QM futures: E-mini crude oil specs, margin and how to backtest it

Contents

1. What are QM futures and how big is one contract?
2. How does QM differ from CL and MCL?
3. What is the QM margin, and is QM right for a small account?
4. What are the QM trading hours?
5. How do liquidity, costs and rolls change a QM backtest?
6. How do you set up a QM backtest step by step?
7. Frequently asked questions
8. References

What are QM futures and how big is one contract?

QM futures are the E-mini crude oil contract on NYMEX, part of CME Group: 500 barrels of WTI crude oil, half the size of the standard CL contract. The minimum tick is $0.025 per barrel, worth $12.50, so a $1 move in the oil price is worth $500 per contract. QM is financially settled against the CL price.

WTI (West Texas Intermediate) is the US benchmark grade of crude oil, and CL, QM and MCL all follow its price. A tick is the smallest price step the exchange allows an order to use. On QM the step is $0.025 per barrel, and one contract covers 500 barrels, so one tick moves the position by 500 × $0.025 = $12.50. A full $1 move in the oil price is 40 ticks, or 500 × $1 = $500. The free futures contract calculator turns any entry, exit and contract count into dollars so you can check these numbers before you trust a backtest report.

Financial settlement means no barrels change hands at expiry: an open position is closed in cash at a final settlement price, and for QM that price is taken from the corresponding CL contract. This is why mini crude oil futures appeal to traders who want crude exposure without the delivery rules of the full contract. Exchanges can change contract values, so read the QM contract page on the exchange's website before you set a backtest.

How does QM differ from CL and MCL?

QM, CL and MCL follow the same oil price, and the five things that differ are barrels per contract, tick size, dollars per tick, settlement type and liquidity; those five decide which contract your strategy can be tested and traded on, not the price chart. The table puts them side by side.

Feature	CL (full)	QM (E-mini)	MCL (Micro)
Contract size	A thousand barrels	500 barrels	100 barrels
Minimum tick	$0.01 per barrel	$0.025 per barrel	$0.01 per barrel
Tick value	$10.00	$12.50	$1.00
Dollars per $1 move	A thousand dollars	$500	$100
Settlement type	Physical delivery	Financial (cash)	Financial (cash)
Typical liquidity	Highest	Lowest of the three	Middle

Two rows need a closer look. First, the tick. CL and MCL can trade at $70.01, $70.02 and so on; QM only trades at $70.000, $70.025, $70.050. A stop or target written in cents from a CL study lands between QM prices and has to be rounded, and a QM backtest run on CL data uses prices QM could never fill at. Second, the settlement type. CL is physically delivered, so a speculative account must exit or roll before the broker's deadline near expiry; QM and MCL settle in cash, so the roll is a question of liquidity, not delivery. For strategy research this means a QM test needs QM data, or at least a model that snaps every fill to the $0.025 grid.

What is the QM margin, and is QM right for a small account?

QM margin sits between CL and MCL because QM is half a CL contract, but the number to size an account around is the backtest's maximum drawdown in dollars at $500 per point, not the exchange margin. Margin is the deposit the clearing house requires per open contract, and it covers the exchange's risk, not yours. The exchange raises and lowers it without a schedule, so read the current figure on the QM contract page rather than from any article, including this one.

The reason drawdown matters more than margin is leverage. Leverage is control of a large position with a small deposit, and it means the same $3 move in crude produces a very different dollar swing on each contract. A small account that can post the QM margin can still be emptied by an ordinary losing streak, because the margin says nothing about how far the equity curve falls before it recovers.

The practical order is: run the backtest on one contract, read the maximum drawdown in dollars, multiply it by a safety factor you choose in advance (many systematic traders use two or more, because a future drawdown is usually deeper than the one in the history), add the current margin, and compare the total with the account. If the total is too large for QM, the same rule can often be tested on MCL at a fifth of the size. If it is comfortable, QM gives five times the exposure of MCL with one contract and far less commission per barrel than five MCL contracts.

What are the QM trading hours?

QM trades on CME Globex in the same energy session as CL and MCL: Sunday to Friday, 5:00 p.m. to 4:00 p.m. Central Time, with a 60-minute break each day from 4:00 p.m. to 5:00 p.m. A QM backtest therefore uses the same session setting and the same Taiwan-time conversion as any crude oil backtest. The trading-hours details, and how the session looks from Taiwan, are covered in crude oil futures trading hours for CL and MCL in Taiwan time; QM follows them exactly.

