September’s $105 billion trade gap narrowed slightly as exports and imports rose. Different treatments of gold help explain why the headline and real goods figures point in different directions.
The United States’ goods and services trade deficit narrowed to $105.0 billion in September from a revised $105.7 billion in August, according to the joint release published by the Bureau of Economic Analysis and Census Bureau on 4 November. Exports reached $321.5 billion and imports $426.6 billion. Both increased, with exports rising faster.
The improvement was modest: the agencies reported a 0.7% reduction in the deficit. The goods shortfall remained partly offset by a services surplus of $30.9 billion. The release also showed a larger three-month average deficit, underlining how a single monthly movement can differ from the recent trend.
A smaller gap can accompany expanding trade
A trade balance is the difference between two large flows. It can improve because overseas demand for American output strengthens, because domestic demand for foreign output weakens, or because the two expand at different speeds. September belongs to that last category at the aggregate level. A narrower gap therefore does not, by itself, describe a retreat in imports.
That distinction matters for businesses reading the figures as a demand signal. An exporter wants to know what customers abroad are buying; a retailer or manufacturer wants to understand its own supply chain. The net figure compresses both decisions into one number. Examining the two sides separately gives a more useful starting point than treating every reduction in the deficit as the same economic event.
Rounding also matters when working with headline totals. Published dollar amounts and monthly changes are rounded independently, so subtracting two displayed figures need not reproduce every reported change exactly. Small discrepancies at that scale should not be mistaken for a different trend.
Why the goods measures diverge
The release contains another important qualification: the price-adjusted goods deficit on a Census basis increased, even as the headline goods and services balance improved. BEA identifies a balance-of-payments adjustment for nonmonetary gold as the reason its nominal goods measure and the Census nominal goods measure moved in different directions.
The agency’s explanation of its international accounts describes the underlying distinction. Census goods statistics primarily track goods moving through the customs system. BEA then makes adjustments to align that information with the coverage, valuation and definitions used in the balance of payments. The result is a related measure built for a different accounting purpose.
Gold illustrates why those adjustments can be material. Some transactions involve ownership changing between foreign official institutions and private American entities while the metal remains in storage at the Federal Reserve Bank of New York. Customs movements alone would not capture those transactions in the same way. The international accounts add the relevant ownership-based information.
This is a measurement issue with an economic consequence for interpretation. A change in the recorded value of gold transactions can alter the trade balance without telling the same story as an equivalent change in shipments of factory machinery or household goods. Readers comparing series need to establish that both use the same accounting basis.
The GDP calculation asks a different question
BEA applies a further distinction when translating international trade into the national accounts used for gross domestic product. Its national accounts guidance on nonmonetary gold separates gold used in production from gold held as a store of wealth. Purchases for investment in the metal are not treated like spending on newly produced equipment or other GDP investment.
For that reason, BEA removes the international-account gold trade flows and substitutes an adjustment based on domestic production and industrial use. Gold used to manufacture products, including jewellery and electronic equipment, has a different role from bullion acquired as a financial holding. The accounting aims to reflect that difference rather than transfer all recorded gold trade directly into GDP.
The practical implication is that this month’s total trade deficit cannot simply be inserted into a growth forecast. Analysts also need price-adjusted flows, a consistent accounting basis and the national-account treatment of particular transactions. The monthly release is an input to that assessment, rather than a complete estimate of trade’s contribution to output.
What the next reading needs to establish
For the outlook, the useful test is whether export demand remains firm across successive releases and whether imports reflect sustained demand for productive inputs, consumer spending or volatile transactions. A one-month improvement offers evidence about September, but cannot settle the direction of the next quarter or identify the effects of a particular policy on its own.
The September figures leave a measured conclusion: American cross-border trade expanded in nominal terms while the aggregate deficit edged lower. Understanding what that means for domestic production requires looking beyond the headline, particularly when gold and price adjustments cause related measures to move in different directions.
Sources
BEA and Census Bureau: September 2026 trade release
