Rating: Neutral | Target Price: RMB 10.9–14.3 (base-case fair value range) | Current Price: RMB 11.91 (closing price on 2026-10-29) Time Horizon: 12 months | Market cap approx. RMB 60.11 billion | Total shares: 5.047 billion | Sector: Special steel materials (cyclical)
Key note: The current price deviates from the base-case fair value center of approx. RMB 12.6 within ±6%, with a probability-weighted expectation of approx. RMB 12.1 (+2%); the current price sits in the lower half of the base-case fair range, with a negative margin of safety (approx. -8%, i.e., the current price is above the base-case fair value floor of RMB 10.9). PB of 1.35x at the lowest percentile of the past 5 years and a dividend yield of approx. 5.4% provide downside buffers, but Q3 2026 single-quarter earnings of -23.6% confirm that per-ton profitability remains in a downtrend — "cheap" alone is not yet a reason to buy; wait for the right-side signal of per-ton profitability stabilizing for two consecutive quarters.
CITIC Special Steel is the world's largest specialized special steel producer (approx. 20 million tons of capacity, 19.5382 million tons of sales volume in 2025, capacity utilization approx. 97.7%), with net profit attributable to shareholders of approx. RMB 303 per ton — significantly above the average of key industry steelmakers. Even in its weakest year for non-GAAP net profit over 2021-2025 (2024), it still earned RMB 4.975 billion; operating cash flow coverage of net profit has exceeded 1.86x for three consecutive years, and the payout ratio has remained approx. 50% since listing. In terms of quality, the company is one of the few steel industry names able to offset industry-wide volume decline through structural upgrading.
On valuation, the current price of RMB 11.91 corresponds to a PB of 1.35x (near the lowest 5-year percentile), PE (TTM non-GAAP) of 10.7x, and a dividend yield of 5.42%; the earnings level implied by market pricing is close to the 2024 trough (back-solving using its own 5-year median PE of 12.2x implies normalized earnings of only approx. RMB 4.93 billion, about 17% below the 2021-2025 non-GAAP center of RMB 5.96 billion), and the zero-growth value (EPV) estimate is approx. RMB 14.5 per share. However, this undervaluation judgment is highly sensitive to the "normalized earnings center" assumption: if long-term industry demand decline pushes the per-ton net profit center from RMB 303 down to around RMB 240, the current price would be roughly fair.
Near-term headwinds are real: Q3 2026 net profit attributable of RMB 1.170 billion (-23.64%) and single-quarter gross margin of 12.76%, the lowest since Q3 2024; H1 2026 export gross margin plunged 4.29 percentage points to 14.82%; the 70% U.S. tariff and EU quota cuts (50% excess tariff) fully suppress exports from H2 2026 onward; this report's own estimate for 2026 net profit attributable is RMB 5.28–5.52 billion, 13–17% below sell-side consensus (RMB 6.33 billion). Overall judgment: Neutral. Downside is buffered by book value per share of RMB 8.81, a dividend yield of approx. 5.4%, and EPV; upside offers unpriced options such as export recovery and Nanjing Steel injection, but the earnings inflection is unconfirmed and no positive margin of safety has emerged — recommend adding to the watchlist, with a shift to Positive conditional on "per-ton gross margin ≥ RMB 700/ton for two consecutive quarters and stabilization of export gross margin."
Most important change over the next 6-12 months: Validation of the per-ton profitability center — whether the company holds per-ton net profit attributable at RMB 280–300 (corresponding to non-GAAP earnings of RMB 5.5–6.0 billion) or reverts toward the 2024 level of RMB 743/ton gross margin or lower, is the first-order valuation sensitivity variable. The validation window runs from Q4 2026 to the Q1 2027 earnings report.
Primary share price driver: Export volume and price (2.4 million-ton annual target × slope of export gross margin recovery) and the consolidated debut of the Stanco global trading platform (2026 annual report, expected to be disclosed in March 2027). The 2025 export gross margin of 22.2% exceeded domestic trade by 8.34 percentage points; exports are the highest-quality portion of the company's profit structure. This advantage plunged in H1 2026, and whether it recovers directly determines the earnings center.
Verifiable expectation gap: A two-way divergence rather than one-sided undervaluation. On one hand, sell-side consensus (FY2026 net profit attributable of RMB 6.33 billion, median target price of RMB 15.90) mostly formed before Q3 disclosure and has not been revised down in the past 90 days; if the annual report confirms net profit attributable of RMB 5.3–5.5 billion, consensus and target prices face systematic downward revisions — "current price 33% below target price" is not a basis for upside (in 2025 there was also a reverse precedent of sell-side underestimating by 11% and the company beating, with errors going both ways). On the other hand, the current price is based on trough earnings, 10–17% conservative relative to the true center, with essentially no premium priced in for the Nanjing Steel injection, export recovery, or Stanco synergies.
Validation catalysts and falsification conditions: Upside validation — 2026 annual exports ≥ 2.4 million tons with export gross margin recovering above 15%, and Q4 per-ton gross margin stabilizing sequentially; falsification conditions — per-ton gross margin ≤ RMB 700/ton for two consecutive quarters, export volume falling below 2 million tons or export gross margin below 10%, payout ratio dropping below 40%, or the stock persistently below RMB 9.58 (the secondary put trigger zone of the convertible bond).
