CSRD Double materiality
This page describes how the Upright data engine assesses double materiality (i.e. impact materiality and financial materiality)
Impact materiality
Summary
The Upright data engine assesses impact materiality from various perspectives, including the company's product and service mix, its geographic reach, and other considerations that reflect the company's capabilities to address specific sustainability matters. To understand the significance of each material matter, a concept of “materiality level” is utilized. Materiality level takes into account the scale, scope, irremediable, and likelihood character of the (potential) impact, noting that the first is only relevant for potential impacts, and the last is only relevant for negative impacts.
Company-level assessment
When determining the materiality of a given sustainability matter for a given company, the following aspects are considered:
Products: The materiality level indicated by the product-level assessment depends on the scale, irremediability, and likelihood within the sustainability matters, along with the revenue share of each product (example: manufacturing of diesel cars can lead to a high materiality level for the matter “climate change mitigation”). For definitions of "scale", "irremediability", and "likelihood", see Section "Materiality score calculation".
Geography: The materiality level indicated by the geographical assessment depends on the countries in which the company or its suppliers are operating (example: operating in Congo can lead to a high materiality level for the matter "child labour").
Specific data points: Specific internal data points or reliable third-party reports can indicate a low/medium/high materiality level for a given sustainability matter (example: reported corruption incidents can lead to a high level of materiality for matter "corruption & bribery").
The exact aspects that are considered differ by sustainability matter.
If there are multiple aspects indicating materiality within the same sustainability matter (e.g. product mix indicating a "high" materiality level, and geographical aspects indicating a "medium" materiality level), the highest of those levels will define the overall materiality level for the company.
Product-level assessment
Product-level data serves as one factor in determining materiality at the company level. Upright has generated thousands of product-level materiality assessments, with the following data sources used as input:
Impact scores (e.g. the negative impact of this product within the impact category "GHG emissions" is more than 10 cents per dollar of revenue, hence the topic "climate change mitigation might be material for this product")
EU taxonomy metrics (e.g. this product is linked to EU taxonomy objective "climate change adaptation", hence the topic "climate change adaptation" might be material for this product)
Categorical associations (e.g. the primary purpose of this product is to increase the energy efficiency of another product; hence the topic "energy" might be material for this product)
UN SDG metrics (e.g. this product is strongly linked to SDG 3, hence the topic "health and safety" might be material for this product)
Details on Upright methodology are available separately for impact scores, EU taxonomy metrics, and UN SDG metrics.
For each product-level assessment, a materiality score is calculated based on four factors: scale, scope, irremediability, and likelihood.
Materiality score calculation
Materiality levels are determined based on a materiality score. The following formula is used to determine the product-and-service-mix-based materiality score:
Materiality score for a product = scale x scope x irremediability x likelihood
The definitions for all factors incorporated in the calculation directly derive from the ESRS legislation:
Scale = How grave the impact is for people or the environment
Scope = How widespread the impacts are
Irremediability = To what extent it is impossible to restore the affected environment or people to their prior state (N.B. only considered for negative impacts)
Likelihood = Likelihood of occurrence
All factors, including scale, scope, irremediability, and likelihood, are assessed on a scale ranging from 1 to 4:
Very low = 1
Low = 2
Medium = 3
High = 4
For assessing “scope”, the revenue share of products and services is used as a proxy.
The theoretical maximum for unnormalized product materiality scores is 256 (4x4x4x4). The maximum score would be obtained by a company producing 100% of revenue from products with high scale, irremediability, and likelihood. To enhance understanding, the scores are normalized, with 100 being the maximum attainable score.
Illustrative example of calculating materiality scores for products
The company produces 2 products: apples and oranges. In the last year, apples accounted for 75% of the revenue, while oranges contributed 25%. See table below.
In the example, the materiality score for “apples” is calculated as follows:
Unnormalized materiality score: 3x(0.75x4)x4x4 = 3x3x4x4 = 144
Normalized materiality score: 144/256*100 = 56.25
The resulting total materiality score for the company would be 56+13 = 69.
| Scale | Scope | Irremediability | Likelihood | Materiality score |
Apples | Medium | 75% | High | High | 56 |
Oranges | Low | 25% | Low | High | 13 |
Example of defining scale, scope, irremediability, and likelihood for “Waste”
As the ESRS matter "Waste" aligns closely with the Upright impact category "Waste", the net impact scores are directly utilized for defining scale:
If the Waste score is larger than 5 cents per dollar or revenue, the scale of the impact is considered to be "high".
If the Waste score is between 2.5-5 cents per dollar of revenue, the scale of the impact is considered to be "medium".
If the Waste score is between 1.25-2.5 cents per dollar of revenue, the scale of the impact is considered to be "low".
