PCAF — Partnership for Carbon Accounting Financials
A bank’s own offices, servers and travel produce a modest carbon footprint. But a bank’s real climate footprint is in what it funds: the factories, power stations, homes and cars behind its loans and investments. For years there was no agreed way to count those emissions — so, in practice, they went uncounted. A single question was missing an answer: how much of a company’s emissions belong to the bank that finances it?
PCAF is the body that answered it. PCAF — the Partnership for Carbon Accounting Financials — created the global standard that lets financial institutions measure the emissions of their loans and investments, by attributing to each financier a proportional share of the emissions it finances.
PCAF (the Partnership for Carbon Accounting Financials) is the industry body behind the Global GHG Accounting and Reporting Standard for the Financial Industry — the most widely used method for measuring financed emissions. Its core idea is attribution: an institution counts a share of each borrower’s emissions, set by how much of that entity it finances. Built on the GHG Protocol, PCAF is how banks and investors quantify their Scope 3 Category 15 emissions, which usually dwarf their own operations.
What PCAF Is
The Partnership for Carbon Accounting Financials is an industry-led initiative that develops and maintains a common standard for measuring and disclosing the greenhouse gas emissions associated with financial activities. It began in 2015 among a group of Dutch financial institutions, expanded to North America and then went global toward the end of the decade, and now numbers several hundred banks, asset managers, insurers and other institutions worldwide. Its output is a single document — the Global GHG Accounting and Reporting Standard for the Financial Industry — that has become the de facto method the sector uses.
The problem PCAF exists to solve is consistency. Before it, financial institutions that tried to measure the emissions behind their portfolios each did so differently, so no two figures were comparable and none could be assured. PCAF fixes the rules: which emissions to count, how to divide a company’s emissions between the many parties that finance it, and how to be honest about the quality of the underlying data. It does not invent a new accounting framework from scratch — it is built on and conforms to the GHG Protocol, extending it to the specific problem of financed emissions.
The Attribution Principle
The heart of PCAF is attribution: the idea that a financier owns a proportional share of the emissions of what it finances. If a bank provides a tenth of a company’s capital, it is attributed a tenth of that company’s emissions. The share is set by the attribution factor — the institution’s outstanding amount divided by the total value of the financed entity — and the institution’s financed emissions are the sum, across its whole portfolio, of each attribution factor multiplied by that borrower’s or investee’s emissions.
A bank holds $10m of the bonds and equity of a listed company whose enterprise value including cash (EVIC) is $200m, and which emits 40,000 tCO₂e a year.
Attribution factor = 10 ⁄ 200 = 5%.
Financed emissions = 5% × 40,000 = 2,000 tCO₂e.
That 2,000 tonnes is the bank’s, counted in its own inventory — even though the bank burned none of it. Sum this across every loan and holding and you have the bank’s financed-emissions total.
Two features of this design matter. First, the denominator is the financed entity’s total value (for listed companies, its EVIC), so the shares attributed to all its financiers add up to the whole — no more, no less. Second, attribution is deliberately blind to intention: a green loan and a loan to a coal plant are attributed the same way, by money, so the resulting number reflects what a portfolio actually finances rather than what it is labelled.
The choice of denominator has a subtle side effect worth knowing about. Because a listed company’s EVIC moves with its share price, a bank’s financed emissions for that holding can rise or fall with the market even when the company’s real-world emissions have not changed at all — a rising valuation shrinks the attributed share, a falling one inflates it. This “inventory fluctuation” can make a portfolio look like it is decarbonising, or backsliding, for reasons that have nothing to do with actual emissions. PCAF’s December 2025 update responds to exactly this, adding recommended fluctuation analysis and an inflation adjustment so that changes driven by valuation and prices can be separated from genuine decarbonisation.
