For the past two decades, global supply-chain design revolved around a single objective: push unit cost as low as possible by concentrating production in the few locations with the greatest cost advantage, compressing inventory toward zero, and timing delivery down to days or even hours. That logic performed extremely well in a world of steady demand, unobstructed transport and predictable policy, and it drove much of the manufacturing cost curve's decline over recent decades. But a run of shocks in recent years — capacity disruption during the pandemic, sudden production stoppages in specific regions, and shortages of critical components caused by geopolitical friction — has exposed the other side of that logic in concentrated form: as the probability of a single node failing rises, betting everything on the one node with the best efficiency means carrying a tail risk that had rarely been priced before.
We call the adjustment now under way supply-chain rewiring rather than supply-chain relocation, because most companies are not moving production away from existing sites wholesale — they are adding redundant nodes alongside the existing network, shortening the physical distance for critical steps, and shifting some inventory from zero-stock toward safety-stock policy. This is addition, not substitution, and understanding that distinction is the premise for judging where capital should go. Read it as large-scale production relocation and you will overstate how fast any given region can absorb new capacity. Read it accurately as a systematic adjustment to network redundancy and inventory policy, and capital demand shows up spread across a wider set of steps, a good number of which have nothing to do with where production moves and everything to do with making the existing network better able to absorb a shock.
Regionalization, nearshoring and the inventory pendulum
The core expression of regionalization is not a rise or fall in any single country's share of production, but a growing tendency for companies to organize supply networks around end markets — a relatively self-contained supply chain for the North American market, another for Asia-Pacific, sharing some core technology and component standards while production and warehousing sit in geographically distinct locations. The motivation behind that split comes mainly from two directions: rising uncertainty around tariffs and compliance costs makes the economics of cross-region transport hard to lock in over the long term, and end markets keep demanding faster delivery, which shortening physical distance addresses on its own, without depending entirely on upgrading how goods are transported.
A pendulum-style shift in inventory policy is the other specific change we track. Over the past decade, zero inventory was close to the standard answer sought across manufacturing and retail, with inventory treated as pure cost and its compression treated as synonymous with efficiency gain. That consensus has visibly loosened in recent years, as more companies deliberately raise safety-stock levels for critical components and higher-margin products, even at the cost of some turnover and warehousing expense. This is not a rejection of the efficiency logic; it is a repricing. Inventory is no longer counted only as holding cost — it is also counted as an insurance premium against the probability of a supply interruption, a premium that had barely entered the decision model before.
- Network redundancy: adding a second or third production node alongside the original site, even at a higher unit cost, to reduce the impact of a single node's failure on overall delivery, rather than to replace existing capacity.
- Higher safety stock: shifting inventory levels for critical components and higher-margin products from just enough to a margin of buffer — in effect, pricing supply-interruption probability into the inventory decision.
- Supplier diversification: shifting core components from a single qualified supplier to at least two qualified suppliers running in parallel, even at the cost of giving up some volume discount from a sole supplier.
- Investment in visibility: companies are increasingly willing to pay for knowing in advance which link in the chain might fail, not only for moving goods faster. The return on this kind of spending shows up in risk avoided, which is hard to see directly in any single quarter's financial statements.
Where the capital lands: networks, cold chain, automation and visibility software
Break this round of capital spending down by category, and the distribution we observe is uneven. The first category is the physical network itself — new warehousing nodes, expanded regional hubs. This is the largest category by dollar value and the easiest for outside coverage to capture, but the return per unit of capital tends to be lower, since it is essentially duplicating existing capability in a new location. The second category is cold chain and temperature-controlled facilities, where capital demand is growing faster than for ordinary warehousing across most sub-sectors, because pharmaceutical, fresh-food and some precision-manufacturing customers keep raising their temperature-compliance requirements, and compliant cold-chain capacity takes years to build and approve — early movers accumulate a relatively scarce asset that is hard to replicate quickly.
The third category is automation equipment, spanning in-warehouse sorting, material handling and parts of the production line. Rising labor cost is a factor here too, but in this particular cycle a more direct driver is uncertainty in labor supply — when a company needs a new node to reach usable capacity within a shorter window, an automated system that can be deployed quickly, without depending fully on a hiring-and-training cycle, becomes more attractive. The fourth category is visibility and supply-chain management software, the smallest in absolute capital terms among the four but the fastest-growing in penetration, because it solves a problem that redundancy itself inevitably creates: once a supply chain moves from one clear primary line to several parallel branches, the complexity that redundancy introduces will consume the very certainty redundancy was meant to deliver, unless management has a system that shows the state of the whole network in something close to real time.
Relative capex scale — Physical network: 42, Cold chain & climate control: 27, Automation equipment: 21, Visibility software: 10
Illustrative framework reflecting a relative directional trend, not a precise statistical measure
Who compounds value, and who is riding a cycle
The capital-spending increase driven by supply-chain rewiring will benefit a considerable number of logistics and industrial companies over the next several years, but the nature of that benefit differs sharply. Revenue growth at some companies will fade once this round of capex peaks; at others, this round of spending will convert into a cash-flow base that persists for years afterward. In separating the two, we weight not current contract value or capacity utilization but three more structural questions: whether a company's assets carry genuine scarcity, whether its customer relationships are deeply embedded in the customer's own operating rhythm, and what share of its revenue comes from one-time construction spending versus recurring service spending.
Take cold-chain warehousing and contract logistics as an example: the way value compounds there differs entirely from an engineering firm that simply captures one-time construction capex. Once a customer's warehousing and delivery rhythm is deeply tied to an operator's network, switching providers carries a real operational-disruption cost, and that stickiness keeps generating recurring service revenue well after the construction peak has passed. A company that purely provides warehouse construction or equipment installation services, by contrast, sees revenue track the capex cycle itself closely — order books are full on the upswing and revenue comes under immediate pressure on the downswing, because such a company is essentially selling cyclical construction capacity rather than accumulating a customer relationship that strengthens over time. The most direct observable for telling which category a logistics or industrial company falls into is the share of revenue coming from renewals and expansions with existing customers, not the absolute size of newly signed contracts.
One concrete window we watch
Mingchuan Logistics is one concrete window through which we track this trend. The company's rising share of pharmaceutical customers in recent years lines up closely with the broader shift toward tightening compliance requirements and rising demand for specialized third-party cold-chain logistics, and the direction of its capital deployment — concentrated in temperature-controlled facilities and warehousing capacity at a small number of hub cities, rather than expanding coverage indiscriminately — fits the kind of asset discipline we associate with value that compounds. When we look at a company like this, rather than asking which new cities it has added warehousing square footage in, we would rather ask whether its long-term contract renewal rate has held steady as cold-chain capacity expands. That measure says more about whether network growth rests on genuine demand than the new capacity figure does on its own.
Supply-chain rewiring is still in an early stage, and most companies' network adjustments will play out over several years rather than complete within one or two financial cycles. That means investment judgment on this theme has to return repeatedly to the same set of structural questions — asset scarcity, depth of customer embedding, whether revenue rests on a renewal base — rather than being pulled along by any single capex peak or an impressive list of newly signed contracts. Our tracking of this theme will keep updating, because the path of regionalization and nearshoring is itself still evolving, and any conclusion reached too early carries the risk of being overturned by the next shift in policy or demand.