What is PCAF? Carbon emission calculation and data preparation for investment and financing of financial institutions

For most financial institutions, office electricity, travel and owned vehicles are not the biggest climate impacts. A larger proportion of emissions often come from enterprises and assets supported by loans, investments and other financial activities, that is, investment and financing carbon emissions. PCAF provides the financial industry with a common greenhouse gas accounting and disclosure method, allowing banks, investment institutions and other financial players to measure these emissions in a more consistent way.

PCAF will release the third edition of “Global GHG Accounting and Reporting Standard Part A: Financed Emissions” in 2025. This article is based on the common logic of this version; the actual calculation still needs to be confirmed based on the asset class method, data availability, and the disclosure requirements adopted by the institution.

What problem does PCAF solve?

The same invested enterprise may obtain funds from multiple financial institutions at the same time. If each company counts all of its company’s emissions under its own name, the results will lose comparability. The basic idea of PCAF is to attribute corresponding emissions according to the financing ratio of financial institutions to enterprises or assets, and then use appropriate methods according to asset categories.

Investment and financing emissions are usually linked to GHG Protocol Scope 3 Category 15 investment activities. It is not a proxy for the emissions of invested enterprises, but presents the emissions associated with financial activities from the perspective of financial institutions.

Basic calculation logic: attribution coefficient multiplied by emissions

Conceptually, emissions from a single investment and financing are equal to the attribution coefficient times the emissions of the investee or asset. The attribution factor represents the financial institution’s share of financing or investment in the object, but the denominator will vary by asset class and may involve enterprise value, total equity plus liabilities, total project cost, asset value, or other rules.

Therefore, it is impossible to create a formula that applies to all positions. Financial institutions must first complete asset classification, and then obtain balances, denominators, emissions and necessary activity data according to each method. If the currency, date, holding ratio and data year are inconsistent, the conversion and processing need to be clarified.

Step One: Define the Inquiry Boundaries and Asset Categories

Before you begin, you should confirm which legal persons, investment portfolios, business units and reporting periods are covered, as well as which financial assets are included. The third edition of PCAF covers a variety of asset classes. When actually adopted, the classification should be based on the official method, rather than just judging by internal product names.

The same customer may have corporate loans, project financing or other positions at the same time, and different positions require different information. Classification errors will affect the attribution factors and results, so business, risk, finance and sustainability units need to jointly confirm the mapping rules.

Step 2: Organize parts and main file information

Financial institutions often already maintain large amounts of data in core systems, investment systems, credit, collateral and financial databases, but the columns have different purposes. Investment and financing inspections need to match customer identification codes, asset categories, balances, holding ratios, industries, countries, financial information, subject characteristics and emission data.

The company name is not a reliable unique key. The same group may have different legal persons, languages and codes. Establishing stable entity identification and versioning rules is the basis for avoiding double counting or mismatching of data.

Step 3: Obtain emission data based on data maturity

The ideal situation is to use emission data that has been verified or publicly disclosed by the investee, but many small and medium-sized enterprises have not yet conducted an inventory, or the data year is inconsistent with the reporting period of financial institutions. At this time, unverified data, energy or production activity data, and emissions estimated from industry and financial data can be used according to the PCAF method.

Estimating does not mean that the data has no value, but it must clearly distinguish between sources, assumptions and quality. If estimates are presented mixed with verified data, managers may misjudge the accuracy of the numbers. Financial institutions should retain source dates, versions, units, coverage and conversion processes.

Why is data quality score important?

PCAF uses a data quality approach to present the reliability of the data used in calculations. The purpose of the score is not to allow companies to just chase a pretty average, but to identify which assets or customers need the most improvement. Even if total emissions fall, the results still need to be interpreted with caution if rougher estimates are used for a large number of locations.

Companies can divide quality improvement into stages: first improve the completeness of master files, then prioritize obtaining actual data from high-emission, high-exposure or strategic customers, and finally establish an annual update and review mechanism. Requiring all customers to provide a complete inventory at once is generally not cost-effective.

