← Back to Blog · 2026-09-26 · 7 min read · Strategy Case Study

Picture this: it's late 2021, thermal coal on the Zhengzhou exchange has gone nearly vertical. Your breakout system is long and printing money. Then Beijing steps in, the exchange raises margins, and within weeks the market gives back most of the move — while your mean reversion friend, who'd been fading every parabolic bar, is suddenly the genius in the room. Same market, same few months, opposite outcomes.

That's the eternal debate in futures trading: do you buy strength or sell exhaustion? Most traders pick a side based on personality, not evidence. In this case study, we'll do it properly — define two simple, mechanical strategies with concrete rules, walk through how they would have behaved on well-known Chinese commodity futures moves you can verify from public price history, and compare where each one earns its keep. No curve-fitted backtest screenshots, no magic parameters. Just logic you can rebuild and test yourself on real data.

Why Chinese Commodity Futures Are a Different Arena

Before we compare strategies, you need to understand the playing field, because the microstructure of Chinese commodity futures shapes which style works.

• Retail dominance and high participation. Unlike CME products dominated by institutions, many Chinese contracts — rebar, methanol, PTA — have heavy retail flow. That flow is notoriously trend-chasing, which historically creates both fat trends (when the crowd piles in) and sharp mean-reverting snapbacks (when the crowd is wrong).
• Policy risk is a first-class variable. The NDRC's intervention in the thermal coal market in 2021, and supply-side reform driving the steel complex higher through 2016-2017, are publicly known events. Chinese commodity markets can reverse violently on policy headlines, not just on price action.
• Session structure. Most contracts trade day sessions plus limited night sessions (rebar's night session ends at 23:00, for example). Gaps between sessions are common, which punishes tight-stop breakout entries and rewards patience.

Know your contracts. Here are the specs that matter for position sizing:

One lot of iron ore moves 50 yuan per tick — that's real money relative to rebar's 10 yuan per tick. Position sizing differences like this decide whether you survive a losing streak, regardless of which strategy style you run.

Strategy A: Mean Reversion — Fading Exhaustion

The core idea: when price stretches too far from its average in a market without a genuine structural driver, it tends to snap back. Here's a simple, honest rule set you can code in an afternoon:

The Rules

• Universe: Methanol, PTA, rebar — contracts with enough liquidity but less single-narrative obsession than iron ore.
• Entry: Daily close more than 2 standard deviations below the 20-day mean (a classic z-score filter), and the daily close is down for at least two consecutive sessions.
• Direction: Long only on oversold stretches. Shorting rallies in Chinese commodities during policy-driven bull phases has historically been dangerous — think of the steel complex's grind higher after supply-side reform kicked in around 2016.
• Exit: Close when price reverts to the 20-day mean, or after a maximum 5-day holding period.
• Stop: 1.5x the 14-day ATR from entry. Non-negotiable — mean reversion without a hard stop is how accounts die in a genuine regime change.
• Filter: Skip entries when the 20-day slope is steeply positive. You don't want to fade a real trend; you want to fade panic.

The logic behind that last filter matters. When methanol sells off 4% in two sessions with no fundamental news, that's usually positioning, not fundamentals. When rebar drops hard because a city announces construction restrictions, that's information — stand aside.

Strategy B: Breakout — Buying Strength

The classic Donchian approach, adapted for Chinese market structure:

The Rules

• Universe: Iron ore, rebar, thermal coal — the contracts that produce the multi-week directional moves traders remember.
• Entry: Buy a close above the highest high of the prior 20 trading sessions. Symmetric for shorts below the 20-day low.
• Confirmation: Require the breakout close to be at least 1 ATR above the channel boundary. This filters out marginal pokes above resistance that reverse overnight.
• Exit: Close below the 10-day low (for longs), or a 2x ATR trailing stop, whichever hits first.
• Sizing: Risk a fixed fraction — say 1% — of account equity per trade, adjusting lot count to the contract multiplier. Iron ore's 100-ton multiplier means one lot can carry far more risk than one lot of rebar.

