June Trade Data Cannot Yet Distinguish Frontloading From a Durable Sourcing Shift
June imports fell, but the data cannot distinguish a frontloading unwind from a durable sourcing shift without product, timing, inventory, and supplier records.
Primary lensTrade policy
Sub-topicPolicy monitoring
Evidence base5 records used
Use casePolicy monitoring
The U.S. goods and services deficit fell to $73.3 billion in June, but the release does not show whether importers were unwinding tariff frontloading or making a durable sourcing change. The monthly deficit fell $4.4 billion while the three-month average deficit rose $5.6 billion. June goods imports declined, yet the products driving the move were uneven. Computers and pharmaceutical preparations fell while some other categories increased. The June headline is a starting point, not evidence of a lasting sourcing shift.
A usable test needs five kinds of evidence: the public baseline, product value and quantity, timing across orders and entry events, inventory movement, and supplier commitments. The legal boundary matters too. June ended before duties under the 60 economy-specific Section 301 actions attached to covered entries on July 24. The June series therefore cannot measure the duties' direct effect, although June orders or arrivals could still reflect anticipation of the expected actions. Import teams should preserve potentially covered consumption entries and warehouse withdrawals on or after July 24, separately identifying product and agreement exemptions and qualifying in-transit goods, together with the commercial decisions associated with those transactions. They can then test whether lower imports reflect stock drawdown, delayed orders, price and product mix, or capacity that actually moved.
The Monthly Result and the Moving Average Disagree
The U.S. Department of Commerce, U.S. International Trade in Goods and Services, June 2026 gives two windows ending in June. The monthly goods and services deficit fell 5.6 percent from May to $73.3 billion. Imports fell $7.3 billion, more than the $2.9 billion fall in exports. The three-month moving average, covering April through June, moved the other way. Average imports rose $4.2 billion and the average deficit increased $5.6 billion to $68.5 billion.
Neither window is the correct one by default. The monthly change is more current and more exposed to shipment timing. The moving average is steadier and slower to register a turn. A frontloading unwind can push the monthly number down after inventory arrived early, while the average still carries the earlier surge. A durable sourcing change can also start with one down month. The difference appears only after the product and business records are matched to the public series.
The recent sequence shows the risk of choosing one month. The deficit was $54.6 billion in April, rose to $77.6 billion in May, then fell to $73.3 billion in June. That is not a clean downward path. It is a volatile three-month run in which imports, exports, gold, petroleum, capital goods, and consumer goods moved differently.
The Baseline Determines the Size of the First-Half Decline
The first-half comparison depends heavily on the year chosen. The official release reports that the January through June 2026 deficit was $371.2 billion, down $189.3 billion or 33.8 percent from the same period in 2025. The U.S. Department of Commerce, U.S. International Trade in Goods and Services, June 2026, Exhibit 1 shows a $401.6 billion deficit for the first half of 2024. Against that earlier base, the 2026 decline is about 7.6 percent.
That evidence is contextual rather than a decomposition of the FT-900 deficit. The WTO series covers North America, merchandise imports, volume, and one quarter. The FT-900 comparison covers the United States, goods and services, nominal dollars, and six months. The WTO result therefore supports frontloading as a hypothesis to test, but it does not measure how much of the 2025 deficit came from frontloading or establish 2024 as an undistorted baseline.
Both FT-900 comparisons should be stated. Product-level U.S. value and quantity data, together with company order and inventory records, are needed to determine whether either difference reflects an early-order surge, a later unwind, or a durable sourcing change.
The second is the gap between the statistical and customs clocks after July 24. FT-900 general imports measure physical arrivals, including merchandise that enters consumption channels, bonded warehouses, or foreign-trade zones. The duty trigger follows a consumption entry or warehouse withdrawal. A monthly arrival series will therefore mix goods with different entry dates and tariff treatments.
The third concerns import values. Customs value generally excludes U.S. import duties, freight, insurance, and other arrival charges. A later fall in import value will not measure the tariff deposit, and a stable value will not show that the deposit stayed the same. The operative USTR Policy Signal belongs beside the statistical series, with the entry population linking the two.
Import Value Does Not Show Whether Fewer Units Arrived
June imports of goods fell $7.9 billion on a balance-of-payments basis. On a Census basis, the decline was $7.7 billion. Capital goods fell $2.1 billion, led by a $3.0 billion drop in computers. Consumer goods also fell $2.1 billion, including a $1.9 billion decline in pharmaceutical preparations.
Those values do not reveal whether fewer units arrived. A price decline, a shift toward cheaper models, or a change in product mix can reduce value without the same change in physical volume. The release's real-goods table helps at the aggregate level. Real imports fell 2.6 percent while nominal Census-basis goods imports fell 2.4 percent. The close rates suggest that the June decline was not solely a broad price effect, but they do not resolve individual products.
