Government data

US goods and services trade deficit, July 2026

What will the first-print U.S. Census Bureau and BEA seasonally adjusted goods and services trade deficit be for July 2026, balance of payments basis, in billions of U.S. dollars?

Forecast

$69B
$56.1B–$81.9B
80% prediction interval
gpt-5.5Run record ↗

Cumulative probability

Chance that the outcome is at or below each value.

0%25%50%75%100%$36.8B$69B$101.3BForecast value

Shaded band: 80% interval ($56.1B–$81.9B). Dashed line: point estimate ($69B).

Derived from the point estimate and 80% interval; the agent did not report a full distribution.

Observed outcomeoutside 80% interval
actual
$88.6B
forecast
$69B with 80% interval [$56.1B, $81.9B]
error
+$19.6B · absolute $19.6B
cdf score
CRPS 13.5 · PIT 0.93
source
bea_census_trade U.S. International Trade in Goods and Services, Exhibit 1

First print for 2026-07 captured from https://alfred.stlouisfed.org/graph/alfredgraph.csv?id=BOPGSTB&vintage_date=2026-09-03 on the official release date named by the cell's resolver.

Analysis

§Forecast: July 2026 U.S. goods and services trade deficit

Framing and exact resolver: this targets the July 2026 first print of the U.S. International Trade in Goods and Services headline goods-and-services balance, seasonally adjusted and on a balance of payments basis, reported in Exhibit 1. The ledger transform turns BOPGSTB's negative millions-of-dollars balance into a positive usd_billions deficit; the agency first print is authoritative over later revisions.

▸ Reported tool use: official.lookupmodel report
Checked BEA 2026 release schedule and the BEA scheduled-release node for U.S. International Trade in Goods and Services, July 2026.
↳ BEA schedule lists U.S. International Trade in Goods and Services, July 2026 on September 3, 2026 at 8:30 a.m.; BEA node 42997 also shows Scheduled Date September 3 08:30 and Release Title July 2026.
▸ Reported tool use: official.lookupmodel report
Fetched the latest official BEA/Census release for May 2026 to anchor the same variant and near-term components.
↳ May 2026 goods and services deficit was 77.6 billion, up 23.0 billion from April's revised 54.6 billion; exports were 317.7 billion and imports were 395.3 billion.
▸ Reported tool use: official.lookupmodel report
Read May 2026 release detail for goods/services decomposition and one-off movement clues.
↳ The May goods deficit increased 23.6 billion to 106.5 billion, the services surplus increased 0.6 billion to 28.9 billion, exports of goods decreased 11.3 billion to 210.6 billion, and imports of goods increased 12.3 billion to 317.0 billion.
▸ Reported tool use: alfred_fred.lookupmodel report
Fetched the public FRED mirror for BOPGSTB recent monthly observations, using it only as a history mirror while retaining BEA/Census as the resolution source.
↳ BOPGSTB observations were May 2026 -77,585 million, April 2026 -54,570 million, March 2026 -56,585 million, February 2026 -54,980 million, and January 2026 -54,185 million; converted positive deficits are 77.585, 54.570, 56.585, 54.980, and 54.185 usd_billions.

Reference class and base rate: the same-variant 2026 monthly first/latest-reference sample after the annual-revision release context has a positive-deficit mean of about 59.6 billion, while the January-April cluster alone averages about 55.1 billion. May's 77.6 billion print says the immediate trade-flow regime is above that base rate, so the point forecast uses May persistence as the primary prior and pulls it partway back toward the pre-May cluster.

Prior/update/interval: base-rate sample = Jan-May 2026 positive BOPGSTB deficits of 54.185, 54.980, 56.585, 54.570, and 77.585, with full-sample mean 59.6 and January-April mean 55.1; chosen prior = May persistence at 77.6 because July is only two months after the latest observed same-variant shock and the May release showed broad goods import/export movement; adjustment components = -8.0 for partial mean reversion toward the 55.1 January-April base-rate cluster, -0.5 for some normalization of the goods shock after nonmonetary gold/capital-goods volatility, and -0.1 for a roughly stable services surplus near 28.9, giving 77.6 - 8.6 = 69.0. For this flow series I use the short-sample value dispersion as an uncertainty proxy: sample sigma = 10.1 usd_billions, so 80% half-width is about 1.28*sigma = 1.28*10.1 = 12.9; final bounds are 69.0 - 12.9 = 56.1 and 69.0 + 12.9 = 81.9.

