IN ONE SENTENCE — A securitisation desk turns a pool of loans that a client cannot easily sell one by one into bonds that investors can buy, by selling the pool to a special-purpose company that pays investors only out of the pool’s cash flows, in an order fixed in advance.
Where it sits. In most European CIBs securitisation lives in Global Markets, inside the credit or “financing” business, next to credit trading and repo. It is a hybrid. The front end is an advisory and underwriting business close to DCM (see C01): the desk is mandated by an originator, structures a deal, obtains ratings, writes the offering circular with lawyers and places the notes through the ABS syndicate desk. The back end is a lending business: before a public deal the desk often finances the assets in a warehouse, a senior secured loan to a special-purpose vehicle that holds the loans until they are numerous enough to be securitised. Some banks book warehouses in a “securitised products financing” or “asset-backed finance” desk; the risk is the same. A third leg trades ABS and CLO paper in the secondary market and makes markets for clients.
Clients. Originators are those who create loans and need funding or capital relief: captive finance companies of car makers (auto loans and leases), consumer finance companies, mortgage lenders and banks themselves (RMBS, SME loans, corporate loans), leasing companies, and, for CLOs, asset managers who buy leveraged loans and need term funding. Investors are bank treasuries (who buy senior STS tranches for liquidity portfolios), asset managers, insurers, pension funds, credit hedge funds, and, for the riskiest tranches, specialist funds. The desk has two client franchises to serve at once, and its value comes from understanding both.
How it earns money, with orders of magnitude. For a EUR 800m auto ABS like the one in the case, the revenues of the arranging bank are, in this module’s case assumptions, of the order of EUR 2m of structuring, arranging and placement fees (around 25 bp of the notes), plus warehouse income of several million euros a year while the pool is built, plus the margin on the interest-rate swap with the SPV, plus secondary trading. A bank that arranges ten to fifteen public European ABS a year and runs a few billion euros of warehouses therefore has a business of a few tens of millions of euros of revenue: not the largest in a CIB, but capital-light on the fee side, sticky (originators come back every year) and strategic, because the same originators borrow, hedge and deposit with the bank.
The team. A typical European securitisation desk of a large bank counts 15 to 40 people:
A realistic week. Monday, the analyst reruns the pool cut of the month for a client, checks it against the eligibility criteria and sends stratification tables to the rating agencies. Tuesday, the structurer has a call with a captive finance company whose warehouse is 85% used and which wants to price a public deal before the summer; the topic is the timing and whether the deal can be labelled green. Wednesday, the warehouse team receives the monthly servicer report and checks the borrowing base: one concentration limit is breached, so the advance rate falls until the originator adds eligible loans. Thursday, the syndicate desk announces a deal to investors and opens books on the senior tranche; the CLO team prices a reset for an asset manager. Friday, the desk’s regulatory capital specialist sends to the supervisor the significant risk transfer notification of a deal closed last month, and the sustainable finance team reviews the factsheet of a green RMBS before external review.
WHO DOES WHAT — Origination owns the client and the mandate; structuring and analytics own the model; syndicate owns the price; Legal owns the documents; the arranging bank’s Credit risk approves the warehouse and any retained position; the originator owns the pool data, the retention and, in a European Green Bond, the use-of-proceeds commitment.
IN ONE SENTENCE — A securitisation sells a pool of loans to a bankruptcy-remote company that finances the purchase by issuing notes of different seniority, so that the pool’s losses are borne first by the most junior notes and last by the most senior.
INTUITION — Think of a cellar of wine bottles poured into a row of glasses placed one below the other. Wine runs into the top glass first; only when it is full does the next glass receive anything. A leak in the cellar is felt first by the lowest glass. Investors choose a glass according to how sure they want to be of filling it.
The European Union defines a securitisation (Regulation (EU) 2017/2402, the Securitisation Regulation or SECR, Article 2) by three features: the credit risk of a pool of exposures is tranched; payments depend on the performance of the pool; and the subordination of tranches determines the distribution of losses during the life of the transaction. Two families exist:
This module’s case is a traditional, public, term securitisation of auto loans: the most common European asset-backed security (ABS) after RMBS and CLOs.
| Party | Role | Typical document |
|---|---|---|
| Originator and seller | Creates the loans, sells them to the SSPE | Receivables purchase agreement |
| Servicer | Collects payments, manages arrears, reports monthly | Servicing agreement, investor report |
| SSPE (issuer) | Buys the pool, issues the notes | Articles, trust deed, terms and conditions |
| Arranger and lead managers | Structure and place the deal | Mandate letter, subscription agreement |
| Security trustee | Holds security for noteholders | Trust deed, deed of charge |
| Cash manager and account bank | Run the waterfall, hold the cash | Cash management agreement |
| Swap counterparty | Converts fixed loan rates into floating note rates | ISDA master and confirmation |
| Rating agencies | Rate each class | Rating letters, presale report |
| STS verifier, external reviewer | Verify STS criteria; review the green factsheet | STS verification report, pre-issuance review |
The true sale is the legal heart of a traditional deal. Lawyers give opinions that the sale would not be recharacterised as a secured loan and that the assets would not be reachable by the originator’s creditors if it went bankrupt. The SSPE is bankruptcy remote: an orphan company (often owned by a charitable trust or a foundation in a European jurisdiction, or a French fonds commun de titrisation managed by a management company), with no employees, limited purpose and creditors who have agreed not to petition for its insolvency. The offering circular (prospectus) describes the pool, the structure, the risks and the documents; the investor report and the loan-level data published monthly in the ESMA templates keep investors informed afterwards.
An auto loan is an amortising annuity: the borrower pays a constant monthly instalment that contains interest on the outstanding balance and a growing share of principal. A pool of 32,000 such loans behaves like one large annuity, modified by two leakages:
FORMULA
is the performing pool balance at the start of month ; the monthly loan rate (annual rate divided by 12); the number of months remaining; the interest collected; the scheduled principal; the prepayment; the single monthly mortality, the monthly equivalent of the CPR.
The most quoted number of an ABS is its weighted average life (WAL): the average time, in years, at which principal is repaid, each repayment weighted by its amount.
FORMULA
is the initial balance and the total principal received in month (scheduled plus prepaid). WAL, not legal final maturity, determines the spread at which a tranche is priced and the swap that hedges it.
INTERACTIVE LAB — An amortising auto loan pool: balance, prepayments and weighted average life
See how the remaining term and the prepayment rate drive the amortisation profile and the WAL of a pool of fixed-rate auto loans.
| Parameter you can move | Range | Default |
|---|---|---|
| Initial pool balance | 100 to 2000 EUR m | 800 EUR m |
| Weighted average customer rate | 1 to 12 % | 5.4 % |
| Weighted average remaining term | 12 to 72 months | 48 months |
| Prepayment rate (CPR) | 0 to 30 % a year | 8 % a year |
| Result at default settings | Value |
|---|---|
| First monthly instalment of the pool | 18.569 EUR m |
| Weighted average life with prepayments | 1.90 years |
| Weighted average life without prepayments | 2.11 years |
| Total interest collected over the life | 82.0 EUR m |
Try it. Raise the CPR to 20%: the WAL shortens by several months. Investors who bought a two-year bond get their money back earlier, which matters when the note was bought above par or hedged with a swap of fixed notional.
The SSPE issues several classes of notes. Each has an attachment point (the share of the pool that must be lost before it loses anything) and a detachment point (the loss at which it is wiped out). The protection a class enjoys is its credit enhancement (CE). Four sources exist:
The priority of payments, or waterfall, sets the order in which cash is used. European ABS usually keep two waterfalls:
Triggers are the automatic switches of the structure: a cumulative net loss above a schedule, a PDL balance above a threshold, a servicer event of default, a breach of the swap counterparty’s rating. When a trigger is hit, the structure protects the senior class: pro rata becomes sequential, the revolving period ends, a backup servicer is activated. Clean-up call: when the pool has amortised below 10% of its initial balance, the originator may buy it back and redeem the notes at par, because the fixed costs of running a small deal are too high.
Rating agencies estimate a base case cumulative default rate and recovery rate from the originator’s historical data (static pools by vintage) and then apply stresses that increase with the rating: for a AAA, the default rate is multiplied several times and the recovery rate cut. A class receives a rating if it is repaid in full under the corresponding stress, through the cash-flow model of the waterfall. Methodologies differ by agency; on the desk, the analyst runs each agency’s stresses in the deal model and the structurer negotiates the class sizes.