The product test needs both value and quantity at the most stable HTS level available. If value falls while quantity holds, examine price and mix. If both fall, check whether orders were moved into an earlier month, delayed, canceled, or shifted to another origin. If an HTS line lacks a reliable quantity measure, the company record becomes more important, not less.
Five Tests for Frontloading and Durable Sourcing
The five tests below join a monthly macro release to product, timing, inventory, and supplier records. No single record can distinguish a frontloading unwind from a durable shift. The evidence must align before either conclusion is supportable.
Test
Public or company record
Pattern consistent with a temporary unwind
Evidence needed for a durable sourcing change
Baseline
Monthly, three-month average, year-over-year, and 2024 comparisons
The decline is concentrated against the elevated 2025 comparison, while product and order records show an earlier surge and later drawdown
The direction persists in later periods and is supported by supplier, qualification, and committed-capacity records
Product
HTS value and quantity, end-use category, unit value
Early surge and later fall in the same product with no supplier change
Quantity migrates to a new origin or product set and remains there
Orders were advanced and later arrivals return toward the prior schedule
New order cadence begins after a documented sourcing decision
Inventory
Receipts, on-hand units, withdrawals, production use, and backorders
Purchases fall while the business draws down stock accumulated earlier
Inventory policy normalizes while new-origin replenishment continues
Supplier
Contract, award date, capacity reservation, tooling, qualification, and cancellation terms
Old supplier and capacity remain in place after a temporary pause
Volume, tooling, qualification, and committed capacity move to the new supplier
A before-and-after chart cannot make this distinction. Frontloading primarily changes timing, while durable sourcing should leave evidence that supplier relationships or capacity have moved. Customs arrival data capture only part of that record.
Durable Sourcing Leaves a Commercial Record
A new country label does not by itself establish a durable shift. The importer should be able to point to a sourcing decision date, approved supplier, qualification file, tooling or process transfer, committed capacity, revised lead time, and purchase orders that continue beyond the first post-tariff shipment. The prior supplier's cancellation rights and retained capacity matter too. A second factory can be an option rather than a replacement.
Traverse's China Plus One Tariff Costs and the USTR Section 301 Continuity Premium separates current landed-cost savings from the cost of keeping alternative capacity ready. The same separation belongs here. A temporary order move can reduce one month's imports without changing the long-run supplier network. A durable move can raise transition costs before it lowers exposure.
Origin and tariff treatment still need their own support, but they are not the controlling question in this test. The task is to determine whether the business changed the location and capacity on which future supply depends. That conclusion should be traceable to decisions made before the later trade release appeared.
What Import Teams Should Do Before the Next Release
For potentially covered consumption entries and warehouse withdrawals on or after July 24, retain the ordinary HTS line, Chapter 99 heading, origin, Customs value, exemption or preference claim, estimated duties deposited, and liquidation status. Classify duty-bearing, product-exempt, agreement-exempt, and qualifying in-transit entries separately.
Join that population to purchase order date, supplier, production location, booking, departure, arrival, warehouse receipt, inventory withdrawal, and final unit use. Keep corrections as dated revisions rather than overwriting the first record. The sequence should show when the commercial decision occurred, when the goods physically moved, when tariff treatment attached, and when inventory reached production or sale.
Price pass-through requires a separate analysis. Traverse's Tariff Impact on Prices Through CBP Entry and Inventory Records shows how to connect a duty-bearing entry to inventory and the later SKU price. That record does not determine whether the import decline reflected a frontloading unwind.
What the September 3 Release Can and Cannot Show
The next FT-900 release, scheduled for September 3, will add July arrivals. It will not isolate entries subject to the 60 Section 301 actions that took effect on July 24, and it will not show when the underlying orders were placed. August will be the first full calendar month after the effective date, but its arrivals will still mix earlier orders, warehouse timing, exemptions, and different product cycles.
Use the next release to extend the public value and quantity series. Use company records to test the order, entry, inventory, and supplier clocks. What would change the calculus is convergence across those records. If lower imports persist while accumulated inventory is exhausted and committed capacity has moved, the durable-sourcing case becomes stronger. If imports rebound when stock returns to normal and the supplier network is unchanged, the earlier decline fits a timing unwind more closely. Until those records converge, the June trade deficit is a starting observation, not a tariff result.
From reading to review
Run the numbers on your lane.
The duty calculator runs the current stack for any HTS code and origin. A free account opens full tool output, AD/CVD detail, Chapter 98 processing, and available exports.