Residential Mortgage Financed Emissions
Step-by-step method for calculating financed emissions from residential mortgage portfolios under the PCAF Global GHG Accounting and Reporting Standard. Covers the attribution factor, the five-level data quality score hierarchy, EPC and energy-label estimation, the property-value-at-origination lock, remortgage resets, portfolio weighting, and worked examples. Aligned to the PCAF Standard Part A and the GHG Protocol Scope 3 Category 15 framework.
A mortgage book is millions of small exposures, each secured against a home whose energy use the lender almost never directly measures. Financed-emissions accounting still demands a number for every one — a fraction of each dwelling’s annual carbon, scaled by how much of the home’s origination value the bank still finances, and graded on how little or much is actually known about that home.
At portfolio scale, the quality of your estimate matters as much as the estimate itself.
PCAF residential mortgage financed emissions = attribution factor × dwelling emissions. Attribution is the outstanding mortgage balance divided by the property value at origination. Dwelling emissions are the home’s Scope 1+2, graded on a 1–5 data quality score.
A methodology page is the execution layer. It takes the concept from the glossary and the rules from the PCAF Financed Emissions Standard and tells you exactly how to translate a residential mortgage into a reported tonnage. If you already know your outstanding balance and the property value at origination, jump to the formula or the worked examples.
Apply this methodology to your own portfolio — open the PCAF Mortgages Calculator.
When to Use This Methodology
Mortgages is Asset Class 5 in the PCAF Standard. It applies to loans secured against residential property occupied or let as dwellings — owner-occupier mortgages, buy-to-let, and residential portions of mixed-use lending. The attribution logic is identical to commercial real estate, but the inputs are residential: residential energy-intensity benchmarks, dwelling-type defaults, and energy-label data such as EPCs. Confirm the asset class before you calculate, because the same building must never appear in two classes.
Pre-check — is mortgages the right asset class?
PCAF defines distinct attribution rules and datasets per asset class. Route the exposure before you calculate.
Use this methodology when
The exposure is a loan secured on a residential dwelling, the property is identifiable, and you can establish (or reasonably estimate) the property value at origination. This covers owner-occupier mortgages, buy-to-let mortgages, and the residential share of mixed-use lending.
Do not use this methodology when
The property is commercial (use commercial real estate), the financing is for construction (use project finance), or the loan is general-corporate to a landlord (use business loans). A mixed-use building needs a documented split so its residential and commercial shares land in the right classes without double-counting.
Provider directory
Need this ready for your auditor?
See who does this work. Every listing names the standards it works to, and paid placements are labelled. Including Guidehouse and SE Advisory Services.
Browse 3 financed emissions providers →Do this work? A listing is US$390 a year. Get listed →
Step 1 — The Two-Part Financed Emissions Formula
Every PCAF financed-emissions calculation is the product of two independent quantities: how much of the asset you finance (the attribution factor) and how much the asset emits (the dwelling emissions). Keep the two parts separate — the attribution factor is a finance number, the dwelling emissions are a carbon number, and they are sourced, verified, and graded differently.
Part A — the attribution factor
The attribution factor is the share of the dwelling’s emissions that belongs to your institution: the outstanding mortgage balance divided by the property value at the time the mortgage was originated.
Part B — dwelling emissions × attribution
Financed emissions are the attribution factor applied to the dwelling’s annual Scope 1 and Scope 2 emissions. The home’s emissions are calculated independently of the mortgage, then apportioned.
A financed home’s Scope 1+2 becomes part of the lender’s Scope 3 Category 15 (investments). The scope label does not transfer — what is Scope 1 to the homeowner (their gas boiler) is Scope 3 to the lender. For a retail bank with a large mortgage book, Category 15 financed emissions routinely dominate the entire group inventory.
