← Back to Blog · 2026-09-27 · 8 min read · Strategy Case Study

Picture this: it's late 2021, and you're short rebar because the Chinese property sector is visibly cracking. You're right — rebar falls hard. But in the same window, iron ore is falling even harder, because steel mills are cutting output and buying less raw material. Your directional trade made money, but a trader on the other side of the value chain — long steel, short ore — made more, with a cleaner story and smaller drawdowns along the way.

That's the quiet appeal of pair trading a value chain. Instead of betting on where steel prices go, you bet on the relationship between steel and its input. In this case study, we'll walk through the rebar-iron ore spread on Chinese commodity futures: what it represents, the exact contract mechanics, how it would have behaved through the major cycles of the past several years, and the practical rules you'd need to trade it yourself.

What the Rebar-Iron Ore Spread Actually Is

Rebar (rebar steel, used in construction) is produced from iron ore in blast furnaces and rolling mills. When you buy rebar futures and sell iron ore futures simultaneously, you're essentially buying steel mill margin — the profit a mill earns per tonne of steel after paying for raw materials.

This is why the trade is often called the "steel profit" or "mill margin" spread. The logic is intuitive:

• Long rebar / short iron ore = betting mill profitability expands. This tends to work when steel supply is constrained (capacity cuts, production restrictions) while demand holds up.
• Short rebar / long iron ore = betting mill profitability compresses. This tends to work when raw material supply is disrupted (mine accidents, export restrictions from major ore producers) or when steel demand collapses faster than mills can cut production.

Because both legs are in the same physical chain, the spread is far less exposed to broad macro shocks than either leg alone. A global risk-off event that dumps every commodity at once tends to hit both legs. What's left is the relative move — which is exactly the thing your thesis is about.

Contract Specs: The Nuts and Bolts

Before any strategy talk, you need the mechanics, because these two contracts are very different animals in size.

Notice the asymmetry: one lot of iron ore controls about ten times the tonnage of one lot of rebar. At typical prices, one lot of iron ore has carried a notional value roughly two to three times that of one lot of rebar. That means a naive 1:1 lot ratio is heavily tilted toward iron ore.

The common practical fix is a 2:1 or 3:1 lot ratio (rebar:iron ore), adjusted as prices move, to roughly balance notional exposure. Some traders instead balance by volatility — iron ore historically moves with larger daily percentage swings than rebar, so they may shade the ratio further toward rebar. There's no single "correct" ratio, but you must pick one deliberately and understand what it implies. Get it wrong and you're not trading a spread at all — you're just running an iron ore position with a rebar hedge that doesn't cover you.

Also note: both exchanges apply margin at levels that change with volatility and policy (often in the high single digits to low teens of notional), and DCE has position limits and, at times, transaction fee adjustments during speculative bursts. Check current exchange notices before sizing anything.

How the Spread Behaved Through Recent Cycles

Here's where it gets interesting. The mill margin spread has gone through dramatically different regimes in the past decade, and each regime punished or rewarded a different assumption.

The supply-side reform era (roughly 2016–2018)

China's supply-side structural reform shut down significant illegal and low-grade steel capacity while infrastructure and property demand stayed firm. Steel prices recovered sharply from the 2015 lows, and mill margins — which had been razor-thin or negative — expanded to historically rich levels. A long rebar / short iron ore spread was the textbook expression of this: mills were printing money, and the spread trended in their favor for an extended period. Traders who understood the policy backdrop had a multi-year tailwind, not just a trade.

The ore supply shock (early 2019)

Then came the Brumadinho dam disaster in Brazil and subsequent supply disruptions from Vale, the world's top iron ore producer. Seaborne ore supply tightened materially, iron ore prices rallied strongly through 2019, and mill margins got squeezed. The long-steel/short-ore spread — the consensus trade of the prior era — bled. This is the regime that teaches the most important lesson: the spread is not a one-way bet. When the raw material leg is the one with the supply story, the spread works against the classic direction, sometimes violently.

The 2020–2021 boom and bust

Post-COVID stimulus in China drove record steel production and strong construction demand into 2021. Rebar rallied to multi-year highs around mid-2021. But two forces then collided: Beijing's push to cap crude steel output (which hit ore demand hard) and the property sector downturn that accelerated in the second half of 2021 (which hit rebar demand hard). Both legs fell — but the relative path was choppy, with ore's decline driven by policy and rebar's by demand destruction. Spread traders who assumed the 2021 rebar peak meant an automatic margin expansion were whipsawed. Directional traders who were simply short rebar did fine; spread traders needed the more nuanced view that output cuts could compress the spread even in a falling market.

