Update: CPB May Trade +1.0% MoM — US Q2 GDP Slows to 1.5% as PCE Hits 5.1%

What changed since the July IMF vintage

Our July growth–trade–prices update answered the forecast revision question: versus the April WEO research print, the IMF’s July Update lifted 2026 world trade volume (goods and services) +0.7 pp to 3.5%, slipped world GDP −0.1 pp to 3.0%, and raised world headline CPI +0.3 pp to 4.7% with oil near $89/bbl. That post was a vintage delta inside the Fund’s own tables. This Q3 note answers the next question markets actually trade on: what did the first wave of official hard data say once that July baseline was on the street?

Three information events force the refresh. The CPB World Trade Monitor for May 2026, released 24 July, printed world merchandise volume +1.0% month-on-month after April’s +0.7% rebound from March’s −2.1% Hormuz/Gulf shock. The BEA advance estimate for Q2 2026, released 30 July, put US real GDP at 1.5% SAAR (down from 2.1% in Q1) while the PCE price index jumped to 5.1% SAAR and the purchases deflator to 5.7%. Eurostat’s Q2 flash (confirmed in the 14 August second estimate) showed the euro area expanding 0.4% QoQ after a flat Q1, with +1.0% YoY. China’s NBS Q2 print slowed to 4.3% YoY (from 5.0% in Q1); Korea’s BOK advance held a still-hot 3.7% YoY on +0.6% QoQ exports. The dashboard above is built as a hard-data check — CPB MoM path, triad tracking bars, US GDP×PCE dual path, forecast-vs-print gaps, Q1→Q2 slopes, and a disclosed growth×price scatter.

The headline table: July forecast path → Q3 hard prints

Read the table as a split triad, not a single scare print. Trade volumes in the CPB monitor have already staged a two-month recovery from the March cliff. US growth is running below the July annual number on a SAAR basis. US prices are running above the July CPI assumption on a PCE SAAR basis — even as core PCE cooled to 3.4%. That is the opposite of a synchronized soft-landing confirmation, and it is a different story from July’s pure forecast revision where trade was the upgrade surprise inside the Fund’s own tables.

Trade: the CPB monitor says the March shock did not stick

July’s IMF path still has world goods-and-services trade volume slowing from 5.0% in 2025 to 3.5% in 2026. That is an annual average, not a monthly run-rate. What CPB adds is the intra-year shape. Merchandise volume fell 2.1% in March as Africa/Middle East trade collapsed and Chinese New Year timing amplified Asia’s export dip. April recovered 0.7%; May added another 1.0%. Staff commentary on the April print already pointed to inventory building and AI-related electronics as offsets to the Gulf shock. The May print extends that rebound for a second consecutive month.

The non-obvious reading: the war shock looks sharp but temporary in the merchandise monitor, at least through May. A rough Mar–May chain leaves cumulative volume only about −0.4% below the February level — not a multi-month freefall. That is consistent with July’s narrative that inventories and rerouting buffered oil shortfalls, and it is friendlier to the Fund’s +0.7 pp trade upgrade than a prolonged Hormuz closure would be. Pair this vintage with our China–US–India GDP comparison for the longer weight shift that still sits under AI-goods trade, and with the April research triangle for the WTO March merchandise baseline (+4.6% in 2025; 1.9% baseline for 2026) that has still not been superseded by a new GTOS vintage.

Growth: US softens; euro area rebounds; China cools on a YoY basis

The US advance estimate is the cleanest single stress test of July’s country table. IMF staff left US 2026 GDP unchanged at 2.3% in July. Q2 SAAR at 1.5% is not a recession print — final sales to private domestic purchasers rose 3.9% — but it is a clear deceleration from Q1’s 2.1% and sits 0.8 pp under the annual baseline if the second half does not re-accelerate. Contributors were consumer spending, investment, and exports, offset by government spending (including SPR sales accounting) and rising imports. That mix is “soft landing with a hotter price impulse,” not “AI boom alone carrying the print.”

The euro area tells the other half of July’s war-vs-tech crosscurrent. July cut EA 2026 growth to 0.9% on Ireland-heavy Q1 carryover and energy drag. Q2’s 0.4% QoQ rebound (after 0.0% in Q1) and 1.0% YoY do not automatically restore the annual path, but they break the flat-line narrative. Spain led among large members; Germany, France, and Italy each posted about 0.2% QoQ. Ireland’s 3.9% QoQ again shows how multinational accounting can dominate the flash.