Breakout systems live and die by the fat tail. The 2021 thermal coal rally is the textbook case: a supply squeeze, a cold-winter demand narrative, and retail FOMO stacked into a move that went far beyond what most models predicted — followed by an equally violent policy-driven reversal. A breakout system with a trailing exit caught a large chunk of the up-move. A breakout system that ignored the margin-raise and position-limit announcements gave much of it back in days.

The lesson from 2021 isn't "don't trade breakouts in China." It's "trade breakouts with an exit that doesn't require the trend to be polite about ending."

The Backtest-Style Comparison: Where Each Style Wins

Instead of cherry-picked equity curves, let's reason through how these two systems behave across the market regimes that any long-term price chart of Chinese commodities will show you:

Regime 1: Sustained trends (steel complex, 2016-2017; thermal coal, 2021)

Breakout wins, and it isn't close. Mean reversion's trend filter keeps it mostly flat or lightly involved, but its occasional fade attempts against a policy-fueled trend produce a string of small stop-outs. Breakout compounds. This is where trend-following on Chinese commodity futures earns its reputation.

Regime 2: Choppy, range-bound markets (methanol in most non-catalyst periods)

Mean reversion wins. Breakout systems get whipsawed — every 20-day high breaks out, fails, and stops out. On a contract like methanol, where the multiplier is modest and moves are frequently positioning-driven, the z-score approach harvests the oscillation the breakout system pays for.

Regime 3: Violent reversals after parabolic moves (thermal coal, late 2021)

Mean reversion can produce its best trades of the year here — fading exhaustion after a vertical run — but only with the strict stop discipline described above, because the first leg of a policy reversal can extend further than any oversold reading suggests. Breakout systems are at their most vulnerable: late longs, wide stops, and a limit-down or two can erase months of profit.

Regime 4: The 2020-style macro shock

When crude went negative in April 2020 and global risk assets seized up, Chinese industrials gapped and trended violently in both directions. Breakout systems with symmetric rules had their best and worst trades within weeks. Mean reversion systems that faded the initial crash too early got hurt; those that waited for stabilization (the slope filter) caught the rebound. The differentiator wasn't the style — it was the filter discipline.

Head-to-Head: The Honest Scorecard

Notice that neither style dominates. The performance difference between a profitable trader and a frustrated one often isn't the entry style — it's matching the style to the contract's character and surviving the style's natural losing environment.

Practical Application: How to Choose and Test

Here's how I'd approach it if I were starting fresh on Chinese futures today:

• Pick one contract and one style first. Iron ore for breakout, methanol for mean reversion. Don't run both systems on both contracts until you understand each market's rhythm.
• Backtest with honest assumptions. Include slippage of at least 1-2 ticks per side, realistic margin, and — critically — model the session gaps. A breakout entry filled at the next day's open, not at the theoretical channel price, is a very different trade.
• Stress-test against known events. Ask: how does my equity curve behave through the 2021 coal intervention, the 2016 steel rally, the 2020 crash? If your backtest only covers calm periods, it tells you nothing.
• Watch the calendar. Chinese commodities have seasonal patterns — construction demand for rebar, heating demand for coal, methanol's winter MTO dynamics. A breakout in iron ore entering the traditional demand season behaves differently than one entering it.
• Size for the worst contract, not the average one. One lot of iron ore or thermal coal carries multiples of the risk of one lot of rebar. Fixed-risk sizing by multiplier is the difference between a losing month and a blown account.

And one more thing: a backtest is a hypothesis, not a result. The gap between a backtest and live execution — fills, discipline, regime shifts — is where most traders actually lose. That gap is exactly what a structured evaluation is designed to expose before real capital is at stake.

Closing: Test It Before You Trust It

Mean reversion and breakout aren't rivals — they're two tools for two different market personalities, and Chinese commodity futures contain both personalities, sometimes in the same quarter. The 2021 thermal coal saga proved that a market can be a trend-follower's dream and a trend-follower's nightmare within weeks.

The traders who last are the ones who test their rules on real historical data, survive their style's losing environments, and know exactly which contract their edge belongs to. If you've built a system on rebar, iron ore, or methanol and want to see how it holds up under evaluation conditions with real market data, that's precisely what we do at XS Select — you can run your system through a China futures evaluation starting from $29, and let the data, not your confidence, make the case.