← Back to Blog · 2026-10-04 · 8 min read · Trading Education

It's 10:47 PM in Shanghai. You're long rebar from the day session, your stop is set, and you're about to close the laptop. Then a US economic print drops, the dollar rips, and within minutes your stop is triggered — at a price forty ticks worse than where you placed it, because the order book thinned out the moment the first wave of traders went to bed. You weren't wrong on direction. You were wrong about where you put your stop and how much you were risking while nobody was watching the tape.

Night sessions are one of the defining features of Chinese commodity futures — and one of the most misunderstood when it comes to risk. Let's fix that.

How Night Sessions Actually Work in China

Most Chinese commodity futures trade a night session starting at 21:00 China Standard Time. The exact close depends on the exchange and product:

• SHFE metals (copper, zinc, nickel, and precious metals): night session runs to 01:00, with gold and silver extending to roughly 02:30.
• SHFE construction and energy products — rebar (RB), hot-rolled coil, rubber, fuel oil: night session ends at 23:00.
• DCE products like iron ore (I) and ZCE products like methanol and PTA: also 21:00–23:00.
• Crude oil (SC) on INE: runs to around 02:30, tracking global energy hours.

After the night session closes, the market goes dark until the day session opens at 09:00. That closed window is where most stop-placement damage happens — but more on that in a moment.

The night session exists for a good reason: Chinese commodity prices are heavily influenced by overnight moves in US and European markets — the dollar index, US Treasury yields, LME metals, and crude. If you trade rebar or iron ore seriously, the night session isn't optional. It's where a large share of the real directional information gets priced in.

The Gap Problem: Your Stop Only Works When the Market Is Open

Here's the uncomfortable truth: a stop-loss order is not a guarantee. It's a conditional order that converts to a market order once your trigger price trades. It protects you only while the exchange is open.

Between 23:00 (or 01:00/02:30 for metals and crude) and 09:00 the next morning, Chinese futures markets are closed. If a shock hits during that window — a surprise policy announcement, a commodity export restriction, a geopolitical event — the day session opens with a gap. Your stop triggers at the open, not at your level. You get the first printed price, which can be far worse.

This isn't hypothetical. During the 2021 thermal coal episode, government intervention to cool the rally led to consecutive limit-down sessions on Zhengzhou thermal coal futures. Traders who were long with stops in place discovered that when a contract opens locked at its downside limit, there are no buyers — your stop order sits in the queue, unfilled, while your loss keeps growing on paper. Similar dynamics played out in energy contracts around the 2020 oil crash, when INE crude hit limit moves alongside the global collapse.

Two structural facts make this worse in China than in many Western markets:

• Price limit expansion. Most contracts have a base daily price limit (commonly in the 4–8% range depending on the product), and exchanges routinely widen the limit when volatility spikes or after consecutive limit moves. The room for a single-day adverse move can double or triple exactly when you least want it to.
• Margin hikes during stress. Exchanges and brokers raise margin requirements during volatile periods. Combined with expanded limits, this forces liquidations into thin books — amplifying the exact moves stops are supposed to protect against.

The professional mindset: a stop defines your intended loss. Position sizing defines your actual worst case. Never let the two be the same number.

Concrete Stop Rules: Structure, Volatility, and Tick Math

Enough theory. Here are the rules that hold up in practice.

Rule 1: Place stops beyond structure, plus a volatility buffer

Your stop belongs on the far side of the most recent swing high or low — not at it. Everyone can see the obvious level, and in thin night-session liquidity, stops clustered at obvious levels get swept. Add a buffer based on volatility, not vibes. A common approach: buffer = 0.5 to 1.0 × ATR(14) on your trading timeframe.

Rule 2: Know your tick value before you place anything

This is where traders get sloppy. Chinese contracts have multipliers that make small tick moves cost real money:

Example: you're long one lot of iron ore. Your structure-based stop is 12 yuan/tonne below entry, and you add a 1.5 yuan/tonne ATR buffer, so your stop distance is roughly 13.5 yuan/tonne. At 100 tonnes per lot, that's about 1,350 yuan of risk on a single lot. If your account risk budget per trade is 1% and your account is 50,000 yuan, one lot of iron ore with that stop already consumes most of it. The math tells you the answer: either a tighter, structurally valid stop, a smaller position, or no trade.

