Climate & nature risk analytics · for European banks
Audit-ready climate & nature risk for the banking book.
refinq turns loan book addresses into supervisory-grade risk data, asset by asset. One dataset serves EBA/GL/2025/01, Pillar 3 ESG disclosures, CSRD/ESRS, EU Taxonomy DNSH and SFDR in one pass.
Anonymised coordinates. Results within days. No core-banking integration.
1 → 6
One asset-level dataset feeds six regulatory frameworks.
Address-level
No sector proxies. The granularity supervisors and DNSH assume.
Days, not quarters
From portfolio file to first full screening.
Regulatory anchors
EBA/GL/2025/01
EBA/GL/2025/04
CRR Art. 449a
CSRD · ESRS E1/E4
EU Taxonomy · SFDR
The supervisory calendar
Most of this is no longer coming. It is in force.
In force
Since 11 Jan 2026
EBA/GL/2025/01: ESG risk management
Applies to all large and other institutions: ESG risks embedded in strategy, governance, risk management and ICAAP, with data and methodology the supervisor can inspect. → refinq: hazard, nature and transition data per exposure, documentation attached
Since 15 Jun 2026
ECB climate factor on collateral
The Eurosystem now adjusts the value of corporate bonds pledged as collateral for their exposure to climate transition risk. Funding conditions, not disclosure. → refinq: transition risk signals per counterparty (early access)
Semi-annual
Pillar 3 ESG disclosures (CRR Art. 449a)
Large, listed institutions disclose today. The final draft ITS of 22 June 2026 extend proportionate ESG disclosure to all institutions, including SNCIs, with physical and transition risk at counterparty level. → refinq: exposure-level physical risk extract, disclosure-ready
Since FY 2024
CSRD / ESRS E1: climate
Wave-1 banks disclose physical risk exposure, portfolio resilience under climate scenarios and anticipated financial effects. → refinq: scenario-based exposure and resilience datapoints
Ahead
1 Jan 2027
EBA/GL/2025/04: scenario analysis
Environmental scenario analysis for all CRR institutions, materiality-based: forward-looking pathways, not point-in-time scores. → refinq: SSP scenario projections per asset, to 2100
11 Jan 2027
EBA/GL/2025/01 reaches SNCIs
Small and non-complex institutions (≤ €5bn) enter scope. The data infrastructure takes longer to build than the time remaining suggests. → refinq: proportionate setup, file-based, no integration project
Earliest end 2027
ECB climate factor reaches credit claims
Decided 24 July 2026: the climate factor extends to loans to non-financial corporations pledged as collateral, with a maximum additional value reduction of 5%. This moves the measure into the loan book. → refinq: transition risk per borrower (early access)
FY 2027
ESRS E4: biodiversity mandatory
Geospatial assessment of financed assets against biodiversity-sensitive areas. E4 cannot be answered without location data. → refinq: protected-area screening and nature metrics per asset
Platform
Physical risk
Nature & biodiversity
Transition & emissions
Cross-reporting
What refinq computes
Four modules. One dataset underneath.
Module · Physical
Physical climate risk
Hazard scores per address under IPCC scenarios, horizons to 2100.
PILLAR 3 ESG
ESRS E1
ICAAP
EBA/GL/2025/04
Details
Flood, heat stress, wildfire, drought, storm and further acute and chronic hazards, scored per address under SSP1-2.6, SSP2-4.5 and SSP5-8.5 across horizons to 2100. Data sources include CMIP6, ERA5, JRC flood maps and Copernicus.
Module · Nature
Nature & biodiversity risk
Protected-area screening, ecosystem dependencies, satellite-based land-cover change. TNFD LEAP aligned.
ESRS E4
SFDR · PAI 7
TAXONOMY · ANNEX D
Details
Proximity to Natura 2000, Key Biodiversity Areas, Ramsar and UNESCO sites; dependencies and impacts per counterparty; change over time from satellite data. Climate and nature in one model, not a bolt-on.
