Work
Individual ProjectAsset-Backed Securities · Modelling · Scenario Analysis

Renewable Energy ABS Waterfall Engine

ABS means asset-backed securities. Cash from 2,500 loans is distributed through a waterfall: senior notes are paid first, while junior notes absorb losses first.

Central findingThe structure protects senior notes in the named stress scenario, but junior losses accelerate once collateral damage exhausts the available protection.

Evidence

The result in context

Synthetic renewable-energy loan pool
£100m
UK solar loans modelled
2,500
Validation tests
52
Stress net collateral loss
26.43%
Class C is fully impaired; Class B loses 34.73%; Class A remains whole.
On this page

Question

How do defaults, prepayments, recoveries, reserves and structural triggers transmit through a securitisation capital structure?

A deterministic cash-flow and priority-of-payment model for a synthetic portfolio of UK solar loans.

What is this?

A loan pool translated into investor cash flows

The engine models defaults, voluntary prepayments, lagged recoveries, reserves and overcollateralisation across a £100m synthetic pool. Cash is then allocated to Class A, B and C notes in strict priority order.

Overcollateralisation, or OC, is the excess of collateral over note balance. It acts as protection until losses and structural leakage consume it.

Architecture

Collateral first, capital structure second

Loan-level events roll into monthly collateral cash flows. The waterfall then applies fees, interest, principal, reserve movements and trigger rules before updating each note balance.

This separation makes it possible to validate the collateral mechanics independently from the payment waterfall.

Interactive evidence

Collateral amortises across the 239-month base case

Ending collateral balance, £m.

Read: The pool pays down over time; the capital-structure charts below show how that cash is allocated rather than creating a second pool balance.

Data table · 239 verified rows
Representative preview rows; the full 239-row local dataset loads with the interactive chart.
DateEnding Balance GbpEnding Balance Gbp M
2026-07-3198,818,658.3822798.818658
2026-08-3197,647,789.6467397.64779
2028-09-3071,525,966.90542171.525967
2030-12-3149,160,937.27300249.160937
2033-02-2832,197,919.0665932.197919
2035-05-3118,384,923.40454418.384923
2037-07-318,994,612.3639848.994612
2039-10-313,298,094.9622753.298095
2041-12-311,359,855.4088581.359855
2044-03-31246,205.5880860.246206
2046-04-3000
2046-05-3100

Interactive evidence

Sequential paydown protects the junior notes until losses arrive

Base-case ending note balance, £m.

Read: Class A amortises first, followed by Class B and Class C under the modelled priority of payments.

Data table · 239 verified rows
Representative preview rows; the full 239-row local dataset loads with the interactive chart.
DateClass AClass BClass C
2026-07-3168.906561128
2026-08-3167.822544128
2028-09-3041.784784128
2030-12-3119.338835128
2033-02-282.314711128
2035-05-3100.4521068
2037-07-31000
2039-10-31000
2041-12-31000
2044-03-31000
2046-04-30000
2046-05-31000

Results

Loss location matters as much as total loss

In the downside case, 9.57% net collateral loss produces no note loss. In the named stress case, 26.43% net collateral loss fully impairs Class C and removes 34.73% of Class B principal, while Class A remains whole.

The point is not that senior notes are always safe. It is to make the sequence and boundary of protection visible under a defined scenario.

Interactive evidence

Waterfall scenario explorer

Change the collateral scenario to see which layer absorbs the loss.

Collateral performanceScheduled performance
Loss outcomeNo note impairment
ReserveAvailable
OC protectionMaintained
TriggersNot breached
Class AWhole
Class BWhole
Class CWhole

Cash follows the normal priority of payments; all note principal remains intact.

Interactive evidence

Stress losses are concentrated in the junior capital

Principal loss as a percentage of original tranche balance.

Read: Under stress, Class C is fully written down and Class B loses 34.73%; Class A remains whole in the tested scenarios.

Data table · 3 verified rows
Complete dataset
ScenarioClass AClass BClass C
Base000
Downside000
Stress034.733451100

Interactive evidence

Class C absorbs the first material loss combinations

Class C principal loss across annual default and recovery assumptions.

Read: The loss surface makes the interaction explicit: higher defaults and lower recoveries push the junior tranche rapidly toward a full write-down.

Data table · 25 verified rows
Representative preview rows; the full 25-row local dataset loads with the interactive chart.
Annual Cdr PctRecovery Rate PctLoss Pct
1.5300
1.5450
1.5750
3300
3750
5450
5750
74521.239346
7900
94579.001667
9750
9900

Validation

The waterfall is checked as an accounting system

Fifty-two automated tests cover cash conservation, allocation order, balance evolution and scenario outputs. The scenario explorer uses the model’s fixed base, downside and stress outputs.

Limitations

What this evidence does not establish

  • The collateral pool is synthetic and is not evidence of realised renewable-energy loan performance.
  • The displayed scenarios are deterministic model states, not forecasts of a marketed security.

Source and reproducibility

Trace the evidence

Source code, evaluation outputs and supporting material are available in the repository.

View repository
  1. Tranche balancesdata/processed/base_tranche_cashflows.csvCommit / evidence ID: 526f4e32e8e6e07c17a328975b0de0ae3a8888de
  2. Tranche lossesdata/processed/tranche_scenario_analytics.csvCommit / evidence ID: 526f4e32e8e6e07c17a328975b0de0ae3a8888de
  3. Collateral and sensitivity outputsdata/processed/loss_sensitivity.csvCommit / evidence ID: 526f4e32e8e6e07c17a328975b0de0ae3a8888de