Pricing follows the market. Senior European auto ABS are floating-rate notes quoted as a discount margin over one-month or three-month EURIBOR; mezzanine classes price wider in a steep curve, and the unrated first-loss class is either retained or sold at a negotiated yield. The fixed-rate pool is hedged with an interest-rate swap: the SSPE pays fixed and receives floating, on a notional that follows the expected amortisation of the notes (see C07). The originator cares about one number: the all-in cost of the notes sold, compared with its unsecured funding.
RMBS securitise residential mortgages; they are long (WAL of five to eight years), often retained by the originating bank to be used as collateral with the central bank, and placed when funding or capital relief is needed. CLOs buy a diversified pool of 150 to 400 leveraged loans (see B05), are actively managed by a CLO manager within eligibility criteria and coverage tests (overcollateralisation and interest coverage), have a reinvestment period of four to five years and an equity tranche held by the manager or by third-party funds. Unlike an auto ABS, a CLO’s pool is not originated by a single seller: the risk retention is held by the manager or a capitalised retention holder. Consumer ABS, credit card ABS, SME ABS, leasing ABS, CMBS and, in the United States, solar ABS and PACE bonds complete the family.
IN ONE SENTENCE — A securitisation can be called green because the loans in the pool finance green assets (green collateral), because the originator commits to spend the money it receives on green assets (green use of proceeds), or both.
INTUITION — A market stall can be “organic” because what it sells today is organic, or because the farmer promises to use the takings to convert his fields to organic. The first is a property of the basket; the second is a promise about the future.
Most deals labelled green in Europe combine both: a pool of green loans, and an originator commitment to reinvest the proceeds in more of the same.
Regulation (EU) 2023/2631, applicable since 21 December 2024, contains a chapter on securitisation. It is voluntary, but an issuer that uses the “European Green Bond” or “EuGB” designation must comply with all of it. For a securitisation bond:
The external reviewer must be registered with ESMA. The use-of-proceeds rules of the regulation (Articles 4 to 8) then apply: allocation to fixed assets, capital or operating expenditure, or financial assets (the originator’s new green loans are financial assets) that meet the Taxonomy’s technical screening criteria, with up to 15% in the flexibility pocket.
PITFALL — “EuGB securitisation” does not mean that the pool is Taxonomy-aligned. A EuGB auto ABS could in principle be backed by loans on petrol cars if the originator allocates the proceeds to aligned assets; Article 19 only requires the pool’s alignment to be disclosed. Conversely, a pool of 100% electric cars sold under a plain ICMA label carries none of the EuGB obligations. Read the factsheet, not the name.
The relevant activity is 6.5, transport by motorbikes, passenger cars and light commercial vehicles, which covers purchasing, financing, renting, leasing and operating M1, N1 and L vehicles. Since 1 January 2026 the substantial contribution criterion for climate mitigation requires zero tailpipe CO2 emissions for cars and vans (until the end of 2025 vehicles below 50 g CO2/km qualified). A plug-in hybrid financed in 2026 is therefore no longer aligned. The DNSH criteria add vehicle-level conditions on circular economy (reusability and recyclability under type-approval rules) and on pollution (tyres meeting rolling-noise and rolling-resistance requirements), plus a climate-risk assessment for adaptation. A battery-electric car meets the mitigation criterion by construction; demonstrating the DNSH criteria loan by loan requires data that lenders do not always collect (tyre classes, for instance). The Commission published draft revisions of the technical screening criteria on 17 March 2026 with adoption planned for the third quarter of 2026; at the date of this module the final transport criteria should be checked before any new framework is written.
The STS label is about credit simplicity, not sustainability, but it carries two sustainability provisions:
An originator of an STS auto ABS must therefore disclose CO2 data on its vehicles whether or not it calls the deal green. For an electric pool the figure is zero tailpipe emissions; the interesting information lies in the PAI templates and in the DNSH evidence.
Investor demand for green ABS comes mainly from asset managers with sustainability mandates and from bank treasuries that also report on their own green investments. Research on European auto ABS (Latino, Pelizzon, Riedel and Wang, SAFE working paper 448, 2025) finds that mutual funds with a green mandate hold more of the deals that disclose sustainability information and allocate more to deals with a higher share of electric vehicles, but show no clear preference for lower average CO2 emissions: investors use the cheapest proxy available. Rating agencies integrate ESG as credit factors, not as a label. Fitch reported in July 2025 that it had assigned ESG relevance scores to 4,691 structured finance transactions and covered bond programmes, about 82% of which had low or no ESG relevance to the rating; the 836 with elevated relevance were driven mostly by governance (asset isolation, payment interruption), then by social factors in consumer credit; environmental factors remained largely immaterial. On the desk the lesson is that “green” does not improve a rating; electric-vehicle residual values, battery technology risk and originator governance can change it.
INTERACTIVE LAB — From warehouse to closing and beyond: the life of a public green auto ABS
Follow the eighteen months from the mandate to the first annual allocation report, with the actors and the documents of each step.
| # | Who | Does what | Deliverable |
|---|---|---|---|
| 1 | Coverage banker and securitisation originator (Meridian) | Pitch a warehouse plus term ABS programme to Lumen’s treasurer; agree on a green use-of-proceeds and green collateral approach | Mandate letter and term sheet of the warehouse |
| 2 | Structured finance credit risk (Meridian) and credit committee | Approve the warehouse: advance rate, eligibility criteria, concentration limits, triggers, hold | Credit approval and warehouse facility agreement |
| 3 | Lumen origination and Meridian warehouse team | Sell eligible battery-electric vehicle loans into the warehouse SPV each month; check the borrowing base | Monthly transfer notices and borrowing base reports |
| 4 | Analytics desk and Lumen data team | Build static pool data by vintage, the loan-level tape and the cash-flow model; collect vehicle data for Taxonomy alignment | Data tape, stratification tables, deal model |
| 5 | Structurer, rating agencies | Present the originator, the data and the structure; negotiate class sizes against agency stresses | Rating committee outcome and presale reports |
| 6 | Sustainable finance advisory and Lumen treasury | Draft the green framework and the EuGB factsheet; obtain pre-issuance review from an ESMA-registered reviewer | Factsheet, pre-issuance review report |
| 7 | Legal counsel, transaction management, STS verifier | Draft the prospectus, purchase, servicing, swap and trust documents; verify STS criteria; legal opinions | Prospectus approved by the regulator, STS verification report, STS notification to ESMA |
| 8 | ABS syndicate | Pre-sound investors, announce the deal, open books with initial price thoughts, tighten, allocate | Order book, final spreads, allocation list |
| 9 | Swap desk and Lumen treasury | Execute the interest-rate swap with the SSPE on the expected amortisation profile | ISDA confirmation |
| 10 | Transaction management, paying agent, settlement | Close: the SSPE buys the pool, notes are issued, the warehouse is repaid, Lumen subscribes its 5% vertical retention | Closing memorandum, repayment notice of the warehouse |
| 11 | Lumen regulatory reporting and Meridian regulatory capital | Notify significant risk transfer to the supervisor; compute retained-position RWA; report the warehouse exit | SRT notification file, COREP securitisation templates |
| 12 | Servicer (Lumen) and cash manager | Collect, run the monthly waterfalls, publish investor and loan-level reports | Monthly investor report, ESMA templates via a securitisation repository |
| 13 | Lumen treasury and external reviewer | Allocate the proceeds to new aligned loans; publish the annual allocation report with the updated pool alignment shares | EuGB allocation report and post-issuance review |
Mandate and warehouse (months 0 to 2). The mandate letter appoints Meridian as arranger and lead manager of a future public ABS and as warehouse lender; fees are payable only on closing of the public deal, which is why the warehouse is the banker’s real commitment. The warehouse facility agreement defines eligible receivables (French retail borrowers, battery-electric vehicles, no arrears at transfer, maximum term 72 months), concentration limits (used cars, single dealer, regions), the advance rate (88%) and the triggers that stop new advances (cumulative net loss above a threshold, servicer downgrade). Credit risk analyses the originator (underwriting, collections, systems, financial strength as servicer), not only the pool; a captive’s credit is linked to the car maker’s, and its servicing continuity matters if the parent struggles.
Data and structuring (months 3 to 8). The analytics desk receives several years of static pool data, by quarterly vintage, from which base-case default, recovery and prepayment curves are derived. The loan-level tape is audited by an accounting firm on a sample (agreed-upon procedures). For the green layer, the data request extends to vehicle identification numbers, models, type-approval CO2 values and tyre data: this is the step that limits Taxonomy alignment in section 5.5.
