← Back to Blog · 2026-10-04 · 7 min read · Trading Education

You've done it on paper a hundred times. Glass futures pull back to a level you like, you size up because the contract looks "cheap," and you wait. Then a supply-side headline drops overnight, the contract opens near limit-down, and your stop — the one you placed 40 ticks away — fills somewhere you didn't plan for. Before lunch, your account is down 8%. Not because your directional read was terrible, but because your position was three times too large for the volatility you were actually trading.

This is the single most common way traders blow up in Chinese commodity futures. Not bad analysis. Not bad luck. Sizing. And glass (FG) and methanol (MA) on the Zhengzhou Commodity Exchange are two of the fastest teachers you'll ever meet. Let's break down why, with the actual contract math, and then build a sizing framework that keeps you alive long enough for your edge to show up.

The Contract Math That Traps You

The trap starts with how small these contracts look. Here are the core specs:

Compare that to iron ore on the Dalian Commodity Exchange, where one lot is 100 tons and the notional is routinely several times larger. Glass and methanol look like the "affordable" contracts. So traders buy ten lots of glass the way they'd buy two lots of iron ore, and feel equally sized.

They are not. Notional per lot is not risk per lot. Risk is stop distance × contract multiplier × number of lots. And because these contracts are small, the temptation is to multiply the lots instead of respecting the multiplier. A 20-tick adverse move in glass costs you 400 RMB per lot — trivial with two lots, account-threatening with twenty.

Run the numbers on a realistic account

Take a $10,000 account and a standard 1% risk rule. That's $100 of risk per trade. Now say your glass setup needs a stop 3% away from entry — a fairly normal distance for a contract that regularly moves 2–3% in a session. Three percent of a ~1,500 yuan/ton price is roughly 45 yuan/ton, which is 900 RMB per lot, or around $125. One lot already slightly exceeds your risk budget.

Most traders in this situation don't take one lot. They take five, or ten, "because the contract is small." Ten lots means roughly $1,250 of risk on a $100 budget. Two or three consecutive losses — entirely normal in a choppy, policy-driven market — and you're staring at a 25–35% drawdown. That's not a bad streak. That's arithmetic.

Why Volatility Here Isn't Like Volatility Back Home

If you're coming from trading rebar, iron ore, or Western index futures, the volatility profile of glass and methanol deserves respect for a specific reason: both are downstream of Chinese industrial policy, and policy moves in steps, not gradients.

• Glass is hostage to the property completion cycle and to environmental production restrictions on float-glass capacity. When property policy shifts or energy-intensive production gets curtailed, glass can trend hard for weeks.
• Methanol is chained to coal. Coal is the feedstock and the energy cost, which means methanol inherits every shock in the coal market. During the 2021 energy crunch, when thermal coal futures on Zhengzhou rallied dramatically through the autumn before the NDRC stepped in and prices reversed violently, methanol swung alongside it. Traders who were comfortably sized in August were overleveraged by October — same position, completely different risk.

This is the core lesson: your position size is a constant, but volatility is not. A size that was conservative at 15% annualized volatility becomes reckless when volatility doubles. Glass and methanol can double their volatility regime in a matter of weeks, and they do it more often than most contracts you're used to.

Limit Moves: The Feature That Makes Oversizing Lethal

Here's the part that catches even experienced traders. Chinese exchanges apply daily price limits — commonly in the range of 4–8% depending on the product and conditions — and exchanges can widen limits and raise margin requirements when markets get disorderly, as they did across the coal complex in late 2021.

What does that mean for you in practice?

• Your stop-loss is not a guarantee. It's a market order waiting to happen. If glass opens locked at limit-down after a weekend policy announcement, your stop fills at whatever price the queue allows — possibly several percent worse than your planned exit.
• Consecutive limit moves are rare but real in policy-driven markets. One locked session can turn a planned 3% loss into a 7–10% loss per lot.
• When the exchange raises margins mid-crisis, oversized positions face forced liquidation at the worst possible moment, converting a paper drawdown into a realized one.

This is why the 2020 oil crash — where WTI went negative in April 2020 — is the mental model to keep in mind, even though it happened in a different market. Tail moves in commodity markets don't politely stop at your stop-loss. The only defense is size small enough that a gap against you is painful, not fatal. If ten lots of methanol gapping 6% against you would take 15% of your account, you don't have a stop-loss strategy. You have a hope strategy.

