Geopolitical Risk Scoring: A Quantitative Model for Strategic Decision-Making

Executive Summary

Geopolitical risk is no longer a soft variable that analysts mention in disclaimers. Between 2020 and 2024, geopolitically driven market dislocations—ranging from the Russian invasion of Ukraine to Red Sea shipping disruptions—erased an estimated $4.7 trillion in global market capitalization across single-event windows of 30 trading days or fewer (BlackRock Geopolitical Risk Dashboard, 2024). Yet most organisations still rely on qualitative country-risk ratings that assign letter grades or ordinal tiers with no explicit weighting logic, no update frequency, and no linkage to financial outcomes.

This article presents the InekIA Geopolitical Risk Score (IGRS), a 12-dimension quantitative model built on publicly available data sources, validated against 15 historical crisis episodes, and designed to produce a single composite score (0–100) that correlates with measurable financial and operational outcomes. We detail the full methodology, data pipeline, weighting rationale, and three applied case studies—Taiwan Strait tension (2022–2024), Sahel military coup cluster (2021–2023), and Poland logistics corridor stress (2022–2024)—with results you can replicate in a standard spreadsheet or Python environment.

1. Why Existing Models Fall Short

1.1 The Qualitative Trap

The dominant commercial products in geopolitical risk assessment—Control Risks' RiskMap, Eurasia Group's GPRI, Oxford Analytica country briefs—are authoritative but structurally limited for quantitative integration:

• Update latency: Most flagship ratings update quarterly or annually. In fast-moving crises, a 90-day lag is operationally catastrophic.
• Opacity: Weighting schemes are proprietary. Analysts cannot audit why Country A scores worse than Country B.
• Binary triggers: Many systems flag 'elevated risk' without scaling severity, making portfolio hedging calculations impossible.
• No outcome linkage: Very few published models report correlation coefficients between their scores and subsequent GDP contraction, FDI withdrawal rates, or equity drawdowns.

1.2 Quantitative Benchmarks from the Literature

Academic work provides a stronger foundation. Caldara & Iacoviello's Geopolitical Risk Index (GPR) (2022), built on newspaper text analysis across 10 languages since 1900, demonstrates that a one-standard-deviation increase in the GPR index precedes:

• A 0.8% decline in global investment in the subsequent quarter
• A 1.2% increase in oil price volatility (30-day window)
• A 14 basis-point rise in sovereign CDS spreads for affected countries within 60 days

The IMF's World Economic Outlook (April 2024) estimated that geopolitical fragmentation could reduce global GDP by 2–7% relative to a baseline integration scenario by 2030, depending on the depth of bloc formation. These numbers justify the investment in rigorous scoring infrastructure.

2. The IGRS Methodology: 12-Dimension Framework

2.1 Dimension Architecture

The IGRS model organises 12 sub-indicators into three thematic pillars, each contributing a defined weight to the composite score:

Scoring convention: Each dimension is scored 0–10 (10 = highest risk). The composite IGRS is the weighted average, scaled to 0–100. Scores above 65 trigger 'Elevated' classification; above 80 trigger 'Critical.'

2.2 Data Sources and Update Frequency

Each dimension maps to at least two independent, publicly verifiable sources to reduce single-source bias:

A1 – Government Effectiveness: World Bank Worldwide Governance Indicators (WGI), annual; ICRG Political Risk Rating, monthly.

A2 – Rule of Law: WGI Rule of Law percentile rank; V-Dem Liberal Democracy Index (quarterly interpolation).

A3 – Political Violence Index: ACLED (Armed Conflict Location & Event Data Project) — event counts per 100,000 population, 30-day rolling window, updated weekly. Uppsala Conflict Data Program (UCDP) for baseline.

A4 – Coup/Succession Risk: Cline Center Coup D'état Project historical base rates + Powell & Thyne (2011) logit model recalibrated on 2010–2023 data. Output: annual coup probability %.

B1 – Active Armed Conflict Intensity: ACLED fatality-weighted event score + UCDP battle-related deaths, 90-day rolling window.

B2 – Territorial Dispute Severity: ICOW (Issue Correlates of War) project + Global Conflict Tracker (CFR), coded on a 0–4 ordinal scale (latent → active militarised).

B3 – Cross-border Spillover Risk: Composite of refugee outflow velocity (UNHCR), arms flow estimates (SIPRI), and geographic contiguity weighting.

B4 – Sanctions Exposure: OFAC SDN list additions per quarter + EU/UN sanctions regime coverage ratio.

C1 – Trade Dependency Concentration: Herfindahl-Hirschman Index applied to export/import partner concentration (UN Comtrade data).

C2 – Critical Resource Exposure: Share of global supply of any resource exceeding 25% single-country concentration (IEA, USGS Critical Minerals data).

C3 – Alliance Alignment Volatility: UN General Assembly voting divergence from Western bloc mean (rolling 3-year trend); bilateral defence agreement churn rate.

C4 – Cyber & Information Threat: Microsoft Digital Defense Report incident counts per country + National Cyber Power Index (Belfer Center) inverse score.

2.3 Normalisation Procedure

Raw data arrives in incompatible units (percentile ranks, fatality counts, HHI values, probabilities). Normalisation follows a min-max rescaling with anchor points set at the 5th and 95th percentile of the 2015–2024 global distribution for each indicator, preventing outlier distortion:

Normalised_Score_i = (Raw_i − P5_i) / (P95_i − P5_i) × 10

Values below P5 are floored at 0; values above P95 are capped at 10. This preserves variance in the critical middle range where most operational decisions are made.

2.4 Dynamic Weighting and Scenario Adjustment

The pillar weights (35/40/25) represent the baseline geopolitical environment. Two adjustment protocols apply:

1. Sector overlay: For energy companies, C2 weight increases to 35% and A-pillar decreases to 25%. For financial institutions, B4 (sanctions) doubles its within-pillar weight. Practitioners should document their overlay logic explicitly.

1. Velocity shock multiplier: When any single dimension changes by more than 2 points within a 30-day window, a 1.15× multiplier applies to that pillar's contribution for the subsequent 60 days. This captures the non-linear risk amplification documented in crisis escalation literature (Beardsley & Asal, 2009).