Ratings, documents, verification (months 7 to 9). Two rating agencies are typical. The structurer negotiates the class sizes; the lawyers write the prospectus, approved by the competent authority (the CSSF or the AMF depending on where the notes are listed); the STS verifier checks some hundred criteria and the deal is notified to ESMA as STS. The ESMA-registered external reviewer gives a pre-issuance review of the EuGB factsheet.
Execution and closing (months 9 to 10). The syndicate desk pre-sounds investors with the green angle, announces the deal, opens books with initial price thoughts, tightens and allocates over two or three days. The swap is executed on pricing. At closing the SSPE pays the purchase price, the warehouse is repaid, and Lumen subscribes its retention.
After closing (years 1 to 4). Monthly: investor report, loan-level data, waterfall payments, trigger tests. Yearly: EuGB allocation report with updated Article 19 shares, reviewed if the allocation changed. Lumen’s regulatory reporting holds the SRT file; Meridian’s market-making desk quotes the notes.
WHO DOES WHAT — The structurer signs nothing alone. The warehouse goes to Meridian’s credit committee, the retained or underwritten positions to the regulatory capital team, the prospectus to Legal and Compliance, the green claims to the sustainable finance team and to Compliance for greenwashing review, and the swap to the derivatives desk with its own credit and XVA approval.
Question. What does Lumen want, and what does Meridian have to offer?
Data (case assumptions). Lumen Auto Finance SA is the French captive finance bank of Lumen Motors, a fictional European car group whose sales are moving fast towards battery-electric vehicles (BEVs). Lumen is a licensed credit institution supervised under the CRR, using the standardised approach for its retail book, with EUR 9bn of auto loans, of which BEV loans are a growing share. It finances itself with unsecured bonds (three-year senior spread about 120 bp over swaps), bank lines and a deposit platform. Its parent wants it to grow BEV lending without raising more equity.
In November 2026 Lumen’s treasurer asks Meridian for three things: fund the BEV loans it will originate over the coming months; refinance them in the public market at a lower cost than its unsecured bonds; and free capital. A fourth wish comes from the group’s sustainability team: the transaction should be green, and if possible a European Green Bond.
Meridian proposes a EUR 700m warehouse at an 88% advance rate, then a EUR 800m public STS auto ABS about nine months later. The target pool at closing:
| Pool characteristic | Value |
|---|---|
| Number of loans | 32,000 |
| Pool balance | EUR 800m |
| Average balance | EUR 25,000 |
| Weighted average customer rate (fixed) | 5.40% |
| Original / remaining term | 60 / 48 months (weighted average) |
| Vehicles | 100% battery-electric, new and used |
| Average vehicle value at origination | EUR 38,000 |
| Base case: default (CDR), recovery, prepayment (CPR) | 0.8% a year, 45%, 8% a year |
The loans are standard amortising loans with no balloon payment: the SSPE does not bear the residual value of the cars directly, only through recoveries after default. This is a deliberate choice of the structurer: a lease or balloon pool of BEVs would add residual value risk, which rating agencies stress heavily for electric vehicles.
CASE — The request combines three motives that every securitisation balances: funding (replace the warehouse with term notes), capital (transfer enough risk to release capital) and franchise (a green label to widen the investor base and support the group’s transition narrative). The structure must serve all three, and they pull in different directions.
Question. When will investors be repaid, and how much will the pool lose in the base case?
Calculation. The pool is a 48-month annuity at 5.40%. The first monthly instalment of the pool is EUR 18.569m (block c04-pool-amort in section 2.3 with the case values). Without prepayments the pool’s WAL would be 2.11 years; with a CPR of 8% it falls to 1.90 years.
Adding defaults and the waterfall (block c04-waterfall below, base case), the pool loses 1.49% of its initial balance to default over its life and, after 45% recovery, a cumulative net loss of 0.82% of the initial pool. The clean-up call is exercised in month 42, when the pool falls below EUR 80m. Because principal pays Class A first (sequential amortisation), Class A’s WAL is 1.69 years, shorter than the pool’s.
Reading. A two-year average life is what makes auto ABS attractive to bank treasuries: short, amortising, collateralised by many small, granular loans. The pricing reference for Class A is therefore the two-year part of the curve, not the legal final maturity (typically pool maturity plus a buffer of a year or more).
PITFALL — WAL depends on an assumption. The prepayment rate of BEV loans has little history; incentives to buy a newer model with a better battery could push it up, while a fall in used BEV prices could keep borrowers in their loans. The prospectus shows WAL tables at several CPRs because the investor bears this uncertainty.
Question. How should EUR 800m be cut into classes, and how much stress can each class survive?
Data. The structurer proposes six classes, sequential pay, no reserve fund, a 10% clean-up call. Spreads are those indicated by the syndicate desk at the time of structuring (case assumptions).
| Class | Size (EUR m) | Subordination | Target rating | Spread (bp) |
|---|---|---|---|---|
| A | 728 | 9.0% | AAA | 55 |
| B | 24 | 6.0% | AA | 90 |
| C | 16 | 4.0% | A | 130 |
| D | 12 | 2.5% | BBB | 200 |
| E | 8 | 1.5% | BB | 375 |
| F | 12 | 0% | Unrated | Residual |
Class F receives no coupon but all the excess spread left after the PDL is cured: it is the residual or “equity” of the deal.
FORMULA
is the credit enhancement of class ; the size of the classes junior to it; the initial pool; the cumulative net loss of the pool; the excess spread used to cover losses over the life. The second expression is a simplification: in the real waterfall excess spread arrives month by month and only covers losses that occur while it is available.
Calculation. Meridian’s analysts mimic the agencies by multiplying the base-case default rate and cutting the recovery rate from 45% to 27% (a 40% haircut) in the stressed scenarios. The waterfall below runs month by month: interest pays senior fees and coupons, then cures the PDL, and the rest goes to Class F; principal and cured amounts repay the classes in order.
INTERACTIVE LAB — The waterfall under stress: who loses what when defaults rise
Run Lumen’s pool through the monthly revenue and principal waterfalls at a chosen default multiple and recovery, and read the cumulative loss, the losses of each class and the break-even stress of each class.
| Parameter you can move | Range | Default |
|---|---|---|
| Default rate as a multiple of the base case | 1 to 15 x | 1 x |
| Recovery rate in the scenario | 0 to 70 % | 45 % |
| Base case default rate (CDR) | 0.2 to 3 % a year | 0.8 % a year |
| Prepayment rate (CPR) | 0 to 20 % a year | 8 % a year |
| Stressed recovery used for the chart and break-evens | 0 to 70 % | 27 % |
| Result at default settings | Value |
|---|---|
| Cumulative gross default (% of initial pool) | 1.49 % |
| Cumulative net loss (% of initial pool) | 0.82 % |
| Class A principal loss | 0.0 % of class |
| Class B principal loss | 0.0 % of class |
| Class C principal loss | 0.0 % of class |
| Class D principal loss | 0.0 % of class |
| Class E principal loss | 0.0 % of class |
| Class F principal loss | 0.0 % of class |
| Residual excess spread paid to Class F over the life | 18.37 EUR m |
| Class A weighted average life | 1.69 years |
| Month of the 10% clean-up call | 42 month |
| Break-even multiple of Class A (at stressed recovery) | 11.00 x |
| Break-even multiple of Class B | 8.20 x |
| Break-even multiple of Class C | 6.35 x |
| Break-even multiple of Class D | 5.05 x |
| Break-even multiple of Class E | 4.15 x |
Try it. Set the multiple to 8 and the recovery to 27%, the stylised AAA stress of the case: Classes A and B are repaid in full, the junior classes are wiped out. Then set the CPR to 0: slower amortisation keeps more interest in the deal and moves the break-evens.
The break-even multiples at a 27% recovery are 11.0x for Class A, 8.2x for B, 6.35x for C, 5.05x for D and 4.15x for E. At an 8x stress, which the case uses as a stylised AAA scenario, cumulative defaults reach 11.34% of the pool and cumulative net losses 8.27%: Classes A and B are repaid in full, Class C loses about 90% of its principal and D, E and F are wiped out. Using stylised multiples of 8x for AAA, 6.5x for AA, 5x for A, 4x for BBB and 3x for BB, each class passes its target with a cushion.