Counter-considerations: upside risk for a larger positive deficit would be another July import surge in consumer goods, autos, semiconductors, crude oil, or a further export slump, which would land above the interval if the deficit exceeds 81.9 billion. Downside risk would be a quick reversal of May's import jump, a rebound in goods exports, or weaker domestic demand for imported goods, which would land below the interval if the deficit is under 56.1 billion.

Review disposition: accepted the critique to make the base-rate and persistence-prior bridge explicit, treating May persistence as the chosen prior because of release proximity while using the 2026 reference-class base rate for the mean-reversion update; retained the resolver, source grounding, release date, and interval arithmetic.

Key drivers

  • May goods deficit jumped to 106.5 while the services surplus was 28.9
  • Recent 2026 deficits before May clustered near 54-57 billion
  • May exports fell 10.5 and imports rose 12.5, so the latest level is partly a broad import/export shock
  • Nonmonetary gold and capital-goods components were unusually visible in the May movement
  • Two-month-ahead July forecast allows partial mean reversion but keeps a wider-than-normal deficit bias

Tool evidence

This run has no captured tool responses. “Reported tool use” in the analysis is the model’s account, not an archived tool response.

Sources and resolution

Official source
U.S. Census Bureau and U.S. Bureau of Economic Analysis, U.S. International Trade in Goods and Services
Resolved
September 3, 2026
Resolution rule
Resolve to the first-print seasonally adjusted goods and services balance in Exhibit 1 of the U.S. International Trade in Goods and Services release for July 2026, balance of payments basis, converted to a positive deficit in usd_billions and rounded to one decimal place. Use the initially released July 2026 value only; ignore later monthly, annual, benchmark, or same-series revisions.
Run details

The analysis is the model’s written report. Tool-use descriptions in that report are model claims; the activity artifacts contain the execution record.

thesis.analyst · gpt-5.5 · fast · v2.3.0

pre-submit review · completed

Draft is publishable after clarifying the prior/update framing; resolver, unit, first-print rule, date, interval arithmetic, and JSON fields are otherwise coherent.

  • warning base_rate: The reasoning states a 2026 Jan-May mean base rate of about 59.6, but the compact math step switches to a May persistence prior of 77.6, so the outside-view prior is not consistently stated before updates.
  • warning prior_update_interval: The compact step includes the requested components, but the adjustment components are framed as inside-view movements from May while the preceding paragraph says the outside-view anchor is the recent 2026 base rate.
  • info optional_suggestion: Consider saying the 5-month level sigma is a short-sample uncertainty proxy rather than realized forecast-error volatility.
  • info optional_suggestion: The future BEA URL is plausible, but the reasoning should rely on the schedule/node and Exhibit 1 resolver rather than assuming the exact URL will resolve before publication.

disposition accepted: Review disposition: accepted the critique to make the base-rate and persistence-prior bridge explicit, treating May persistence as the chosen prior because of release proximity while using the 2026 reference-class base rate for the mean-reversion update; retained the resolver, source grounding, release date, and interval arithmetic.

disposition accepted: Review disposition: accepted the critique to make the base-rate and persistence-prior bridge explicit, treating May persistence as the chosen prior because of release proximity while using the 2026 reference-class base rate for the mean-reversion update; retained the resolver, source grounding, release date, and interval arithmetic.

disposition not applicable: Review disposition: accepted the critique to make the base-rate and persistence-prior bridge explicit, treating May persistence as the chosen prior because of release proximity while using the 2026 reference-class base rate for the mean-reversion update; retained the resolver, source grounding, release date, and interval arithmetic.

disposition not applicable: Review disposition: accepted the critique to make the base-rate and persistence-prior bridge explicit, treating May persistence as the chosen prior because of release proximity while using the 2026 reference-class base rate for the mean-reversion update; retained the resolver, source grounding, release date, and interval arithmetic.

