity, until the class takes its first dollar of loss. The lifetime cumulative net loss at that point, divided by the base case’s, is the coverage multiple. In this model Class E breaks at a 16.88% lifetime loss against a 12.2% base case, 1.39x.
Should loss coverage use lifetime or remaining cumulative net loss?
Both are useful on a seasoned deal. Lifetime coverage counts losses already taken on both sides of the ratio; remaining-loss coverage counts only losses still to come. Class E covers 1.39x on the lifetime measure and 1.69x on the remaining measure, so a 1.5x test passes or fails depending on which one the investor uses.
What do you compare a trust's losses with when the shelf has no prior deals?
The prospectus static pool annex. Exeter Select 2025-1 was the first deal on its shelf, so the model uses the 20 quarterly Exeter origination vintages the prospectus presents as comparable, with data through December 31, 2024. At month 18 the trust’s 5.35% sits just under the 5.48% vintage average.
How do you back out a voluntary ABS prepayment speed from servicer reports?
Take the trust’s cash principal collections as a share of the beginning balance, subtract scheduled amortization to get a monthly voluntary rate, and convert it to ABS at the pool’s age. Here the trailing six-month cash principal rate of 2.722% a month implies a 1.21% ABS, which the model rounds to 1.20%. Static pool ABS speeds usually include defaults, so they overstate voluntary prepayments.
Why does excess spread matter so much in subprime and near-prime auto ABS?
The pool earns far more than the notes cost. With a 17.84% weighted average APR against coupons of 4.69% to 8.44% and a 2.00% servicing fee, excess spread absorbs losses each month before they reach overcollateralization. That is how Class E, with 8.6% of hard credit enhancement, can absorb a 16.88% lifetime loss at the base prepayment speed.





