Validation

Backtesting & Historical Validation

We applied the WTM methodology retroactively to five historical crises. This is out-of-sample validation β€” the model was not trained on these outcomes.

🌑️
Key finding: Score above 70 β†’ real economic impact in all backtested cases
Five historical events tested. All scores above 70 preceded measurable inflation upticks, supply disruptions, or market corrections within 90 days.
Score timeline vs key events
Estimated WTM score had the model been running during these periods. Dashed line = 70-point threshold above which economic impact was observed.
COVID Supply Shock β€” March 2020
Est. score: 82 βœ“ Strong signal

Global lockdowns collapsed supply chains overnight. Oil demand evaporated β€” Brent futures briefly went negative ($-37/barrel). Central banks cut to zero. The model's energy and trade signals would have spiked simultaneously, with media sentiment at extreme levels.

βˆ’$37
Oil price at peak panic (futures)
βˆ’32%
S&P 500 peak-to-trough
85.5
VIX peak (highest since 2008)
Q2 2020
Global recession confirmed
Signal drivers: Energy 95, Finance 88, Trade 82, Media 90, Conflict 45
Russia-Ukraine Invasion β€” February 2022
Est. score: 85 βœ“ Strong signal

The largest land war in Europe since WWII. Russia is the world's largest gas exporter and 2nd largest oil producer. Sanctions and energy supply collapse drove inflation to generational highs across Europe and the US. All five WTM signals would have been simultaneously elevated β€” the defining test case for the model.

+300%
European natural gas price spike
9.1%
US CPI peak (Jun 2022) β€” 40yr high
11.1%
UK CPI peak (Oct 2022) β€” 41yr high
βˆ’20%
S&P 500 decline in 2022
Signal drivers: Conflict 92, Energy 88, Trade 85, Finance 72, Media 89
Red Sea / Houthi Crisis β€” January 2024
Est. score: 71 ~ Moderate signal

Houthi attacks on Red Sea shipping forced vessels to reroute around the Cape of Good Hope β€” adding 10–14 days and significant fuel costs to Asia-Europe routes. Supply chain stress was real but geographically contained. The WTM energy and trade signals elevated but conflict and finance remained moderate β€” correctly reflected in the mid-range score.

+150%
Container shipping rate spike
+14 days
Asia–Europe route extension
+0.3%
Estimated CPI impact (modest)
Q1 2024
Supply chain delays β€” electronics, autos
Signal drivers: Energy 70, Trade 82, Conflict 68, Finance 55, Media 74
US Tariff Escalation β€” April 2025
Est. score: 74 ~ Moderate signal

Sweeping US tariffs on goods from China, EU, Canada, and Mexico. The trade signal would have spiked sharply; finance signal elevated as markets fell; media sentiment high. Energy and conflict signals remained moderate β€” correctly capturing this as an economic rather than military crisis.

+8–14%
Import price increase (affected categories)
βˆ’15%
S&P 500 from peak
Revised ↓
US GDP growth forecasts
Accelerated
Supply chain diversification from China
Signal drivers: Trade 88, Finance 74, Media 78, Energy 52, Conflict 62
US-Israel-Iran War β€” March 2026
Live score: 93–98 ⟳ Active β€” ongoing

Direct military conflict involving a nuclear-capable state near the world's most critical oil chokepoint (Strait of Hormuz). All five signals simultaneously at extreme levels β€” the first time this has occurred since model launch. IEA warns this is the worst energy shock since 1973. Outcome still unfolding.

$108+
Brent crude price (Apr 2026)
$4,682
Gold β€” all-time record
3 nations
Asia energy emergencies declared
20%
Global oil supply at Hormuz risk
Signal drivers: Conflict 90, Energy 93, Trade 90, Finance 48, Media 92

Methodology note: Backtested scores are estimates, not exact historical values. The model was applied retroactively using the same formulas, weightings, and signal definitions currently in production. Results should be read as directional validation, not precise calibration.

The false-positive rate (score above 70 without measurable economic impact) has not yet been observed in this dataset but cannot be ruled out in lower-stress environments. The model is weakest at detecting slow-burn deterioration (structural factors that build over years without acute events).

Full methodology β†’ Download whitepaper β†’ Raw data (JSON) β†’
Related analysis
πŸ“– Iran War & Oil Prices
Latest active event analysis
πŸ“– WTM Score at 98
What the record reading means
πŸ“‹ Methodology
Full scoring documentation
🌍 Live Score
Current World Tension Score
πŸ“Š Score History 2026
Daily dataset β€” open JSON
πŸ“– All Analysis
21 geopolitical risk articles