Step 2 — The Data Quality Score Hierarchy (Options 1–3, Scores 1–5)
PCAF requires the best data available and a disclosure of how good it is. For mortgages the standard defines three calculation options mapping to a five-level data quality score, where Score 1 [GreenCalculus reporting.pcaf.mortgages.score_1 · PCAF 2024 · v2026.284] is the highest quality and Score 5 [GreenCalculus reporting.pcaf.mortgages.score_5 · PCAF 2024] is the lowest. Most mortgage books begin at the lower end and climb as EPC and energy data are matched to the loan tape.
| Score | PCAF option | Required input | Emission factor basis | Typical source |
|---|---|---|---|---|
| 1 | Option 1a | Actual energy (kWh, m³ gas) | Supplier-specific / verified | Smart-meter / utility data with consent |
| 2 | Option 1b | Actual energy (kWh, m³ gas) | Average (grid / DEFRA) | Utility data + national grid factors |
| 3 | Option 2a | Floor area + EPC band + location | Average (grid / DEFRA) | EPC register matched to loan tape |
| 4 | Option 2b | Partial attributes / estimated floor area | Average | EPC band only, or estimated floor area × benchmark |
| 5 | Option 3 | Dwelling type / count only | Average dwelling-type default | Property register, dwelling-type average |
Score values resolve live from the MasterBrain PCAF dataset (PCAF Global GHG Accounting and Reporting Standard for the Financial Industry, Part A). These are data quality ranks, not emission factors — a Score of 1 means highest data quality, not 1 tCO₂e.
Most mortgage portfolios start at Score 4 because the loan tape carries an address but no energy data. Matching that address to a national EPC register moves the loan to Score 3 — a large, cheap data-quality gain across an entire book, without contacting a single borrower. Smart-meter data (Score 1–2) requires consent and rarely scales, so EPC matching is where most portfolios concentrate effort.
Converting dwelling energy to emissions uses the same factors as any inventory — open the Scope 2 electricity calculator for the grid step, or the natural gas combustion methodology for the heating step.
Step 3 — Attribution Factor Mechanics
The attribution factor is simple as an equation and easy to misapply in a mortgage system. Two rules govern it.
Property value at origination — the lock
The denominator is the property value at origination, and it stays fixed for the life of the mortgage. It is not re-marked when house prices rise, and not re-marked when they fall. This is the most misapplied rule in mortgage financed emissions, because lending systems hold an indexed current valuation for capital and LTV purposes and it is tempting to reuse it. Reusing it lets the attribution factor drift with the housing market every year for reasons unrelated to the loan or the home’s carbon.
If the property value at origination was 400,000 and the home later indexes to 500,000, the attribution denominator is still 400,000. Re-marking to 500,000 would silently cut your reported financed emissions by 20% with no change in the loan or the home — an artefact, not a reduction. The numerator (outstanding balance) moves with repayment; the denominator does not move at all unless a genuine remortgage resets it.
Repayment, remortgage, and the >100% cap
As the borrower repays, the outstanding balance falls and the attribution factor falls with it — correctly, because your institution finances a shrinking share of the home. Three mechanics matter:
Step 4 — Calculating Dwelling Emissions
Dwelling emissions are the home’s own Scope 1 and Scope 2, calculated as the homeowner would calculate them — then attributed. The lender does not invent a new boundary; it borrows the home’s boundary and takes a slice.
| Energy source | Dwelling scope | Converted with | Notes |
|---|---|---|---|
| Mains gas heating / hot water | Scope 1 | Gas combustion factor | The dominant Scope 1 source in most housing stock. See the natural gas combustion methodology |
| Heating oil / LPG (off-grid) | Scope 1 | Fuel combustion factor | Material for rural and off-gas-grid dwellings |
| Purchased electricity | Scope 2 | Grid emission factor | See the Scope 2 electricity methodology. Rising share as heat pumps replace boilers |
| District / communal heat | Scope 2 | Heat network factor | Common in flats; use the network operator’s factor where available |
Location-based vs market-based for the dwelling
The location-based vs market-based choice applies to the dwelling’s electricity. Most homes are on a standard grid tariff, so location-based is the realistic default across a mortgage book; a market-based treatment is only justified where the household holds a contract that meets the market-based criteria. The trap is the same as in commercial real estate — mixing methods across dwellings and summing them into one portfolio total that is not comparable.