Per-ton profitability basis: The company's 2025 net profit attributable per ton was approx. RMB 303 (RMB 5.929 billion ÷ 19.5382 million tons), while key statistical steel enterprises tracked by CISA earned approx. RMB 120 per ton of crude steel in total profit over the same period — the two denominators differ (own sales vs. national crude steel output), so the ratio is for order-of-magnitude reference only, but the size of the gap suffices to show the company's profitability sits at the top of the industry distribution. Two dilutions must be acknowledged: first, non-GAAP net profit cumulatively declined 34.9% from 2021 to 2024 (RMB 7.646 → 4.975 billion) — the company experienced "a slower decline" rather than "immunity to the cycle"; second, operating cash flow / net profit attributable for the first three quarters of 2026 fell to 1.37 (-36.8%), with the Stanco trading platform consolidation driving accounts receivable +77% and inventory +21.5%, eroding cash conversion quality; annual report aging and collection data are the next validation point.
This report's bear-case scenario and sensitivity analysis must be presented alongside: EPV is highly sensitive to the normalized center assumption — excluding the 2021 super-cycle year (four-year average of RMB 5.54 billion) lowers EPV to approx. RMB 13.6; using the three-year average (RMB 5.36 billion) gives approx. RMB 13.2; if the per-ton net profit attributable center shifts down to RMB 240/ton (implying non-GAAP earnings of approx. RMB 4.7 billion), EPV is only approx. RMB 11.9, and the current price would be roughly fair. In other words, "18% undervalued" is the conclusion when the center assumption is RMB 5.96 billion; every 10% downward shift in the center narrows the margin of safety by approx. RMB 2 per share. Net assets of RMB 8.81 per share (2026-09-30) provide an asset-floor reference.
Two cautious amendments: first, "single-quarter weakness" has false-positive precedent — after Q3 2024's single-quarter gross margin of 12.51%, Q4 2024 stabilized and 2025 non-GAAP rebounded +17.4%; hence this report adopts "two consecutive quarters" as the trend confirmation discipline; second, the H1 2026 export gross margin plunge occurred before the EU's new rules took effect (July 1), more a reflection of overseas price centers and order-mix changes; the new rules' volume-price impact will only fully manifest from H2 2026. Sell-side consensus (FY2026 net profit attributable of RMB 6.33 billion, implying a Q4 single quarter of RMB 2.17 billion or +35.7% YoY) is clearly aggressive; this report's own estimate is RMB 5.28–5.52 billion (Q4 single quarter of RMB 1.12–1.36 billion); as of 2026-10-29 (only 6 days after the Q3 report), no institution has revised down forecasts after Q3 disclosure — the downward revision is this report's inference, not a consensus that has already occurred.
Ordinary steel total supply is controlled (2025 crude steel -4.4%, continued cuts in 2026, CISA's "three fixed, three no" self-discipline), supporting a floor in steel prices — a weak industry-level support. But it must be noted: competitors such as Fushun Special Steel, Yongxing Materials, and Baowu Special Metallurgy can equally expand high-end special steel capacity via the equal-replacement channel; the transmission of "supply freeze → profits concentrating to leaders" does not necessarily hold in the special steel niche — the company's high-end barriers come from certification cycles and process accumulation (bearing steel production and sales ranked No.1 globally for over a decade, flexspline steel entering the supply chains of leading reducer makers), not policy exclusivity. Also to clarify: the "30-million-ton-class integration platform" often cited by the market is the CITIC Group figure (listed company approx. 20 million tons + Nanjing Steel's approx. 10 million tons at the group level); the listed entity's 2025 sales were 19.5382 million tons, and figures should not be mixed before the Nanjing Steel injection.
The convertible bond game is an ongoing constraint that has already begun, not a future event: to promote conversion, the stock must stand above RMB 17.80 (13.69 × 130%, +49% vs. current price) — low probability during an earnings downturn; the more realistic path is a further downward revision before maturity (revision floor near book value per share of RMB 8.81, implying maximum dilution of approx. 11%) or cash redemption of approx. RMB 5 billion in February 2028 (at which point net debt/EBITDA will jump). The stock price is framed by three thresholds: RMB 9.58 (secondary put trigger), RMB 10.95 (revision trigger), RMB 17.80 (forced redemption trigger).
| Metric | 2023FY | 2024FY | 2025FY | 2026Q1-Q3 (latest period) |
|---|---|---|---|---|
| Revenue (RMB 100M) | 1140.19 | 1092.03 | 1073.73 | 807.06 (-0.62%) |
| Net profit attributable (RMB 100M) | 57.21 | 51.26 | 59.29 | 41.60 (-3.94%) |
| Non-GAAP net profit (RMB 100M) | 52.63 | 49.75 | 58.41 | 40.74 (-5.00%) |
| Gross margin | 13.14% | 12.85% | 14.94% | 14.09% (Q3 single quarter 12.76%) |
| Net margin (attributable basis) | 5.02% | 4.69% | 5.52% | 5.15% |
| Operating cash flow (RMB 100M) | 106.58 | 107.41 | 146.36 | 57.12 (-36.81%) |
| Free cash flow (OCF-CapEx, RMB 100M) | 86.61 | 101.89 | 138.66 | 51.84 |
| Steel sales volume (10K tons) | 1856.9* | 1888.98 | 1953.82 | 992.62 (H1, +1.05%) |
| Gross margin per ton (RMB/ton, estimated) | 793 | 743 | 821 | approx. 799 (H1 basis) |
| Cash and cash-like assets (RMB 100M) | — | — | — | 76.21 (2026-09-30) |
| Interest-bearing debt (RMB 100M) | — | 356.83 | 332.34 | 332.32 (2026-09-30) |
| Debt-to-asset ratio | — | 60.53% | 58.78% | 58.17% (2026-09-30) |
*2023 sales volume estimated from the YoY base disclosed in the 2024 annual report. R&D spending: RMB 4.403 billion in 2025 (4.10% of revenue, 0% capitalization rate), RMB 4.602 billion in 2024. Net assets per share RMB 8.81 (2026-09-30); total shares 5.047 billion.
