Cross-Border Resilience After the Efficiency Era

What optimisation actually removed

The efficiency programmes of the last two decades were not mistaken on their own terms. Inventory is expensive, dual sourcing costs margin, and geographic concentration produces genuine scale economies. Firms that pursued these outperformed those that did not, for a long time.

What the exercise removed, however, was not waste in general but buffers specifically — and buffers were performing a function that only became visible when they were gone. The industry drew the conclusion that the slack was unnecessary, when the accurate conclusion was that the variation it absorbed had not appeared recently.

Three misdiagnoses

Post-disruption responses have converged on three ideas, each partially wrong.

Reshoring: moving production home reduces geopolitical exposure and frequently increases concentration risk, since a single domestic facility is a single point of failure regardless of jurisdiction. Location and concentration are different variables and are consistently confused.

Dual sourcing: a second supplier who shares the first's sub-tier supplier, port or power grid provides far less protection than the contract implies. Correlation matters more than count, and most dual-sourcing arrangements are not mapped past tier one.

Inventory: holding more of everything is expensive and mostly wasted, because disruptions are specific. The useful version is selective depth on components with long qualification lead times, where the recovery period is set by requalification rather than by shipping.

What actually helps

Four measures show a consistently better return.

Multi-tier visibility: knowing your suppliers' suppliers. Most firms cannot see past tier two, which is where correlated exposure typically hides. This is difficult, unglamorous data work and it is the foundation for everything else.

Design for substitution: qualifying components so alternatives exist before they are needed, since qualification during a shortage takes months you do not have.

Selective inventory depth, sized by requalification lead time rather than uniformly.

And contractual flexibility — the ability to shift volume between suppliers without penalty, negotiated in advance and therefore cheaply.

Making the trade explicitly

There is no configuration that is both maximally efficient and maximally resilient. The choice is real and should be made consciously, priced, and reviewed.

The practical method is to quantify the cost of disruption per product line — lost revenue, penalties, share loss, recovery cost — and let that figure set the resilience budget. Product lines differ by an order of magnitude, and uniform policies therefore over-protect some and under-protect others.

What is not defensible is a resilience posture arrived at through accumulated local efficiency decisions that nobody totalled. That is how most firms got here.