Improving Data Quality Over Time

Improving data quality is not a one-time event. It is a multi-year programme that requires deliberate strategy, stakeholder engagement, and investment in data infrastructure. PCAF expects financial institutions to demonstrate a trajectory of improvement in their weighted average data quality score over successive reporting periods.

Why Data Quality Improvement Matters

A weighted average data quality score of 4 or 5 signals high uncertainty in the reported financed emissions figures. Such estimates may be accurate or they may be systematically biased by 30 to 100 percent in either direction. This matters for:

• Target setting: If the baseline financed emissions figure is unreliable, sector-level targets built on that baseline will be similarly unreliable
• Engagement: Engaging borrowers on decarbonisation requires knowing approximately how large their emissions are and where they come from
• Regulatory disclosure: Regulators and auditors increasingly require financial institutions to explain how their disclosed financed emissions are calculated and how accurate those calculations are

Strategy 1: Direct Engagement with Borrowers

The most direct route to improving data quality is to ask borrowers for their emissions data. This is most feasible for:

• Large corporate borrowers that have sustainability teams and are accustomed to ESG data requests from investors
• Relationship banking clients where the bank has ongoing dialogue and leverage
• Borrowers in emission-intensive sectors (energy, steel, cement, agriculture) where emissions are most material

Financial institutions can embed emissions data requests into:

• Annual loan review processes and covenant compliance checks
• ESG due diligence questionnaires at origination
• Sustainability-linked loan structuring (where the borrower's emissions must be measured and disclosed as a condition of the facility)

Strategy 2: Partnering with Data Providers

Third-party data providers (CDP, Bloomberg, MSCI, Sustainalytics, S&P/Trucost) cover thousands of companies globally, particularly in the listed equity and corporate bond segment. A financial institution that has not yet integrated third-party emission data into its systems can significantly improve data quality for its listed-company exposures by:

1. Subscribing to a data provider with coverage relevant to its portfolio geography and sector mix
2. Configuring its portfolio management system to automatically pull emissions data for each holding
3. Applying the provider's data quality flags to assign PCAF scores

For a bank with significant listed equity or bond exposures, this integration alone can improve the weighted average data quality score materially for those asset classes.

Strategy 3: Improving Coverage Year-Over-Year

Financial institutions should track not just their weighted average data quality score but also the percentage of their portfolio covered by different data quality levels. A clear improvement trajectory shows:

• Share of portfolio at Score 1 (verified reported): Growing from (say) 5% to 20% over three years
• Share of portfolio at Score 5 (sector average): Declining from 60% to 30% over three years

This tracking creates accountability and provides the evidence base for PCAF's disclosure review process.

Strategy 4: Sector-Specific Deep Dives

Not all sectors are equally important for a financial institution's financed emissions. Emission-intensive sectors (energy, power, steel, cement, chemicals, aviation, shipping, agriculture) typically account for a disproportionate share of financed emissions even if they represent a small share of outstanding loans.

PCAF recommends conducting sector-specific data quality improvement programmes starting with the highest-emission sectors. For a bank with heavy exposure to commercial real estate, for example, a programme to collect actual building energy consumption data (moving from EPC-label estimates to actual utility bills) can dramatically reduce the uncertainty in the CRE portion of financed emissions.

Base Year Recalculation

As data quality improves, historical financed emissions figures will change. PCAF provides a recalculation protocol based on GHG Protocol principles:

Financial institutions shall recalculate base year emissions when:

• There is a significant change in methodology (switching from Option 3 to Option 1 for a major sector)
• There is a structural change in the portfolio (major acquisition or disposal)
• There is a significant correction of an error

"Significant" is defined as a change that materially alters the trend of financed emissions or the achievement of targets. PCAF recommends that financial institutions establish their own significance thresholds (typically 5% of total financed emissions) and disclose them publicly.

Tracking Progress

PCAF recommends that financial institutions publish a data quality improvement roadmap alongside their annual financed emissions disclosure. This roadmap should include:

• Current weighted average data quality score by asset class
• Target data quality score for each asset class within 3 to 5 years
• Key initiatives and milestones for achieving those targets
• Resources allocated to data quality improvement (staff, systems, third-party services)

Publishing this roadmap signals credibility and commitment to stakeholders, regulators, and PCAF itself.