How to avoid double counting and misinterpretation?

Investment and financing emissions in the financial system may be attributed separately by different institutions. This is part of each institution’s disclosure of its own financial activities. The results of different institutions cannot be directly added up as emissions from the real economy. Financial institutions should also pay attention to whether the boundaries of funds, entrusted management, subsidiaries and mergers overlap.

In addition, an increase in investment and financing emissions does not necessarily mean a deterioration in management. It may result from asset growth, expansion of data coverage, quality improvement, or changes in market valuation. When disclosing, the boundaries, methods, data quality, and reasons for changes should be explained at the same time. It is not appropriate to publish only a single total amount.

From calculation to financial decision-making

The results of the inventory can be used to identify high-emitting industries and customers in a portfolio, but should not be used solely for mechanical divestments. Financial institutions also need to decide on negotiations, credit conditions, product design or risk management methods based on customer transformation plans, reduction targets, technical paths, policy risks and physical risks.

Originally, sustainable units might export data to various systems before annual disclosure and then merge it manually. After establishing a traceability process, asset classifications, calculation rules, coefficients, emission sources, and quality scores can be retained, and abnormalities can be returned to their original locations. For business personnel, they can see which customers require supplementary information or negotiation; for managers, they can compare the impact of different industries and asset portfolios.

How should cross-department work be divided?

Sustainable units: management methods, boundaries, coefficients and disclosure standards.
Investment and credit granting units: Confirm the characteristics of products, customers and assets, and promote customer negotiations.
Risk management: Link climate information to credit, industry and concentration risks.
Finance and Accounting: Provides balances, valuations, consolidation boundaries and reporting period information.
Information unit: Integrate master files, permissions, data lineage and batch operations.
Internal audit or second line of defence: checks whether methods are consistent and controls are effective.

Common mistakes

Calculate all positions using a single formula without classifying assets first.
Only the final result is retained, the denominator, data source and version are not saved.
Treat estimates as exact data and ignore quality scores.
Only pursues to reduce the total amount, without explaining asset growth and boundary changes.
Complete the audit without linking customer discussions, goals or financial decisions.

Which financial institutions are good to start with?

Has disclosed its own operational emissions and is preparing to expand to Scope 3 investment activities.
Facing investment and financing emission requirements from supervisors, investors or international initiatives.
Data is scattered in credit, investment, finance and customer master files, making it difficult to integrate.
Want to identify high emission exposures and establish customer negotiation priorities.
Plans to further set portfolio reduction or net zero transition targets.

Questions often asked when calculating for the first time

If there is no actual discharge from the customer, can we check it first?

Available proxies or estimates may be used in accordance with the PCAF method, but the source, year, assumptions and quality score must be recorded. In the first year, you can first establish complete positions and classifications, and then prioritize high-risk, high-emission or strategic customers for reinforcement. If the check is not started at all because it is waiting for all customers to complete the check, the financial institution will not be able to see the distribution of data gaps.

If investment and financing emissions are high, should we withdraw from the industry immediately?

Emissions are important information, but they should not be the only decision. Financial institutions also need to evaluate customers’ transformation capabilities, reduction plans, technical paths, financial risks and negotiation effectiveness. If only assets are sold, emissions from the real economy may not decline. A more complete approach is to use the inventory results for prioritization, and then match them with credit conditions, negotiation and transformation financial strategies.

Why is asset classification the most important task before calculation?

Internal product names of financial institutions are usually designed for business, accounting or supervision purposes and may not directly correspond to PCAF asset classes. Enterprises need to establish mapping rules to explain when each product is classified as listed stocks and corporate bonds, corporate loans, project financing, real estate, home loans, car loans, or other methods. When there is insufficient compound products, funds or information, a judgment and approval process must be set up.

Classification tables should be kept versioned and record how new products are incorporated. If the same client has multiple positions, necessary fields must be retained at the transaction or asset level to avoid losing method differences after merging. Finance, investment, credit and sustainability personnel can jointly check positions to confirm that the internal system fields are consistent with PCAF definitions, and then perform a large number of calculations.