China’s 4.3% YoY in Q2 (H1 4.7%) is a step down from Q1’s 5.0% and sits a tenth or so under July’s 4.6% annual number if the second half does not firm. Sequential growth of 0.9% QoQ is still positive. Korea remains the AI-hardware outlier in levels: 3.7% YoY and 0.6% QoQ keep the economy well above the IMF’s 2.6% annual print even after July’s large upgrade. Toggle the dashboard’s region filter to see how Asia’s YoY prints sit proud of July while the US SAAR print sits soft.

Prices: disinflation is not the Q2 story in the United States

July already said world CPI would rise to 4.7% in 2026 before easing to 3.9% in 2027, with US CPI assumptions at 3.6% / 2.4%. Q2 US data do not contradict the “pause in disinflation” headline — they sharpen it. Headline PCE at 5.1% SAAR (from 4.6% in Q1) and the gross domestic purchases price index at 5.7% (from 3.6%) show the energy/conflict impulse still in the quarterly tape. Core PCE at 3.4% SAAR (down from 4.4%) is the partial offset: underlying demand-side pressure cooled even as headline stayed hot.

That split matters for policy. July argued policy rates would be less supportive than in April because inflation was more visible. A quarter where headline PCE is 5.1% and core is 3.4% is exactly the configuration that keeps “look-through the supply shock” contested inside central banks. It is also why the dashboard’s price bridge puts the July $89/bbl oil assumption next to the US deflator impulse rather than pretending they share a unit.

Country trajectories: who is tracking July — and who is not

Cross-country gaps in the dashboard use a deliberately rough comparator (SAAR or YoY versus annual, QoQ×4 for sequential economies). Treat the ranking as a signal board, not a forecast error decomposition:

• United States: Q2 SAAR 1.5% vs IMF 2.3%soft on growth, hot on PCE.
• Euro area: sequential rebound 0.4% QoQ after a flat Q1; annual path still requires H2 follow-through to hit 0.9%.
• China: Q2 YoY 4.3% vs IMF 4.6%modest undershoot after a strong Q1.
• Korea: YoY 3.7% vs IMF 2.6%still the clearest upside AI-export tracker.
• India / Brazil / UK / Japan rows in the gap panel include estimated sequential or YoY placeholders where full national accounts sets were not yet aligned for this desk; disclosed anchors remain US, EA, China, and Korea.

Exposure geometry from July still helps interpret the tape: AI-hardware exporters and economies with strong electronics orders look firmer on activity; energy-importing advanced economies show the price impulse more clearly in US deflators than in EA sequential GDP so far.

What would rewrite this hard-data vintage

1. CPB June/July monitorsif merchandise MoM turns negative again, the two-month rebound narrative fails.
2. BEA second estimate (26 Aug)a large GDP or PCE revision would move the US gap versus July’s 2.3% / 3.6% assumptions.
3. Hormuz / Gulf logisticsa renewed closure would hit CPB Middle East/Africa rows first and oil assumptions second.
4. AI-goods trade momentuma fade in semiconductor and server-related shipments would undercut both Korea’s YoY outperformance and July’s trade upgrade logic.
5. China domestic demandif H2 YoY stays near 4.3% without import recovery, partner CPIs keep absorbing diversion effects.

Until those resolve, the shareable framing is narrow: versus the July IMF baseline, May–August hard data say merchandise trade bounced, US growth undershot on a SAAR basis, and US headline prices overshot — the triad is no longer co-moving inside the quarter.

Caveats and methodology

• Hard data ≠ annual forecast. CPB MoM merchandise volume is not interchangeable with IMF goods+services annual volume (3.5%). WTO still has not issued a post-March GTOS vintage in this window.
• SAAR / QoQ / YoY mismatch. Country gap bars mix units; QoQ×4 is a crude annualiser, not an IMF staff projection.
• US advance estimate will be revised on 26 August and again with the annual update path; treat 1.5% / 5.1% as the 30 July vintage.
• Eurostat flash → second estimate held at 0.4% QoQ; third estimate is still ahead.
• Estimated rows (Japan, UK, India, Brazil sequential/YoY placeholders) are marked in the data moduledo not treat them as disclosed national accounts.
• Oil $89/bbl is the July IMF assumption, not an August futures settle; the price bridge is directional.
• India remains fiscal-year basis in IMF tables; calendar-quarter YoY comparisons are approximate.

The shareable takeaway

In the Q3 2026 hard-data window, CPB merchandise trade printed +1.0% MoM in May after +0.7% in April and −2.1% in March; US Q2 GDP advanced at 1.5% SAAR with PCE at 5.1%; the euro area rebounded 0.4% QoQ. Relative to the July IMF update, the tape is trade firmer into mid-year, US growth softer than the annual path, US prices hotter than the CPI assumption. The soft landing is still not canceled — but the three meters are not confirming July in the same direction.

Related reading: July IMF vintage delta on growth, trade & prices and April research triangle.