Consensus ESG Score Methodology

Purpose

What the consensus score is for

Many investors treat the consensus earnings estimate as a valuable data point, because it shows what the market has already priced in for a company. The consensus ESG score applies the same idea to sustainability: it gives an indicative view of how a wide range of ESG analysts and rating firms currently assess a company.

It also lets investors set Integrum ESG’s own grade for a company against the view of the wider market. Where the two diverge, the subcategory breakdown shows where the difference comes from.

The consensus score is reported separately from Total ESG Risk, our in-house rating, which is set out in the Integrum ESG Scoring Framework. It does not apply our Awareness and Performance scoring rules and does not feed into our own grades.

Sources

Sources and coverage

CSRHub has aggregated more than 600 million data values from more than 900 sources. These include ESG analysis and rating firms, non-governmental organisations such as foundations, associations, unions and activist groups, government databases, research reports and input from CSRHub users. Its sources range from the major ESG rating providers to specialist datasets on single issues.

• Depth per company: for fully rated companies, the score draws on an average of 13.5 sources. Some large companies have more than 50, but most of these assess only one aspect of a company and carry less weight as a result.
• No dominant source: each source tracks a different universe of companies, regions or industries, and no single source covers more than 60% of the entities CSRHub rates.
• Coverage: CSRHub publishes data on more than 50,000 entities across more than 150 countries and more than 130 industries, including public companies, private companies, government organisations and not-for-profits. Industries are based loosely on the North American Industry Classification System (NAICS), with some additions.
• History: sources that no longer track a company still count towards its historical ratings, so the consensus view shows how a company’s rating has changed over time.

Structure

Four categories and twelve subcategories

Every data value is mapped to one or more of 12 subcategories, grouped into four categories. Environment covers the E factor, Community and Employees together cover the S factor and Governance covers the G factor.

Sources measure the same issue in different ways. A company’s community performance might be measured by its charitable donations or by its employees’ volunteering hours, and both map to Community Development & Philanthropy.

Method

How the score is built

1. Mapping Each data value from each source is mapped to one or more of the 12 subcategories.
2. Conversion Sources rate companies in different ways, from numerical scores to plus and minus signs to rankings such as “Top 50”. Each is converted to a common 0 to 100 scale, where 100 is the most positive rating, by fitting the source’s score distribution to a beta distribution. Yes or no answers start from an initial score, such as 75 for yes and 25 for no, which is then recalibrated against the other data in the subcategory.
3. Normalisation Scores from different sources for the same company are compared to find sources that are systematically harsh or lenient, and those biases are adjusted for.
4. Weighting Sources that agree more closely with the consensus of all other sources carry more weight, as do sources that provide several data elements in a subcategory. Major sources, meaning those with analyst teams, a written methodology, a defined update frequency and an established reputation, can carry double weight; other sources can carry a third. A source’s weight can also vary by subcategory, so a climate-focused source counts for more in an environmental subcategory than in a governance one.
5. Aggregation Weighted data is combined into subcategory ratings, which are then combined into category ratings and an overall rating.
6. Trimming and benchmarking Ratings are dropped where there is not enough information, and each company is assigned industry and country benchmarks to compare against.

Category weights

The overall rating weights the four categories using CSRHub’s consensus weightings, which reflect the weightings chosen by users of its platform.

By factor, that is approximately 30.8% E, 45.0% S and 24.2% G. These weights are separate from the default weights in our in-house rating.

Outputs

Ratings, rankings and grades

The consensus score is expressed in three ways:

• Rating: an absolute score from 0 to 100, shown overall and for each category and subcategory.
• Percentile ranking: a relative score from 0 to 100 that shows where a company’s rating sits against other companies in the same industry and country. A ranking of 100 means the best rating among those peers.
• Grade: the 0 to 100 score expressed as an A to E grade.

Worked example: percentile ranking

Take an industry with five companies rated 20, 25, 50, 55 and 80. Their percentile rankings are 0, 25, 50, 75 and 100 respectively. The rating is absolute, while the ranking shows each company’s position relative to its industry peers.

Industry benchmark ratings are tracked over time alongside each company’s own rating, so investors can see whether a company is improving faster or slower than its peers.

Rating rules

When a company receives a rating

A score is only published where there is enough reliable data behind it. Extreme results are treated as a sign of poor-quality data and filtered out.

Of the more than 50,000 entities with published data, roughly half have full ratings and more than 10,000 have partial ratings. Requiring several sources for a full rating means no single provider determines a company’s overall consensus view.

Our approach

Why a consensus view matters

ESG ratings from different providers often disagree, because each uses its own data, scope and weighting. Relying on a single provider leaves investors exposed to that provider’s blind spots and gives no sense of how the rest of the market sees a company.

Related methodologies

• Integrum ESG Core Methodologies
• Integrum ESG Scoring Framework
• Controversial Activities Methodology
• ESG Controversies and Sentiment Methodology
• Company ESG Scores and Glass Box Methodology
• Private Markets ESG Data Capture Methodology
• EDCI Data Collection and Reporting Methodology
• SFDR Classification and PAI Methodology
• EU Sustainable Investment Audit Methodology
• EU Taxonomy Alignment Methodology
• TCFD Reporting and Climate VaR Methodology
• Portfolio Footprint Calculator Methodology
• Fund ESG Ratings Methodology
• Golden Retriever Data Extraction Methodology
• Heatstrips and Peer Benchmarking Methodology
• Proxy ESG Scores Methodology
• ESG Screening Methodology
• SDG Impact from Operations Methodology
• Sovereign ESG Scores Methodology
• Temperature Score and Implied Temperature Rise Methodology
• SDG Revenue Alignment Methodology