The Three Parts: Financed, Facilitated and Insurance
PCAF’s standard is organised into three parts, each covering a different way a financial institution is connected to emissions:
| Part | Covers | Whose activity |
|---|---|---|
| Part A — Financed emissions | Emissions of the loans and investments an institution holds on its balance sheet | Lending & investing |
| Part B — Facilitated emissions | Emissions behind capital-markets deals an institution arranges but does not retain (underwriting) | Capital markets |
| Part C — Insurance-associated emissions | Emissions attributed to a re/insurer’s underwriting portfolio | Insurance underwriting |
Part A, financed emissions, is the largest and most established, and it is what most people mean when they refer to “PCAF”; its financed-emissions guidance reached a third edition in December 2025. Part B recognises that a bank arranging a bond issue enables emissions even though it does not hold the bond, and to avoid overstating that looser connection those facilitated emissions are weighted (at 33% of the attributed amount). Part C extends the same attribution logic to the underwriting side of insurance. Together they aim to capture the full range of a financial institution’s climate connection, not just its balance sheet.
The Asset Classes
Because a mortgage, a corporate bond and a sovereign bond are financed and valued in very different ways, PCAF defines a separate attribution method for each asset class, differing mainly in what goes in the denominator — the measure of the financed entity’s total value. The core asset classes are:
| Asset class | Attribution denominator (financed entity’s value) |
|---|---|
| Listed equity & corporate bonds | Enterprise value including cash (EVIC) |
| Business loans & unlisted equity | Total company equity + debt |
| Project finance | Total project equity + debt |
| Commercial real estate | Property value at origination |
| Mortgages | Property value at origination |
| Motor vehicle loans | Total value of the vehicle |
| Sovereign debt | Purchasing-power-parity-adjusted GDP |
The December 2025 third edition expanded the standard to ten asset classes, adding use of proceeds, securitisation and sub-sovereign debt, and introduced further guidance on transition-related metrics. Each asset class has its own detailed method, which GreenCalculus implements as a dedicated tool — for example the listed equity and corporate bonds calculator, following the corresponding methodology — so a portfolio can be built up asset class by asset class.
The Data Quality Score
Financed emissions almost always rest on incomplete data — a bank rarely has audited emissions for every company it lends to — so PCAF requires every figure to carry a data quality score from 1 to 5. Counter-intuitively, 1 is the best: it marks a figure built on reported, verified emissions, while 5 marks one estimated from very rough proxies.
| Score | Basis of the emissions figure |
|---|---|
| 1 (best) | Audited or verified emissions reported by the company |
| 2 | Unverified emissions reported by the company |
| 3 | Estimated from physical activity data (e.g. energy use, production volumes) |
| 4 | Estimated from economic activity (e.g. revenue × a sector emission factor) |
| 5 (weakest) | Estimated from very limited data (e.g. total assets or turnover with regional/sector averages) |
The score is not a footnote but a headline output: it tells users how much to trust a portfolio’s number, and it turns data improvement into a measurable goal. A bank can lower its portfolio’s average score over time by gathering better data from its clients, and disclosing the score is what stops a precise-looking financed-emissions figure from hiding how much of it is guesswork.
Where PCAF Sits in Accounting
For a financial institution, PCAF-measured emissions are its Scope 3 Category 15 — investments — under the GHG Protocol. This is usually by far the largest part of a bank’s or investor’s inventory: the financed emissions of a lending book routinely run hundreds or thousands of times its operational Scopes 1 and 2. PCAF is the measurement layer that makes this category reportable at all.
It is important to be clear about what PCAF does and does not do. It is an accounting standard — it tells an institution how to measure its financed emissions, consistently and with honest data quality. It does not set targets, decide what a “good” number is, or dictate strategy; that is the domain of initiatives like the Science Based Targets initiative and of the institution’s own transition planning. PCAF provides the measured baseline on which those targets, disclosures and decarbonisation plans are then built — which is why it has become foundational infrastructure for climate accountability in finance, and why measuring financed emissions is now widely expected of, and increasingly required from, financial institutions.
Common Confusions
- Confusing PCAF with the things it measures. PCAF is the standard and the body behind it; financed emissions are the metric, the attribution factor is the ratio, and the data quality score is the grade. PCAF is the umbrella over all three.
- Thinking financed emissions are a bank’s operational footprint. They are Scope 3 Category 15 — the emissions of what the bank finances — and typically dwarf its own Scopes 1 and 2.
- Reading a low data quality score as bad. The scale runs 1 (best, verified) to 5 (weakest, proxy-estimated); a lower number is higher quality.
