When Nearshoring Rewrites the Network: Cross-Border Capacity Before Peak
Nearshoring is shifting freight toward Mexico gateways while Laredo capacity stays tight. What asset, broker, and hybrid leaders should decide before peak.
The failure mode is not “we lost a load at the border.” It is accepting a domestic tender that strands a truck on the wrong side of a network that has already moved south, then discovering your direct B-1 cover and your domestic empty plan no longer match the same map.
That is the September 2026 decision for North American operators. Freight is not vanishing. It is relocating. DAT’s dry van report for the week ending September 11 (published September 15) put the signal in plain language: the freight is moving, just not where it used to. ISM manufacturing panelists described customers relocating production from U.S. plants to Mexico and pushing more sourcing offshore. That pull concentrates volume on cross-border lanes, especially through Laredo and South Texas, and it reshapes domestic linehaul rather than simply adding to it.
At the same time, capacity has not rebuilt into a soft market. For that same DAT week, dry van spot linehaul averaged $2.20 per mile excluding fuel, still 34.2% above the year-ago week and 21.0% above the nine-year seasonal average of $1.81. Load posts fell 18.0% week over week on the holiday short week, but truck posts remained 40.2% below a year earlier. The load-to-truck ratio eased to 10.95 from 11.47, yet stayed more than double the 5.33 print from the same week in 2025. A national average that looks “stable” still hides a network that is expensive to mis-position.
Uber Freight’s Q3 Market Update (published around September 10–11, 2026) rated truckload and Mexico as high exposure heading into late October. That is the planning window. Leaders who treat cross-border as an exception lane will keep paying for it as a recurring surprise.
What nearshoring actually changes in ops
Nearshoring does not only add Mexico-origin freight. It changes where empty miles get paid, which origins stay bid-worthy, and which “domestic” tenders are really the last mile of a border plan.
DAT’s top dry van origin rates for the week ending September 11 ran from Ohio River at $2.52 and Great Lakes at $2.51 down to South Central at $1.95 and Florida–South Georgia at $1.44. The top 10 origins carried 87.7% of U.S. outbound loads that week. A desk that still prices empties and protect lists off a national $2.20 print will overpay into weak destinations and underprice exits from industrial corridors that still clear.
Layer the ISM nearshoring comments on top of that spread. Volume tied to Mexico production and data-center or electrical-infrastructure buildout does not distribute evenly across the old consumer replenishment map. Industrial and project freight can stay firm while consumer-goods cycles get choppier. That is a network design problem, not a one-week rate problem.
For carriers, the decision is whether a lane earns its keep after the empty into the next headhaul. For brokers, it is whether a customer’s “standard” routing guide still reflects where their inventory actually sits. For hybrids, it is whether asset miles and brokered cover are solving the same geography or fighting each other.
Why Laredo can ease and still stay tight
Cross-border capacity loosened from its Q2 peak, which tempts teams to relax. The floor is still wrong for that read.
Uber Freight estimated about 20,000 Mexican drivers lost U.S. visas between April 2025 and April 2026, with active Mexican-domiciled southern border carriers 6.3% lower in late June versus late December. Mid-August Laredo dry van load-to-truck sat between 8.0 and 8.5, down from roughly 10-to-1 in Q2, but still 61.9% higher year over year. Mexico-to-U.S. long-haul spot rates were still 8% to 15% above mid-February levels on many moves, with critical corridors up as much as 30%. Produce exports through Laredo rose 8% year over year in Q2.
That combination (fewer compliant drivers, sticky gateway ratios, elevated corridor pricing) means “eased from extreme” is not the same as “available on demand.” Tariff and inspection volatility made the planning problem worse in Q3. Uber Freight noted repeated policy shifts (Section 122 expiring, Sections 301 and 338 taking effect) that pulled freight forward ahead of changes, then slowed orders for weeks. Gateway dwell and appointment discipline become capacity, not paperwork.
Canada is not a free pass either. Domestic Canadian manufacturing looked healthier in mid-2026 (S&P Global Canada Manufacturing PMI at 53.5 in July), but Canada–U.S. cross-border capacity remained more carrier-favorable and volatile, with visa and enforcement limits still thinning the driver pool. Uber Freight rated Canada medium severity into late October, below Mexico’s high rating, not zero.
Transload is no longer a workaround
The operational tell is how serious shippers are rebuilding the border move itself.
Uber Freight’s report framed transloading as moving from workaround to network design. Shippers that relied only on direct-trailer B-1 capacity are adding cross-dock options in Laredo and starting to explore El Paso. One beverage manufacturer example in the report had been missing appointments in Nuevo Laredo on a pure direct model, then rerouted time-sensitive freight through a Laredo cross-dock while keeping direct B-1 for more flexible freight. That hybrid is a product decision: cost versus speed, appointment risk versus handling risk, one handoff versus two.
For asset fleets, the choice is whether to own border dwell and customs exception work, or to partner and keep power on domestic pivots. For brokers, the choice is whether your cover playbook can flip between direct and transload without a phone tree and a spreadsheet. For hybrids, the choice is which entity holds the appointment clock, the detention dispute, and the customer promise when the gateway stalls.
