How to trade gold futures: GC vs MGC, tick value and a tested rule

Contents

1. How do you trade gold futures with a tested system?
2. What are gold futures, and how do GC and MGC differ?
3. How do you buy a gold futures contract?
4. How much is one tick worth, and how do you size risk per trade?
5. What does a trend rule look like when it is actually tested?
6. How do rollover and trading sessions change a gold backtest?
7. Frequently asked questions
8. References

How do you trade gold futures with a tested system?

How to trade gold futures with a tested system: pick GC or MGC by account size, write one entry and one exit rule, backtest it with commission and slippage, validate it on data the search never used, size each trade from tick value and stop distance, then trade one contract and roll on schedule.

Most new gold traders do these steps in the wrong order. They open the account first, buy a contract second, and think about a rule last, usually after the first loss. The order above puts the test before the first order, which costs nothing but time and tells you whether the rule deserves money at all. A trend rule is the natural first candidate for gold, because it is the simplest of the five types of futures algo trading strategies to write down and to check. Moskowitz, Ooi and Pedersen (2012) found that past returns over several months carried information about the next month across a broad set of futures, including metals, which is a reason to test a trend rule rather than a reason to trade one untested.

What are gold futures, and how do GC and MGC differ?

Gold futures are exchange-traded contracts to buy or sell a fixed amount of gold at a set date, and GC and MGC follow the same gold price while differing only in size, tick value and margin. A gold futures contract is an agreement, standardized by the exchange, in which the buyer agrees to take and the seller agrees to deliver a fixed quantity of gold in a stated month at the price agreed today. Almost all traders close or roll the position before delivery, so in practice the contract is a way to hold a position in the gold price with a deposit, called margin, instead of the full value.

The CME gold futures contract size is 100 troy ounces for GC, and 10 troy ounces for MGC, the micro contract. A troy ounce is the weight unit used for precious metals, about 31.1 grams. Both contracts quote in US dollars per troy ounce and both move in steps of 0.10 dollars per ounce, called a tick. Because GC is ten times larger, one tick is $10 in GC and $1 in MGC. You can check tick size, tick value and point value for either contract with the free futures contract calculator; exchanges can change contract terms, so the CME contract page is the final word.

Feature	GC (gold futures)	MGC (micro gold futures)
Contract size	100 troy ounces	10 troy ounces
Tick size	0.10 dollars per ounce	0.10 dollars per ounce
Tick value	$10	$1
A $10 per ounce move in gold	one thousand dollars per contract	$100 per contract
Margin	Set by the exchange, about ten times MGC	Set by the exchange, about a tenth of GC
Who it fits	Larger accounts, one contract at a time	Smaller accounts, or anyone who wants finer sizing steps

The practical reading: MGC is not a different market or a weaker signal. It is the same price with a smaller step, which matters when you size a position in the fourth section. The MGC tick value of $1 per tick on the 10-ounce contract is the number that makes most of the arithmetic below easy.

How do you buy a gold futures contract?

Buying a gold futures contract is a fixed sequence: open a futures account with access to CME, post the margin, choose the contract month with the most volume, place a limit or market order for one contract, and roll before the position enters the delivery period. There is no shortcut in the list, and the two steps traders most often skip, margin and the roll date, are the two that produce surprises.

1. Open a futures account with a broker that offers CME metals. Gold futures trade on COMEX, part of CME Group, so the broker needs access to that exchange. If you are used to Taiwan cash day trading rules and costs, the first difference you meet is here: a futures buyer does not pay for 100 ounces of gold, only a deposit.
2. Post the initial margin. Initial margin is the deposit the exchange requires to open one contract; the broker may ask for more. The exchange changes it as gold becomes more or less volatile, so read the current figure from your broker rather than from an article.
3. Pick the active contract month. Gold lists many months, but volume concentrates in one at a time. Look at the volume column in your platform and trade the month with the most contracts traded that day.
4. Send the order. A limit order is an instruction to buy at your price or better, and may not fill. A market order is an instruction to buy now at the best available price, and fills at whatever the market offers. For a tested system, use the order type the backtest assumed, so that live fills can be compared with the test.
5. Hold the position under maintenance margin. Maintenance margin is the minimum account value the exchange requires while the position is open; if the account falls below it, the broker asks for funds or closes the position.
6. Roll before first notice day. First notice day is the first day on which a holder of a long position can be told to take delivery. Close the expiring month and open the next active month before that date. Your broker publishes the date for each month.

How much is one tick worth, and how do you size risk per trade?

Risk per trade is a multiplication of three things, and the one that changes between GC and MGC is tick value. One tick is $10 in GC and $1 in MGC, so with MGC you can hold the same dollar risk with a wider stop or a smaller account. Think of tick value as the size of the coin the market pays or takes on every price step; GC uses a coin ten times heavier.

The only formula you need:

risk per trade = stop distance in ticks × tick value × number of contracts

Here stop distance in ticks is how far your exit sits from your entry, counted in 0.10 dollar steps; tick value is $10 for GC or $1 for MGC; and number of contracts is how many you hold. Decide the dollar risk first, decide the stop distance from the rule or from volatility second, and let the formula tell you the contract count. The free position size calculator does this arithmetic for any account risk and stop distance.

The same tick value belongs in the backtest. Commission and slippage are paid per contract and per side, so a test of an MGC rule with GC costs, or the reverse, gives an average trade that does not match what the account will see.