- Assuming PCAF sets targets. PCAF measures; it does not set climate targets. Target-setting for finance is the role of the SBTi and similar, using PCAF numbers as the baseline.
- Treating attribution as double counting. The same real-world emission is counted by the company (its Scope 1) and by its financiers (their Scope 3.15) — normal value-chain accounting from different vantage points, not an error.
Frequently Asked Questions
PCAF, the Partnership for Carbon Accounting Financials, is an industry-led body that created and maintains the Global GHG Accounting and Reporting Standard for the Financial Industry — the most widely used method for measuring the greenhouse gas emissions associated with loans and investments. It began in 2015 among Dutch financial institutions and has since become a global partnership of several hundred banks, asset managers and insurers. Its central idea is attribution: a financial institution counts a proportional share of the emissions of each company or project it finances, set by how much of that entity it funds. Built on the GHG Protocol, PCAF is how financial institutions quantify their Scope 3 Category 15 (investments) emissions, which usually far exceed their own operational footprint.
Through attribution. For each loan or investment, PCAF calculates an attribution factor — the outstanding amount the institution has provided, divided by the total value of the financed entity — and multiplies it by that entity’s emissions. Summed across the whole portfolio, this gives the institution’s financed emissions. For a listed company, the total value used is its enterprise value including cash (EVIC); for a private company it is total equity plus debt; and other asset classes such as mortgages or sovereign debt use their own appropriate measures. For example, a bank holding $10 million of a company worth $200 million that emits 40,000 tonnes of CO₂e is attributed 5%, or 2,000 tonnes. Every figure is also assigned a data quality score from 1 to 5 to reflect how reliable the underlying emissions data is.
As of its December 2025 third edition, PCAF’s financed-emissions standard (Part A) covers ten asset classes. The long-established core are listed equity and corporate bonds, business loans and unlisted equity, project finance, commercial real estate, mortgages, motor vehicle loans, and sovereign debt; the 2025 update added use of proceeds, securitisation and sub-sovereign debt. Each asset class has its own attribution method, differing mainly in how the financed entity’s total value is measured — enterprise value for listed companies, property value for mortgages, GDP for sovereign debt, and so on. Two further parts extend beyond the balance sheet: Part B covers facilitated emissions from capital-markets activity such as underwriting, and Part C covers the insurance-associated emissions of underwriting portfolios.
The PCAF data quality score is a rating from 1 to 5 that every financed-emissions figure must carry, indicating how reliable the emissions data behind it is. A score of 1 is the best, meaning the figure is based on audited, reported emissions; 2 is unverified reported data; 3 is estimated from physical activity such as energy use; 4 is estimated from economic activity such as revenue multiplied by a sector factor; and 5, the weakest, is estimated from very limited data such as total assets with regional averages. Because financed emissions usually rely on incomplete information, the score is a headline output rather than a footnote — it tells users how much of a portfolio’s number is measured versus estimated, and it turns improving data into a concrete, trackable goal for the institution.
They are the three parts of the PCAF standard, covering three different ways a financial institution connects to emissions. Financed emissions (Part A) are those of the loans and investments an institution holds on its own balance sheet — the largest and most established category. Facilitated emissions (Part B) are those behind capital-markets deals an institution arranges but does not keep, such as underwriting a bond issue; because that connection is looser than holding the asset, these are weighted down (to 33% of the attributed amount) to avoid overstating them. Insurance-associated emissions (Part C) apply the same attribution logic to the underwriting portfolio of an insurer or reinsurer. All three use the core PCAF idea of attributing a proportional share of a client’s emissions, adapted to the nature of the financial relationship.
PCAF is a voluntary standard, but it has become the expected — and increasingly required — method for measuring financed emissions, referenced by disclosure regimes and net-zero commitments across the sector. It is built on and conforms to the GHG Protocol: financed emissions measured with PCAF are reported as a financial institution’s Scope 3 Category 15 (investments). PCAF is strictly an accounting standard, concerned with how to measure emissions consistently and with honest data quality; it does not set targets or judge whether a number is good. Target-setting is the role of the Science Based Targets initiative and the institution’s own transition planning, which use PCAF-measured emissions as their baseline. In short, PCAF provides the measurement, the GHG Protocol provides the accounting framework it fits into, and the SBTi provides the goals set against it.