PCAF Financed Emissions: What Australian Banks and...

If you're a sustainability lead at one of the major Australian banks or insurers, your Scope 3 Category 15 financed emissions number is now larger than the rest of your group emissions combined by an order of magnitude. CBA, ANZ, NAB, Macquarie, and QBE are all PCAF signatories, which means they've committed to measuring and disclosing their financed emissions following the Partnership for Carbon Accounting Financials standard.

The 2025 PCAF standard expansion materially raises the bar. New methodologies for use-of-proceeds structures, securitisations, structured products, sub-sovereign debt, treaty reinsurance, and project insurance mean financial institutions are now expected to measure financed and insurance-associated emissions across asset classes that were previously out of scope.

For Australian banks reporting under AASB S2, this is the disclosure that drives the headline emissions number. For insurers, the PCAF Insurance Associated Emissions standard is now the relevant framework for underwriting emissions disclosure.

What PCAF actually is

PCAF is a methodology for measuring greenhouse gas emissions associated with financial activities. It sits inside the GHG Protocol Scope 3 framework as Category 15 (investments) for banks and asset managers, and it provides the accounting rules for how to allocate counterparty emissions to the financial institution.

The standard has three parts:

Part A: Financed Emissions. Methodology for loans, equity investments, debt securities, project finance, and similar lending and investment activities.

Part B: Facilitated Emissions. Methodology for capital markets activities including underwriting and advisory services where the financial institution facilitates a transaction without holding the resulting position.

Part C: Insurance-Associated Emissions. Methodology for insurance underwriting, including commercial lines and treaty reinsurance.

Each part assigns emissions from the underlying counterparty to the financial institution based on an attribution factor. For loans, the attribution factor is typically the outstanding loan balance divided by the borrower's enterprise value (or total equity plus debt). For listed equity, it's the proportion of shares held. For project finance, it's the share of project debt.

The data quality score

PCAF's defining feature is the data quality score system. Each financed emissions calculation is rated 1 to 5 based on the quality of the underlying data:

Score	Description	Example
1	Reported emissions, verified	Counterparty discloses verified Scope 1 and 2
2	Reported emissions, unverified	Counterparty discloses but no third-party assurance
3	Physical activity-based estimates	Counterparty reports physical activity (tonnes produced, MWh consumed); emissions calculated using emission factors
4	Economic activity-based estimates	Counterparty reports economic activity (revenue); emissions estimated using sector intensity factors
5	Asset-class average estimates	No counterparty data; estimated using asset class averages

Most Australian banks today carry portfolios where the average data quality score is around 4 to 4.5. Score 5 dominates for SME lending, score 3 to 4 for corporate lending, score 1 to 2 for the largest listed counterparties that publish their own emissions.

The disclosure target for most banks is to move the weighted average score down (toward 1) over time. That requires either more data from counterparties or better physical activity estimates from the bank's own systems.

The counterparty data problem

A typical Australian major bank has tens of thousands of corporate counterparties and hundreds of thousands of SME and retail counterparties. Most of those counterparties do not publish their own emissions data and never will.

For PCAF score 1 or 2, the bank needs the counterparty's own audited or unaudited emissions disclosure. This is realistic only for the largest few hundred counterparties.

For score 3, the bank needs counterparty-reported physical activity. For a major bank, this means engaging with thousands of mid-market counterparties and asking them to provide kWh of electricity, litres of fuel, tonnes of product, square metres of building. Most banks are running supplier or client engagement programs for exactly this purpose.

For score 4, the bank uses revenue and applies sector intensity factors. The data exists in the bank's lending systems already, but the spend-based approach has well-known accuracy issues.

For score 5, the bank uses asset class averages with very limited counterparty data.

The trajectory the major Australian banks are on is to move retail and SME exposures from score 5 toward score 4, mid-market exposures from score 4 toward score 3, and large corporate exposures from score 3 toward score 1 or 2. Each step requires more data discipline.

The 2025 PCAF expansion

The PCAF Standard updates in 2025 expanded coverage in several directions:

Use-of-proceeds structures. Green bonds, sustainability-linked bonds, and other instruments where proceeds are tied to specific projects now have a defined allocation methodology. The allocation matches the proceeds to the underlying project, rather than spreading across the issuer's whole emissions profile.

Securitisations. Mortgage-backed securities, asset-backed securities, and similar structures now have a methodology that allocates the underlying portfolio's emissions to the security holder.

Structured products. Derivatives and complex structured products have specific allocation rules.

Sub-sovereign debt. Lending to states, provinces, municipalities, and similar entities now has a methodology that allocates emissions from the underlying jurisdiction.

Treaty reinsurance. The insurance methodology now extends to treaty reinsurance, where the reinsurer assumes a portfolio of risks rather than individual policies.