If the Waste score is lower than 1.25 cents per dollar of revenue, the scale of the impact is considered to be "very low".
The scope of the impact is evaluated at the company level, by examining the revenue shares of products.
The irremediability of "Waste" is defined by the nature of the impact. For instance, in the case of hazardous waste, which can have detrimental or even deadly impacts on people and nature, the irremediability level is considered “high”, whereas, for non-hazardous waste, it can be defined as 'low'.
For waste, the likelihood of occurrence significantly depends on how waste is treated or how consumers typically use the product. For example, disposable plastic waste, such as plastic bags, does not always end up in the environment, resulting in a “low” likelihood of this type of impact.
No netting of impact scores
Using net scores or other similar metrics would be inappropriate for determining materiality; it is clear, for example, that the topic "climate change mitigation" would be material for a company creating a large amount of GHG emissions while removing the exact same amount.
This is also the position explicitly taken by the CSRD regulation, which states that
"information shall not be netted or compensated to be neutral".
When Upright uses impact scores for assessing materiality scales, this is based on either the positive score or the negative score for a given category, but never their net.
Materiality levels and thresholds
The materiality levels are determined based on the materiality scores using the following thresholds:
Low: < 15
Medium: 15-50
High: 50
For negative impacts, the concept of materiality level is equivalent to “severity” described in the ESRS regulation.
The materiality level thresholds are calibrated by taking advantage of Upright's database of 50,000+ companies and their materiality results, such that a sensible amount of sustainability matters end up being material for each company (on average). This calibration is updated annually to match the prevailing interpretation of how "sensitively" materiality ought to be triggered for each topic.
Consideration of all parts of the value chain
The input data used for product-level assessments incorporates value chain information from all parts of a product's value chain. Hence, all parts of the value chain are automatically considered. How value chain information is represented differs based on the type of product-level assessment. For example, in product-level assessments based on impact scores, the impact scores used as input data include an explicit subdivision of the share of the impact that is produced internally, upstream, and downstream. In product-level assessments based on categorical association, the relevant value chain part is a property of the category.
Example of value chain information within categorical associations
As an example of a product-level assessment based on a categorical association, the sustainability matter topic 'energy' is considered to have the relevant value chain part 'downstream' material for products that belong to a category whose primary purpose is to improve energy efficiency of other products.
Regardless of the underlying factors, the underlying value chain information from all product-level assessments (and their input data) is incorporated into a final result that states the relevant value chain parts on the sustainability-matter-level. This works additively, i.e. if one product-level assessment is based on a downstream impact, and another triggers materiality based on an internal impact, the value chain parts listed for the topic will be "internal and downstream".
Consideration of all time horizons
Similar to value chains, the input data used for product-level assessments incorporates impacts from all relevant time horizons including the short-term, medium-term, and long-term time horizons defined in the ESRS.
Similar to value chains, regardless of the underlying factors, the underlying temporal information from all product-level assessments (and their input data) is incorporated into a final result that states the relevant time horizons on the topic level. This works additively, i.e. if one product-level assessment materiality is based on a short-term impact, and another triggers materiality based on a medium-term impact, the value chain parts listed for the topic will be "short-to-medium-term".
Financial materiality
Summary
Financial materiality (FM) is assessed in terms of risks and opportunities. Financial materiality for a topic is triggered when risks or opportunities related to a given topic are deemed sufficiently significant.
There are two types of risks and opportunities:
Impact-driven risks and opportunities: These are caused by a (material) impact that the company has (e.g. company creates negative impacts on biodiversity and there is a risk that demand for their products will reduce when consumers gain more awareness on the importance of biodiversity, hence the topic "biodiversity" is material for the company).
Dependency-driven risks and opportunities: These relate to dependencies on natural, human and social resources (e.g. company has HQ in area that is at risk of flooding due to climate change, "hence the topic climate change adaptation" is material for the company).
In Upright's data engine, the former are identified by patterns that translate impacts into risks and opportunities, whereas the latter are caught by product-level or company-level triggers that directly trigger risks and opportunities.
Risk patterns
Impacts are translated into risks and opportunities using risk patterns. Each risk pattern has:
Criteria on when it is relevant: Criteria on when the risk pattern is relevant at all. Typical criteria are that the risk pattern is relevant only when a company has a positive (or negative) impact related to a given topic. Relevance does not mean that a risk is significant, as a relevant risk might still have a very low probability and scale.
Intrinsic risk probability for pattern: This defines how likely the risk is for a company meeting only the relevance criteria, but without taking yet into account any specifics.
Factors modulating risk probability: Factors that increase or lower the risk probability. For example, risks related to negative impacts harming demand are lessened if a company is (mainly) B2B.