Reasons for metric changes (items with YoY ≥ ±20%): ① Q3 2026 accounts receivable +77.0% vs. year-start (52.31 → 92.60 RMB 100M), inventory +21.5% (162.39 → 197.25 RMB 100M), short-term borrowings +70.0% (68.18 → 115.89 RMB 100M) — mainly due to Stanco (Fujingte) consolidation at end-August bringing in trading-related assets and liabilities, plus Q3 stockpiling; the company's official explanation for the cash flow decline is "mainly due to changes in accounts receivable and inventory," without separately explaining the surge in receivables; ② 2025 operating cash flow +36.25%, mainly due to improved profitability and lower working capital occupation (receivables -3.2%, inventory -3.0%); ③ 2025 income tax expense +100.5% (effective tax rate 9.54% → 15.76%), not specifically explained in the annual report; the direction is conservative (over-accrual of tax).
Q3 single-quarter revenue of RMB 27.034 billion (+2.05%), net profit attributable of RMB 1.170 billion (-23.64%), non-GAAP of RMB 1.134 billion (-25.38%), single-quarter gross margin of 12.76% (a clear sequential decline from H1's 14.77%), single-quarter net margin of 4.33%. Stable volume, weak price is the main theme: sales volume in the first three quarters was roughly flat YoY (H1 +1.05%), but per-ton profitability was squeezed by three factors — a plunge in export gross margin (H1 14.82%, -4.29pp), strength in coke and other raw material costs ("weak ore, strong coal and coke"), and ordinary steel's shift toward premium grades intensifying low-to-mid-end competition. Operating cash flow of RMB 5.712 billion for the first three quarters (-36.81%) was the weakest in recent years; beyond consolidation effects, it also reflects weakening collection pace. Compared with the high base of Q3 2025 (RMB 1.532 billion), part of the YoY decline is a base effect, but the 12.76% single-quarter gross margin is already below the 2024 full-year level, confirming a downward sequential trend. Sell-side has not yet revised down earnings forecasts after disclosure (most forecasts were formed before October 23); this report estimates Q4 single-quarter net profit attributable of approx. RMB 1.12–1.36 billion in the traditional off-season, with full-year net profit attributable of RMB 5.28–5.52 billion.
Business model: Asset-heavy, full-process integrated special steel smelting and rolling (approximately 20 million tonnes of capacity, with multi-site coastal/riverside layout across Jiangyin, Daye, Qingdao, Jingjing and other bases), 100% direct sales, with products including bearing steel, automotive steel, and energy-grade seamless tubes — no subscription or recurring revenue, but certification lock-in with top customers (automotive/bearings/wind power) creates de facto long-term contract stickiness; after consolidating Stanco in August 2026, a global steel trading platform is added (network covering 75+ countries), extending the business model toward "materials manufacturing + global distribution services."
Cash content of earnings: Excellent long-term, under pressure short-term. OCF/net profit attributable to parent for 2023–2025 was 1.86/2.10/2.47, and FCF/net profit attributable to parent was 1.51/1.99/2.34 — book profits convert almost fully into freely disposable cash, with depreciation (approx. RMB 4.9 billion annually) far exceeding capex; cash generation on the balance sheet is first-tier among manufacturers. However, OCF/net profit attributable to parent fell to 1.37 in the first three quarters of 2026, as trading-related working capital assets brought in by the Stanco consolidation (receivables +77%) began to dilute cash quality; the gap between non-recurring-excluded and attributable net profit was only 1.5%, with no one-off items smoothing results.
Return on capital: Calculated ROIC of approx. 8.6% (2025, NOPAT RMB 6.77 billion ÷ invested capital RMB 78.7 billion), roughly in line with the steel industry WACC (8–10%) — scale and synergy have created above-average asset returns, but have not significantly exceeded the cost of capital; weighted ROE of 14.08% (2025 disclosed figure) is significantly higher than peers, mainly due to leverage amplification from net debt operations (net debt RMB 25.6 billion, net debt/equity approx. 55%).
Capex quality: This is the report's biggest reservation on long-term competitiveness. CapEx collapsed from RMB 1.997 billion in 2023 to RMB 770 million in 2025, and only RMB 528 million in the first three quarters of 2026; CapEx/depreciation was only 0.11–0.16, and construction in progress was only RMB 2.04 billion (1.8% of total assets). Two-sided interpretation: in the short term, this is a harvest-period dividend — FCF is passively amplified, supporting high dividends and M&A; over the long term, net fixed assets stand at only 53% of the original value, and high-end special steel upgrading (special metallurgy forging, aviation bearing steel) depends on sustained heavy asset investment — investment running below depreciation for consecutive years will erode equipment precision and new-product capability over a 3–5 year horizon. A portion of the "high FCF" is a temporary manifestation of "underinvestment"; no disclosure of maintenance vs. growth capex split, so precise quantification is impossible.
Moat and quality warnings: The moat is real — certification cycles measured in years, bearing steel production and sales ranked #1 globally for over a decade, and export gross margins consistently 8 pp higher than domestic sales, demonstrating product-side pricing power. Two quality warnings: ① legacy inefficient lines (plate/pellet/beneficiation) have accumulated RMB 1.42 billion in impairments, and "impairment of specific production lines" has been repeatedly flagged by the auditor as a key audit matter; ② in the first three quarters of 2026, overall net profit fell -7.13%, a larger decline than the -3.94% attributable to parent, with the gap from collapsing minority-interest income (-92%, with profitability deteriorating more sharply at non-wholly-owned units such as Tianjin Pipe and Qingdao Special Steel) — cycle resilience on the attributable-to-parent basis is partially buffered by the minority-interest structure; the true downturn on a consolidated basis is steeper.