The property drag (2022 onward)

Since 2022, China's property construction downturn has been the dominant rebar story, with new starts falling for an extended stretch and steel demand shifting toward manufacturing and infrastructure. Mill margins have generally lived in a lower, tighter regime, with the spread oscillating rather than trending. Repeated policy attempts to support property produced sharp but short-lived rebar rallies — good conditions for disciplined spread traders fading extremes, frustrating conditions for trend followers.

The meta-lesson: the rebar-iron ore spread is really a trade on Chinese industrial policy plus the ore supply cycle. If you can't articulate which of those two forces is currently dominant, you don't have a thesis — you have a coin flip.

A Practical Rule Set for Trading It

Let's get concrete. Here's a framework in the spirit of how experienced spread traders approach this market — not a holy grail, but a real starting point you can refine and test.

Position construction

• Core ratio: start at 2 lots rebar per 1 lot iron ore, rebalanced when the notional ratio drifts beyond roughly ±25% from target. This keeps you genuinely spread rather than accidentally directional.
• Define the spread in CNY terms: e.g., spread value = (rebar price × 10 × rebar lots) − (iron ore price × 100 × ore lots). Track it daily. All entries and exits should be defined against this series, not against eyeballing the two charts.

Entry logic

• Regime filter first: before any technical entry, write one sentence on the dominant policy/supply driver. Long margin needs a supply-constraint story for steel or a comfortable ore supply picture. Short margin needs an ore disruption or demand collapse story.
• Seasonality as a tilt, not a trigger: Chinese construction has well-known seasonal rhythms — the "Golden March, Silver April" spring demand window and the winter heating-season production restrictions in northern China (which curb steel output and can support the long-margin side, while also cutting ore demand, which cuts the other way). Use seasonality to bias timing, never as the sole reason to enter.
• Technical trigger: enter when the spread series pulls back against its medium-term trend — for example, a retracement toward a rising 20-day average on the spread, confirmed by a momentum turn. Keep it simple; the edge here is structural, not in indicator complexity.

Risk management

• Stop on the spread, not the legs: set your maximum adverse excursion on the spread value series (a common starting point is a multiple of the spread's recent average true range). Never widen it because "the story is still intact."
• Cap single-trade risk at a small fixed fraction of account equity — 1% or below is a sane default for spreads with this much policy risk.
• Watch for forced convergence risk: policy announcements (output cuts, export measures, exchange margin changes) can gap the spread overnight. Size so that a limit-move day on one leg doesn't blow through your stop.

What would have mattered most, historically?

Looking back across the cycles above, three things separated survivable spread trading from painful spread trading:

1. Respecting regime changes. The 2019 ore shock and the 2021 policy pivot each flipped the spread's character. Traders who treated the spread as a permanent structural long got run over.
2. Honest notional balance. The 10:1 tonnage asymmetry between the contracts is the single most common mechanical mistake in this pair.
3. Patience around policy windows. Major Politburo meetings, annual output-cut announcements, and winter restriction schedules repeatedly created the spread's biggest moves. Trading blind through them is optional; most people shouldn't.

The Risks Nobody Puts in the Brochure

• Policy risk is the tail risk. Chinese commodity futures live inside an industrial policy ecosystem. A single announcement about steel output caps or ore import measures can move the spread several percent in a session. There is no stop-loss that fully protects against a gap through limit positions.
• The spread can stay irrational. Mill margins have sat at both extreme compression and extreme expansion for far longer than relative-value traders expected. "It must mean-revert" is a timing claim, not a timing strategy.
• Contract roll complexity. Rebar and iron ore have different dominant contract months and liquidity profiles. Rolling the two legs on different schedules introduces basis risk you need to manage deliberately.
• Correlation isn't static. In risk-off episodes the two legs can correlate toward 1 and the spread becomes pure noise; in policy-driven episodes they can decouple hard. Know which environment you're in.

How to Actually Test This Before Risking Money

Everything above is testable, and it should be tested — on real data, with real contract specs, including the ratio drift and roll mechanics. The rebar-iron ore spread is one of the best educational trades in Chinese commodity futures precisely because it forces you to think about notional balance, policy regimes, and relative value all at once.

If you want to put a ruleset like this through its paces, you can run it against real-market China futures data in a structured evaluation at XS Select — evaluations start from $29. It's a straightforward way to find out whether your spread logic survives contact with live conditions before any meaningful capital is on the line.

The steel value chain will keep cycling — policy will tighten, ore supply will wobble, property will surprise in both directions. The traders who profit from the next regime won't be the ones who predicted it. They'll be the ones whose process was built to detect and adapt to it. Build the process first.