A list of impact-driven risks is produced by evaluating the impact materiality results against the risk patterns. This yields a list of risks, along with their probabilities and scales. The scale of impact-driven risks is driven by the share of revenue from products with impacts driving a given risk.
Example of a risk pattern that translates negative impacts related to biodiversity to a financial risk
Description: Risk of decrease of demand to the company's product and services due to consumers becoming increasingly averse to negative impacts related to "biodiversity"
Relevant when: Company has negative impacts related to the topic "biodiversity"
Intrinsic risk probability: Medium (consumers' awareness of biodiversity-related impacts is high and growing)
Factors modulating risk probability:
The risk is reduced, if the company's products are B2B
The risk is increased, if the company's products are B2C
The risk is reduced, if the products have an overall net positive impact
The risk is increased, if the relevant products have an overall net negative impact
The risk is reduced, if the products are necessities (such as medicines)
The risk is increased, if the products are non-necessities
Note
This example above was related to consumers and negative impacts. Similar risk patterns exist related to investors, employees, and regulators (governments), and both positive and negative impacts.
Opportunity patterns
Opportunity patterns work similarly to risk patterns. Each opportunity pattern has:
Criteria on when it is relevant: Criteria on when the opportunity pattern is relevant at all. Typical criteria are that the opportunity pattern is relevant only when a company has a positive impact related to a given topic. Relevance does not mean that an opportunity is significant, as a relevant opportunity might still have a very low opportunity level.
Intrinsic opportunity level: This defines the (average) opportunity level for a company meeting only the relevance criteria, but without taking yet into account any specifics about the company.
Factors modulating opportunity level: Factors that increase or lower the opportunity level, based on specifics about the company. For example, the opportunity level related to positive impacts helping to recruit highly qualified workers increases if the company is in a business that relies on a highly educated workforce.
A list of impact-driven opportunities is produced by evaluating the impact materiality results against the opportunity patterns. This yields a list of opportunities, along with their opportunity levels.
Capturing dependency-driven risks and opportunities
Dependency-driven risks and opportunities relate to dependencies on natural, human, and social resources.
As detailed in the European Sustainability Reporting Standards (ESRS), dependencies may trigger effects in two possible ways:
They may influence a company's ability to continue to use or obtain the resources needed in its business processes, as well as the quality and pricing of those resources
They may affect a company's ability to rely on relationships needed in its business processes on acceptable terms.
Examples:
A company has HQ in an area that is at risk of flooding due to climate change, hence the topic "climate change adaptation" is material for the company.
A company manufactures weapons. Climate change increases the amount of armed conflicts in poor countries, increasing the demand for the company's products. Hence, there may be (grotesk) opportunities for the company related to "climate change".
A company grows vegetables and is dependent on ecosystem services such as pollination and nutrition cycling. Ecosystem services face threats from climate change and pollution, restricting the company's ability to leverage these services. Hence, "Impacts and dependencies on ecosystem services" is material for the company.
Dependency-driven risks are captured by product-level and company-level materiality triggers that seek to capture key dynamics around how companies are affected by changes in environmental and social conditions. Of the examples, the first one would be captured by a company-level trigger, while the second one would be captured by a product-level trigger.
Risk and opportunity assessment scales
Risk probabilities, magnitudes, and opportunity levels are assessed on a 4-point scale:
Scale for risk probabilities
| Probability | Indicative description |
|---|---|
Very low | Remote (i.e. probability <5%) |
Low | Unlikely (i.e. probability ~5%-25%) |
Medium | Possible (i.e. probability 25%-50%) |
High | From likely to almost certain (i.e. probability is >50%) |
Scale for risk magnitudes
| Magnitude | Indicative description |
|---|---|
Very low | Potential damage relates to a negligible portion of the company's business |
Low | Potential damage relates to a small portion of the company's business |
Medium | Potential damage relates to a significant portion of the company's business |
High | Potential damage relates to a major portion of the company's business |
Scale for opportunity levels
| Magnitude | Indicative description |
|---|---|
Very low | Potential likely upside equivalent in size to a negligible portion of the company's business |
Low | Potential likely upside equivalent in size to a small portion of the company's business |
Medium | Potential likely upside equivalent in size to a significant portion of the company's business |
High | Potential likely upside equivalent in size to a major portion of the company's business |
Triggering materiality from risks and opportunities
Materiality for a given topic is triggered if there are any risks or opportunities relating to the topic that exceed the thresholds.
Thresholds for triggering financial materiality based on risks
Thresholds for triggering financial materiality based on risks are defined by a combination of risk probability and risk magnitude as outlined in the triggering matrix below:
Thresholds for triggering financial materiality based on opportunities
Financial materiality is triggered when there are any opportunities with an opportunity level higher or equal to medium.
Last updated