Rule 3: Never place a stop inside the noise

If your stop distance is smaller than typical night-session noise for that contract, you will get stopped out on spread and slippage alone. For rebar, a stop that's only 10–15 yuan/tonne away (1,000–1,500 yuan per lot in notional movement terms — wait, that's 100–150 yuan per lot in risk) can be legitimate on a scalp, but it needs to be justified by the contract's recent range, not by how much risk you wish you were taking.

Rule 4: Respect round numbers — on both sides

Rebar at 3,500, iron ore at 800 — these psychological levels attract order flow in both directions. Don't hide your stop at the round number. Either place it clearly beyond it (give it room to be tested) or treat the round number as your invalidation point and exit with a limit order before price gets there.

Position Sizing Comes First — Always

Here's the sequence most retail traders get backwards: they pick a position size, then hunt for a stop distance that produces an acceptable risk number. That's how you end up with stops inside the noise, getting swept nightly.

The correct order:

• Define the structural invalidation level and add your volatility buffer. That gives you the stop distance.
• Convert stop distance to yuan risk per lot using the tick value table above.
• Divide your per-trade risk budget (1% of account is a sane default; 0.5% for night-session holds) by that per-lot risk. That's your position size, rounded down.
• If the answer is zero lots, the trade doesn't exist. Move on.

This one change — sizing derived from the stop, not the stop from the size — eliminates most of the self-inflicted losses in Chinese commodity futures.

Night-Session-Specific Rules

Now layer the overnight context on top:

Cut your size, not your stop

Liquidity in the 22:00–23:00 window and the 01:00–02:30 metals window is a fraction of day-session depth. Spreads widen, and a market order can slip several ticks. If you hold through the night, halve your size relative to an equivalent day-session setup. Same stop logic, smaller exposure — because your stop will fill worse at night than it would at 10 AM.

Know the overnight calendar

US CPI, FOMC decisions, and major energy inventory reports land in the middle of, or just after, Chinese night sessions. You have two honest choices: flatten before the release, or hold with deliberately reduced size and accept that your stop may fill with meaningful slippage. What you cannot do is hold full size with a tight stop and pretend you're protected. Around the 2020 oil crash, traders learned this lesson in the most expensive way possible — stops in fast markets are executed at whatever price exists, not the price you wrote down.

Set a hard time cutoff

Decide in advance: no new night-session entries after a certain hour (many experienced traders use roughly 23:00 for the 23:00-close contracts, since the remaining liquidity is thin). And no adding to positions at night to 'average' a day-session entry. Night-session additions to a losing day trade are how small losses become account-threatening ones.

Assume the gap before it happens

Before holding anything overnight, ask: what happens at 09:00 if this opens 2% against me? With expanded limits, a 2–3% adverse open is entirely plausible in stressed conditions. If that loss would exceed your normal risk budget by several times, either reduce size or take the trade only intraday. This single question, asked honestly, would have saved many accounts during the thermal coal limit-down sequence in late 2021.

Use the night session for information, not just execution

Even if you trade rebar or iron ore primarily in the day session, watching how price behaves overnight — does it hold a level when US markets are moving, or does it bleed? — tells you whether your day-session levels are still valid. Many Chinese traders mark their levels off night-session behavior, not day-session closes.

Putting It Together: A Pre-Trade Checklist

Before any overnight hold in Chinese commodity futures, run this list:

• Stop placed beyond structure with an ATR-based buffer — not at the obvious level, not at the round number.
• Stop distance converted to yuan using the correct tick value and multiplier for the contract.
• Position size derived from the stop and a risk budget of 0.5–1% of account.
• Overnight calendar checked — no full-size exposure into major US data or known policy-risk windows.
• Gap stress test answered: what does a 2–3% adverse open cost me, and can I absorb it?
• Hard cutoff time set for entries and additions.

None of this is glamorous. That's the point. Stop placement in night sessions isn't about finding the perfect level — it's about accepting that your stop is a tool with limits, and sizing your position so that the tool's failure doesn't end your career.

Test It Before You Trust It

Rules like these sound obvious until you're the one staring at a thin order book at 11 PM. The only way to know whether your stop logic and sizing hold up is to run them against real market conditions — real tick data, real contract specs, real slippage — over a sustained stretch.

That's exactly what we built XS Select for. It's a futures evaluation platform focused on Chinese commodity markets, where you can test your system on real historical data in a structured evaluation — starting from $29. No promises of easy profits; just an honest environment to find out whether your night-session risk rules survive contact with the market. If you trade China futures seriously, that's a question worth answering before real money does it for you.