Consistency between words and deeds: prudent on volumes, systematically optimistic on exports. Sales volume commitments have a high fulfillment rate: 2024 plan 19.2 million tonnes vs. actual 18.89 million tonnes (98.4%); 2025 plan 18.6 million tonnes vs. actual 19.5382 million tonnes (+5% above plan). But export targets have missed for two consecutive years: 2024 target 2.8 million tonnes vs. actual 2.2 million tonnes (78.6%); 2025 target 2.65 million tonnes vs. actual 2.3 million tonnes (86.9%) — export commitments have been systematically optimistic amid deteriorating tariff conditions; the 2026 target was lowered to 2.4 million tonnes (still implying +4.2%). Another signal: investors pressed for quantitative answers on capacity utilization and revenue share of steel for robotic reducers on investor platforms, and the company's replies contained no numbers (2026-07 minutes, 2026-09 Interaction Platform) — disclosure is selectively conservative.
Shareholder friendliness: stable dividends, no buybacks. Cash payout ratio of approx. 50% for five consecutive years (actual 49.69%/49.86%/49.94% for 2022/2023/2024, and 54.97% in 2025 with the first interim dividend), cumulative dividends of approx. RMB 22.7 billion since listing; aside from the RMB 5 billion convertible bond issued in 2022 (minimal conversion, constituting a repayment obligation due 2028), no share placements or rights issues. Verdict: shareholder-friendly (on the dividend dimension), but cash-like assets are only RMB 7.62 billion vs. annual dividends of approx. RMB 3.26 billion plus the RMB 5 billion convertible bond awaiting repayment — dividend sustainability depends on cash flow conversion rather than reserves.
Risk signals: Large related-party transactions — RMB 9.594 billion of daily related-party transactions in 2025, including RMB 4.295 billion of purchases of Australian iron ore from related party SINO IRON (largest supplier is a CITIC Group affiliate); both deposits and borrowings are partially intermediated through CITIC Finance Company (deposits RMB 1.946 billion, 26.7%; loans RMB 5.291 billion, 20.8%, with regulatory metrics compliant); shareholder loan framework quota was reduced from RMB 5 billion to RMB 2.5 billion (balance at end-2025 was 0). Group dependence is a "normal but requires continuous monitoring" structural exposure. Governance action: the removal of "Taifu" from the company name was completed in September–October 2026 (ticker symbol unchanged), a brand-level adjustment.
| Segment (2025FY) | Revenue Share | Gross Margin | YoY | One-line Business Logic |
|---|---|---|---|---|
| Special steel bars | 40.61% | 16.94% (+2.48pp) | +1.35% | Bearing/automotive steel, global bearing steel leader, main profit driver |
| Special seamless steel pipes | 26.67% | 16.62% (+1.56pp) | +5.70% | Tianjin Pipe OCTG, driven by energy cycle; the only segment with positive growth in 2025 |
| Special steel wire rods | 13.33% | 12.11% (+2.18pp) | -5.12% | Fasteners/tire cord/spring steel; export volume +72% in 2026H1, lowest gross margin |
| Special steel plates | 9.90% | 14.36% (+1.54pp) | -7.21% | Includes impaired plate lines; weakening profitability (2026H1 gross margin 12.94%) |
| Other businesses (pellets/coke, etc.) | 9.49% | 6.20% | -18.79% | Raw-material self-supply extension, low-margin internal transfers |
| By region: Exports | 12.88% | 22.20% (+3.14pp) | -0.62% | Exports of 2.3029 million tonnes; 2026H1 gross margin plunged to 14.82% |
Profit drivers: By "revenue share × gross margin," bars contribute approx. 41% × 16.94% ≈ a dominant share of gross profit (2026H1 bars gross margin continued rising to 18.17%), with seamless tubes second — together contributing approx. 70% of gross profit, both "high-end products + certification barrier" businesses. Gross margin structure gap: the highest (exports 22.2%, bars 16.9%) vs. the lowest (other businesses 6.2%) differ by approx. 16 pp, essentially the difference between "deeply processed high-end materials" and "raw-material self-supply byproducts"; plates (including impaired lines) sit in the middle-to-weak range and are the part of the structure most in need of contraction. Trend watch: in 2026H1 gross margins fell across all segments (seamless tubes 14.73%, plates 12.94%), with only bars rising against the trend — the "high-end product mix offsets industry downturn" narrative holds for bars, but one pillar cannot support the whole.
Accounting red flags (by severity): ① Impairment testing of specific production lines relies on management assumptions (key audit matter; cumulative impairments RMB 1.92 billion; forecast gross margins/discount rates highly subjective) — Medium; ② Consolidation-driven surge in receivables/inventory and deterioration in operating cash flow (receivables +77%; trading asset quality and bad debt policy to be verified in the annual report) — Medium; ③ Effective tax rate jumped to 15.76% in 2025 without specific explanation (direction is conservative, not profit embellishment) — Low; ④ High both-way deposits and loans at the related finance company (pricing disclosed and deliberated) — Low; ⑤ Supplier letters of credit for trade financing of RMB 5.33 billion booked in short-term borrowings (inflates apparent interest-bearing debt) — Low.
Cross-period consistency: ① Capex vs. depreciation (0.45→0.11→0.16→approx. 0.1; company explains "more transfers to fixed assets and relatively little new capital expenditure," consistent with construction in progress -48.7% — internally consistent, but long-term implications see the two-sided interpretation in the "Business Model and Earnings Quality" chapter); ② Non-recurring-excluded net profit series (RMB 7.646→6.079→5.263→4.975→5.841 billion) is directionally consistent with industry profit fluctuations, and the official attribution of the 2025 rebound to "raw material/fuel price declines exceeding steel price declines" is self-consistent; ③ Divergence between export gross margins and the export narrative: the 2025 annual report emphasized "export product gross margins above domestic" (22.2%), yet 2026H1 plunged to 14.82% without specific explanation — inconsistent; the company has not explained; whether the high export premium was a one-off window (quota front-running/price window) awaits verification in the annual report; ④ Inventory write-down provisions of RMB 764 million vs. reversals of RMB 776 million largely offset, no "profit reservoir" operations observed. Audit opinion: KPMG Huazhen, unqualified (for consecutive years), effective internal control audit.
