Update: China H1 Steel Share 53.7% (−1.3 pp YoY); Newbuild Orders 66% China GT

What changed since the August update

Our August heavy industrial capacity update reframed the builder-base map as a vintage delta: UNCTADstat’s June 2026 insight put 91% of 2025 merchant GT completions in the China–Korea–Japan trio (−4.2 pp versus the research peak), World Steel’s 2025 annual totals eased China’s crude share to 52.0%, and Cirium’s H1 2026 wrap showed Airbus at 54% of the Airbus–Boeing duo. Dry docks (~62% China) and the six-shop RPV-class forge club were held. That post answered “what moved in the newest official print.” This Q3 print answers the same question against that baseline.

Three newer vintages force the rewrite. World Steel’s June 2026 release (23 July) prints H1 2026 crude steel for the 70-country pool at 931.5 Mt (−0.7% YoY), with China at 500.0 Mt (−3.0%) and India at 87.0 Mt (+7.1%). China’s share of that pool lands at 53.7% — −1.3 percentage points versus a back-solved H1 2025 share of ~55.0%. Separately, the Japan Ship Exporters’ Association (JSEA) 2025 status tables disclose country GT completions (China 52.6%, Korea 27.6%, Japan 14.0%) and, more importantly for forward yard load, 2025 new orders that put China at 66.0% of booked GT — +13.4 pp above same-year completions. On the hangar side, OEM year-to-date July 2026 deliveries put Airbus at 418 and Boeing at 367 (785 duo; Airbus 53.2%), a mild −0.8 pp versus the Cirium H1 54% frame.

What still did not move remains the industrial-capacity punchline. Tracked VLCC-capable dry docks and the six-shop ultra-heavy forge inventory stay held — there is still no audited global registry that adds a US RPV-class forge or rewrites the dock map. The dashboard above is built as a Q3 vintage delta: diverging pp bars, JSEA completions-versus-orders flow, World Steel H1 share/YoY/region toggles, Airbus–Boeing monthly composed delivery pace, a cross-sector capacity radar, and a held docks-and-forges panel.

The headline table: Aug update vs Q3 2026 print

Open Vintage Δ and toggle Moved only versus Held / flat. The thesis is not “China stopped building.” It is that the newest mid-year steel and orderbook prints show movable feedstock and contracting layers shifting while scarce dock and forge capital stay frozen — and that JSEA’s orderbook is more China-concentrated than same-year completions.

Steel H1 2026: China still half the pool — India’s YoY is the story

World Steel’s June release is the hard Q3 vintage for feedstock geography. The 70-country pool printed 931.5 Mt in January–June 2026 (−0.7% versus H1 2025). China produced 500.0 Mt (−3.0%), India 87.0 Mt (+7.1%), the United States 42.8 Mt (+6.3%), Japan 40.4 Mt (−0.4%), and Korea 31.7 Mt (+2.1%). Viet Nam remains a small-base rocket: 15.2 Mt, +26.9% YoY.

Share math matters for the capacity question. China’s 53.7% of the H1 pool is higher than the Aug update’s FY2025 China share of 52.0% — that is seasonality and vintage mix, not a rebound thesis. The clean YoY comparison is H1-to-H1: back-solving the disclosed −3.0% / −0.7% rates puts China’s H1 2025 share near 55.0%, so the Q3 print is −1.3 pp on a like-for-like half-year. India rises from ~8.7% to 9.3%. North America is the regional bright spot (+5.7% YoY on 56.5 Mt); CIS+Ukraine remains the soft region (−8.0%).

Toggle Steel H1 across world share, YoY, million tonnes, and region YoY. Crude steel is still not ultra-heavy forging and not VLCC dock concrete. A one-point China share dip does not unlock reactor-pressure-vessel capacity in Ohio. It does mean the plate and section feedstock geography is slightly less China-weighted on a YoY H1 basis — with India and the US absorbing relative gain while Asia/Oceania as a bloc still printed 690.1 Mt (−0.9%).

Yard flow: completions look oligopolistic; orders look more China-heavy

The Aug update’s ship headline was UNCTAD’s 91% Asia-trio completions print. That UNCTAD vintage has not been superseded by a newer UNCTADstat insight card — so the trio’s official UNCTAD number remains held at 91%. What is new for Q3 is JSEA’s disclosed 2025 country GT table: China 52.6%, Korea 27.6%, Japan 14.0%, Europe 2.8%, others ~3%. Summed, the Asia trio is 94.2% under JSEA’s scope — a parallel measurement, not a clean replacement for UNCTAD’s 91%. Treat the +0.4 pp China move versus the Aug update’s derived 52.2% as a refinement across sources, not a boom.

The forward-looking signal is sharper. JSEA’s 2025 new orders put China at 66.0% of GT booked, Korea 19.6%, Japan 9.0%, Europe 3.2%. China’s orders-minus-completions gap is +13.4 pp; Korea and Japan sit negative (they completed more share than they booked). That is the industrial-capacity tell for physical capital stock: who is loading the berths for the next delivery wave. Completions can ease at the margin while the orderbook re-concentrates.

Open Yard flow and toggle completions, orders, and the order gap. Pair with the shipbuilding GT concentration companion when you need the longer UNCTAD/RMT path; this update’s job is the Q3 flow delta, not a full historical rewrite.

Aircraft YTD July: Airbus still leads, duo tempo is the news

The Aug update’s aircraft frame was Cirium H1 (649 duo; Airbus 54%). OEM prints through July 2026 extend the recovery: Airbus 418, Boeing 367, combined 785. Airbus’s duo share eases to 53.2% (−0.8 pp). Both OEMs are ahead of the same point in 2025 (Airbus was 373, Boeing 328 through July last year) — roughly a ~12% combined YoY lift on the YTD stack.

