Corridor Disruption and Supply Chain Vulnerability: A NASCDI-NARDL Analysis of Kashmir’s Apple Sector

Abstract Agricultural supply chains that depend on a single transport corridor are structurally exposed to disruption, yet price-transmission studies rarely measure disruption intensity directly, relying instead on binary event indicators or case narratives. We construct the Nonlinear Asymmetric Supply-Chain Disruption Index (NASCDI), a weekly, reproducible measure of highway-disruption intensity built from GDELT Global Knowledge Graph metadata, and apply it to Kashmir's apple trade, where nearly all output moves through a single highway, NH-44, to the Azadpur terminal market in Delhi. Using a nonlinear autoregressive distributed lag (NARDL) framework across eight producer-market-variety-grade chains (1,276 matched weekly observations, 2015–2025), we find universal long-run cointegration between producer and terminal prices, but adjustment speed differs sharply by origin: Shopian chains re-equilibrate roughly twice as fast as Sopore chains. Asymmetric transmission, intensification versus easing, is limited and chain-specific rather than pervasive, significant in only one chain at the long-run horizon. A composite vulnerability index combining disruption sensitivity, adjustment speed, and model support, validated by a 10,000-draw bootstrap, shows individual-chain rankings are only moderately stable but that Shopian is more exposed than Sopore as a market in 74% of simulations. We translate these findings into a real-time monitoring design and an illustrative parametric-insurance trigger, and show the NASCDI methodology is transferable to other corridor-dependent agricultural systems.