China's Late-October Steel Prices: Flats, Not Rebar, Did the Falling

The headline from China's National Bureau of Statistics for October 21–31 is a modest one: 20 mm HRB400E rebar averaged RMB 3,167.3/t, down RMB 20.5/t, or 0.6%, from mid-October. But the composition of the monthly dataset matters more than the rebar number, because the NBS ten-day series is one of the few genuinely consistent, nationwide, sample-based price benchmarks published for Chinese steel. Its late-October reading signals a market where flat products are correcting faster than longs, and where construction steel is flatlining rather than collapsing.

What the official data actually show

The NBS release (English, Chinese) monitors 50 products across nine categories from roughly 2,000 wholesalers, agents and dealers in more than 300 markets in all 31 provinces. It is compiled with the China Economic Monitoring and Analysis Center and SCI (Zhuochuang). Steel is one slice of a much larger basket — 17 products rose, 28 fell and 5 were flat versus mid-October — but the six ferrous benchmarks are widely followed because they are the most methodologically transparent national series available.

Source: NBS, Market Prices of Important Means of Production in Circulation, October 21–31, 2026. USD conversions in trade press use RMB 6.718/$; at that rate rebar is roughly $471.5/t.

The real story is the flat-long divergence

Three observations stand out from the table, and none of them is the rebar percentage that drove the headline coverage.

First, the weakness is concentrated in flats. Medium plate fell 1.2% and HRC 1.0%, roughly double the 0.6% rebar decline and far more than wire rod's 0.3%. In absolute terms plate lost RMB 43.4/t against rebar's RMB 20.5/t. The NBS series is a selling-price benchmark, so this is not a question of grade mix: the same survey that collects rebar prices also collects plate and coil prices, and it found flat products softening materially faster over the same ten days.

Second, the plate-over-rebar spread narrowed. In mid-October the medium-plate premium over rebar was RMB 322.1/t; by late October it was RMB 299.2/t — a compression of roughly RMB 23/t in ten days. This is a classic sign of flat-product supply running ahead of demand while construction steel, already depressed, has less room to fall. Trade reporting for the period supports that reading: HRC output rose 0.9% week on week for October 19–25 as rolling lines in North China restarted, even as HRC inventories slipped only 0.7% (Promisteel weekly, citing Mysteel/NBS).

Third, the angle-steel increase is almost certainly noise. A 0.2% gain, or RMB 5.2/t on a base of about RMB 3,388/t, sits within rounding and sampling tolerance for a wholesale-price survey. It follows a 0.7% fall in mid-October, so it reads more like a small lagged correction than a structural signal.

Why a 0.6% move still matters

It would be easy to dismiss the release as a routine statistical update. For anyone actually trading or hedging Chinese steel, however, the NBS series carries information that mill list prices and index assessments do not.

The first point is level, not direction. Rebar at RMB 3,167/t is close to the bottom of the range seen through 2026, and the mid-October print (RMB 3,187.8/t) showed a small rebound before this late-October slip. The rebar price is therefore oscillating around a floor rather than in a downtrend. For integrated mills, the key question is whether that floor sits above or below the cash cost of the marginal producer. Industry profitability data for October showed only about 11% of mills profitable — improving, but still a minority — which is consistent with rebar prices hovering near break-even for many operations.

Second, methodology matters. NBS wholesale prices run materially below widely quoted spot assessments — a national average rebar assessment for October 19–25 was put at RMB 3,311/t, versus the NBS's RMB 3,167.3/t for roughly the same period. That gap is not a contradiction; it reflects different samples, specifications and market points. Analysts comparing NBS against Mysteel or Platts assessments without adjusting for this will draw the wrong conclusion about the direction of travel.

Third, timing distorts the series. NBS cancelled its early-October (October 1–10) report because of the National Day holiday, so the mid-October release compared against late September rather than early October. This means the late-October comparison is effectively the first clean ten-day read after the holiday disruption — and it captured a market that had lost momentum rather than gained it.

The demand backdrop that explains the numbers

Rebar's small decline is not a signal of resilience in construction. It is a story of a market already at a low plateau. Apparent rebar demand in late October was reported at roughly 2.01 Mt — the weakest for the comparable week in almost four years — while national rebar inventories fell 4.8% to about 5.22 Mt. Inventories are falling because mills have cut output, not because demand is strong. Rebar production earlier in the autumn reached an intra-year low, with rebar-mill capacity utilisation around 36%, according to Mysteel data cited by SEAISI.

That combination — deep output cuts, inventories still falling, but prices unable to rally — is the defining feature of China's current construction-steel market. Supply has adjusted; demand has adjusted further. The result is a price that grinds sideways in a narrow band rather than trending lower, which is why the rebar print moved only 0.6% while flats, where output is rising, moved twice as fast.

The flat-product side faces a different problem. HRC output is recovering as mills shift capacity toward coil, but downstream demand from manufacturing and export-oriented sectors is not accelerating enough to absorb it. That is why the plate premium over rebar is narrowing and why HRC fell 1.0%. If the pattern persists, the flat-long spread will continue to compress — a margin squeeze for mills heavily weighted toward flats and a relative reprieve for rebar producers who have already cut deepest.

What to watch next

The NBS early-November release (covering November 1–10) will be the first clean test of whether the late-October softening was a post-holiday adjustment or the start of a wider move. Two things deserve attention: whether the plate premium over rebar narrows further or stabilises, and whether rebar breaks below approximately RMB 3,150/t — the level that would signal the autumn floor is giving way. On current evidence, the more likely outcome is continued flat-product underperformance rather than a broad-based rebar selloff.

All prices are as reported by the National Bureau of Statistics. Spreads, basket averages and directional interpretations marked as analysis are the author's own calculations from the NBS table and secondary market reporting; they are not official indices.