Central Time shifts twice a year, Taiwan time does not, so the Taiwan clock times move by one hour between winter and summer:

Event	Central Time	Taiwan time, US standard time (winter)	Taiwan time, US daylight time (summer)
Weekly open	Sunday 5:00 p.m.	Monday 7:00 a.m.	Monday 6:00 a.m.
Daily close	4:00 p.m.	6:00 a.m.	5:00 a.m.
Daily break	4:00 p.m. to 5:00 p.m.	6:00 a.m. to 7:00 a.m.	5:00 a.m. to 6:00 a.m.
Weekly close	Friday 4:00 p.m.	Saturday 6:00 a.m.	Saturday 5:00 a.m.

The session setting decides where a daily bar begins and ends, so set the trading day to end at the 4:00 p.m. CT close rather than at midnight in any time zone. A daily bar cut at midnight Taiwan time splits the US afternoon in half and produces highs, lows and closes that no exchange session ever showed. The same 5:00 p.m. to 4:00 p.m. CT pattern with a one-hour break applies to the metals, so the session work in gold futures backtesting with specs, margin and trading hours carries over to QM with only the contract numbers changed. Holiday schedules and early closes differ by year; the free futures market hours tool shows the current session in your own time zone, and the exchange's contract page is the final word when hours change.

How do liquidity, costs and rolls change a QM backtest?

QM trades far less volume than CL and MCL, and its $12.50 tick is a bigger share of contract value than CL's $10 tick, so a QM backtest needs more slippage per trade, a fixed monthly roll rule, and a check that the average trade after costs is several times the round-trip cost. Slippage is the difference between the price a backtest assumes and the price a real order gets. On a thin book that difference is larger, and on a coarse tick grid even the smallest difference is 2.5 cents per barrel.

Look at it per barrel. A $0.05 move in crude is five CL ticks of $10, or two QM ticks of $12.50. One tick of slippage on CL costs 1.0 cent per barrel; one tick of slippage on QM costs 2.5 cents per barrel, two and a half times as much for the same oil. A strategy that was priced at one tick of slippage on CL is paying two and a half times that per barrel the moment it is traded on QM, before the thinner book adds anything.

A workable cost model for QM research, which you should adjust to your own fills once you have them:

1. Commission and exchange fees per side, from your broker's schedule, charged on entry and exit.
2. Slippage of at least one QM tick per side ($12.50 in, $12.50 out) as the base case for market and stop orders.
3. A stress run at two ticks per side ($50 round trip). A rule that only works at one tick was depending on cheap fills.
4. The average trade test: the gross average trade should be several times the round-trip cost. How to set each cost and read this test is explained in trading costs in backtests: commission, slippage and the average trade.

Fast intraday rules fail this test first, because they have the smallest move per trade and the most trades. The same pattern appears on the index futures in whether an ES emini scalping strategy survives costs in a backtest, and on QM the coarse tick makes it sharper. Among the rule families in futures algo trading strategies and how to backtest each type, the ones that hold for hours or days have the best chance of clearing QM's cost bar.

Rolls are the third adjustment. Rolling is closing the expiring contract month and opening the next one, and QM has a new contract every month, so a multi-year test contains dozens of rolls. Each roll produces a price gap between the two months that is not a market move; a continuous series that ignores it books a fake profit or loss on every open position. Fix one roll rule before the first run, either a set number of business days before expiry or the day the next month's volume passes the current month's, and apply it to every roll in the history. Because QM volume is thin, roll on the same day or earlier than you would roll CL, so the backtest does not sit in a month that real traders have already left.

How do you set up a QM backtest step by step?

Set the contract specs, session, costs and roll rule before the first run, then judge the result on completed trades and drawdown in dollars; nothing about the price chart should be looked at until those six settings are fixed. The order matters because each later setting depends on an earlier one: slippage is counted in ticks, ticks depend on the contract spec, and the drawdown you judge is only meaningful after costs and rolls are in.

