PCAF data quality scores explained

Updated 28 July 2026

What is a PCAF data quality score?

A PCAF data quality score, usually shortened to DQ score, rates the reliability of the emissions data behind a financed emissions figure, on a scale from DQ 1 (reported and verified) to DQ 5 (estimated from economic activity). The scale comes from the Partnership for Carbon Accounting Financials (PCAF) Standard, and auditors and regulators increasingly expect the score to be disclosed alongside the emissions number itself.

The hierarchy, as PCAF publishes it:

• DQ 1: reported emissions, verified by an independent third party.
• DQ 2: reported emissions, not verified.
• DQ 3: emissions calculated from primary physical activity data, such as energy consumed.
• DQ 4: emissions estimated from production or revenue using sector averages.
• DQ 5: emissions estimated from economic activity alone, such as turnover or assets with a sector proxy.

A lower score means better data. The score matters because two institutions can publish the same portfolio emissions number and mean very different things: one is quoting what its counterparties actually reported, the other is quoting a sector average multiplied by exposure.

Why do most portfolios sit at DQ 4 and DQ 5?

Because most counterparties are private companies, and private companies rarely publish emissions where a data vendor can see them. When no reported figure is available, the calculation falls back to proxy data: a sector average applied to revenue or assets. That is a DQ 4 or DQ 5 by definition, however carefully the arithmetic is done.

The problem compounds in the long tail. A loan book or portfolio of thousands of counterparties cannot be surveyed by hand, so teams end up with reported data for a handful of large listed names and proxies for everyone else, held together in spreadsheets that are out of date before the reporting cycle closes.

What moves a portfolio from DQ 5 to DQ 2?

Reported data that already exists but has not been found. A meaningful share of any book's counterparties disclose emissions somewhere: an annual report, a sustainability report, a regulatory filing, a customer's supplier programme. Each disclosure that is found, matched to the right legal entity and normalised to a consistent standard replaces a proxy, and the DQ score for that counterparty moves from 4 or 5 to 2, or to 1 where the disclosure was independently verified.

The work is therefore data work, not modelling work: finding disclosures, resolving entities across parent and subsidiary structures, normalising units and boundaries, and keeping the source attached to the figure.

How do auditors treat DQ scores?

An auditor's question is rarely the number itself. It is whether the number is auditable: where each figure came from, what changed since last year, and why. DQ scores give the auditor a structured way to ask, so a defensible submission carries a source and a change history for every counterparty figure, and shows its coverage gaps rather than hiding them.

This is where audit-ready means something specific. DitchCarbon's calculator is verified to ISO 14064-3, limited assurance, by UL Solutions, renewed annually, and every figure carries its source and change history. The verification documents are available in the trust centre.

How does DitchCarbon improve a PCAF data quality score?

DitchCarbon provides verified emissions data for over 2 million organisations, so procurement, sustainability and finance teams can measure and act on supply chain and portfolio emissions from one source. For a portfolio, that means matching counterparties against found disclosures rather than defaulting to sector proxies: entity resolution runs against DUNS, LEI and ISIN identifiers, private companies are covered alongside listed ones, and each record shows the DQ score behind it.

The practical effect is a data quality distribution that shifts toward DQ 1 to 3, with numbers you can defend within 2 weeks rather than a reporting cycle spent chasing spreadsheets.

If someone is asking you for this data

Many readers of this page are on the other side of the request: a bank or customer has asked for your emissions. You can claim your organisation's profile free, see what requesters already see, and answer once rather than survey by survey.

See your own distribution

The fastest way to understand your book's data quality is to look at it. Request a walkthrough and we will show the DQ distribution on a sample of your own counterparties.