Current market data: Share price RMB 11.91 (2026-10-29 closing price), market cap RMB 60.11 billion, PB 1.35x (lowest quintile in the past 5 years), PE (TTM, non-recurring excluded) 10.7x (approx. 1–2% percentile over the past 5 years), dividend yield 5.42%. For cyclical stocks, PB-ROE is the primary anchor: ROE 14.0% ÷ PB 1.35x = implied equity return of approx. 10.4%, above the steel industry cost-of-capital band of 9–10% — i.e., the market is demanding an additional risk premium; the PE percentile is reference-only as earnings are at a mid-cycle position (TTM attributable net profit of RMB 5.758 billion is approx. 97% of the midpoint, not at an extreme trough, so the percentile remains reasonably informative).
| Peer Comparison | PE (TTM) | PB | 2025 ROE | 2026 Trend |
|---|---|---|---|---|
| CITIC Special Steel 000708 | 10.7 (ex non-recurring) | 1.35 | 14.0% | Q1-3 attributable -3.9% |
| Nanjing Iron & Steel 600282 | 10.5 | 0.95 | 10.67% | Q1-3 -17.4% |
| Jiuli Special Metals 002318 | 22.7 (TTM distorted by profit collapse; static 10.5) | 2.02 | 18.31% | Q1-3 -64.1% |
| Fushun Special Steel 600399 | Loss, N/A | 1.52 | -13.57% | Turned from profit to loss in 2025 |
Market-implied expectations: The current price of RMB 11.91 implies a mid-cycle PE of approx. 10.1x (÷ RMB 5.96 billion midpoint); pricing at its own 5-year PE median of 12.2x back-solves an implied normalized profit of approx. RMB 4.93 billion ≈ the 2024 trough. In one line: the current price implies normalized earnings reverting to the worst level of 2024 (approx. RMB 4.9 billion), whereas reality is that 2025 has recovered to RMB 5.84 billion and TTM stands at RMB 5.76 billion — expectations are conservatively low by approx. 10–17%. However, if industry demand enters a long-cycle downturn pushing normalized per-tonne attributable profit down to RMB 240/tonne (implying approx. RMB 4.7 billion ex non-recurring), the current price would be close to fair — the crux of the debate is "whether the leader's alpha can keep outpacing the industry's beta."
Three-layer value (EPV): Asset value (floor) RMB 8.81/share (book value per share as of 2026-09-30); EPV zero-growth value approx. RMB 14.5/share (normalized FCFE RMB 6.95 billion ÷ 5.047 billion shares ÷ 9.5% cost of equity; adding back depreciation of RMB 4.93 billion, deducting maintenance capex of RMB 3.94 billion, working capital approx. 0); EPV is approx. 22% above the current price — the growth option is negatively priced — the current price does not even cover the zero-growth value — no identifiable premium has been awarded to the Nanjing Iron & Steel injection, export recovery, Stanco synergies, or high-end product ramp-up. Sensitivity must be presented alongside: normalized midpoint at RMB 5.54 billion (ex 2021) → EPV RMB 13.6; at RMB 5.36 billion (three-year average) → RMB 13.2; at RMB 4.7 billion (bear-case midpoint) → RMB 11.9, with the current price close to this scenario. EPV is not a layer in a sum-of-the-parts, but an independent reference: the asset floor of RMB 8.81 plus a 5.4% dividend yield form the first layer of support; the lower bound of the EPV range (RMB 11.9–14.5) forms the second.
Three scenarios and odds (normalized EPS basis: RMB 5.5–6.0 billion ÷ 5.047 billion shares = RMB 1.09–1.19):
| Scenario | Probability | Fair Range | Midpoint vs. Current Price | Key Driver |
|---|---|---|---|---|
| Bear | 30% | RMB 7.1–8.9 | approx. -33% | Per-tonne gross margin falls below the RMB 700/tonne band (ex non-recurring RMB 4.5–5.0 billion) × trough multiple of 8–9x; dividend yield 8%+ and net assets of RMB 8.81 provide a cushion |
| Base | 47% | RMB 10.9–14.3 | approx. +6% | Normalized earnings RMB 5.5–6.0 billion × 10–12x (anchors: own 5-year median 12.2x, static median 11.7x, NIS current 10.5x, taking a conservative band) |
| Bull | 23% | RMB 15.5–17.8 | approx. +40% | Exports ≥2.4 million tonnes with gross margin recovering to 15%+; ex non-recurring RMB 6.5–7.2 billion × 12–12.5x (recovery between Jiuli's static 10.5x and sell-side cited comps at 14x) |
Exit multiple anchoring note: the base case's 10–12x band is anchored to the company's own 5-year PE median of 12.2x and NIS's current 10.5x (two anchors independent of the current price), avoiding the circular anchoring of "using the current multiple to justify the current price"; the current price corresponds to ex non-recurring TTM of 10.7x,恰 in the lower half of the base band. Normalized P/E ±5x sensitivity: base-case EPS of RMB 1.09–1.19 × (median 11x ± 5x) = approx. RMB 6.5–17.9, showing conclusions are similarly sensitive to multiple assumptions — hence this report presents odds via a scenario probability distribution rather than a point target price. Probability-weighted expectation is approx. RMB 12.1 (+2% vs. current price); the odds distribution is mildly positively skewed: downside -33% (cushioned) vs. upside +40% (requires export recovery to materialize).