The dwelling-emissions step is ordinary corporate carbon accounting at household scale. The grid factor that converts a home’s kWh to tCO₂e is the same factor a Scope 2 inventory uses. Do not build a parallel mortgage-specific energy model — feed dwelling energy (measured or EPC-estimated) through the established Scope 1 and Scope 2 methodologies and attribute the result.
Step 5 — Worked Examples
Four complete calculations. Each shows the inputs, the formula, and the result, and each reconciles. All emission factors and EPC-band intensities are fixed with stated sources so the arithmetic can be re-verified against the figures that applied on the review date — worked examples are audit records, not live references.
Outstanding balance: 240,000 | Property value at origination: 400,000
Actual dwelling energy: 4,200 kWh electricity · 13,500 kWh gas Energy from consented smart-meter data (Option 1a). Electricity factor 0.20705 kg CO₂e/kWh; gas factor 0.18254 kg CO₂e/kWh (illustrative average factors, fixed for this worked example).
Outstanding balance: 300,000 | Property value at origination: 375,000
Floor area: 95 m² · EPC band: D · band-D intensity: 180 kWh/m²/yr Option 2b: EPC band and floor area known, energy estimated from a band-specific intensity. Band-D intensity 180 kWh/m²/yr and factor 0.20705 kg CO₂e/kWh fixed as illustrative figures for this worked example. Scored 4 here because floor area is partially estimated from the EPC record.
Original origination value: 400,000 → remortgage origination value: 460,000 (RESET)
Outstanding at remortgage: 240,000 · dwelling emissions unchanged: 3.3339 tCO₂e A genuine remortgage is a new origination event, so the denominator resets to the value at the remortgage date — unlike a partial repayment, which leaves the denominator alone.
Asset A (metered): 2.00 tCO₂e, score 1 | Asset B (EPC estimate): 2.83 tCO₂e, score 4 | Asset C (remortgaged): 1.74 tCO₂e, score 1 PCAF requires the portfolio data quality score to be weighted by financed emissions, not averaged across loan count.
| Mortgage | Financed tCO₂e | Score | tCO₂e × score |
|---|---|---|---|
| A — metered | 2.00 | 1 | 2.00 |
| B — EPC estimate | 2.83 | 4 | 11.32 |
| C — remortgaged | 1.74 | 1 | 1.74 |
| Total | 6.57 | — | 15.06 |
Step 6 — Edge Cases and Governance
The formula is stable; the boundary decisions around it are where mortgage portfolios diverge. Five recurring situations need a documented policy before they appear at scale.
- Buy-to-let and multi-dwelling exposures. Each financed dwelling is its own asset with its own attribution. A single loan secured against several units is decomposed per dwelling where data allows, or treated with a portfolio-of-dwellings estimate at a lower data quality score.
- EPC age and reliability. An EPC is a point-in-time estimate that can be a decade old and may predate a retrofit. Document the EPC vintage policy; a stale label can mis-estimate energy in either direction.
- Mortgage vs commercial real estate double-counting. A dwelling must sit in exactly one asset class. Mixed-use buildings with both residential and commercial financing require a documented split rule so no tonne is counted twice and none is dropped.
- Data quality improvement over time. Matching EPCs or onboarding meter data moves loans up the ladder and will usually change reported tonnage. Disclose score migration alongside emissions change so a reader does not mistake a data-quality improvement for real-world decarbonisation.
- Heat-pump and electrification transitions. A dwelling switching from a gas boiler to a heat pump moves emissions from Scope 1 to Scope 2 and changes the grid-factor sensitivity of the estimate. Refresh the energy basis when a known retrofit occurs rather than carrying the pre-retrofit label.
The most expensive portfolio error in financed emissions is double-counting a property across the mortgages book and the commercial real estate book, or against a corporate loan to the same landlord. Maintain a single asset-to-class mapping table and reconcile it every reporting cycle. A property that appears in two classes inflates the institution’s total and fails verification under the PCAF Standard.
