Physical climate risk platform for banks
UrClimate Next is a map-centric analytics platform that measures physical climate risk across a bank’s credit portfolio and translates it into regulation-ready, auditable financial impact. It takes the bank’s own portfolio data as input and produces a quantitative answer to: “where, to what, when and how much is this loan exposed?”
The problem it solves
Today
Under TSRS S2, EBA Pillar 3 and the BDDK Climate Guidance, banks must now answer their portfolio’s physical climate exposure quantitatively. Today that answer is typically:
- Not available — data is scattered, hazard-to-sector links are missing.
- Produced manually — not reproducible, not auditable.
- Left qualitative — “we may be exposed” is not the number a regulator wants.
With UrClimate Next
That answer is produced from the bank’s existing data — automated, reproducible and audit-tracked.
Asset location, sector (NACE) code and loan parameters go in; an auditable number in currency, differentiated by scenario and horizon, comes out.
The risk framework: H×E×V
Physical climate risk comes in two forms: acute hazards that damage through a single event (flood, storm, wildfire) and long-term chronic shifts (drought, extreme-heat trend, sea-level rise). Six hazards are measured with peer-reviewed climate indices and normalized to a 0–1 range (methodology aligned with IPCC AR6). Hazards are treated as dependent — compound structures are built into the model.
The projected climate index at the location, by scenario and year; normalized 0–1.
The loan amount and asset nature at that location.
The sector’s (NACE) sensitivity to that hazard.
Compound hazard structures included; prioritised on a 0–100 score.
Financialization: three channels to expected loss
This is the core of the product: physical hazard is converted into measurable financial impact through three credit-risk channels.
Probability of default; all loans, by the sector’s hazard sensitivity.
Loss given default; real-estate-collateralized loans, via collateral impairment.
Debt-service coverage; project finance, via revenue/opex sensitivity.
- ΔPD, ΔLGD, ΔEL — loan and portfolio level
- Collateral value loss
- IFRS 9 ECL impact
- Capital impact (RWA, ICAAP)
Analytical coverage: seven dimensions
Risk definition
Acute/chronic risk profile per asset, under the selected scenario.
Exposed loans
Amount and share of exposed loans in the portfolio.
Geographic distribution
Concentration by province/district/coastline: pins, clusters, heatmap, choropleth.
Sectoral distribution
Exposure by NACE Section; concentration alerts.
Time horizon
The year risk begins; short/medium/long maturity bands.
Risk rating
Prioritization on a 0–100 score.
Financial impact
The three channels and derived outputs; drill-down to a single loan, audit-trailed.
Scenarios and time horizon
Regulatory alignment
TFRS / TSRS S2
Climate reporting (§13(b), 15–17, 29(c)); Turkish and English, Word + PDF output.
EBA Pillar 3 — Template 5
Climate-exposure disclosure under CRR3 Article 449a; CSV + Excel output.
BDDK Climate Guidance + İSEDES/ICAAP
Nine-section appendix; Article 16 audit alignment.
BCBS d239 and NGFS
A methodological foundation consistent with these scenario frameworks. Regulatory reports are derived outputs: the underlying computation and data are traceable in the audit log.
On-prem deployment and data sovereignty
Deployment model
- Container-based — the compute service (API) and web interface ship as Docker images, inside the bank’s own network.
- Data never leaves the bank — air-gapped installation is supported.
- Handles a ~400,000-location portfolio with server-side query/aggregation.
- PostgreSQL read-model and audit store; retention configurable.
- Enterprise security — JWT, multi-tenant structure, CORS restriction, health checks.
Why deterministic?
The computation engine is deterministic: the same input always yields the same output.
- Reproducibility — a regulator or internal audit can reproduce the exact result.
- Traceability — source, formulation and timestamp behind every number, in the audit log.
- Defensibility — not “the model said so” but a step-by-step, demonstrable computation.
What it is NOT — honest boundaries
- It does not compute GAR/BTAR — physical-risk measurement, not taxonomy alignment.
- Transition risk is not its primary focus — the focus is physical climate risk.
- It does not replace the bank’s data — the bank brings its own data; the platform measures, visualizes and financializes it.
At a glance
| Physical hazards | 6 — flood, drought, extreme heat, wildfire, storm, sea-level rise |
|---|---|
| Financial channels | 3 — PD / LGD / DSCR → ΔEL, ECL, capital |
| Scenarios | SSP2-4.5 and SSP5-8.5 (configurable), 2015–2100 |
| Maturity bands | Short / Medium / Long |
| Regulation | TSRS S2 · EBA Pillar 3 T5 · BDDK İSEDES · BCBS d239 · NGFS |
| Sector framework | NACE Rev. 2.1 (full taxonomy) |
| Scale | ~400,000-location portfolio |
| Deployment | On-prem, container-based, air-gap capable |
| Engine | Deterministic, audit-tracked |
Turn your portfolio’s climate exposure into a number
Contact us to evaluate UrClimate Next for your institution.