ρᵢⱼ = Σₖ wₖ · min(βᵢₖ, βⱼₖ) / Σₖ wₖ · max(βᵢₖ, βⱼₖ)
Scenarios
9
Drivers
7
Avg ρ
—
Div. benefit
—
ρᵢⱼ = Σₖ [ wₖ · min(βᵢₖ, βⱼₖ) ] / Σₖ [ wₖ · max(βᵢₖ, βⱼₖ) ]
wₖ = driver importance weight (0–1)
βᵢₖ = scenario i exposure to driver k (0, 0.3, 0.6, 1.0)
min / max = shared overlap / total exposure
Intuition: how much of the combined driver exposure is shared?
0 = no overlap → ρ = 0  ·  perfect overlap → ρ = 1
Driver weights amplify the most important root causes
1
Scenarios
2
Driver exposure — click to set level
None Low Med High
3
Driver weights

Correlation Matrix

Derived directly from driver overlap — hover any cell for the breakdown
Scale:
<0.15
0.15–0.30
0.30–0.50
0.50–0.65
>0.65
4
Capital aggregation (£m)
Correlation floor sensitivity
5
Ranked pairwise correlations
Pair Shared drivers ρ Magnitude