Step 7 — Reporting Metrics and Disclosure Logic
Mortgage financed emissions are reported as more than a single tonnage. PCAF and the frameworks built on it expect a set of complementary metrics, each answering a different question.
Absolute financed emissions
The headline tCO₂e attributed across the mortgage portfolio. Answers “how much carbon does this book finance?” Required by PCAF and the basis for financial-sector target tracking.
Economic emissions intensity
Financed tCO₂e per million of outstanding (tCO₂e/$M or /€M), or per dwelling. Lets you compare books of different sizes and track intensity decline independent of book growth.
Data quality score (weighted)
The emissions-weighted portfolio score from Step 5, Example 4. A mandatory PCAF disclosure that travels with the tonnage. For mortgages, the EPC-match rate is the leading indicator of where this score will move.
The data quality score is not optional colour — it is a required field. A reported financed-emissions figure without its accompanying weighted data quality score is incomplete under the PCAF Standard. Where the institution also reports under CSRD ESRS E1, the mortgage financed-emissions line feeds the Scope 3 Category 15 disclosure, and the EPC-distribution narrative supports the ESRS data-quality and estimation-uncertainty requirements. Many lenders pair the tonnage with an EPC-band distribution of the book, which doubles as a transition-risk indicator.
The spread of EPC bands across a mortgage book tells two stories at once: how confidently emissions are estimated, and how exposed the book is to tightening minimum-energy-efficiency regulation. A book weighted toward low EPC bands faces both a data-quality challenge and a forward regulatory and collateral-value risk. Report the distribution alongside financed emissions where the framework allows.
Error Traps — With Calculable Magnitudes
Each error below produces a specific, quantifiable distortion. Magnitudes use the Example 1 dwelling (240,000 outstanding, 400,000 origination value, 3.3339 tCO₂e dwelling footprint) unless stated otherwise.
| Error | What happens | Magnitude | How to avoid |
|---|---|---|---|
| Re-mark denominator to indexed value | Attribution denominator uses indexed current value instead of origination value. | Understates by 20% at +25% indexation 240k ÷ 500k = 0.48 vs correct 0.60; financed emissions 1.60 vs 2.00 tCO₂e. |
Lock the denominator at the origination value. Reset only on a genuine remortgage. |
| Loan-count-weighted portfolio score | Portfolio data quality score is a simple mean across loans, not emissions-weighted. | Misstates score by 0.29 (2.00 vs 2.29) See Example 4 — the largest-emitting dwelling is also the lowest quality. |
Weight the score by financed emissions, per PCAF. |
| Double-count across asset classes | A property sits in both the mortgages and commercial real estate books. | Inflates total by the full asset The whole dwelling footprint counted twice — a verification failure. |
Maintain one asset-to-class mapping table; reconcile every cycle. |
| Reset denominator on a product transfer | A rate switch on the same facility is wrongly treated as a remortgage. | Spurious attribution change Denominator moves with no new origination — fabricated emissions movement. |
Reset only for a genuine new facility; product transfers keep the original denominator. |
| Carry a stale EPC after retrofit | Pre-retrofit EPC band drives the energy estimate after the home is upgraded. | Mis-estimates dwelling energy A band-F label on a now-band-C home overstates; the reverse understates. |
Refresh the energy basis on known retrofits; document the EPC-vintage policy. |
| Omit the data quality score disclosure | Absolute tonnage reported without the weighted score. | Disclosure incomplete Figure is uninterpretable; fails PCAF and weakens CSRD ESRS E1 alignment. |
Report the emissions-weighted score alongside every absolute figure. |
| Forget the >100% cap | Outstanding exceeds origination value and attribution is reported above 1.0. | Overstates beyond the whole dwelling Attribution of 1.15 reports 115% of a footprint you cannot finance. |
Cap attribution at 100%; investigate the underlying valuation. |
What the Calculator Handles vs What You Decide
A financed-emissions calculator automates the mechanical steps. This methodology page covers the upstream judgements a calculator cannot make for you.