Reading. Three things carry the senior class: subordination of 9%, excess spread of more than 1.7% a year in the early months, and sequential amortisation, which makes Class A’s subordination grow as it is repaid. In a real deal the agencies would also stress prepayments, the timing of defaults (front-loaded or back-loaded), recovery delays, the swap counterparty and the servicer. The chart shows the textbook pattern: each class goes from no loss to total loss over a narrow band of stress, the “cliff” that makes thin junior tranches so sensitive to assumptions.
WHY IT MATTERS ON THE DESK — The class sizes are where the structurer creates value for the originator: every euro moved from Class B into Class A saves 35 bp of spread. The rating agencies fix how far this can go. The negotiation with the agencies on the base case and the stress multiples is the core of the structurer’s job.
Question. How much does the pool earn above its costs, and is the ABS cheaper than an unsecured bond?
Calculation. The SSPE pays the swap rate (2.50% on the flat curve) plus each class’s spread. Weighted by size, the coupons of Classes A to E cost 3.0835% of the pool. Subtracting senior fees of 0.60% (0.50% servicing paid to Lumen as servicer, 0.10% to third parties) from the pool yield gives the excess spread at closing:
is the excess spread a year as a share of the pool. In the base case it adds up to EUR 18.37m paid to Class F over the life of the deal (waterfall block) after covering net losses of 0.82% of the pool on the way.
For Lumen the relevant comparison is the all-in spread of the notes. The weighted spread of Classes A to E is 63.05 bp. Third-party fees add 10.15 bp per unit of notes and EUR 3.2m of upfront costs (Meridian’s 25 bp fee, legal, rating, verification and listing costs), spread over a notes’ WAL of 1.9 years, add 21.05 bp. The all-in cost is 94.25 bp, a saving of 25.75 bp against Lumen’s unsecured three-year spread of 120 bp.
INTERACTIVE LAB — From pool yield to excess spread and to Lumen’s all-in funding cost
Compute the weighted cost of the notes, the excess spread at closing and the all-in spread of the notes sold, and compare them with Lumen’s unsecured funding.
| Parameter you can move | Range | Default |
|---|---|---|
| Pool yield (weighted customer rate) | 2 to 10 % | 5.4 % |
| Senior fees (servicing, trustee, cash manager) | 0.2 to 1.5 % a year | 0.6 % a year |
| Swap rate paid by the SSPE (flat curve) | 0 to 5 % | 2.5 % |
| Class A spread | 20 to 150 bp | 55 bp |
| Greenium on Class A | 0 to 10 bp | 0 bp |
| of which third-party fees (trustee, cash manager, agents) | 0 to 0.5 % a year | 0.1 % a year |
| Upfront costs (arranging fees, legal, rating, verification) | 0 to 8 EUR m | 3.2 EUR m |
| Weighted average life of the notes, for amortising upfront costs | 1 to 4 years | 1.9 years |
| Lumen’s 3-year unsecured senior spread | 50 to 300 bp | 120 bp |
| Result at default settings | Value |
|---|---|
| Note coupons (Classes A to E) as % of the pool | 3.0835 % |
| Excess spread at closing | 171.7 bp a year |
| Weighted spread of Classes A to E | 63.0 bp |
| Third-party fees per unit of notes | 10.15 bp |
| Upfront costs amortised over the notes’ life | 21.05 bp |
| All-in spread of the notes (coupons, third-party fees, upfront costs) | 94.3 bp |
| Saving against unsecured funding | 25.7 bp |
Try it. Lower the swap rate to 1%: the excess spread widens because the pool rate is fixed while the note coupons float. The swap exists precisely to remove this exposure.
Reading. The servicing fee is not a cost for Lumen, which is the servicer and keeps the 0.50%; it is however a real cost for the deal, sized to pay a replacement servicer if Lumen failed. The saving of about 26 bp is real but modest: securitisation is not mainly cheaper funding for a captive with decent unsecured access; it is diversified funding, matched in maturity, and, as stage 5.6 shows, capital. Lumen also sells most of the excess spread by selling Class F: the price it obtains for that class is part of the negotiation.
Question. Which green label, what must be disclosed, and what is it worth?
The choice of label. Meridian’s sustainable finance team sets out two options to Lumen’s treasurer.
Lumen chooses the EuGB, for two reasons: its BEV origination (about EUR 1.5bn a year in the case) can absorb the allocation of the EUR 800m of proceeds to new aligned loans within twelve months, and its parent wants the strongest available label. The pool contains no fossil-fuel exposure, so the Article 18 exclusion is met by construction and explained in one paragraph of the factsheet.
Article 19 disclosure. All 32,000 loans finance activity 6.5 and are therefore Taxonomy-eligible. The substantial contribution criterion (zero tailpipe emissions) is met by every BEV. The DNSH evidence is the problem: for 4,500 loans, mostly used cars and vehicles imported by a distribution partner, Lumen cannot document the tyre data required by the DNSH criteria. Only 27,500 loans are shown as aligned:
| Article 19 metric | Share of the pool |
|---|---|
| Taxonomy-eligible activity | 100% |
| Taxonomy-aligned (evidence complete) | 85.94% |
| Eligible, alignment not demonstrated | 14.06% |
The 14.06% is not “harmful”: it is undocumented. The factsheet says so, and explains that the use-of-proceeds commitment is met separately by the allocation to new loans for which the evidence is collected at origination.
Greenium. The syndicate desk expects green-mandated investors to make Class A about 2 bp tighter than an equivalent non-labelled deal. There is no reliable public measure of a greenium in European auto ABS; the 2 bp is a case assumption, consistent with the small greeniums observed in other bond markets (see C01). Its value is simple to estimate, because it is paid on the Class A balance for as long as Class A is outstanding:
is the value of the greenium over the life (undiscounted); the greenium (2 bp); the size of Class A (EUR 728m); its WAL (1.69 years, stage 5.2).
INTERACTIVE LAB — The green layer in numbers: Article 19 disclosure, greenium and the cost of the label
Compute the pool shares disclosed under Article 19 of the EU Green Bond Regulation and compare the value of a greenium on Class A with the cost of the label.
| Parameter you can move | Range | Default |
|---|---|---|
| Loans in the pool | 10000 to 60000 number | 32000 number |
| Loans with complete DNSH evidence | 0 to 60000 number | 27500 number |
| Greenium on Class A | 0 to 10 bp | 2 bp |
| Class A weighted average life (from stage 5.2) | 1 to 3 years | 1.69 years |
| Pre-issuance external review and framework | 0 to 200 EUR k | 60 EUR k |
| Annual allocation report and review | 0 to 60 EUR k a year | 20 EUR k a year |
| Data collection and loan tagging (one-off) | 0 to 200 EUR k | 50 EUR k |
| Years of reporting | 1 to 6 years | 4 years |
| Result at default settings | Value |
|---|---|
| Share of the pool financing Taxonomy-eligible activity | 100.0 % |
| Share demonstrated Taxonomy-aligned | 85.9 % |
| Share eligible but not demonstrated aligned | 14.1 % |
| Value of the greenium over the life of Class A | 246.1 EUR k |
| Total cost of the green label | 190.0 EUR k |
| Net value of the label for Lumen | 56.1 EUR k |
| Class A spread cost over its life (55 bp) | 6,767 EUR k |
| Greenium as share of the Class A spread cost | 3.64 % |
Try it. Set the greenium to 0: the label costs Lumen EUR 190k net. That is the honest base case for many green ABS, and the reason why the decision rests on investor reach and on Lumen’s own targets rather than on price.
Reading. The label costs about EUR 190k over four years (pre-issuance review and framework EUR 60k, data collection EUR 50k, four annual allocation reports and reviews at EUR 20k), leaving a net benefit of about EUR 56k. The greenium is 3.64% of the spread Lumen pays on Class A over its life (EUR 6.77m). In other words, it is real, small, and not the reason for the label: with a greenium of zero the label costs Lumen EUR 190k net. The rationale lies elsewhere: a broader and stickier book (green funds and bank treasuries that report their green holdings), better execution in a weak market, and consistency with the group’s own transition plan. In the funding block, a 2 bp greenium lowers the weighted spread of the notes from 63.05 to 61.20 bp and the all-in cost to 92.40 bp.
PITFALL — The greenium accrues to Lumen. Meridian’s fee is unchanged; the green structuring fee is separate (stage 5.7). A banker who presents the greenium as the arranger’s achievement, or who promises it before books open, sets up a mis-selling complaint.