1. Contract specs: 500 barrels, tick $0.025, tick value $12.50, $500 per $1 move. Check them on the exchange's contract page on the day you set them.
2. Session: Globex energy hours, trading day ending at 4:00 p.m. CT, with the hour break.
3. Commission and slippage per trade: your broker's fees plus at least one QM tick per side, with a second run at two ticks.
4. Roll rule: monthly, by volume or by a fixed number of days before expiry, applied to every roll.
5. Next-bar execution: a signal computed on a bar's close is filled on the next bar, because the close itself was not available to trade at.
6. Judge on completed trades and dollar drawdown: count the trades, read the maximum drawdown at $500 per point, and compare it with the account as in the margin section.

In AlphaPrime (Valynox) the session, commission, slippage and next-bar fill rule live in the backtesting settings and are charged inside every candidate from the first evaluation, so a QM search is scored at QM's cost rather than cleaned up afterwards; the speed of that search depends on the processor or graphics card and the length of the data.

To show what step six looks like in practice, here is a real run from AlphaPrime. It is not a QM test. It tested ES daily bars from 2019-01-02 to 2025-06-30, long and short with one contract, using an Average True Range rule as the only entry and exit, with no stop, target or filter, as a stand-in rule built from a standard indicator. The result shown is the one candidate the search kept out of the 1,024 it tried, chosen for its return against its drawdown.

Read it the way you would read a QM run. The drawdown of $57,590 is the number that sizes the account, and it is a drawdown at ES's $50 per point; on QM the same curve would be read at $500 per point of crude. The 165 trades over six and a half years are enough to compute an average trade, and the average trade of $713.64 is many times the $45 round-trip cost, which is the test from the previous section. What the numbers do not say is what the rule earns: this candidate was picked after the fact as the best of 1,024, and picking the winner after the race flatters it. Bailey, Borwein, López de Prado and Zhu (2014) show that the best backtest in a large search is inflated by the search itself, and Harvey and Liu (2015) argue that any result should be discounted for the number of tests behind it. A QM candidate that comes out of a search gets the same treatment: out-of-sample grading on data the search never touched, then the stress run at two ticks of slippage, before any contract is traded.

The next step is concrete: open the exchange's QM contract page, confirm the 500-barrel size, the $0.025 tick and the current margin, then enter one past QM trade of yours, or an imagined one, into the futures contract calculator and check that the dollar result matches 500 × the price move. If it does, your first QM backtest can start from settings you have verified yourself.

Frequently asked questions

What are QM futures contract specs?

QM is the E-mini crude oil future on NYMEX, part of CME Group. One contract is 500 barrels of WTI crude oil, the minimum tick is $0.025 per barrel worth $12.50, and a $1 move in oil is worth $500 per contract. It is financially settled, with the final price taken from the CL contract. The exchange can change these values, so confirm them on the QM contract page.

QM or CL for a small account?

Decide by the backtest's maximum drawdown in dollars, not by the margin. QM moves $500 per $1 of crude and CL moves a thousand dollars, so the same rule shows half the dollar drawdown on QM. Many small accounts start on MCL at $100 per $1, where the drawdown is a fifth of QM's, then step up once the account covers a drawdown deeper than the backtest plus the margin.

Does QM settle physically like CL?

No. QM is financially settled: an open position at expiry is closed in cash at a final settlement price based on the CL contract, and no barrels are delivered. CL is physically delivered, so brokers require speculative accounts to exit or roll it before a deadline near expiry. For a backtest the difference shows up in the roll rule, not in the price data.

Can I backtest on CL data and trade QM?

Only as a rough first look. QM prices move in $0.025 steps, so stops, targets and fills from 1-cent CL data land on prices QM cannot trade at, and QM's thinner book means more slippage. A backtest meant for QM should use QM data, or at least snap every fill to the $0.025 grid and charge at least one QM tick of slippage per side.

References

• Bailey, D. H., Borwein, J., López de Prado, M., & Zhu, Q. J. (2014). Pseudo-Mathematics and Financial Charlatanism: The Effects of Backtest Overfitting on Out-of-Sample Performance. Notices of the American Mathematical Society, 61(5), 458–471.
• Harvey, C. R., & Liu, Y. (2015). Backtesting. The Journal of Portfolio Management, 42(1), 13–28.