Monthly composition matters for FAL risk. June was the peak duo month (153 handovers; Airbus 89 / Boeing 64); July cooled to 120 (67 / 53) while Airbus’s monthly duo share stayed elevated at 55.8%. The industrial fact from the research and Aug posts survives: narrowbody campuses on both sides of the Atlantic still dominate throughput, and a single-program bottleneck (MAX line, A320neo family) can still move the global delivery total. This Q3 print does not re-census FAL sites — for campus geography, the commercial aircraft FAL companion still governs.

Switch to Aircraft YTD. The composed chart stacks monthly OEM bars under a duo area and an Airbus share line so you can see tempo and mix in one pane — a different visual grammar than the Aug update’s H1 bar strip.

Capacity radar: movable shares vs immobile capital

Cross-sector dashboards invite fake indices. The Capacity radar panel is deliberately a shape comparison, not a composite score: Asia ship trio GT, China steel H1 share, China newbuild orders, Airbus duo share, China VLCC docks, and forge shop count. Prior (Aug / H1 frames) versus Q3 print. The rose overlay fattens on orders and JSEA trio intensity; it shrinks on China steel H1 share and edges in on Airbus duo; docks and forges sit unchanged.

That shape is the post’s argument in one glance. Contracting and feedstock layers move every half-year. Dock concrete and certified ultra-heavy forges do not. Anyone writing industrial policy that treats “steel capacity” as a synonym for “forging capacity,” or “ship GT share” as a synonym for “dry-dock slots,” is reading the wrong radar axis.

What stayed frozen: docks and the six-shop forge club

Industrial-capacity commentary loves to extrapolate from steel or ship GT into “manufacturing power.” The held layers refuse that leap. Tracked VLCC-capable dry docks remain China-heavy at ~62%, with Korea ~18%, Japan ~8%, and the rest of world ~12%. The ultra-heavy forge strip is unchanged: Japan Steel Works, Doosan Enerbility, Shanghai Electric’s heavy complex, OMZ/AEM, Sheffield Forgemasters, and Framatome/Creusot. North America still has no RPV-class member in that inventory.

Open Docks & forges. Equal-weight shop counts are not throughput weights — JSW and Doosan matter more than a strip implies — but the phone book did not get a seventh number in Q3. Nuclear new-build, naval propulsion, and any strategy that hoped a steel-share dip implied forge relief still face the same bottleneck the research map flagged.

Who is exposed — and what the Q3 delta buys

Exposed to the wrong headline: readers who treat “China steel share rose versus FY2025” as evidence of re-acceleration without checking the H1-to-H1 −1.3 pp YoY share move; readers who treat JSEA’s 94.2% trio as overturning UNCTAD’s 91% without reading the scope caveat; readers who see India’s +7.1% H1 steel and conclude Western merchant yards are back. US commercial GT remains noise (~0.04%).

Exposed to orderbook concentration: shippers and financiers who underweight China’s 66% of 2025 newbuild GT when planning 2027–29 delivery slots; Korea/Japan yards whose completion shares still outrun their fresh order shares; European specialty yards that punch above weight in GT but do not rewrite the VLCC dock map.

Exposed to frozen scarcity: nuclear and naval programs that hoped feedstock YoY implied forge relief; airlines whose delivery risk still clusters on a handful of FAL campuses even as duo YTD totals rise ~12%.

What the delta buys: a clearer Q3 separation between movable shares (H1 steel, OEM delivery mix, newbuild orders) and immobile capital (large docks, RPV forges, FAL campuses). China produces slightly less of the H1 steel pie than a year earlier, books a larger share of new merchant GT than it completed in 2025, and still anchors docks and the scarce forge club. India is again the steel YoY story. Airbus leads the YTD duo, but the gap narrowed versus Cirium H1 as Boeing’s recovery continued into July.

Caveats and methodology

• Sectors are not additive. Do not average ship GT with steel share into a fake industrial-capacity index.
• UNCTAD’s 91% Asia-trio print is heldno newer UNCTADstat completions insight superseded it. JSEA’s 94.2% trio is a parallel scope, not a clean vintage replacement.
• JSEA China 52.6% completions versus the Aug update’s derived 52.2% is a cross-source refinement; confidence is disclosed for JSEA, derived for the prior.
• China H1 steel share prior (~55.0%) is derived by back-solving World Steel’s disclosed −3.0% / −0.7% YoY rates on the H1 2026 levels.
• FY2025 China steel share (52.0%) is not directly comparable to H1 2026 (53.7%) without a seasonality adjustmentuse H1-to-H1 for YoY claims.
• Aircraft figures are duo-only (Airbus+Boeing OEM). COMAC and other OEMs are outside the 785 YTD total.
• Monthly OEM bars are tuned to disclosed YTD July totals; individual months may differ slightly from secondary trackers.
• Dry-dock and forge counts are held inventories, not a 2026 re-census.
• GT ≠ dwt. Orderbook and completions shares in GT will not match deadweight tables.
• Naval and cruise yards remain strategically important and deliberately outside the merchant GT leaderboard.

The shareable takeaway

The Q3 official vintage did not redraw who can build physical capital stock — it moved the feedstock and contracting layers while the scarce capital stayed frozen. China’s H1 steel share is 53.7% (−1.3 pp YoY) on a soft world pool; India printed +7.1% YoY to 9.3% share; JSEA 2025 new orders put China at 66% of GT booked, +13.4 pp above same-year completions; Airbus’s duo lead eased to 53.2% YTD July (−0.8 pp) as both OEMs outran 2025. Dry docks (~62% China) and the six-shop forge club did not move. Watch vintage deltas by layer — steel, orders, deliveries, docks, forges — not a single “manufacturing” headline, to see who still has the yards, docks, hangars, and forges.