The 2023–2024 Red Sea shipping crisis rerouted container traffic from the Suez Canal to the Cape of Good Hope, reducing canal tonnage by 82% and extending Asia–Europe transit times by 10–14 days. This paper identifies the disruption severity thresholds at which three classical inventory policies—order-up-to (S), (s,S), and periodic-review (R,S)—transition from stable operation to operational failure. An analytical framework for the order-up-to policy under stochastic lead times is derived, calibrated to UNCTAD transit data, SCFI freight rate indices, and M5 retail demand patterns, and validated against a single-echelon discrete-event simulation. Per-cycle service level agreement between analytical predictions and simulation holds within 3 percentage points across disruption severities k=1.0 to 3.0 (R^2 > 0.99). Three principal findings emerge. First, all three policies breach the 85% cycle service level at disruption severities near k=1.04—equivalent to approximately one additional transit day on a 28-day baseline voyage—demonstrating that no classical policy possesses a meaningful buffer against maritime disruption. Second, probability-based service metrics degrade before expectation-based cost metrics, establishing a metric sensitivity hierarchy with a narrow early-intervention window between service degradation (k≈1.04) and cost doubling (k≈1.11). Third, a reactive base-stock policy with rolling parameter re-estimation recovers the target 91% service level in steady state but is constrained by an adaptation lag of 56–70 days, during which it performs identically to the static policy.