Activity artifacts

Complete original trace

§Forecast: July 2026 U.S. goods and services trade deficit

Framing and exact resolver: this targets the July 2026 first print of the U.S. International Trade in Goods and Services headline goods-and-services balance, seasonally adjusted and on a balance of payments basis, reported in Exhibit 1. The ledger transform turns BOPGSTB's negative millions-of-dollars balance into a positive usd_billions deficit; the agency first print is authoritative over later revisions.

▸ Reported tool use: official.lookupmodel report
Checked BEA 2026 release schedule and the BEA scheduled-release node for U.S. International Trade in Goods and Services, July 2026.
↳ BEA schedule lists U.S. International Trade in Goods and Services, July 2026 on September 3, 2026 at 8:30 a.m.; BEA node 42997 also shows Scheduled Date September 3 08:30 and Release Title July 2026.
▸ Reported tool use: official.lookupmodel report
Fetched the latest official BEA/Census release for May 2026 to anchor the same variant and near-term components.
↳ May 2026 goods and services deficit was 77.6 billion, up 23.0 billion from April's revised 54.6 billion; exports were 317.7 billion and imports were 395.3 billion.
▸ Reported tool use: official.lookupmodel report
Read May 2026 release detail for goods/services decomposition and one-off movement clues.
↳ The May goods deficit increased 23.6 billion to 106.5 billion, the services surplus increased 0.6 billion to 28.9 billion, exports of goods decreased 11.3 billion to 210.6 billion, and imports of goods increased 12.3 billion to 317.0 billion.
▸ Reported tool use: alfred_fred.lookupmodel report
Fetched the public FRED mirror for BOPGSTB recent monthly observations, using it only as a history mirror while retaining BEA/Census as the resolution source.
↳ BOPGSTB observations were May 2026 -77,585 million, April 2026 -54,570 million, March 2026 -56,585 million, February 2026 -54,980 million, and January 2026 -54,185 million; converted positive deficits are 77.585, 54.570, 56.585, 54.980, and 54.185 usd_billions.

Reference class and base rate: the same-variant 2026 monthly first/latest-reference sample after the annual-revision release context has a positive-deficit mean of about 59.6 billion, while the January-April cluster alone averages about 55.1 billion. May's 77.6 billion print says the immediate trade-flow regime is above that base rate, so the point forecast uses May persistence as the primary prior and pulls it partway back toward the pre-May cluster.

Prior/update/interval: base-rate sample = Jan-May 2026 positive BOPGSTB deficits of 54.185, 54.980, 56.585, 54.570, and 77.585, with full-sample mean 59.6 and January-April mean 55.1; chosen prior = May persistence at 77.6 because July is only two months after the latest observed same-variant shock and the May release showed broad goods import/export movement; adjustment components = -8.0 for partial mean reversion toward the 55.1 January-April base-rate cluster, -0.5 for some normalization of the goods shock after nonmonetary gold/capital-goods volatility, and -0.1 for a roughly stable services surplus near 28.9, giving 77.6 - 8.6 = 69.0. For this flow series I use the short-sample value dispersion as an uncertainty proxy: sample sigma = 10.1 usd_billions, so 80% half-width is about 1.28*sigma = 1.28*10.1 = 12.9; final bounds are 69.0 - 12.9 = 56.1 and 69.0 + 12.9 = 81.9.

Counter-considerations: upside risk for a larger positive deficit would be another July import surge in consumer goods, autos, semiconductors, crude oil, or a further export slump, which would land above the interval if the deficit exceeds 81.9 billion. Downside risk would be a quick reversal of May's import jump, a rebound in goods exports, or weaker domestic demand for imported goods, which would land below the interval if the deficit is under 56.1 billion.

Review disposition: accepted the critique to make the base-rate and persistence-prior bridge explicit, treating May persistence as the chosen prior because of release proximity while using the 2026 reference-class base rate for the mean-reversion update; retained the resolver, source grounding, release date, and interval arithmetic.

calibrated forecast · 80% CI
$69B[$56.1B · $81.9B]
Target metadata

Data point: bea.trade.goods_services_deficit.2026_07.first_print

More government data forecasts