Module · Transition
Transition risk & financed emissions
Financed emissions following PCAF Part A, alignment, EBA transition plan output.
EBA/GL · §110 PLAN
ESRS E1-6
PILLAR 3 ESG
Details
Financed emissions per asset class with a data-quality score per position, alignment against sector pathways, and the transition plan required under EBA/GL/2025/01 §§84 to 110. In development with pilot banks. PCAF accreditation in progress.
Layer · Reporting
Cross-reporting outputs
Every module writes into the formats your teams file.
ONE DATASET
SIX FRAMEWORKS
AUDIT TRAIL
Details
Disclosure-ready extracts for Pillar 3 ESG, ESRS E1/E4 datapoints, DNSH and CRVA evidence, SFDR PAIs and ICAAP scenario inputs, plus methodology documentation an auditor can follow end to end.
The reconciliation problem
Advisors report six-figure coordination costs at individual large institutions for exactly this exercise. Structurally, it is a mapping problem. refinq removes it: risk management and disclosure read from one asset-level dataset, so the numbers agree because they are the same numbers.
Build vs. buy, itemised
Derived from EBA/GL/2025/01, Pillar 3 ESG, ESRS, EU Taxonomy
Without refinq
Twelve workstreams you would run in-house.
The obligations apply either way. Building means:
01
Geocode every borrower and collateral address, keep it current.
02
License and validate hazard data for every lending geography.
03
Turn hazard layers into a defensible, versioned scoring methodology.
04
Classify the banking book by hazard and geography.
05
Build scenario projections for ICAAP and supervisory analysis.
06
Screen every asset against protected-area registers (ESRS E4).
07
Assess nature dependencies and impacts along TNFD LEAP.
08
Run site-level Taxonomy CRVA checks for every DNSH assessment.
09
Compute PCAF-conformant financed emissions with data-quality scores.
10
Reconcile risk figures with CSRD disclosures, every cycle.
11
Track EBA, ECB and national supervisory change continuously.
12
Produce methodology documentation that survives on-site inspection.
Or: one file of addresses and exposures. refinq returns the rest, documentation attached.
How it works
Excel in. Evidence out.
STEP 02
refinq computes
Geocoding, hazard scoring, nature screening, scenario projection. Every value carries provenance.
STEP 03
Outputs land in your formats
Disclosure extracts, ICAAP inputs, DNSH evidence. Your risk framework keeps the decisions.
Three commitments
The claims we make, and the ones we deliberately don’t.
Modelled, not measured.
Uncertainty ranges, methodology cards and versioned models. That is what survives supervisory review.
E, deliberately.
S and G are largely solvable in-house. Only the E structurally requires external geodata and models.
Foundation, not engine.
refinq is the asset-level data layer your ICAAP, credit process and disclosures consume.
Outsourcing readiness
DORA Art. 28–30
EBA Outsourcing GL
EU data residency
Security & vendor due diligence
Built to pass your outsourcing review.
EU data residency
Microsoft Azure, region Germany West Central.
DORA-ready contracting
ICT third-party flowdown, Articles 28–30.
Outsourcing documentation
EBA-aligned pack: exit strategy, subcontractor register.
ISO 27001
Implementation in progress.
Data processing (DPA/AVV)
Standard agreements; pilots run anonymised.
Methodology & audit
Versioned models, end-to-end data provenance.
Before you ask
Four questions every risk committee asks.
Which institutions is this for?
From small and non-complex institutions to significant institutions across the EU, Switzerland and the Nordics. The supervisory framing differs by jurisdiction and size. The underlying data requirement does not.
Does this mean an IT integration project?
No. refinq works file-based (CSV, Excel) or via API and runs standalone beside your core systems. The security-review scope stays deliberately small.
What does a pilot look like?
A scoped screening of a defined sub-portfolio with agreed deliverables, run on anonymised coordinates and priced as a project. Structured to stand in front of your risk committee, not a feasibility study.
