Financed Emissions for Australian Banks: The Honest...

An Australian bank's operational footprint (branches, data centres, corporate travel) is a rounding error. The real number sits in the loan book, the investment portfolio, and the insurance underwriting file. AASB S2 Scope 3 Category 15 pulls all of it into disclosure.

Many carbon accounting vendors now describe a "PCAF-aligned" financed emissions module. That phrase can mean very different things, so ask what sits behind it. The methodology is public. The disclosure obligation is the bank's. The question worth asking is what the platform actually stores.

Here is our honest answer for Carbonly. It stores primary counterparty data with the source document attached and a data quality tier your own analyst assigns, and it does not score PCAF for you. That suits some Australian banks and does not suit others.

What we do not have

We will lead with the disclosure that most vendor decks bury.

Carbonly does not have a proprietary PCAF data quality scoring engine. We do not have a dedicated financed emissions module with a branded PCAF output screen. We do not auto-fill the CDP Financial Services questionnaire. We do not have a signed data partnership with S&P Trucost or MSCI ESG that pipes their scored PCAF outputs into our ledger by default.

If any of those are the shortlist criteria for your bank's procurement process, we are not the right platform, and we would rather you knew that in paragraph four than paragraph forty.

If your procurement question is different (can the underlying counterparty ledger stand up to reasonable assurance under ASIC surveillance) then keep reading.

What financed emissions actually is under AASB S2

AASB S2 requires disclosure of Scope 3 emissions where material. For a bank, Category 15 (Investments in the GHG Protocol Scope 3 Standard, extended to loans and other financial services under PCAF) is almost always material. The Partnership for Carbon Accounting Financials methodology is the industry reference, and APRA prudential guidance CPG 229 refers to it for the Australian banking sector.

PCAF is a methodology document. It is not a software product. It tells you:

How to attribute counterparty emissions to a lender or investor (usually equity share, loan-to-value, or outstanding balance divided by enterprise value including cash)
How to score data quality on a 1-to-5 tier (1 = counterparty-verified primary data, 5 = sector-average with region proxy)
Which asset classes to cover (listed equity, corporate bonds, business loans, project finance, commercial real estate, mortgages, motor vehicle loans, sovereign debt)

The methodology tells you nothing about what your ledger should look like. That is a platform choice. And that is where the honest question lives.

The trade-off in plain terms

There are two paths for a bank preparing financed emissions disclosure.

Path A is the turnkey PCAF-branded platform. Counterparty data goes in, a scored PCAF output comes back, the vendor screen shows a portfolio-weighted average data quality tier and a formatted disclosure block. The bank submits it. If the assurance provider or ASIC asks how a specific counterparty number was derived, the answer is "the vendor's algorithm using inputs A, B, C". The methodology is defensible because PCAF is a public standard. The specific number depends on how the vendor treated missing fields, applied proxies, and interpolated.

Path B is the primary-data ledger. Every counterparty exists as a set of records with methodology labels, source documents, attribution factors, and a data quality tier the bank populates. There is no vendor black box. When the assurance provider samples a record, the walk-through goes straight to the counterparty's annual report, EPD, or primary-data submission. The disclosure is a query on top of the ledger, not a vendor output.

Path A is faster to stand up. Path B is more auditor-defensible when the assurance requirement steps up from limited to reasonable, and when ASIC's surveillance priorities under RG 280 start walking disclosures back to source.

Carbonly is a Path B platform. That is a deliberate design choice, not a gap we are working to close.

What Carbonly actually stores for financed emissions

The counterparty ledger sits inside the same emission record model that carries every other emission on the platform. That matters because a separate substrate for financed emissions makes the assurance story harder. One ledger, one audit trail, one factor pinning policy.

What we can point to today, per record: the source documents attached to it (including counterparty submissions through the Supplier Portal), the emission factor version it was calculated with, typed audit events on ledger writes, and period locking with restatement reasons. How PCAF asset classes, attribution factors and data quality tiers would be captured for your book is something to have us demonstrate before you rely on it.

A PCAF tier your analyst can defend

This is where our approach differs most from a scoring engine.

We do not compute the PCAF data quality tier for you. In our view the bank's financed emissions analyst should assign it, per record, based on what actually arrived. If a listed corporate counterparty submitted its own audited Scope 1 and 2 disclosure, that record gets tier 1 or 2. If the bank is using sector-average intensity from an external database with regional proxy, that record gets tier 4 or 5.

Why do it this way. Because the tier is a judgement call embedded in evidence, not a formula. Two vendors can score the same counterparty record differently depending on how they weight "verified" versus "reported but unverified", how they treat partial coverage, and how they apply defaults. When ASIC or the assurance provider asks why a record is tier 2 rather than tier 3, the bank's analyst has to answer. Making the tier a bank-owned judgement puts a name and a date against it.

The trade-off is real. If your team wants an automatic tier assignment they can hand to management, we do not do that. If your team wants a tier it can defend, with the reasoning written down, we think this is the right shape.

Counterparty data arrives with the source attached

The Supplier Portal that Carbonly built for Scope 3 upstream data collection doubles as a counterparty engagement surface for financed emissions. The mechanics are the same. Invite the counterparty. They log in and submit their data, scoped to the relevant project, and the submission lands in the bank's ledger with the source attached.