In-house earnings forecast (reconciliation anchor):
| Period | Revenue (RMB 100M) | Net Profit Attributable to Parent (RMB 100M) | Key Drivers |
|---|---|---|---|
| 2026E | 10.40–11.20 billion | 5.28–5.52 billion | Volume 19.1 million tonnes (plan, -2.2% YoY); per-tonne gross margin falling from RMB 821 to RMB 780–800; Q4 single quarter RMB 1.12–1.36 billion; limited profit contribution from 4 months of Stanco consolidation |
| 2027E | 10.60–11.60 billion | 5.5–6.3 billion | Export target of 2.4 million tonnes achieved + full-year Stanco consolidation; export gross margin recovering toward 15%; per-tonne profit reverting to midpoint from the Q3 low; price elasticity from supply-side clearing |
Comparison: sell-side consensus for 2026E is RMB 6.33 billion / 2027E RMB 6.76 billion (n=6, most formed before the Q3 report); management guidance provides only volumes (19.1 million tonnes / exports 2.4 million tonnes), no profit figures. This report's forecast is 13–17% below sell-side and 7–11% below the 2025 actual — upside to revision exists if Q4 per-tonne gross margins stabilize sequentially; otherwise sell-side faces downward revisions.
Conclusion: reasonably cheap. Assess quality and price separately: quality is strong (global leader, top-tier cash conversion, 50% payout ratio), and the price sits in the 5-year trough zone with trough-level expectations priced in — but positive margin of safety has not yet emerged (current price of RMB 11.91 is above the base-case fair lower bound of RMB 10.9; margin of safety approx. -8%), and the earnings inflection is unconfirmed. The degree of undervaluation (10–17%) is insufficient to cover the cost of a wrong call; neutral patience is the better risk-adjusted choice.
Market size: Special steel refers to steel with special properties derived from special composition, processes, or microstructure (structural alloy steel, bearing steel, gear steel, spring steel, tool and die steel, superalloys, special stainless steel, etc.), as distinct from ordinary steel (commodity materials such as rebar and hot-rolled coil). Authoritative data under the China framework (China Special Steel Enterprises Association): in 2025, output of key statistical member units of the Association was 109.06 million tonnes, accounting for 11.35% of national crude steel output (961 million tonnes), of which quality and special steel products output was 75.91 million tonnes. Growth history: from 2021 to 2025, special steel output grew at a low single-digit rate (1–5%), while national crude steel output fell from 1.05 billion tonnes to 961 million tonnes (−8.6%) — special steel outperformed total ordinary steel by roughly 10 percentage points; structural substitution is the main theme. Outlook: the Association expects major enterprises' special steel output to remain broadly flat at about 83 million tonnes — volumes flat, structure upgrading — with industry targets for 2030 of a 15% sales share for high-end products, >95% self-sufficiency in high-end special steel, and a 25% share of global special steel for China (Association framework); third-party forecasts (e.g., Sealand Securities) project high-end special steel volume CAGR of about 13% over the next 10 years (directional reference). Benchmark potential: special steel accounts for roughly 11% of China's crude steel output versus 20% in Japan and 45% in Sweden — the penetration-upgrade logic holds over the long term, though the pace depends on manufacturing upgrading.
Value chain and value distribution: Upstream comprises iron ore (high import dependence; Platts 62% index down 14.4% YoY in 2025H1), coking coal and coke (strengthening into 2026, a core cost variable), scrap steel, and the ferroalloys specific to special steel (chromium, molybdenum, vanadium, nickel, tungsten — high-purity requirements lift their value; ferromolybdenum was at a high level of about RMB 338,000/tonne in August 2026). Midstream smelting and rolling is where value accrues: special steel gross margin per tonne far exceeds ordinary steel (CITIC Special Steel's overall gross margin about 14.4–14.9% vs. the sales profit margin of key ordinary steel enterprises of only 1.89%; Nanjing Steel's advanced steel products gross margin 20.88% vs. 2.84% for its rebar). Downstream, automobiles account for roughly 40% of special steel demand, industrial manufacturing 20%, metal products 17%, with the remainder from energy, petrochemicals, wind power, aviation, and nuclear power. The Company sits in the integrated midstream and is extending into global trading: it holds relatively strong bargaining power over upstream suppliers (concentrated procurement at the 20-million-tonne scale plus long-term contracts plus captive coking plants and self-owned power) and strong bargaining power over downstream customers (certification barriers plus customization, evidenced by an 8-percentage-point export premium) — placing it in a relatively strong position within a landscape where the industry overall is squeezed by ore prices.