Question. How much capital does Lumen release, and which retention option should it use?
Before. Lumen’s loans are retail exposures under the standardised approach, risk-weighted at 75%: the pool carries RWA of EUR 600m and, at the course’s 13% capital convention, EUR 78m of capital.
The SEC-SA. After a traditional securitisation that achieves significant risk transfer, Lumen holds securitisation positions instead of loans. As a standardised-approach bank with no external-rating route chosen, it applies the SEC-SA (CRR Articles 261 and 262).
FORMULA
is the capital charge of the pool under the standardised approach (8% × 75% = 6%); the share of defaulted loans in the pool (zero at closing); the resulting pool charge; the supervisory parameter (1 for non-STS, 0.5 for STS); and , with and the attachment and detachment points of the tranche. A tranche that lies entirely below is weighted 1,250%; the floor is 15% for non-STS positions and 10% for senior STS positions.
INTUITION — Below a tranche absorbs the losses the regulator already expected the pool to bear, so it carries a full deduction (1,250%). Above capital falls off exponentially with thickness of protection, faster when is small. The parameter is a surcharge for “model risk” of securitisation; the STS label halves it.
Calculation. With = 6%, Class A (attaching at 9%) gets 15.16% in STS form, Class B (6% to 9%) 790.15%, Classes C to F, which lie below 6%, 1,250%. The whole capital structure carries RWA of EUR 900m, 1.5 times the RWA of the pool: this non-neutrality is the price the framework charges for slicing. Lumen keeps a vertical slice of 5% of every class: its RWA becomes EUR 45m, capital EUR 5.85m, and the capital released is EUR 72.15m.
INTERACTIVE LAB — Lumen’s capital relief: SEC-SA risk weights, retention options and STS
Compute the SEC-SA risk weight of each class, Lumen’s RWA before and after the securitisation with a 5% retention, the capital released and the mezzanine test for significant risk transfer.
| Parameter you can move | Range | Default |
|---|---|---|
| Transaction is STS | yes / no | yes |
| Retention option (Article 6 SECR) | vertical 5% / first loss 5% | vertical 5% |
| Standardised risk weight of the pool (retail) | 35 to 100 % | 75 % |
| Subordinated reserve loan granted by Lumen | 0 to 20 EUR m | 0 EUR m |
| Class A notes bought by an investor bank (Meridian Treasury) | 0 to 200 EUR m | 50 EUR m |
| Result at default settings | Value |
|---|---|
| KA: capital charge of the pool (8% x risk weight) | 6.00 % |
| Risk weight of Class A | 15.16 % |
| Risk weight of Class B | 790.15 % |
| Risk weight of Classes C to F | 1,250 % |
| Lumen’s RWA on the pool before securitisation | 600.0 EUR m |
| RWA of the whole capital structure | 900.0 EUR m |
| Non-neutrality: whole structure / pool | 1.50 x |
| Lumen’s RWA after securitisation (retained positions) | 45.00 EUR m |
| Capital before (13% of RWA) | 78.00 EUR m |
| Capital after (13% of RWA) | 5.850 EUR m |
| Capital released | 72.15 EUR m |
| Share of mezzanine RWA retained (SRT test: at most 50%) | 5.0 % |
| Investor bank: RWA on its Class A holding | 7.58 EUR m |
| Investor bank: capital on its Class A holding (13%) | 0.985 EUR m |
| Investor bank: LCR value as Level 2B (25% haircut, STS only) | 37.50 EUR m |
Try it. Switch the retention to ‘first loss 5%’: Lumen keeps the 1,250% layers and releases only EUR 13m of capital instead of EUR 72m. Then add an EUR 8m subordinated reserve loan with the vertical slice and watch EUR 13m of relief disappear.
Retention options compared. Had Lumen retained a 5% first-loss piece instead (Class F, E, D and half of C), it would have kept EUR 40m of 1,250% positions, RWA of EUR 500m, and released only EUR 13m. The vertical slice transfers the risk proportionally and is the natural choice for an originator seeking relief; the first-loss retention aligns incentives more strongly and is the natural choice for an originator who only wants funding. If the deal were not STS, Class A would be weighted 49.99%, the stack 2 times the pool, and the release EUR 70.2m: STS matters little for a vertical retainer’s relief, but a great deal for the investors in Class A, whose capital charge falls by two thirds. A subordinated reserve loan of EUR 8m from Lumen, a common feature, would add EUR 100m of RWA and cut the release to EUR 59.15m: the structurer removed it for that reason.
Significant risk transfer. Under the current CRR mechanical test (Article 244), risk is deemed transferred if the originator holds no more than 50% of the RWA of the mezzanine positions. Here the only mezzanine class under the SEC-SA is Class B (weighted below 1,250% and junior to the senior class); Lumen holds 5% of it. Lumen notifies the transaction to its supervisor, which may still object if the reduction in RWA is not justified by a commensurate transfer of risk. The reform agreed in October 2026 is expected to replace these tests with a principle-based assessment of the transfer of unexpected losses (see B16).
WHY IT MATTERS ON THE DESK — For a captive bank the capital relief is the main economic benefit: EUR 72.15m of capital, at any plausible cost of equity, is worth several million euros a year, against a greenium worth EUR 246k over the whole life of Class A. The green label decides who buys; capital decides whether the deal is done.
Question. Is the mandate worth Meridian’s capital?
Data (case assumptions). The warehouse is a EUR 700m commitment, outstanding nine months in the year of the deal, average drawn EUR 528m. Margin 125 bp over EURIBOR; FTP liquidity premium 12 bp (two-year committed funding at about 6 bp a year); commitment fee 35 bp on the undrawn amount; HQLA carry cost of 40 bp on a liquidity buffer of 100% of the undrawn amount, because an undrawn facility to an SSPE is treated as a full outflow by Meridian’s Treasury. Arranging and placement fee 25 bp of EUR 800m. Green structuring fee EUR 0.1m; additional green costs EUR 0.05m. Expected loss: Meridian’s structured finance rating of the senior warehouse tranche maps to grade A (PD 0.07%), with LGD 40% on the commitment. The senior warehouse tranche is a non-STS securitisation position attaching at 12%; exposure value is the full commitment (100% conversion factor for off-balance-sheet securitisation positions). Fees attract the course’s ancillary convention of EUR 2.5 of RWA per euro of revenue.
INTERACTIVE LAB — Meridian’s economics: warehouse, arranging fees and RAROC
Compute the revenues, costs, expected loss, RWA and RAROC of Meridian’s warehouse facility and arranging mandate over the year of the deal, with and without the green features.
| Parameter you can move | Range | Default |
|---|---|---|
| Warehouse commitment | 200 to 1200 EUR m | 700 EUR m |
| Average drawn amount | 100 to 1200 EUR m | 528 EUR m |
| Months the warehouse is outstanding in the year | 1 to 12 months | 9 months |
| Warehouse margin over EURIBOR | 50 to 250 bp | 125 bp |
| FTP liquidity premium (2-year funding) | 0 to 40 bp | 12 bp |
| Commitment fee on the undrawn amount | 0 to 80 bp | 35 bp |
| Attachment point of the senior warehouse tranche (1 - advance rate) | 6 to 25 % | 12 % |
| Arranging and placement fee | 0.05 to 0.5 % of notes | 0.25 % of notes |
| Green structuring fee | 0 to 0.5 EUR m | 0.1 EUR m |
| Additional cost of the green work (ESG team, data) | 0 to 0.5 EUR m | 0.05 EUR m |
| Result at default settings | Value |
|---|---|
| Net interest margin over FTP | 4.4748 EUR m |
| Commitment fees | 0.4515 EUR m |
| HQLA carry cost on the undrawn amount | 0.5160 EUR m |
| Lending revenues net of HQLA cost | 4.4103 EUR m |
| Fee revenues | 2.10 EUR m |
| Operating costs | 3.0741 EUR m |
| Expected loss | 0.147 EUR m |
| SEC-SA risk weight of the senior warehouse tranche (non-STS) | 31.35 % |
| Average RWA over the year | 169.86 EUR m |
| Allocated capital (13%) | 22.08 EUR m |
| Net profit after tax | 2.467 EUR m |
| RAROC | 11.17 % |
| RAROC without the green fee and cost | 11.22 % |
Try it. Move the attachment point from 12% to 13%: the risk weight of the senior warehouse tranche falls and the RAROC crosses the hurdle; beyond 17% the 15% floor binds and nothing more is gained. This is why the warehouse advance rate is the most negotiated number between the originator and the arranging bank.