The Correlation Trap: Glass and Methanol Are Cousins

A subtler drawdown killer: traders size each position correctly in isolation, then stack glass and methanol together and think they're diversified. They are not. Both sit on the coal-and-construction side of the Chinese industrial economy. Energy policy, environmental inspections, and property demand hit them in the same direction, often in the same week.

If you run 1% risk on glass and 1% on methanol, your real exposure in a policy shock is closer to 1.5–2% correlated risk, not 2% independent risk. During a broad commodity unwind — think of the sharp reversals across Chinese commodity markets after the 2021 coal intervention — correlated positions draw down together, and your equity curve feels a single 4% hit, not two separate 2% hits.

Practical rule: treat glass, methanol, and other coal-chain products (thermal coal, PVC, even soda ash as a glass-chain cousin) as one risk bucket. Cap total bucket risk, not just per-trade risk.

A Sizing Framework You Can Actually Run

Here's the framework I'd hand any trader moving into Chinese commodity futures. It's not clever. It's designed to be boring, because boring sizing is what survives.

Rule 1: Risk per trade is fixed in currency, not in lots

Decide your risk per trade first — 0.5% to 1% of account equity for anything in the volatile industrial complex. Lots are the output of the calculation, never the input:

Lots = (Account × Risk%) ÷ (Stop distance in yuan/ton × Contract multiplier × USD/CNY rate)

If the formula gives you 0.7 lots, you trade one lot with a wider stop or you skip the trade. You never round up to "make it worth it."

Rule 2: Cap notional, not just risk

Keep total open notional at no more than roughly 2–3× account equity across all positions. With 10–12% exchange margins and broker add-ons, this keeps your effective leverage in a range where a limit-move gap is survivable. Ten lots of glass at ~30,000 RMB notional each is ~300,000 RMB — around $42,000 of exposure. On a $10,000 account, that's already over the cap before you've added a second position.

Rule 3: Scale stops to ATR, not to round numbers

"40 ticks because that's a nice number" is how stops end up inside the noise. Use a multiple of recent average true range — commonly 1.5–2× daily ATR for swing holds — and let that distance drive the lot count. When ATR doubles, your size halves automatically. This is the mechanical fix for the volatility-regime problem described above.

Rule 4: Set a drawdown circuit breaker

Define it before you trade: at a 5% account drawdown, cut position size in half. At 8–10%, stop trading for the week and review. This isn't weakness — it's acknowledging that drawdowns cluster, and that your sizing was calibrated for a market that has temporarily stopped cooperating.

Rule 5: One risk bucket, one budget

Coal-chain products share a bucket. Give the bucket a total risk cap — say 2% — and split it among positions. Glass plus methanol plus thermal coal is not three trades. It's one trade with three tickets.

Putting It Together: A Worked Example

Let's close the loop. $10,000 account, 1% risk = $100. Methanol at roughly 2,500 yuan/ton, daily ATR implying a 2.5% stop (~62 yuan/ton). Per-lot risk: 62 × 10 = 620 RMB ≈ $87. Formula says 1.1 lots → trade one lot. Feels small? Good. Now suppose volatility doubles after a policy headline and ATR implies a 5% stop. Per-lot risk doubles to ~$175 — more than your budget. The framework forces you to stand aside or wait for a tighter setup. That's the system protecting you precisely when your instincts are screaming to press.

Contrast that with the trader who holds ten lots through the same event. Even without a gap, a 5% adverse move costs them roughly $1,750 — 17.5% of the account in one trade. Two of those, and the drawdown conversation becomes a survival conversation.

Test the Discipline Before the Market Tests You

None of this is theoretical, but it is only real once it's been executed against live prices — the gaps, the limit moves, the margin changes. The honest way to find out whether your sizing rules hold up is to run them against real Chinese market data in a structured evaluation, where drawdown limits are enforced rather than merely intended. That's exactly what we built at XS Select: a China futures evaluation on real data, starting from $29, where glass, methanol, rebar, and iron ore will happily show you whether your position sizing is as disciplined as you think it is. Bring your rules. The contracts will do the rest of the teaching.

Size for the move you can't see coming. In Chinese commodity futures, it arrives more often than you expect — and faster.