Intermodal sits next to that decision, not outside it. U.S. intermodal volume was up 3.8% year to date through August 22, and providers applied peak-season surcharges in early July, about two months earlier than usual. Uber Freight reported rate increases of 10% or more in constrained markets such as Los Angeles and Laredo, with Los Angeles peak surcharges in a $500 to $1,000 per box range expected through year-end. Modal mix that looked like cheap relief in spring can become another constrained lane by peak. Re-evaluate lane by lane. Do not assume last quarter’s crossover still holds.
Asset vs broker vs hybrid: different clocks, same geography
Asset carriers own the empty and the hours. A nearshoring shift that concentrates headhauls into South Texas or industrial Midwest origins forces a harder protect list: which destinations you will not exit without a prepaid reposition or a contracted outbound. Border dwell burns HOS and next-day capacity the same way a slow DC does. If your statuses treat customs hold, cross-dock queue, and live unload as the same “delay,” you will mis-price the next day’s truck.
Brokers own cover optionality and customer trust. Primary tender acceptance on Uber Freight’s book improved from 76% in July to 78% in August after lane repricing, still well below the 90% to 94% range of the prior three years. SONAR’s U.S. Outbound Tender Rejection Index was 13.45% on September 10, higher than the same window in each of the prior three years. On cross-border, optionality means pre-cleared direct capacity, named transload partners, and a defined spill path when the gateway rejects your first plan. Week-to-week spot buying into Mexico-rated high exposure is how peak turns an ordinary tender into an all-in surprise.
Hybrids get hurt when asset and brokerage systems disagree on the same load. If the asset side will not deadhead into Laredo without a floor, and the brokerage side sold a customer a direct appointment promise, the company has encoded a conflict, not a network. Peak is a bad time to discover that in the chat thread.
Encode uniqueness where the exceptions live
Every freight business runs a different border story: some customers insist on direct trailers, some accept transload, some need both by SKU or by service level. Serious shops encode that uniqueness in rules, not in tribal knowledge.
That usually means customized workflow logic more than a new org chart. Border and customs statuses that are distinct from generic dwell. Routing rules that choose direct versus transload by customer, commodity, and appointment criticality. Effective-date handling for tariff and policy changes so sales and ops stop arguing from last month’s playbook. Permissions that control who can flip mode or add a cross-dock without waiting for a VP. Visibility that shows gateway queue and document holds to the people who can still recover the appointment. Capacity logic that blocks domestic accepts which strand power away from contracted cross-border commitments.
None of that is sales theater. It is how you keep two networks (domestic industrial and cross-border gateway) from silently taxing each other when September still looks “quiet” and late October does not.
What to decide in the next 30 to 60 days
Map the real geography. Separate Mexico-gateway dependent freight from pure domestic. Score lanes on empty risk out of weak destinations, not only on inbound rate.
Pick a default border model per customer. Direct, transload, or hybrid with explicit flip rules. Write the appointment and detention ownership into the rate con path.
Secure baseline and backup before peak. Uber Freight’s advice for the September–October window applies hardest where exposure is rated high: truckload and Mexico. Named backups beat heroic phones.
Reprice the crossover. Truckload versus intermodal versus partial or LTL spill should be rechecked where intermodal peak pricing and gateway tightness have already moved the math.
Align systems to the exception path. If your statuses, permissions, and visibility still assume a single domestic live-load world, the next policy or inspection surge will recreate the same fire drill.
The market can look sideways in national averages and still punish anyone whose network map is a year behind their customers’ production map. Nearshoring is already rewriting where freight starts. Peak will test whether your cover, empties, and border playbooks were rewritten with it.
FAQ
What is the main cross-border risk for truckload operators heading into Q4 2026?
Gateway capacity can ease from a Q2 extreme and still stay structurally tight. Uber Freight’s mid-August Laredo dry van load-to-truck ratio of 8.0 to 8.5 remained 61.9% higher year over year after large visa-driven driver losses, so peak demand hits a thinner compliant pool.
How is nearshoring changing domestic dry van networks?
It pulls volume onto Mexico-linked gateways such as Laredo and South Texas and reshapes domestic linehaul instead of simply adding nationwide demand. DAT’s September 15 dry van analysis tied ISM panel comments on Mexico relocation and offshore sourcing to that geographic rewrite.
Should shippers still lean on intermodal to offset truckload cost?
Sometimes, but not on last spring’s assumptions. Intermodal volume was up 3.8% year to date through August 22, peak surcharges started about two months early, and constrained markets such as Los Angeles and Laredo already saw rate pressure, so modal mix needs a lane-level reset before peak.
Why does primary tender acceptance around 78% matter?
It means routing guides are still failing often enough to force spot cover. Uber Freight’s August primary tender acceptance of 78% improved from July’s 76% but remained far below the 90% to 94% range of the prior three years, which is fragile heading into peak.
What should asset carriers change first?
Protect lists and empty rules around gateway and industrial origins. If a domestic accept burns the truck that was supposed to protect a contracted border or high-RPM industrial exit, the “won” load is a network loss.
What should brokers change first?
Pre-agreed direct and transload options with clear flip authority, plus customer-specific service rules. Cross-border is where phone-tree cover collapses first when inspections or tariff timing swing volume.