Result. Net interest margin over FTP EUR 4.4748m, commitment fees EUR 0.4515m, minus HQLA carry EUR 0.516m: lending revenues EUR 4.4103m. Fees EUR 2.1m. Costs EUR 3.07412m (40% of lending revenues, 60% of fees, plus the green cost). Expected loss EUR 0.147m. The SEC-SA risk weight of the senior warehouse tranche is 31.35%, average RWA EUR 169.86m, capital EUR 22.08m, net profit after tax EUR 2.467m, RAROC 11.17%, just below the 12% hurdle.
Reading. The warehouse carries most of the capital and most of the revenue; the fee is earned only if the public deal closes. One percentage point of advance rate (attachment at 13%) lifts the RAROC to 12.98%; beyond 17% the 15% floor binds. Without the green fee and cost the RAROC would be 11.22%: the green work is slightly dilutive for Meridian, because the fee hardly covers the extra cost and adds RWA through the ancillary convention. Meridian accepts the deal on a relationship basis (the swap, Lumen’s deposits and its future deals are not in this RAROC) and negotiates an 87% advance rate for the next warehouse.
Question. Why does a bank treasury buy a 55 bp AAA auto ABS, and what does it cost in capital and liquidity?
Data. Meridian’s Treasury, which runs the bank’s liquidity portfolio separately from the securitisation desk, puts in an order of EUR 50m for Class A. Its investment committee requires that the notes be STS, rated at least AA-, eligible as Level 2B liquid assets and, for its green investment target, labelled under a recognised standard. Information barriers apply: the Treasury sees the same prospectus as other investors and is allocated by the syndicate desk under the same allocation policy (see C01 on allocation).
Calculation. With the capital block above, the holding is risk-weighted at 15.16% under the SEC-SA for STS positions: RWA EUR 7.58m, capital EUR 0.985m. If the deal were not STS, RWA would be EUR 24.99m and capital EUR 3.249m, and the notes would not count in the liquidity buffer at all. As senior STS auto ABS meeting the conditions of the LCR Delegated Regulation, the notes count as Level 2B assets with a 25% haircut: EUR 37.5m of liquidity value. Income is 55 bp over EURIBOR, compared with a few basis points for most Level 1 assets, which is why treasuries anchor senior auto ABS books.
Reading. STS is worth more to the investor than to the originator in this case: it divides the investor’s capital by more than three and makes the notes count for liquidity. The green label adds a third motive: the Treasury reports the holding in its green investment portfolio and, from 2027, in the bank’s own disclosures (section 6.5). The pricing benefit observed in green ABS, when there is one, comes from this kind of buyer with several reasons to hold the same paper.
PITFALL — A bank buying its own client’s deal faces a conflict of interest: allocation to the arranger’s Treasury must follow the published allocation policy and be documented, and the credit view must come from the Treasury’s own analysis, not from the structurer’s marketing.
Question. What happens after closing when the environment changes?
Scenario (case narrative, no new figures). Eighteen months after closing, a price war on new electric cars depresses used BEV prices. Repossessed vehicles sell for less: recoveries fall from 45% towards 30% on recent defaults. Two consequences follow.
For the notes. The monthly investor report shows a rising PDL, still cured each month by excess spread. The cumulative net loss trigger is not hit: the base case was 0.82% over the life, and the deal would need a much larger deterioration before even Class F lost principal (the chart of the waterfall block shows Class F starting to lose principal between two and three times the base default rate with a 27% recovery). Rating agencies publish a sector comment and place no class on watch. The Class F holders, who earn the residual, see their expected return fall: they are the ones who bear residual-value risk in an amortising-loan deal.
For the green label. Lumen’s treasury must publish its first annual allocation report. Lumen has allocated EUR 800m to new BEV loans, of which, thanks to the data collection process put in place at origination, 97% are documented as aligned; the remainder is allocated within the flexibility rules of the regulation. The report also updates the Article 19 shares of the pool: the aligned share of the remaining balance has moved because loans amortise and prepay at different speeds. The external reviewer issues a post-issuance review because the allocation changed.
Reading. In an amortising-loan ABS the green collateral itself does not change; what moves is the credit of the residual holders and the evidence behind the label. A structurer who sells the deal as green “by construction” forgets that the label is a stream of reports for four years, signed by the originator.
| Item | Value | Source |
|---|---|---|
| Pool: first instalment, WAL (CPR 8%) | EUR 18.569m; 1.90 years | c04-pool-amort |
| Base case: cumulative default, net loss | 1.49%; 0.82% | c04-waterfall |
| Class A WAL; clean-up call | 1.69 years; month 42 | c04-waterfall |
| Break-even multiples A, B, C, D, E | 11.0, 8.2, 6.35, 5.05, 4.15 | c04-waterfall |
| 8x stress: default, net loss | 11.34%; 8.27% | c04-waterfall |
| Residual paid to Class F (base) | EUR 18.37m | c04-waterfall |
| Note coupons; excess spread | 3.0835%; 171.65 bp | c04-funding |
| Weighted spread; all-in cost; saving | 63.05 bp; 94.25 bp; 25.75 bp | c04-funding |
| Taxonomy-aligned share; not demonstrated | 85.94%; 14.06% | c04-green-label |
| Greenium value; label cost; net | EUR 246k; EUR 190k; EUR 56k | c04-green-label |
| Greenium as share of Class A spread cost | 3.64% of EUR 6.77m | c04-green-label |
| RW Class A STS; Class B; C to F | 15.16%; 790.15%; 1,250% | c04-capital |
| Lumen RWA before; stack; after | EUR 600m; 900m; 45m | c04-capital |
| Capital released (vertical; first loss; non-STS) | EUR 72.15m; 13m; 70.2m | c04-capital |
| Investor holding EUR 50m Class A: RWA, capital, LCR value | EUR 7.58m; 0.985m; 37.5m | c04-capital |
| Meridian warehouse RW; capital | 31.35%; EUR 22.08m | c04-raroc |
| Meridian RAROC; at 13% attachment; without green | 11.17%; 12.98%; 11.22% | c04-raroc |
IN ONE SENTENCE — A bank that holds a tranche does not hold loans: it applies a separate capital framework that looks at the pool’s capital charge, the tranche’s position in the structure and a surcharge for slicing.
Three approaches exist (CRR Articles 254 to 264). SEC-IRBA is used by a bank with IRB permission for the type of underlying exposures and enough data to compute of the pool; SEC-SA uses the standardised charge of the pool, as in the case; SEC-ERBA maps the external rating and maturity of the tranche to a risk weight, and is available, under conditions, where the other two cannot be used. CRR3 (Regulation (EU) 2024/1623, applicable since 1 January 2025) did not rewrite this chapter but matters in two ways. First, through the output floor: an IRB bank’s total RWA may not fall below a percentage of the standardised RWA (phased in from 50% in 2025 to 72.5% in 2030), and for securitisations the standardised comparator is the SEC-SA, which is more conservative than SEC-IRBA, so IRB banks lose part of the benefit of their models. CRR3 provided a transitional reduction of the factor for this floor calculation until the end of 2032. Second, through the standardised risk weights of the pool, which feed .
For Meridian the two positions of the case are the senior warehouse tranche (non-STS, SEC-SA, 31.35%) and the Treasury’s Class A holding (STS, 15.16%). The warehouse is private: under the reform agreed in October 2026, private securitisations will keep lighter disclosure and the recalibrated floors could reduce capital for senior positions of good-quality pools; until publication of the text, the current figures apply.
All figures are outputs of block c04-raroc (stage 5.7), for the year of the deal.
| Line | EUR m | Comment |
|---|---|---|
| Net interest margin over FTP | 4.4748 | 528 × (125 − 12) bp × 9/12 |
| Commitment fees | 0.4515 | 172 × 35 bp × 9/12 |
| HQLA carry cost | −0.516 | 172 × 100% × 40 bp × 9/12 |
| Fees (arranging 2.0 + green 0.1) | 2.1 | At closing |
| Operating costs | −3.07412 | 40% lending, 60% fees, + 0.05 green |
| Expected loss | −0.147 | 0.07% × 40% × 700 × 9/12 |
| Net profit after tax (25%) | 2.467 | |
| Average RWA | 169.86 | 700 × 31.35% × 9/12 + 2.5 × 2.1 |
| Allocated capital (13%) | 22.08 | |
| RAROC | 11.17% | Hurdle 12% |
The sustainable feature changes three lines: fees (+EUR 0.1m), costs (+EUR 0.06m from the fee’s cost ratio and +EUR 0.05m of specific green work) and RWA (+EUR 0.25m through the ancillary convention). Its net effect is −0.05 points of RAROC (11.17% against 11.22%). Capital, expected loss and funding are unchanged: CRR3 contains no green supporting factor for securitisation positions (see E02), and the rating of the warehouse tranche does not depend on the label. The levers that matter are the advance rate and the STS status of the future term deal.