Supply-demand and competitive landscape: On the demand side, total volume fluctuates with manufacturing capex (auto production down 3.4% in January–July 2026), but the special steel share of crude steel output is rising secularly; structurally, new demand from EV lightweighting, wind power (about 100 GW of onshore installations in 2025, a record high), nuclear power, aviation, and robotics is growing faster than traditional demand. A quantitative assessment of new demand such as robotics (important): the first year of mass production of humanoid robots (about 19,100 units shipped globally in 2026H1) is indeed a genuine increment, but special steel usage per robot is only about 14 kg (alloy steel plus bearing steel combined, per industry estimates) — at 50,000 units in 2026, incremental demand would be only about 700 tonnes, less than 0.04% of the Company's sales volume; even 10 million units in the long run corresponds to a market of only about RMB 1.49 billion, just 1.4% of the Company's annual revenue. This direction is a "high-margin product option plus valuation label," not an earnings driver, and should not be built into base-case forecasts — the Association has likewise made clear that emerging demand "overall represents a product-mix adjustment and does not greatly affect total special steel demand." On the supply side, the capacity tiers (Association framework): the CITIC Special Steel group, with 30 million tonnes+ (including the group's Nanjing Steel platform), stands alone at the top; Baowu and Shagang groups at 10 million tonnes+; HBIS, Jianlong, and Hunan Iron & Steel at 5 million tonnes+. The listed company's own sales volume accounts for about 25.7% of member units' quality and special steel output (own calculation: 19.5382/75.91 million tonnes). In the bearing steel niche, CITIC Special Steel (Xingcheng, Daye, Qingdao) plus Jiyuan and Juneng together account for about 65% of the bearing steel output of major quality/special steel producers (Association/Mysteel sample, January–April 2025). Entry barriers: certification cycles measured in years, accumulated refining expertise for cleanliness (bearing steel requires oxygen content <10 ppm via vacuum degassing; flexspline steel imposes extreme purity requirements), capital intensity, and breadth of product categories — mid-to-low-end quality/special steel faces oversupply against weak demand (conversion from ordinary to special steel intensifies price wars), while high-end niches ("three-highs and one-special"/superalloys) are in tight balance with import-substitution upside. The special steel industry contributes about 16% of steel industry profits on only about 11% of output; its earnings resilience is structurally higher than ordinary steel's.
Cycle and regulation: Ordinary steel is in a weak-recovery-under-renewed-pressure stage after a deep downturn (2025 key steel enterprises' profit of RMB 115.1 billion, up 140% YoY, was a recovery off an ultra-low base; 2026H1 industry profit fell 25%, bottoming again, with the traditional "golden September, silver October" failing); the special steel sub-sector is relatively buoyant. Regulatory direction is broadly favorable to incumbent leaders: ① continuation of crude steel output controls (−4.4% in 2025); ② new capacity-swap rules (1.5:1 reduction; new capacity strictly prohibited in key regions) plus suspension of swap approvals since August 2024 — ordinary steel total supply is frozen; ③ "anti-involution" (July 2025 Central Financial and Economic Affairs Commission meeting) plus CISA-led self-disciplined output restraint; ④ completion of ultra-low-emission retrofits plus carbon quota management will eliminate inefficient capacity; ⑤ RMB 200 billion of special treasury bonds for equipment renewal supporting high-end upgrades; ⑥ risk factors: export license management from 2026, EU quota cuts of 47% plus 50% tariffs on excess volumes (effective 2026-07-01), and US tariffs of 70% — the policy mix protects domestic profitability while tightening exports; the net effect depends on the ability of companies with high export exposure (exports are about 12% of the Company's sales volume) to respond.
Company positioning: Clear leader. The world's largest specialized special steel producer; No. 1 globally in bearing steel production and sales for more than a decade running; leading domestic share in high-end automotive steel and mooring chain steel; 2025 attributable net profit per tonne of about RMB 303 versus about RMB 120 for key industry steel enterprises, with gross margins of 14.4–14.9% through the ordinary steel downturn (2025 attributable profit +15.7% vs. an industry profit margin of just 1.89%). Share trend steady to rising: group consolidation of Nanjing Steel in 2023, sales volume +3.4% in 2025, acquisition of the Staco global trading platform in 2026, and the 15th Five-Year Plan explicitly centered on overseas M&A integration. Sources of moat: ① scale and multi-site procurement synergies; ② breadth of six product groups hedging single-downstream volatility; ③ certifications and customer lock-in (flexspline steel entering the supply chain of a leading global reducer maker; aerospace bearing steel entering European and American supply chains); ④ R&D (RMB 4.4 billion in 2025, 4.1% of revenue; Xingcheng Special Steel won a Second Prize of the State Science and Technology Progress Award); ⑤ industrial and M&A synergies with controlling shareholder CITIC Group (four key integrations — Nanjing Steel, Tianjin Pipe, Shanghai Electric pipe assets, and Staco — all leveraged the group). Framework note: the company's annual report claims global No. 1 under the framework of "capacity/output of specialized special steel producers," which differs from the third-party framework (Association) of "about 25.7% of member units' quality/special steel output"; the former is a global dimension, the latter a Chinese member dimension — the two are not contradictory but citations must specify the framework.
Historical cycle template: The last two full cycles — ① 2016–2021 upcycle (triggers: supply-side reform output cuts plus property/infrastructure demand plus global inflation), with 2021 the Company's earnings peak (deducted non-recurring profit of RMB 7.646 billion, adjusted framework); ② 2022–2024 trough (triggers: deep property adjustment; apparent crude steel consumption falling for five consecutive years, cumulatively −20.9% from the 2020 peak), with the Company's deducted non-recurring profit bottoming at RMB 4.975 billion in 2024 (cumulative −34.9%); ③ 2025 policy-driven recovery (RMB 5.841 billion, +17.4%, driven by falling raw material/fuel costs plus mix upgrading); ④ 2026 renewed bottoming (Q3 single quarter −23.6%). Cycle length: downturns about 3 years (2021H2–2024); bottoms in 2015 and 2024 about 9 years apart; CSPI steel price index peak-to-trough amplitude about 40–50%. The Company's earnings cycle amplitude (−34.9%) is markedly smaller than the industry's (key steel enterprises' profit down 56.4% in the first three quarters of 2024), though directionally in sync.
Current cycle position: CSPI at about 91–93 points (2026Q1 average of 91.39, the lowest for the period in years), around the 10–15th low percentile since 2015; key steel enterprises' sales profit margin 1.89% in 2025 (2026H1 profit −25%; listed steel companies' net margin about 0.7%), around the 15–20th percentile since 2015. Company framework: gross profit per tonne of RMB 821 in 2025 (four-year high) → RMB 799 in 2026H1 → Q3 single-quarter gross margin of 12.76% implies roughly RMB 760–770 — a position "above the mean but trending down," not yet at the 2024 extreme low (RMB 743/tonne), let alone a 2015-style deep trough.
