Financed emissions from motor vehicle loans.
Every vehicle loan, lease and hire purchase agreement in the book carries the same methodological requirement. The method is fixed. The work is in the loan file.
Updated October 2026 · 9 min readWhat the section requires.
Section 5.6 covers loans, leases and hire purchase agreements where the financed asset is a vehicle. The bank reports the share of each vehicle's annual emissions that its money represents.
A book of two hundred vehicles and a book of twenty thousand face the same requirement. What changes is how much of the book is measured in the first period, and that coverage is disclosed.
- The formula
- Attribution factor × vehicle annual emissions
- Attribution factor
- Outstanding balance ÷ vehicle value at origination, capped at 1.0
- Boundary
- Tank-to-wheel for petrol and diesel. Well-to-wheel for battery electric and plug-in hybrid vehicles.
- Period alignment
- The balance and the vehicle data have to describe the same reporting period.
- Coverage
- The share of the vehicle book measured, reported with the plan to widen it.
The attribution factor, worked through.
One vehicle, one reporting date. Figures are shown in the local currency unit.
| Vehicle value at origination | 3,200,000 |
| Outstanding balance at reporting date | 1,600,000 |
| Attribution factor | 1,600,000 ÷ 3,200,000 = 0.50 |
| Vehicle annual emissions | 2.8 tCO2e |
| Financed emissions, bank share | 0.50 × 2.8 = 1.4 tCO2e |
The factor declines as the loan amortises. A facility that is fully repaid has a factor of zero and leaves the portfolio, which is why the calculation is repeated at each reporting date rather than stored once.
The data behind each score.
Three measurement options, each mapped to a band on the data quality scale.
| PCAF option | Score | What you hold | What it means |
|---|---|---|---|
| Option 1 | 1 to 2 | Reported fuel use, fuel cards or telemetry | The best evidence a vehicle book can hold. Still uncommon for retail lending. |
| Option 2 | 3 | Make, model, year and VIN | A type-approval factor stands in for measured fuel use. |
| Option 3 | 4 to 5 | Fuel type and vehicle category | A category average. Allowed, provided the proxy and its reason are recorded. |
Most vehicle books start at score 4 or 5, because the balance is reliable and the vehicle detail is patchy. That is a governed starting point, not a failure. What a reviewer penalises is an undocumented score, not a low one.
Mapping the fields you already hold.
The extract is the same exercise every bank runs: pull the vehicle book, then join the fields that decide the score.
Balance and origination value
The outstanding balance comes from the core banking system. The vehicle value at origination usually sits in collateral records rather than the main loan table, which is the first join most teams have to build.
Vehicle identity
Make, model, year and VIN decide whether a type-approval factor can be used at all. A VIN that decodes cleanly moves an exposure from a category average to a vehicle-specific figure.
Fuel and distance
Fuel type is needed for every option. Annual distance travelled is the field most often missing; where it is absent, a published distance proxy is applied and the score reflects that choice.
What a missing field costs
No origination value means the exposure falls back to an economic proxy. No vehicle detail means a category average. The score is not a judgement on the bank; it is a description of the evidence behind that one line.
Vehicles the global factor sets miss.
The common published sets were built around European and American fleets. Several vehicle types that dominate East African lending are absent.
| Vehicle | Where | Annual distance | What has to be handled |
|---|---|---|---|
| Matatu | Kenya, Uganda, Tanzania | ~72,000 km | Not covered by the UK or US factor sets. Needs a regional factor. |
| Bodaboda | Kenya, Uganda, Tanzania, Rwanda | ~25,000 km | High volume in the book. The factor moves with engine size. |
| Tuk-tuk | Kenya, Tanzania, Ethiopia | ~20,000 km | Petrol and LPG units behave differently, so fuel type has to be recorded. |
| Minibus taxi | South Africa | ~65,000 km | High annual distance. Load weight shifts the real-world figure. |
Any regional adjustment has to name its source, be versioned and frozen at approval, and carry the score impact it causes. An adjustment without a source is a guess with a decimal point.
Where electric vehicles change the boundary.
The tailpipe figure. Fuel combustion only, converted with a dated fuel factor.
Distance and fuel consumption are usually the only variables that move.
The grid's emissions are counted, using an electricity factor for the same period.
A plug-in hybrid without charging or fuel records is flagged as uncertain and sent for review rather than split by assumption.
That reduction is measured on its own ledger. It is never subtracted from the financed emissions total, because the two answer different questions.
From a proxy to a meter reading.
Improvement is part of the disclosure. Each year's score is explained by the data that became available, and earlier periods are left as they were reported.
| Period | Expected score | What changes |
|---|---|---|
| Year 1 | 4 to 5 | Baseline set. Governance in place. Most exposures measured on a proxy, with the reason written down. |
| Year 2 | 3 to 4 | VIN enrichment from origination records and the vehicle registry. Type-approval factors replace fleet averages. |
| Year 3 | 2 to 3 | Telemetry partnerships for high-value fleets. Published factors replace internal working assumptions. |
The same layer carries the next asset class.
The ingestion schema and the factors change per asset class. The approval, the record and the replay do not. That is why motor vehicles are a sensible first step rather than a side project.
Questions we get asked.
Does a small vehicle book need to be measured?
Yes. Section 5.6 sets one method regardless of book size. A small book can be measured in full sooner, and the coverage percentage is disclosed either way.
What happens when the loan is repaid?
The attribution factor falls to zero and the vehicle leaves the portfolio. That is why the factor is recalculated at each reporting date rather than fixed when the loan is written.
Can a category average be used for every vehicle?
It can be used, but every exposure that relies on it carries a lower data quality score and a written reason. The score is disclosed alongside the figure, not hidden in a workbook.
How are electric vehicles treated?
The boundary moves from tank-to-wheel to well-to-wheel, so the grid's own emissions are counted. Plug-in hybrids need usage data to split electric and combustion distance; without it the exposure is flagged as uncertain.
Put one vehicle book in front of the engine.
Send the extract you already have. We return the mapped exposures, a score for each one and the record behind the total.