Improving PCAF Data Quality Scores: Practical Strategies for Banks

Why Data Quality Matters So Much

Calculating financed emissions has become a routine process for many financial institutions. But a question equally important as the calculation itself remains: How reliable is this estimate?

The PCAF (Partnership for Carbon Accounting Financials) Global Standard defines a five-level data quality scale. Score 1 represents the highest quality (verified primary data), while Score 5 represents the lowest (sector averages) (PCAF Global GHG Accounting and Reporting Standard, Part A, 2022). Many banks rely on Score 4 or 5 data for the vast majority of their portfolios. This creates significant uncertainties in emission estimates and weakens target-setting processes.

Typical Score Distribution

Based on experience from global PCAF implementations, many banks' first-year portfolio distributions look like this:

This distribution shows that more than half the portfolio relies on sector averages — meaning the bank can only approximately estimate its actual climate impact.

Why Scores Are Critical for Target-Setting

PCAF data quality matters not just for reporting accuracy, but also for strategic decisions:

Net-zero targets. Frameworks such as the Net-Zero Banking Alliance (NZBA) and SBTi Financial Institutions require reliable emissions data for target-setting. Targets set using Score 5 data may deviate significantly from actual emission levels.

Sectoral decarbonization pathways. Tracking the decarbonization trajectory of bank portfolios by sector requires client-level data. Sector averages do not reflect individual client performance.

Regulatory expectations. The European Central Bank and other regulators are demanding higher-quality data for climate risk stress tests. Risk assessments based on low-quality data undermine regulatory confidence.

Improvement Strategies by Asset Class

PCAF defines methodologies for seven distinct asset classes, each with different dynamics for improving data quality.

Listed Equity and Corporate Bonds

This asset class is the easiest area for improving data quality:

• CDP data: Emissions data from publicly listed companies reporting to CDP can be used directly. The CDP-PCAF alignment work explains how CDP data maps to PCAF methodology (CDP-PCAF Alignment, 2023)
• Sustainability reports: GRI, ESRS, or ISSB-aligned reports can provide Score 2 or 3 quality data
• Commercial data providers: Company-level emission estimates from providers such as MSCI, ISS, and Sustainalytics can be rated between Score 3-4

Strategy: Collect CDP and sustainability report data directly for the top 50-100 positions in the portfolio. These positions typically represent 60-80 percent of the portfolio and will significantly improve the weighted average score.

Business Loans

Business loans are the largest and most challenging asset class for many banks:

• Large corporate clients: Direct data can be collected from large companies that publish sustainability reports
• SMEs: Small and medium enterprises generally do not report emissions data — Score 4 or 5 is often unavoidable here

Strategy: Apply a tiered approach:

1. Request direct data from the top 100-200 clients that constitute approximately 70 percent of the portfolio
2. For the mid-segment, calculate at Score 3 using energy consumption data (utility bills)
3. Continue using sector averages for the small segment, but disaggregate at the NACE code level

Mortgages

Improving data quality for residential loans is directly tied to energy performance data:

• Energy Performance Certificates (EPC): If a building's energy class is known, Score 3 calculations are possible
• Building age and type: If the energy class is unknown, estimates based on building age and type yield Score 4
• Actual energy consumption: If energy consumption data can be obtained with client consent, Score 2-3 is achievable

Strategy: Start by matching EPC data with loan files. In many countries, EPC data is publicly available or can be obtained from national databases.

Commercial Real Estate

• Tenant energy data: Request energy data from tenants through green lease agreements
• Building management systems: Smart building systems can provide real-time energy data
• GRESB data: Data from real estate funds reporting to GRESB can be utilized

Borrower Engagement: The Most Effective Lever

The most sustainable way to improve data quality is direct engagement with borrowers. Banks can leverage their lending relationships to encourage borrowers to share emissions data.

Engagement Strategies

Data request integration. Incorporate emissions data requests into loan application and renewal processes. This does not have to be a mandatory condition — even an optional data-sharing invitation increases response rates.

Capacity building. Offer simplified emission calculation tools or guides, especially to SME clients. Enabling borrowers to calculate their own emissions directly improves the bank's data quality.

Incentive mechanisms. Offering improved loan terms to clients that share emissions data or set reduction targets (sustainability-linked loans) enhances both engagement and data quality.

Sector events. Organize information sessions for clients in specific sectors. Targeted engagement in high-emission sectors (energy, cement, steel, chemicals) yields more efficient results.

Leveraging CDP Data

CDP is one of the most comprehensive platforms, collecting environmental data from over 23,000 companies globally. The alignment work between PCAF and CDP clarifies how CDP data maps to the PCAF scoring system (CDP-PCAF Data Integration, 2023):

• Companies reporting to CDP with verified data: Score 1
• Companies reporting to CDP without verification: Score 2
• Companies that have not responded to CDP but have sectoral estimates in the CDP database: Score 4-5

Systematically matching CDP data with portfolio data can provide a rapid quality improvement, particularly for listed companies and large corporate clients.

Roadmap: From Score 5 to Score 1

It is not possible to leap forward in data quality all at once — this is a multi-year process. A recommended roadmap:

Year 1: Map the current portfolio distribution, collect CDP and publicly available emissions data for the largest positions. Target: Improve the weighted average score by 0.5 points.

Year 2: Launch a borrower engagement program, integrate data requests into lending processes. Match EPC data with the mortgage portfolio. Target: Bring the Score 1-3 share to 40 percent of the portfolio.

Year 3: Evaluate engagement results, develop targeted strategies for sectors with low response rates. Target: Bring the weighted average score below 3.0.

Key Takeaway: Data quality is not a destination but a continuous improvement process. Each score improvement strengthens the bank's climate risk assessment and target-setting capacity. The most effective first step is to start with the clients that represent the largest share of the portfolio.

See how financial institutions measure and manage financed emissions.