Sourcing Risks

Sourcing risks

Purpose

Starting from the commodity inventory, this module identifies nature-related risks embedded in a company's supply chain by flagging commodities with known high biodiversity impact potential.

Flagging approach

Two flags are applied to each commodity in the inventory:

HICL flag — SBTN High Impact Commodity List

Based on the SBTN HICL v1, a science-based reference list of 47 commodities identified as material drivers of nature loss across three socioeconomic systems:

• Food system: e.g. cattle, cocoa, soybean, oil palm, timber, sugarcane, aquaculture
• Energy and extractives: e.g. crude oil, coal, copper, gold, lithium, nickel
• Built environment: e.g. cement, sand

For each commodity, material pressures are documented across 9 pressure categories (from the SBTN literature review and ENCORE). The number displayed on the HICL flag indicates how many of these 9 categories are material for that commodity. The 9 categories are:

• Land use (terrestrial ecosystem use & use change)
• Freshwater ecosystem use & use change
• Marine ecosystem use & use change
• Water use
• Climate change
• Air pollution (non-GHG)
• Soil / terrestrial pollution
• Freshwater pollution
• Marine pollution

EUDR flag — EU Deforestation Regulation

Flags commodities subject to the EU Deforestation Regulation, which applies to cattle, cocoa, coffee, palm oil, soya, wood, rubber and their derived products. Companies placing these commodities on the EU market must demonstrate they are not linked to recent deforestation.

Examples

Sourcing Risk score

Beyond the flags, the module scores where a company's supply-chain nature risk concentrates. The unit of analysis is a commodity × sourcing-country pair (e.g. cocoa from Côte d'Ivoire).

The score is built in two distinct stages — selection (which drivers count) and scoring (how the selected drivers form the number). Keeping them apart matters: materiality only selects a driver, it does not weight the score.

Stage 1 — Selection: which drivers count

A risk driver (indicator) enters a pair only when it passes both gates:

• Materiality gate — the driver is material for the commodity. This is a presence test (yes / no), satisfied by either source:

  • flagged by ENCORE materiality — each commodity is mapped to an ENCORE economic activity, i.e. its sector (e.g. sawlog and veneer log → forestry), and inherits that activity's materiality ratings; the driver counts when the activity is scored ≥ 4 (High / Very High) for that indicator at Scope 1; or
  • flagged by an SBTN HICL flag that maps to a scorable risk driver — 6 of the 9 HICL categories (land use, water use, air pollution, freshwater pollution, soil pollution, freshwater ecosystem use). Three categories are not mapped: climate/GHG and marine ecosystem use have no matching risk indicator at all, and marine pollution is deliberately left unmapped because the available pollution layers are terrestrial (air, pesticides, soil & water) and scoring marine pollution against them would be wrong. A commodity flagged only on those three gains no driver here.

  The ENCORE materiality level beyond the ≥ 4 threshold is not carried forward — selection is pass/fail, not graded. Three cultural ecosystem services (spiritual & artistic, educational & scientific, recreation) are also excluded from ENCORE selection: they score Very High almost universally and would not discriminate between commodities.

• Geography gate — the driver's own wired layer(s) are aggregated over the sourcing country (for a raster layer, an area-weighted mean of its values) and bucketed into a risk level — the worst level across a driver's layers when it has more than one; the driver is selected when that country-level risk is High or Very High. Each driver reads its own layer(s), not the worst level of its whole risk dimension — a dimension carries many drivers, so a per-dimension level would let one severe driver make the others read as severe too.

Both gates apply identically, whichever materiality source admitted the driver.

Stage 2 — Scoring: how the selected drivers form the number

Only drivers that passed both gates contribute. A pair's score is a pure environmental severity:

score(pair) = Σ ( severity + HICL bonus )   over the selected drivers

  severity     from the GEOGRAPHY level:  Very High = 2  ·  High = 1
  HICL bonus   +1 when the driver was admitted by BOTH ENCORE and HICL
               (an agreement of two independent sources), else 0

Two things follow from this:

• The severity weight comes from the geography level, not from the ENCORE materiality rating. Materiality selected the driver; the sourcing country's risk weights it. A driver barely material (ENCORE = 4) and one strongly material (ENCORE = 5) score the same if their country level is the same.
• The only materiality-derived term in the number is the HICL confirmation bonus. It rewards agreement between two independent sources — so it does not saturate the way a standalone HICL term would. The weights (Very High 2 / High 1 / HICL bonus 1) are a methodology-owned calibration.

EUDR is a tier, not a score term

EUDR exposure is regulatory, not an environmental severity, so it is deliberately kept out of the numeric score. Folding it in with an arbitrary weight would let a regulated-but-low-risk pair outrank a genuinely exposed one. Instead it applies as a hierarchical sort tier: EUDR-regulated pairs rank above non-EUDR pairs, and the environmental score orders pairs within each tier.

Displayed risk level

A pair's displayed level is its worst country driver — Very High, else High, else no material risk. EUDR does not raise this level: regulatory exposure is surfaced as a separate ⚖ compliance chip, so an EUDR-only pair shows an empty risk badge while still sorting to the top through the EUDR tier.

From pairs to a commodity

A commodity's level and score are those of its worst pair (the maximum), so ranking commodities by risk reflects their most-exposed sourcing origin.

Sorting

The rollup can be ordered by:

• Risk — the EUDR tier first, then the environmental score.
• Impact — the assessed environmental impact.
• Unit impact — impact ÷ quantity.