The warehouse is a committed revolving facility to an SSPE. Treasury charges the drawn amount the liquidity premium of a two-year commitment (12 bp in the case) and charges the undrawn amount the carry of the HQLA buffer (40 bp) on a 100% outflow assumption, which Meridian applies to undrawn facilities to SSPEs. In the case, this carry (EUR 0.516m) is larger than the commitment fee (EUR 0.4515m): undrawn warehouse lines are not free options for the bank. On the asset side, the Class A notes bought by Treasury are Level 2B HQLA (stage 5.8); within the LCR, Level 2B assets may not exceed 15% of the buffer, so the bank’s appetite for ABS in its buffer is bounded. For Lumen, the ABS replaces short warehouse funding by term funding matched to the amortisation of the pool, which improves its net stable funding ratio.
Originator (Lumen, IFRS). Two questions are asked in sequence. Consolidation (IFRS 10): does Lumen control the SSPE? It is servicer at a market fee and holds 5% of every class; it does not hold the residual (Class F is sold) and its variable returns are small. In the case, Lumen’s auditors conclude that it is an agent, not a principal, and the SSPE is not consolidated. Derecognition (IFRS 9, section 3.2): Lumen has transferred the contractual rights to the cash flows and substantially all the risks and rewards on 95% of the pool, and retains a fully proportionate 5%. It derecognises 95% of the loans and keeps the vertical slice as notes measured under IFRS 9. Had Lumen kept Class F, it would have retained substantially all the risk of the expected losses and the excess spread, and the loans would likely have stayed on balance sheet: accounting derecognition and regulatory risk transfer are separate tests with different results.
Arranger and investor (Meridian, IFRS 9). The warehouse loan and the Class A notes are tranches of contractually linked instruments. To be measured at amortised cost (warehouse, held to collect) or at fair value through other comprehensive income (Treasury’s notes, held to collect and sell), they must pass the look-through test: the underlying pool must have SPPI cash flows and the tranche’s credit risk must not exceed the pool’s. The IFRS 9 amendments applicable from 1 January 2026 clarified the characteristics of contractually linked instruments and of non-recourse features. A senior tranche over a pool of amortising consumer loans passes; an equity tranche does not and would be at fair value through profit or loss. Expected credit losses are recognised on both positions (12-month ECL in stage 1). Fees: the arranging fee is recognised under IFRS 15 when the service is complete, at closing; the commitment fee of the warehouse is part of the effective interest of the facility under IFRS 9.
Green Asset Ratio. The GAR counts Taxonomy-aligned exposures in the bank’s banking book. For Lumen, BEV loans to households are a natural GAR asset; once 95% are derecognised, they leave its GAR, while the new loans financed with the proceeds enter it. For Meridian, the warehouse loan is an exposure to an SSPE, which is not itself subject to sustainability reporting. Under Delegated Regulation (EU) 2026/73, applicable from 1 January 2026, exposures to undertakings outside CSRD scope leave both the numerator and the denominator of the GAR, but special purpose vehicles that finance groups within scope stay in. Whether an orphan securitisation SSPE falls in that carve-out is a methodology choice Meridian documents (see E06); in its base case the warehouse is treated as out of scope, so a green warehouse does not raise Meridian’s GAR. Meridian may report it in its own voluntary green-financing metrics, with that caveat (see E06).
Pillar 3 ESG. Securitisation positions appear in Meridian’s Pillar 3 templates on climate transition risk under the sector of the underlying where it can be identified; the EBA’s revised technical standards on ESG disclosures, published in final draft on 22 June 2026, are still subject to Commission adoption (see A02).
Financed emissions (PCAF). The PCAF standard of December 2025 applies a look-through to the collateral and chains attribution factors: a collateral attribution factor for each loan (current outstanding over vehicle value at origination, capped at 1), a tranche factor (tranche over deal, regardless of seniority) and an investment factor (holding over tranche). Emissions follow the money: a tranche holder reports, whether or not the originator still carries the loans. For a holding in Class A the chain simplifies to the holder’s share of the deal multiplied by the pool’s financed emissions.
INTERACTIVE LAB — Financed emissions of the pool and of Meridian’s Class A holding (PCAF look-through)
Estimate the vehicle-use emissions of the BEV pool, attribute them with the PCAF chain to the Class A holding and to Lumen’s retention, and compare with a petrol-car pool as a counterfactual.
| Parameter you can move | Range | Default |
|---|---|---|
| Vehicles in the pool | 5000 to 60000 number | 32000 number |
| Annual distance per vehicle | 5000 to 25000 km | 13000 km |
| Electricity consumption | 12 to 25 kWh per 100 km | 17 kWh per 100 km |
| Grid emission factor | 0.02 to 0.6 kgCO2e per kWh | 0.06 kgCO2e per kWh |
| Average outstanding loan | 5000 to 50000 EUR | 25000 EUR |
| Average vehicle value at origination | 15000 to 80000 EUR | 38000 EUR |
| Meridian’s holding of Class A | 0 to 200 EUR m | 50 EUR m |
| Counterfactual petrol car emissions | 80 to 200 gCO2 per km | 120 gCO2 per km |
| Result at default settings | Value |
|---|---|
| Annual emissions per vehicle (electricity) | 0.1326 tCO2e |
| Collateral attribution factor | 0.6579 ratio |
| Financed emissions of the whole deal | 2,792 tCO2e a year |
| Meridian’s share of the deal (tranche x investment factors) | 6.25 % |
| Financed emissions attributed to Meridian | 174.5 tCO2e a year |
| Financed emissions attributed to Lumen’s 5% retention | 139.6 tCO2e a year |
| Counterfactual: same pool of petrol cars (tailpipe only) | 32,842 tCO2e a year |
Try it. Raise the grid factor to 0.25 kgCO2e/kWh, closer to a coal-heavy grid: the same cars carry four times the emissions. Financed emissions of an electric pool depend on where it is charged.
Reading. With the case assumptions (13,000 km a year, 17 kWh per 100 km, a low-carbon grid factor of 0.06 kgCO2e per kWh), each car causes 0.1326 tCO2e a year; the collateral attribution factor is 0.6579; the deal’s financed emissions are about 2,792 tCO2e a year, of which 6.25%, about 174.5 tCO2e, are attributed to Meridian’s EUR 50m holding, and 139.6 tCO2e to Lumen’s retention. A pool of petrol cars of the same size would carry about 32,842 tCO2e of tailpipe emissions. That comparison is an avoided-emissions estimate: under PCAF it may be reported only separately from financed emissions, never netted. As arranger, Meridian may also report facilitated emissions under PCAF Part B for the notes it placed, a separate metric with its own weighting (see E07).
| Risk | Example in the case | Control | Owner |
|---|---|---|---|
| Pool credit | Defaults above base case | Stresses, triggers, sequential pay | Structuring, rating agencies |
| Residual value | Used BEV prices fall | Amortising loans, recovery stress | Credit risk, structuring |
| Servicer | Lumen’s parent in difficulty | Backup servicer, fee sized for replacement | Credit risk, Legal |
| Warehouse | Deal cannot be placed | Advance rate, stop triggers, term-out | Warehouse team, credit committee |
| Data tape | Errors in eligibility or vehicle data | Agreed-upon procedures, loan-level checks | Analytics, Lumen |
| Regulatory capital | SRT challenged by supervisor | Early notification, SRT file | Regulatory capital |
| Label | Alignment overstated | EuGB review, Article 19 figures | Sustainable finance, Compliance |
| Conduct | Treasury favoured in allocation | Allocation policy, records | Syndicate, Compliance |
| Swap | SSPE swap out of the money | Collateral, rating triggers, XVA | Derivatives desk |
Greenwashing red flags in green structured credit
INTERACTIVE LAB — Securitisation and green structured credit: self-assessment
Check the mechanics, the green rules and the bank-side economics of the module.