The Supplier Data Chase Agent follows up outstanding submissions on the bank's configured cadence, which helps the bank show its assurance provider that reasonable effort was made to collect primary data before falling back to sector averages.

This is not a bespoke financed-emissions capability. It is the same supplier engagement pipeline every other Scope 3 workflow uses. That is the point. One engagement model, one ledger, one audit story.

Keeping the scoring provider you already have

If your bank has already procured S&P Trucost, MSCI ESG, or another PCAF-Aligned Data provider, Carbonly is not asking you to rip it out.

Carbonly has a REST API, so counterparty data collected through the engagement workflow could be passed to a scoring provider as inputs, with its output brought back for internal reconciliation. Ask us to scope that integration for your provider before relying on it. The aim is that the primary-data record and the scored output can be reconciled internally rather than in a disclosure crisis.

A bank that already uses a scoring provider could layer Carbonly underneath as the collection and evidence substrate. A bank that wants a single platform for financed emissions with a scored output on top would usually pick a scoring-first vendor. Both are legitimate choices.

Ask the ledger a question without going through a data team

The same ledger is queryable from an AI assistant such as ChatGPT or Claude via the Carbonly MCP server, so the Head of ESG can ask plain-language questions of the data.

That is not a chat gimmick. It is a way for finance leadership to interrogate the ledger without going through a data team.

What AASB S2 assurance actually samples

Under ASSA 5010, limited assurance extends to Scope 3 from an entity's second year, and reasonable assurance applies from periods starting on or after 1 July 2030. For banks in Group 1, financed emissions land inside the assurance scope early, given materiality.

Under reasonable assurance, "the vendor's algorithm produced this number" is unlikely to be enough. The assurance provider samples records and walks the emission through to source. For financed emissions that means:

Show me the counterparty record
Show me the source document (annual report page, EPD, primary submission)
Show me the attribution factor calculation
Show me the PCAF tier assignment and the reasoning
Show me the audit trail on any change to the record in the last two years

A bank that has primary data with source documents on 60 percent of its counterparties by financed exposure, and industry-average data with a clearly documented tier assignment on the remaining 40 percent, has an auditable story. A bank that has a vendor-scored PCAF output with no primary data behind it has an argument.

The gap between those two positions is where the disclosure risk lives. And it is the reason we designed Carbonly to store primary data with its source documents, rather than to score anything ourselves.

NGER and AASB S2 interaction for banks

Most banks will not trip NGER thresholds on operational emissions. But NGER reporters in the loan book do report their own Scope 1 and 2 into the Clean Energy Regulator's system. Where the bank can get that filing (it is not always public at counterparty level, but disclosure is trending in that direction), it is the highest-quality PCAF tier 1 or 2 input available for corporate lending counterparties. Building the ledger so that NGER-source counterparty data can be attached to the record with provenance intact is the right architecture. That is what we do.

For AASB S2, the bank's own disclosure is separate. Financed emissions sit in the bank's Scope 3 Category 15. The counterparty's own Scope 1 and 2 becomes the bank's Scope 3 input. Same ledger, different lens. Carbonly calculates on AR5. Where the bank's AASB S2 disclosure needs AR6, treat that conversion as a separate, documented step.

Consultants who run financed emissions engagements

Carbonly is designed to be a ledger that sustainability and financed emissions consultants who work with banks (specialist advisory practices doing counterparty data remediation, PCAF methodology implementation, and assurance readiness) can run their bank engagements on.

The pattern is simple. The consultant runs the engagement (counterparty prioritisation, data quality strategy, PCAF tier scoring methodology, assurance liaison). Carbonly is the workshop equipment: the ledger, the engagement portal, the audit trail, the API for integration. The consultant owns the client relationship and the judgement calls. We are infrastructure.

Consultants are buyers, not competitors. If your bank has already engaged a financed emissions consultant, ask them to look at Carbonly as the ledger they run the engagement on top of.

Practical starting point

The trap in financed emissions is trying to boil the whole portfolio in year one. Do not do that.

Pick your top twenty loan counterparties by financed exposure. Set each one up with basic identity, asset class, and attribution factor. Invite them to the Supplier Portal. Give them thirty days. See what comes back. Assign the PCAF data quality tier for each counterparty based on what actually arrived, and write down the reasoning.

You will end with twenty records with evidence behind them, a realistic sense of how counterparty engagement performs on your book, and honest data on your tier distribution before you scale. From there, expand to the next tier of counterparties, and the next. The disclosure improves as the data quality improves, and the audit trail carries the story of how it got there.

That is a different shape of program than "buy the PCAF module, run it, disclose the output". It is slower to headline. It is faster to defend.

If the honest disclosure is a dealbreaker

We accept that some banks will read this post and decide they want the turnkey scored output. That is a reasonable decision, and there are vendors who sell it well.

If your team wants primary counterparty data with source documents and a bank-owned tier assignment with the reasoning written down, that is the trade we designed for. Per-project pricing with a $100 per month workspace minimum. Email hello@carbonly.ai and we can walk through the top-twenty counterparty setup with your team.

Financed Emissions for Australian Banks: The Honest View Without the PCAF Marketing