Understand attribution calculations with simplified examples

Assume that the financial institution’s exposure to a company in compliance with the method is 100 million yuan, and the calculation denominator is 1 billion yuan, and the conceptual attribution coefficient is 10%. If the enterprise’s relevant emissions under the same boundary and year are 50,000 metric tons of carbon dioxide equivalent, the emissions attributable to the financial institution are conceptually 5,000 metric tons. This is just illustrative logic, the practical denominator and emission range must be chosen according to the asset class regulations.

Calculations check for balance date, enterprise value or financial year, currency, negative values, zero values, and gaps. The emissions of invested enterprises may include different categories. Financial institutions should clearly indicate the scope of adoption and cannot directly combine the results of different boundaries. If you use estimates, you also need to save industry, revenue, activity, coefficients and source versions.

What stages can data quality improvement be divided into?

The first stage is to improve the financial position and customer master file to ensure that legal persons, industries, countries, balances and asset classes are complete; the second stage is to organize public emissions and financial information and establish automatic or semi-automatic matching; the third stage is to negotiate for high-risk, high-emission and strategic customers to obtain activity data or verified emissions; the fourth stage is to incorporate data requirements into credit, investment and annual reviews.

Each update should retain the original value, source, date and modifier to avoid direct overwriting of new information. Quality indicators can be presented by asset class, industry and exposure weight, not just a simple average. Managers can therefore see that improving a small group of key customers may be more effective than requiring all low-exposure customers to fill out complex questionnaires.

How to put the results into customer negotiation?

Before negotiating, you can first understand the customer’s emission hot spots, data quality, reduction targets, capital expenditures and transformation constraints, and then design issues based on the industry. For customers who have not yet been audited, the first step may be to establish organizational boundaries and energy information; for customers who have been audited, scope 3, reduction plans and management can be discussed. Using the same long questionnaire for all customers often makes it difficult to get useful information.

The outcome of the negotiation should record commitments, deadlines, evidence and follow-up actions, and be linked to credit or investment review. Financial institutions must also explain the purpose of data and how to keep it confidential to prevent customers from providing only unverifiable numbers for submission. If data or plans do not improve for a long time, conditions, risk assessment, or resource allocation can be adjusted according to policies.

How can inventory results support management without being misused?

Investment and financing emissions can be used to identify high-emission exposures, improve setting data, plan negotiations, and assess portfolio transformation, but are not suitable for individually comparing the environmental performance of different financial institutions. Asset size, product mix, valuation, data quality and methodological differences will all affect the results. Description of methods, boundaries, exclusions, quality and annual changes should be provided when disclosing to the public.

Emission intensity, transformation plans, target coverage, green and transformation finance and climate risk indicators can be combined internally to form a more complete view. Major decisions still need to combine financial, risk and social impacts, and a carbon emission number cannot be turned into an automatic approval or rejection rule. When methods are updated, it is also important to evaluate whether to recalculate the basis and how to account for changes.

Self-inspection before starting investment and financing carbon inventory

Are the legal entities, investment portfolios, asset classes and reporting periods clearly understood?
Does the internal product already have a reviewable PCAF asset classification mapping?
Are customer, location, financial and emissions data concatenated using stable identifiers?
Does each result retain attribution denominator, emission source, year and quality scores?
Are estimated, actual and verified data clearly distinguished?
Are calculation results linked to data improvements, customer negotiations and financial decisions?

If the above foundation has not yet been completed, you can first target coverage and traceability in the first year, and there is no need to rush to set performance with a single total amount. When the classification and data are stable, and the proportion of actual emission data is gradually increased, annual comparisons will be more meaningful.

Conclusion

PCAF does not just produce a huge carbon emission number, but helps financial institutions establish the relationship between assets, customers, emissions and data quality. Only by first establishing boundaries, classifications and data lineage, and then gradually increasing the proportion of actual data, can investment and financing emissions become an effective basis for risk management and transformation decisions.

##Official reference material

Data access date: July 20, 2026.