Q1. In a traditional securitisation, what does the originator receive at closing?
Answer: B. In a true sale the SSPE buys the pool with the proceeds of the notes; the originator gets cash (funding) and, if risk is transferred, capital relief.
Q2. Which is NOT one of the five retention options of Article 6 SECR?
Answer: C. Retaining only senior notes would not align incentives; the options are vertical, seller’s share, random selection, first-loss tranche and first loss on each exposure.
Q3. Under the EU Green Bond Regulation, who bears the use-of-proceeds obligations in a securitisation?
Answer: D. Article 16 reads ‘issuer’ as ‘originator’ for these obligations; the proceeds are what the originator receives from the sale.
Q4. Can a synthetic SRT transaction be issued as a European Green Bond?
Answer: A. Bonds issued for a synthetic securitisation cannot use the EuGB designation.
Q5. Since 1 January 2026, which cars meet the Taxonomy mitigation criterion of activity 6.5?
Answer: C. The 50 g threshold applied until the end of 2025; from 2026 only zero-emission cars qualify.
Q6. In the case, why is the Taxonomy-aligned share of the pool 85.94% and not 100%?
Answer: B. All vehicles are BEVs and eligible; the DNSH evidence could not be documented for 4,500 loans.
Q7. Under the SEC-SA, a tranche that lies entirely below KA receives a risk weight of
Answer: C. Such a tranche absorbs the losses the pool’s capital charge already covers, so it is weighted 1,250%.
Q8. The STS label lowers capital mainly through
Answer: A. STS halves p and lowers the senior floor from 15% to 10%.
Q9. Why did Lumen choose a vertical slice rather than a first-loss retention?
Answer: D. In the case the first-loss option releases EUR 13m against EUR 72.15m for the vertical slice.
Q10. What does excess spread do in the waterfall?
Answer: B. Revenue left after fees and coupons cures the principal deficiency ledger; only the remainder goes to the residual holder.
Q11. In the case, what is the main economic benefit of the deal for Lumen?
Answer: C. EUR 72m of released capital dwarfs the EUR 246k greenium and the 26 bp funding saving.
Q12. Why is Meridian’s warehouse RAROC below the hurdle?
Answer: A. The advance rate drives the SEC-SA risk weight; at 13% attachment the RAROC reaches 12.98%.
Q13. Under PCAF, who reports the financed emissions of a securitised pool?
Answer: D. Emissions follow the money, with a tranche factor that ignores seniority.
Q14. Does a green warehouse to an SSPE raise Meridian’s Green Asset Ratio?
Answer: B. Since 2026 such exposures are removed from the denominator and cannot be counted in the numerator.
A EUR 500m consumer ABS pays a pool yield of 7.20%, senior fees of 0.80%, and note coupons of EURIBOR (2.50%) plus a weighted spread of 90 bp on EUR 485m of notes; a EUR 15m residual class has no coupon. Compute the excess spread at closing in bp of the pool, and the number of years of a 2% annual net loss it could absorb in the first year.
Solution. Coupons: 485 × (2.50% + 0.90%) = EUR 16.49m, or 3.298% of 500. Excess spread = 7.20 − 0.80 − 3.298 = 3.102%, about 310 bp, EUR 15.51m a year. A 2% net loss is EUR 10m a year: excess spread covers it 1.55 times in the first year, before any subordination is touched. As the senior notes amortise the coupon bill falls faster than pool interest only if the deal pays sequentially; under pro rata pay the ratio stays roughly constant.
A standardised bank securitises EUR 1,000m of SME loans weighted 100% (KSA = 8%), non-STS (p = 1). Tranches: senior 12% to 100%, mezzanine 8% to 12%, junior 0% to 8%. (a) Compute the senior risk weight. (b) (b) Compute the mezzanine risk weight. (c) The bank retains a vertical 5% slice. Compute RWA before and after.
Solution. (a) a = −1/(1 × 0.08) = −12.5; u = 1 − 0.08 = 0.92; l = 0.12 − 0.08 = 0.04. K = (e^(−11.5) − e^(−0.5)) / (−12.5 × 0.88) = (0.00001 − 0.60653)/(−11) = 0.05514; RW = 12.5 × 0.05514 = 68.9%. (b) The mezzanine attaches exactly at KA: l = 0, u = 0.04, K = (e^(−0.5) − 1)/(−12.5 × 0.04) = 0.39347/0.5 = 0.78694, RW = 983.7%. The junior tranche lies below KA: 1,250%. (c) Stack RWA = 880 × 68.92% + 40 × 983.7% + 80 × 1,250% = 606.5 + 393.5 + 1,000 = EUR 2,000m. Retained: 5% × 2,000 = EUR 100m against EUR 1,000m before: RWA falls by EUR 900m, EUR 117m of capital at 13%. Note the non-neutrality: with p = 1 the stack carries twice the pool’s RWA.
A EUR 1,200m green RMBS has a Class A of EUR 1,080m with a WAL of 5.5 years. The greenium is 1.5 bp. The label costs EUR 120k upfront and EUR 25k a year for six years. Is the label worth it on price alone? What greenium would break even?
Solution. Value of the greenium = 0.00015 × 1,080 × 5.5 = EUR 891k (undiscounted). Cost = 120 + 6 × 25 = EUR 270k. Net EUR 621k: on a long RMBS a small greenium pays for the label, unlike the short auto ABS of the case. Break-even greenium = 270 / (1,080 × 5.5 × 1,000) × 10,000 = 0.45 bp. Long WAL multiplies the value of each basis point; that is also why green labels are more common on RMBS than on short consumer ABS.
1. Walk me through a securitisation waterfall. Collections are split into revenue and principal. Revenue pays senior fees and the swap first, then interest class by class from the top, then cures the principal deficiency ledger, then tops up any reserve, and the rest goes to the residual holder. Principal, with recoveries and cured amounts, repays the notes sequentially or pro rata depending on triggers. The order protects the senior notes: losses climb from the bottom, cash runs from the top.
2. What makes a securitisation “green”? Either the pool finances green assets, or the originator commits to allocate the proceeds to green assets, or both. The EU Green Bond Regulation follows the second route: the originator is responsible for allocation and reporting, and the pool’s alignment is only disclosed. A pool of electric-car loans under an ICMA label and a EuGB securitisation are therefore different claims, and an investor must read the factsheet.
3. Why does STS matter if the originator keeps a vertical slice? For the originator with a vertical slice the capital released barely changes (EUR 72.15m against 70.2m in our case). STS matters to investors: it divides the capital charge on a senior tranche by about three and makes it eligible for the liquidity buffer of a bank. That widens demand and tightens the senior spread, which is where most of the funding cost sits.
4. A client asks you to label its SRT deal green. What do you say? A synthetic securitisation cannot be a European Green Bond. Under voluntary market standards the bank could commit to use the freed capital for green lending, but capital is fungible and the commitment is hard to verify. I would suggest disclosure of the reference pool’s characteristics and a separate, verifiable green lending target rather than a label on the trade.
5. How do you size Class A? I start from the base case defaults, recoveries and prepayments derived from the originator’s static pools, apply each rating agency’s AAA stress through the cash-flow model, and find the largest Class A that is repaid in full. Excess spread, sequential amortisation, triggers and the swap all enter. Then I check investor demand and the capital treatment, because a larger Class A lowers the cost of funding but leaves thinner mezzanine tranches.
6. What is non-neutrality in securitisation capital? The sum of the capital charges of all tranches exceeds the capital charge of the pool. Under the SEC-SA with p = 1 the whole stack costs about twice the pool, with p = 0.5 about 1.5 times. It reflects model and agency risk, and it is the main target of the 2025 to 2026 reform.
7. Is there a greenium in green ABS, and should the client care? Evidence is thin; where it exists it is a few basis points, similar to other bond markets. In our case 2 bp on Class A is worth about EUR 246k, against EUR 190k of label costs and EUR 72m of capital released. The client should care about the investor base and the credibility of its reporting, not about the greenium.
8. How does a falling used-EV price affect an auto ABS? In a lease or balloon deal the SSPE bears residual value directly, so losses at maturity rise and rating agencies apply specific haircuts. In an amortising-loan deal it hits recoveries after default, so net losses rise and the residual holder’s excess spread falls first. Triggers and subordination protect senior notes; the junior classes and the originator’s retention take the hit.
Regulation and supervisory texts
Standards and market sources
Academic