United Bank for Africa (Ghana) Limited

1 · Portfolio risk map

124 assets pinned by vulnerability band. Low Moderate High Critical · switch to Satellite/Hybrid (top-right) to see building footprints.

2 · Geographic concentration of exposure

Where the book — and its high-risk exposure — clusters.

3 · Hazard exposure across the book

Number of assets at High or Critical for each hazard. An asset exposed to several hazards is counted under each, so these counts (and the per-hazard exposures elsewhere) can overlap and sum above the portfolio total — they are not a partition.

4 · Financial impact — physical damage, collateral & climate ECL

Two channels, kept separate. (1) Physical damage — an Average Annual Loss (AAL) per hazard from a depth-damage × frequency proxy: GHS 883k/yr of expected (mostly uninsured) damage. (2) Market repricing — a forward collateral devaluation that raises stressed LTV and loss-given-default, driving GHS 9.50M of potential collateral shortfall and an GHS 139k/yr indicative climate ECL uplift. A single severe event in Greater Accra could cause GHS 1.57M of correlated, simultaneous loss.

⚠ Indicative annual marginal overlay — not a lifetime/12-month staged IFRS 9 ECL, not a valuation, not the bank's PD/LGD model. All parameters (AAL, repricing, downturn LGD, PD, ~90% uninsured) are illustrative and exposed for calibration — see §Methodology and the gap register.

Top 12 by exposure × vulnerability. Full per-asset detail is in the CSV export.

5 · Forward-looking scenario projection (IPCC AR6)

How the share of exposure at High+ risk migrates over time under SSP2-4.5 — Middle of the road. Each hazard is intensified by an AR6-grounded factor (Rx1day rainfall, hot-day count, consecutive dry days, sea-level rise) and applied to today's score with a bounded saturating projection — a screening estimate, not a downscaled climate forecast. The interactive dashboard lets the user switch between the SSP1-2.6 / SSP2-4.5 / SSP5-8.5 pathways and 2030–2100, and pull any single asset through all three.

6 · Most-exposed assets — ranked register

Top 30 by exposure × vulnerability, with hazard fingerprint, data-confidence flag and the priority action. Hazard chips: Flood Coastal / sea-level Extreme heat Riparian proximity Land / contamination Drought / water stress.

7 · Regulatory & disclosure conformance matrix

How this screen maps to the frameworks a Ghanaian bank must satisfy. Needs bank data / sign-off marks requirements that depend on the bank's own governance, models or disclosure — see the gap register (§8).

Regulatory basis (verified June 2026): the Bank of Ghana Climate-Related Financial Risk Directive was issued November 2024 (under Act 930 s.92(1) and related Acts) and is effective for banks from January 2026 (SDIs/NBFIs from January 2027); it requires disclosure aligned to IFRS S2 / ISSB. ICAG's IFRS S1/S2 roadmap (28 March 2024) is voluntary 2024–2026 and mandatory for Significant Public Interest Entities from 1 January 2027. Confirm the latest text against the BoG primary source before filing.

BoG Climate-Risk Directive

Ghana Sustainable Banking Principles

IFC SBFN Measurement Framework

ISSB IFRS S2 / TCFD

NGFS scenarios

PCAF (data quality)

PCAF (financed emissions)

8 · Regulatory return summary (BoG semi-annual)

The figures a Ghanaian bank files under the BoG Climate-Related Financial Risk Directive — material risk + exposure to vulnerable segments. Parameter set: BoG baseline v1 (2026-06) [baseline] · seed screen · as of 2026-10-03. Headline: 20% of exposure at High+ risk; GHS 139k/yr climate ECL uplift; GHS 9.50M stressed collateral shortfall; largest single-event loss GHS 1.57M (Greater Accra). Exportable as a filing CSV.

9 · Gap register — what this screen does not replace

6 of 12 items are reliance-blocking. PortfolioCheck is screening-grade triage that flags collateral for closer review — it complements the bank's credit process; it does not certify it.

Methodology (screening-grade)

Vulnerability. Composite = 50% weighted hazard load (Flood 34% · Coastal / sea-level 18% · Extreme heat 16% · Riparian proximity 12% · Land / contamination 10% · Drought / water stress 10%) + 50% of the single worst hazard, so one extreme hazard that can impair the security is not averaged away (bands: Low <25 · Moderate 25–50 · High 50–75 · Critical ≥75). A single-peril severity floor completes the blend: a worst hazard ≥0.85 (e.g. an observed Sentinel-1 flood) floors the composite at Critical, and ≥0.70 at High, so a near-certain peril carries the asset even when the other hazards are low. Each hazard is scored 0–1 by the built-in analyser (Ghana location model) and, with the live engine overlay mode, overlaid with SafeGround (flood / river / coastal / land) and FloodGuard (flood susceptibility), FloodGuard felt-heat (ERA5) and NASA POWER rainfall climatology (drought / water-stress).

Hazard → money (two channels). (1) Physical damage — an Average Annual Loss (AAL) per hazard = max-AAL × score^1.5 (building-collateral ceilings: flood 5%; financially supported coastal risk 5.5%). Riparian proximity is represented through flood to avoid double counting; heat, land and drought remain due-diligence scores without a standalone building-damage AAL until a case-specific impairment model is supplied. Hazard AALs are combined across hazards; damage is assumed mostly uninsured in Ghana (~90% assumed — penetration is very low, so the loss falls on the collateral). (2) Market repricing — a forward collateral devaluation = ~35% of the capitalised AAL (cap rate 9%) + an insurability premium, capped at 40%; this raises the stressed LTV. ECL = PD × LGD × EAD (indicative annual marginal): LGD from realisable collateral with a smooth downturn liquidation, floored at 5% (Basel-style, so no LTV kink); PD = 2% base, rising with negative equity (strategic default) and uninsured damage. Not a lifetime/12-month staged IFRS 9 ECL, not a valuation. The single-event figure is a correlated-tail proxy (one severe event hitting all High+ assets in a region at once). Data quality is flagged on the PCAF 1–5 scale (screening is never better than 3). Scenario multipliers are illustrative — anchor to IPCC AR6 SLR / NGFS before disclosure. All parameters are exposed for recalibration against the bank's own loss experience.