Project insurance. Insurance for specific projects (construction, infrastructure, energy) now has a methodology aligned with the project finance treatment under Part A.

Each addition expands the scope of what financial institutions have to measure. Australian banks and insurers active in these areas now have more PCAF data to gather.

How AASB S2 and PCAF interact

For Australian banks reporting under AASB S2, financed emissions sit in Scope 3 Category 15. AASB S2 paragraph 29(a) requires Scope 3 emissions disclosure including all relevant categories. For banks, Category 15 is the most relevant category by an order of magnitude.

The AASB S2 requirement essentially mandates PCAF (or equivalent) methodology, because there's no other widely-accepted way to measure financed emissions consistently. The disclosure must include:

Total financed emissions in tonnes CO2-e
Breakdown by asset class (corporate loans, mortgages, project finance, etc.)
Methodology disclosure including data quality scoring
Comparability over time, with restatements if methodology changes

For insurers, the same logic applies for insurance-associated emissions, although the AASB S2 disclosure for insurance underwriting is still being interpreted in practice.

The connection to investor pressure

Australian super funds and listed asset managers are increasingly asking the banks they invest in for financed emissions disclosure that supports their own portfolio carbon footprint reporting. The Investor Group on Climate Change (IGCC) and the Australian Council of Superannuation Investors (ACSI) have both pushed for greater PCAF disclosure from listed financial sector entities.

This creates a cascade: super funds want PCAF data from banks; banks want PCAF data from corporate counterparties; corporate counterparties want emissions data from their own suppliers. The same Scope 3 supplier data collection problem repeats up the financial chain.

The discipline that breaks the cascade is one continuous emissions ledger at the corporate level. Counterparties that maintain audit-grade emissions data provide it once and have it consumed by their bank, their reinsurer, their auditor, and their disclosure under AASB S2. Counterparties that don't have to redo the work for each requester.

Practical implementation for an Australian bank

The pattern we see in mid-tier Australian banks (regional banks, mutuals, specialist lenders) implementing PCAF for the first time:

Asset class inventory. List every type of financial activity (loans, mortgages, project finance, equity holdings, government bonds, structured products). Map each to a PCAF asset class.
Counterparty data engagement program. Reach out to major counterparties with a request for emissions data. Tier the engagement by exposure size.
Calculation engine. Build (or buy) a system that applies PCAF methodology consistently across asset classes. The same outstanding balance number drives both the lending system and the financed emissions calculation.
Data quality scoring. Tag every calculation with its score. Track the weighted average score over time.
Disclosure preparation. Map outputs to AASB S2 Scope 3 Category 15 disclosure. Document methodology in basis of preparation.

For the major Australian banks, this is essentially an industrial-scale data engineering problem with hundreds of millions of underlying records. For mid-tier banks, the problem is more tractable but the methodology discipline is the same.

What the auditor will check

Under ASSA 5010 assurance for AASB S2, financed emissions are one of the highest-risk disclosure areas. The auditor's likely questions:

Is the PCAF methodology applied consistently across asset classes?
Does the data quality score align with the underlying counterparty data?
For score 1 and 2 counterparties, is the source disclosure traceable?
For score 4 and 5 counterparties, are the sector intensity factors current and appropriate?
Are restatements properly documented when counterparty data improves?
Does the calculation tie back to the underlying lending or investment ledger?

The reasonable assurance year is when this gets hardest. Most banks will be on limited assurance for several years before reasonable assurance applies, but the data discipline has to be in place from the start.

The insurance case

QBE has been a PCAF signatory since 2022 and was an early adopter of the Insurance Associated Emissions methodology. The treatment for insurance is structurally different from lending: instead of attributing emissions based on outstanding balance, the methodology attributes based on premium volume relative to the insured entity's enterprise value or revenue.

For a commercial property insurer, this means tracking the insured entity's emissions, the premium written, and the entity's financial size. For motor insurance, the methodology relies on vehicle-level emissions data. For health and life insurance, the relevance is limited but not zero.

Australian insurers are at varying stages of implementation. Most are using the methodology for select commercial portfolios and treating retail portfolios with simplified approaches. The 2025 PCAF expansion to treaty reinsurance and project insurance pulls more activities into scope.

The bottom line

PCAF is the methodology that turns "we lend a lot to fossil fuel companies" into a number. The number is ugly for many Australian financial institutions today, and it will get more visible as data quality improves. The lever for change is the same lever that drives data quality up: counterparty engagement, methodology discipline, and continuous accounting rather than year-end estimation.

If you're a financial institution implementing PCAF for the first time or refining an existing implementation, email hello@carbonly.ai or join the waitlist. Happy to talk through how the same data discipline that supports AASB S2 carries over to financed emissions, and how the engagement workflow with corporate counterparties shapes the data architecture you'll need.