Supply response: The key difference in this cycle is that supply is under strong policy constraint — output cuts to crude steel continuing, capacity frozen via the swap regime (near-zero new additions), licenses plus tariffs forcing export volumes back home, and carbon quotas eliminating inefficient capacity. Supply cadence over the next 2–3 years: modest total decline, ordinary steel additions frozen, but an approval channel for equal-volume swaps in high-end special steel (differentiated terms) — total-volume clearing benefits incumbents' profitability, but supply elasticity in high-end special steel niches exceeds that of ordinary steel; the timing of the cycle bottom depends on when demand stabilizes rather than on supply contraction itself.
Through-cycle (normalized) earnings: Mid-cycle deducted non-recurring profit center of RMB 5.5–6.0 billion (range between the five-year average of RMB 5.96 billion and the three-year average of RMB 5.36 billion), corresponding to normalized EPS of RMB 1.09–1.19 and attributable profit per tonne of RMB 275–300); trough EPS RMB 0.986 (2024 actual of RMB 4.975 billion); normalized P/E of 10–12× (±5× sensitivity: RMB 6.5–17.9). Qualitative view of current P/E: mid-cycle pricing leaning toward trough — TTM deducted non-recurring P/E of 10.7× is neither a "peak low-P/E trap" (earnings are not at a peak) nor trough high-P/E (earnings are not in loss), but the market-implied earnings expectation (about RMB 4.9 billion) is already priced at trough levels. Normalized valuation, two tiers: industry-center 10–11× → RMB 10.9–13.1; leader premium at 12× → RMB 13.1–14.3 (a premium for export barriers and integration capability, requiring validation via export gross margin recovery). Asset-value floor: book value per share of RMB 8.81 (PB 1.0× corresponds to a share price of RMB 8.81, near the upper bound of the bear-case range — a hard-floor reference).
Downside stress test: Gross profit per tonne back to the RMB 743/tonne tier (2024 actual) → attributable profit about RMB 4.3 billion (−27%), net debt/EBITDA rising from about 2.2× to 2.7×; RMB 700 tier → about RMB 3.6 billion, 2.9×; RMB 650 tier → about RMB 2.8 billion, 3.25×; RMB 600 tier → about RMB 2.0 billion, 3.7× (framework: change in gross profit × 0.85 after-tax attributable factor). Liquidity: cash of RMB 7.62 billion covers short-term debt (RMB 17.89 billion) at 0.43×, or 0.77× including receivables financing of RMB 6.17 billion — in the trough scenario, liquidity can hold up (annualized OCF of about RMB 7.6 billion covering capex of RMB 0.7 billion plus dividends of about RMB 3.3 billion), but the buffer thins markedly; if the RMB 5 billion convertible bond maturing February 2028 is also added, deleveraging and dividend maintenance become hard to reconcile, the dividend-yield anchor (5.4%) could fail, and valuation could slide toward book value of RMB 8.81. Earnings-trough alignment note: 2024 was the Company's trough year for deducted non-recurring profit (RMB 4.975 billion), coinciding with the low point of gross profit per tonne; if 2026 declines further, Q4 2026–H1 2027 is the window to observe whether there is a "second bottoming or stabilization".
Management's cycle discipline: Defensive tilt plus countercyclical integration. At the cycle bottom (2024–2026), sharply curtailed organic capex (CapEx/depreciation 0.11–0.16), maintained a payout ratio of about 50–55% without buybacks, executed countercyclical acquisitions of the Staco trading platform (RMB 1.51 billion, settled August 2026) and establishment of a new advanced materials fund (RMB 389 million LP commitment); shareholder loan facilities were instead cut from RMB 5 billion to RMB 2.5 billion. Verdict: capital allocation discipline is sound, but the "cut organic capex + high payout" combination carries an implicit cost to long-term competitiveness; whether capex rebounds in 2027 is the key validation point.
Overall rating: Neutral, confidence 0.58, horizon 12 months. CITIC Special Steel is of high quality (global special steel leader, top-tier cash conversion, stable dividends); the current price sits at the lowest PB percentile in five years, with a 5.4% dividend yield and EPV of RMB 14.5/share about 18% above the current price — but earnings have yet to bottom (Q3 single quarter −23.6%, abrupt drop in export gross margin, sell-side expectations likely to be cut), and the current price falls within the lower half of the base-case fair-value range (RMB 10.9–14.3) with no positive margin of safety. Strategy: add to the watchlist without increasing the position; wait for a right-side signal — conditions for upgrading to positive are "gross profit per tonne ≥ RMB 700/tonne for two consecutive quarters and export gross margin stabilizing and recovering" or "launch of Nanjing Steel injection at a reasonable price"; if gross profit per tonne falls below the RMB 700/tonne tier or the payout ratio drops below 40%, shift to defensive.
Risks and monitoring points (next 12 months):
Falsification conditions (reversal of stance): gross profit per tonne ≤ RMB 700/tonne for two consecutive quarters; annual export volume <2 million tonnes or export gross margin <10%; payout ratio <40%; underlying share price persistently below RMB 9.58. Upside re-rating conditions: gross profit per tonne ≥ RMB 800/tonne for two consecutive quarters plus export gross margin back above 15%; Nanjing Steel injection plan announced at ≤1.3× PB; industry profit margin back above 3%.
The price anchor of this report is the 2026-10-29 closing price of RMB 11.91 (CNY); financial data are as of the Q3 2026 report (disclosed 2026-10-23); industry data sources include the China Special Steel Enterprises Association, China Iron and Steel Association, MIIT, Mysteel, etc. This report is independent research and does not constitute investment advice.