- Attribution factor: outstanding ÷ origination value, with the 100% cap applied
- Energy-to-emissions conversion using grid and fuel factors
- EPC-band and floor-area energy estimation
- Data quality score assignment from the inputs you actually supply
- Financed emissions = attribution × dwelling Scope 1+2
- Portfolio roll-up with the emissions-weighted data quality score and EPC distribution
- Asset class: is this genuinely a mortgage, or commercial real estate / project finance / business loan?
- Origination value: have you captured the value at origination, not the indexed current mark?
- Best available data: is there meter data (Score 1–2) or an EPC match (Score 3–4) before you fall back to a dwelling-type default (Score 5)?
- Remortgage events: any genuine remortgages that should reset the denominator — and any product transfers that should not?
- Double-count check: does this property appear in any other asset class?
- EPC vintage: is the label recent and post-retrofit?
Methodology Metadata — For Financed-Emissions Documentation
Copy into your financed-emissions methodology statement for audit transparency. The data quality option and Scope 2 method lines are selected per asset.
Across the asset classes. This is one of ten PCAF methods. For which class applies to a given exposure, and how the attribution denominator changes between them while the equation stays the same, see PCAF Financed Emissions.
Frequently Asked Questions
Financed emissions equal the attribution factor multiplied by the dwelling’s Scope 1+2 emissions. The attribution factor is the outstanding mortgage balance divided by the property value at origination. The dwelling’s emissions are its on-site combustion (Scope 1, typically gas heating) and purchased electricity and heat (Scope 2), calculated as the homeowner would and then apportioned. The reported figure must carry a data quality score from 1 (actual energy) to 5 (dwelling-type estimate).
An EPC (or equivalent energy label) lets you estimate a dwelling’s energy from its band and floor area, moving the loan to Option 2 — typically Score 3 or 4 depending on how much of the floor area and attributes are confirmed. Matching addresses on the loan tape to a national EPC register is the fastest, cheapest way to lift a mortgage book’s data quality, because it requires no borrower contact. Smart-meter data would reach Score 1–2 but needs consent and rarely scales, so most portfolios concentrate on EPC matching first.
PCAF fixes the denominator at origination so the attribution factor reflects the financing relationship, not house-price movements. Lending systems hold an indexed current valuation for capital and LTV purposes, but reusing it would make reported financed emissions rise and fall with the housing market for reasons unrelated to the loan or the home’s carbon. The outstanding balance in the numerator does move — it falls with repayment — but the denominator stays locked unless a genuine remortgage creates a new origination event.
Both use the same attribution logic — outstanding amount over property value at origination — but they are separate asset classes with separate datasets. Mortgages covers residential dwellings and uses residential energy-intensity benchmarks and EPC-style labels, while commercial real estate covers offices, retail, and industrial buildings with commercial benchmarks. A single property must sit in exactly one class; a mixed-use building needs a documented split so no emissions are double-counted across the two books.
Yes, if it is a genuine remortgage. A new facility is a new origination event, so the denominator resets to the property value at the remortgage date. A product transfer or rate switch on the same loan is not a remortgage and leaves the original origination value in place. The distinction matters because resetting the denominator on a simple rate switch would create an attribution change with no real economic origination behind it.
Weight each mortgage’s score by its financed emissions, not by loan count. Multiply each loan’s financed tCO₂e by its score, sum those products, and divide by total financed emissions. A simple average across loans is incorrect under PCAF because it ignores which dwellings carry the most carbon. As the worked Example 4 shows, the emissions-weighted score can be better or worse than the loan-count mean depending on whether the largest emitters are well-measured or estimated.
For the lender, they fall under Scope 3 Category 15 (investments). What is Scope 1 or Scope 2 to the homeowner becomes Scope 3 to the financing institution — the scope label does not transfer with the emissions. For a retail bank with a large mortgage book, Category 15 financed emissions routinely dominate the entire group inventory, which is why mortgage financed emissions are a central line item in financial-sector climate disclosure.