Decision Resilience — economic interpretation

The 38% resilience number from chapter 1.3.7 isn't an abstract metric. It's the probability the decision still holds across the realistic distribution of futures. The 62% of futures where it breaks each have a dollarized cost — the expedite needed to recover, the SLA penalty triggered, the displaced higher-margin order, the supplier-relationship cost of the last-minute reroute. The expected cost-of-break is the probability-weighted sum. A probability times a dollar cost is a dollar number.

Resilience-aware auto-commit composes three axes. The Decision Policy's autonomy threshold isn't just "auto-commit if value at risk is below $X." It's "auto-commit if value at risk is below $X and resilience is above Y% and EVW is below $Z." Three axes, all dollarized, all configurable per decision class. The decisions that survive all three gates are the ones the system signs on its own; everything else surfaces.

Planning resilience and execution resilience are distinct. Planning resilience is about not signing brittle plans — the lookahead score before commit. Execution resilience is about the plan absorbing shocks gracefully once committed — the event-driven loop's memory and scoping. VYAN handles both, and the two interact: a high-planning-resilience commit is usually also high-execution-resilience because the characteristics that made it survive simulated futures also make it survive actual ones. But not always — and the execution resilience score updates as the plan ages.

The CVaR connection makes resilience-aware planning rigorous. CVaR (Conditional Value at Risk) is the expected value conditional on being in the lower tail of the distribution — the average of the worst N% outcomes. A policy that maximizes expected EVA without considering tail risk produces plans that break catastrophically in the bad scenarios; a CVaR-aware policy gives up some expected EVA for tail-risk insulation. The risk posture's percentile choice (P75 vs P90 vs P95) operationally encodes a CVaR preference. The CFO who prefers $4M of expected EVA with a P5 of $1M over $5M of expected EVA with a P5 of −$3M is making a CVaR choice. VYAN makes that choice explicit and dollarized rather than buried in master-data settings.