IFRS S2 for commercial banks: the industry metrics behind financed emissions

Your credit team can tell you the outstanding balance on every facility to the cent. Ask the same team for the emissions attributable to those facilities and you get a spreadsheet with a methodology note attached. That gap, between a number the bank controls and a number the bank estimates from somebody else's data, is what the industry guidance for commercial banks is about.

Most coverage of this topic stops at the main standard. But IFRS S2 has a second layer for banks, and it is the layer that decides which metrics a reasonable investor will expect to see.

The second layer most banks skip

IFRS S2 sets out cross-industry metrics that every entity discloses: gross Scope 1, 2 and 3, and the rest. Those are the same for a bank and for a cement plant.

Paragraph 32 then adds something else. It requires an entity to refer to and consider the applicability of the industry-based metrics in the Industry-based Guidance on Implementing IFRS S2. The wording matters. Refer to and consider is a requirement. Applying the metrics is not a condition of stating compliance.

That distinction gets flattened constantly, usually into a claim that the industry metrics are mandatory. They are not. What is mandatory is that you looked, and can say what you concluded.

The guidance itself is derived from the climate content of the SASB Standards and covers 68 industries, set out in Appendix B. You find yours through SICS, the Sustainable Industry Classification System, which is how the volumes are organised. A commercial bank lands in Financials, under Commercial Banks.

What the amendment changed on the SASB side

The December 2025 amendments to IFRS S2 are covered elsewhere on this site: the permission to limit Category 15 to financed emissions, the subtotal that permission triggers, the derivatives exclusion, and the end of GICS as the only classification. If you need those mechanics, read Category 15 after the financed emissions relief rather than the summary here.

What is less widely noticed is that the ISSB amended the SASB standards at the same time. The consequential amendments align the financed emissions metrics in three industry standards with the amended IFRS S2: Asset Management & Custody Activities, Commercial Banks, and Insurance. Three industries, not the whole financial sector.

The aligned metrics ask for absolute gross financed emissions disaggregated by Scope 1, Scope 2 and Scope 3, and for the methodology used, including the allocation method that attributes the bank's share. They also carry the same derivatives permission into the SASB metric. The effective date is the same: annual reporting periods beginning on or after 1 January 2027.

So a bank reading only the main standard sees the relief. A bank reading the industry guidance sees the relief and the metric it is attached to.

PCAF is how the number actually gets made

Neither IFRS S2 nor the SASB standard tells you how to calculate an attributed emission. They tell you to disclose the methodology. In practice the methodology is almost always the PCAF Global GHG Accounting and Reporting Standard for the Financial Industry, whose third edition was published in December 2025.

PCAF does two things that matter for a disclosure record.

The first is the attribution factor. The bank's share of a counterparty's emissions is its outstanding amount over the counterparty's total value, applied to that counterparty's emissions. Every input in that fraction is a number someone has to be able to produce again later.

The second is the data quality score. PCAF scores the basis of each counterparty's emissions from 1 to 5, where 1 is reported and verified and 5 is estimated from economic activity alone. A portfolio number carries a weighted average of those scores.

This is the part worth sitting with. The score is not a footnote about confidence. It is a statement, per counterparty, about where the number came from. A book that is mostly 4s and 5s is not a worse bank. It is a bank whose disclosure rests on estimation, and that has to be visible rather than averaged away.

The three exclusions are three different standards

The amendment lets a bank exclude derivatives, and separates financed emissions from facilitated and insurance-associated emissions. Read alongside PCAF, that separation stops looking arbitrary.

PCAF publishes them as different parts of the same framework. Part A covers financed emissions. Facilitated emissions, from capital markets activity, sit in their own part. Insurance-associated emissions have a separate standard again.

The IFRS S2 subtotal is not carving up one number for convenience. It is drawing the line where the measurement methods already diverge. A bank that also arranges bond issues or writes insurance is running two or three measurement exercises, and only one of them belongs in the subtotal.

What a lender has to be able to show

Here is where this stops being a standards question. A financed emissions figure is an aggregate of thousands of per-counterparty estimates. An assurance provider will not test the aggregate. They will pull a sample of counterparties and ask the same four things about each.

What was the outstanding amount at the measurement date, and does it tie to the ledger. What was the counterparty's total value, and where did that figure come from. What emissions were used, and were they reported by the counterparty or estimated. What data quality score was assigned, and who assigned it.

Most banks can answer the first. The fourth is where the record usually does not exist. The score gets produced inside a model run, lands in a portfolio summary, and is never written back against the individual exposure with a date and a name attached. Eighteen months later, when the sample lands, the team can reproduce the total but cannot show how one counterparty's score was reached.

That is a control gap, not a data gap. The measurement was fine. What is missing is the record that it happened, which is the same problem in every other part of a climate disclosure, arriving here at portfolio scale.

If you are scoping a first cycle in a bank, the useful question is not which metric to disclose. It is whether your counterparty-level evidence survives being asked for one name at a time.

Common questions

Are the IFRS S2 industry-based metrics mandatory for banks?

No. Paragraph 32 requires an entity to refer to and consider the applicability of the industry-based metrics. Applying them is not a condition of stating compliance with IFRS S2. A bank that considers them and concludes some do not apply has met the requirement, provided it can support that conclusion.

Which industry standard applies to a commercial bank?

The Commercial Banks volume of the Industry-based Guidance, found through SICS rather than GICS. Banks with material asset management or insurance activity should also look at the Asset Management & Custody Activities and Insurance volumes, since the December 2025 consequential amendments touched all three.

Does IFRS S2 require a bank to use PCAF?

No. The standard requires disclosure of the methodology used, including the allocation method. PCAF is the approach most banks apply and the one most assurance providers expect to see, but it is not named as a requirement in IFRS S2.

What evidence should a bank keep per counterparty?

The outstanding amount at the measurement date and its tie to the ledger, the counterparty value used in the denominator and its source, the emissions figure and whether it was reported or estimated, and the data quality score with the date and person behind it. The last of those is the one most often missing.

Where do you stand against IFRS S2?

The readiness diagnostic covers governance, evidence and controls across the four pillars, and names the gaps. For the paragraph-by-paragraph structure of the standard itself, see the IFRS S2 reference.