China's Late-October Steel Board: A Price Picture Dated One Day Earlier Than It Claims

Mysteel's "普钢实时行情" (carbon steel real-time quotations) page is one of the most-quoted screens in the Chinese steel market, and it is routinely misread. The version carrying the 29 October 2026 update stamp displays rebar at 3,313.35, HRC at 3,208.54, CRC at 3,664.79 and medium plate at 3,412.98 yuan/t. Those are not 29 October prices. They are Mysteel's 28 October index levels, the last complete absolute-price release; the following day's release published rebased index points (comprehensive 121.15, +0.17) without yuan-per-tonne equivalents.

That distinction matters less for the level than for the direction, because both days tell the same story: long products are holding while flats sag.

Two days, one signal

(Sources: Mysteel daily index releases for 26, 28 and 29 October 2026; 29 October figures are rebased index points, not yuan/t, and are therefore shown only as direction.)

The 0.07% gain quoted on the board is not a Mysteel publication. It is arithmetic — 2.26 ÷ (3,406.02 − 2.26) = 0.066% — computed by whoever assembled the page. Readers citing "the composite rose 0.07%" should know they are quoting a derived percentage from a single day's absolute change, not an official growth rate.

The more useful observation is the spread. Over the two days to 28 October the HRC index fell 13.78 yuan/t while rebar lost 1.58 — a roughly nine-to-one ratio of weakness. On 29 October the pattern persisted in index-point terms, with plate the only negative line on the board. This is the familiar "长强板弱" (long strong, flat weak) configuration, and it is being driven by two independent forces: infrastructure and construction demand that is weak but no longer collapsing, against manufacturing demand — autos, white goods, machinery — that is visibly deteriorating.

The briefing archive is the real tell

Mysteel's rolling feed on the same page carries three regional briefings that the 29 October date stamp implicitly presents as current. None of them is:

The Hubei note remains analytically useful because it describes the same long-versus-flat divergence seen nationally — Wuhan rebar and wire rod up 10 yuan/t week-on-week, while HRC, plate, CRC and H-beam each fell 10 yuan/t. It also flags a detail worth watching: total Wuhan carbon steel inventory rose 1.54% w/w even as construction steel stocks dipped, meaning the build-up was concentrated in coil and plate. Egang's ex-works prices were unchanged, but provincial mill finished-goods inventories rose notably — the classic signature of mills defending list prices while product backs up behind them.

Subsequent official data complicate the picture. The Hubei Provincial Department of Commerce's weekly report for 19–25 October shows the provincial steel average at 3,522.5 yuan/t, down 0.7% w/w, with bar and rod at 3,492.5 (−1.3%) and plate at 3,565 (−0.3%). That reverses the earlier long-outperforms-flat pattern within Hubei specifically, and the report notes cold-rolled coil demand as particularly weak with merchant inventories above normal. (Caveat: the commerce department's basket covers a different product mix and sales channel than Mysteel's index, so the two are not strictly comparable — but the divergence in direction is worth flagging as a regional caveat rather than a contradiction.)

Where the export signal actually sits

The page's export rows — dated daily at roughly 14:00–15:20 Beijing time for H-beam, CRC and HRC, ahead of the domestic index close around 18:00 — contain only one visible USD number in the retrieved text: Chinese H-beam at USD 511/t FOB Tianjin, up USD 1 for 28 October. HRC and CRC export values were not exposed.

That asymmetry is worth noting for anyone building models off this page: section export prices are being published with a specific spec and basis, while flat-product export assessments sit behind a link. Do not infer HRC or CRC export levels from this board.

The single USD 1/t move is itself unremarkable, but the sustained daily publication of three export rows in the middle of a soft domestic market reflects the structural reality — Chinese mills are actively pricing into export markets to absorb volume that domestic flat demand cannot take.

What the numbers imply for margins

The 29 October Mysteel commentary (the correct attribution for that day's narrative) notes HRC average at 3,229 yuan/t (+5), iron ore at 652 yuan/wet tonne (+5), and Shanghai Futures Exchange rebar at 3,056 yuan/t (+13). The important line: raw materials range-bound, finished steel falling faster than cost, mill margins squeezed, some mills cutting output, heavy Q4 maintenance expected.

Crude arithmetic on the board's own index levels illustrates the bind. A mill producing HRC at a 3,208.54 index against iron ore at 665.15 and coking coal at 2,022.60 has little headroom; add the 1,390 metallurgical coke line (a different spec basis from the 1,953.5 coking coal figure and not directly comparable), and the cost stack leaves thin conversion margins. Steel mills do not publish their cost curves, so this is directional analysis rather than a computed margin — but the qualitative conclusion matches Mysteel's own assessment: output cuts and Q4 maintenance are the rational response.

Reading the board correctly

Three practical rules for anyone consuming this page:

1. Check the valuation date, not the page timestamp. The 2026-10-29 stamp reflects page assembly. The absolute yuan figures are 28 October; the 29 October release uses rebased index points.
2. Ignore the briefing dates at your peril. Hubei (10 Oct), Henan (11 Sep) and Wuxi stainless (24 Aug) are archived items, not same-day news. The Wuxi note's 304 price cuts of 200 yuan/t at Qingdao Steel (青山) — five-foot 14,550 yuan/t, wide 14,750 yuan/t — relate to the August market, not October.
3. Do not mix index bases. The 3,406.02 series is yuan/t; the 121.15 series is base-100. They measure different things and cannot be spliced into a single time series.

The takeaway

Strip out the aggregation noise and a coherent picture remains: China's steel market in late October 2026 is in a grinding, low-margin equilibrium where construction-linked long products are marginally better supported than manufacturing-linked flats, inventories are building in the wrong products, mills are defending list prices while output backs up, and Q4 production cuts are becoming the default response. The most consequential signal for the next quarter is not the daily price move but whether the long-flat spread reverses — because if flats keep underperforming into the winter, mill maintenance schedules will likely lengthen, and import markets will feel the displacement.

Note on sourcing: the primary source is an aggregation page and the regional briefings it carries are selectively dated. Where price levels or product-specific figures are cited, the valuation date and basis are stated. Margin commentary is analytical, not computed from published cost data.