What does a trend rule look like when it is actually tested?

A tested rule shows its return next to its drawdown, trade count and costs, and it says how many other candidates were tried before this one was kept. A headline profit on its own tells you almost nothing. The number of trades says whether the result rests on evidence, the drawdown says what you would have lived through, the costs say whether the average trade is real, and the count of candidates says how much the result was flattered by picking the best after the fact.

The run below is a stand-in, not a gold result: it tests a Linear Regression trend rule on NQ daily bars, because that is the run we have, and the subject of this article is gold. A Linear Regression rule is a trend rule that fits a straight line through recent closing prices and reads the slope of that line as the trend direction. The run used that rule as its only entry and exit, with no stop, no target and no filters, on NQ daily bars from 2019-01-02 to 2025-06-30, long and short, one contract, with $20 commission and $10 slippage charged on every trade. It is the one candidate AlphaPrime (Valynox) kept out of 1,024 it tried, chosen for its return against its drawdown.

Read it the way you would read any candidate. The max drawdown of $111,615 is most of the net profit of $150,865, so the equity curve spent long stretches giving back what it had made. Time in market of 86.09% says the rule was almost always positioned, which is what a trend rule with no stop does: it stays until the slope turns. The 127 trades over six and a half years are enough to read the average trade of $1,187.91 against the $30 of costs per trade, but not enough to be sure of a small edge. None of this says the rule earned that money in a way a reader can expect, on NQ or on gold. Picking the kept candidate after the fact flatters its numbers; Bailey, Borwein, López de Prado and Zhu (2014) show that the best backtest out of many tries is inflated even when none of the rules has any edge, and their later paper (2017) gives a way to estimate how likely the chosen one is to be overfit.

What turns a candidate into a rule you would trade on gold is the step that follows: run the same rule, unchanged, on gold bars the search never touched, in walk-forward windows, and judge it against thresholds written before you look. Pardo (2008) describes this walk-forward design as the standard way to find out whether an optimized rule holds up outside the data that chose it. When a search in AlphaPrime keeps a candidate like this one, its strategy validation runs the walk-forward and the other checks against the thresholds you set beforehand and stops at the first check the candidate fails, so the unseen data does the judging rather than the headline profit.

How do rollover and trading sessions change a gold backtest?

Gold trades almost around the clock and changes its active month every two months, so a gold backtest must use a continuous series built with a stated roll rule and a session setting that matches when you will actually place orders. Get either one wrong and the test is of a market you cannot trade.

Rollover first. The active gold months are the even months, February, April, June, August, October and December, so a position held for a year rolls about six times. Each expiring month trades at a slightly different price from the next month, and when a data vendor stitches the months into one series, that difference appears as a price gap on the roll date. A rule that reads the gap as a real move will open or close trades that never existed. A back-adjusted continuous series is a stitched series in which earlier prices are shifted by the roll gap so the join is smooth, and it is what the backtest should see. The roll rule, for example "roll five trading days before first notice day" or "roll when the next month's volume passes the current month's", must be written down and must be the one you will follow live.

Sessions second. Gold futures trade on CME Globex for close to 23 hours a day, from Sunday evening to Friday afternoon US Eastern time, with a one-hour break each day; exchanges can change these hours, so the CME contract page is the reference, and the free futures market hours by exchange and session shows them in your own time zone. The question for the backtest is which part of that day your orders will be in the market. A trader in Asia who places orders in the Asian morning is trading the quietest part of the gold day; a rule tested on full-session bars but traded only in that window will not behave the way the test did. Set the session in the backtest to the hours you will trade, or test on full-session bars and trade full-session. The same two decisions, roll rule and session window, come up on every futures contract with its own calendar; the FTSE China A50 futures backtest setup on SGX walks through them for a contract with a different exchange and a different roll schedule, and the reasoning carries over to gold unchanged.

Before any gold rule reaches a live order, run two checks on your own gold bars: count the roll dates in the series and confirm each one shows no gap in the back-adjusted close, and confirm that the session hours in the backtest settings are the hours in which you will actually be awake and placing orders.

Frequently asked questions

What are gold futures in simple terms?

Gold futures are standardized contracts traded on an exchange to buy or sell a fixed amount of gold at a set date at a price agreed today. The full-size CME contract, GC, covers 100 troy ounces; the micro contract, MGC, covers 10 troy ounces. Almost all traders close or roll before delivery, so the contract works as a way to hold a position in the gold price with a margin deposit instead of the full value.

Is MGC a different market from GC?

No. MGC follows the same gold price and uses the same 0.10 dollar tick size; it is one tenth the size of GC, so each tick is worth $1 instead of $10 and the margin is about a tenth. A rule tested on gold applies to either contract, as long as the backtest uses the tick value and the per-contract costs of the contract you will actually trade.

How do you buy a gold futures contract?

Open a futures account with a broker that offers CME metals, post the initial margin, pick the contract month with the highest volume, and send a limit or market order for one contract. Once filled, keep the account above maintenance margin and roll to the next active month before first notice day. Your broker publishes margin figures and roll dates for each contract month.

Can a rule tested on NQ be trusted on gold?

Not without a gold test. A rule such as the Linear Regression trend rule in this article can be written once and run on any market, but a result on NQ is evidence about NQ. Run the same rule, unchanged, on gold bars the search never used, in walk-forward windows, and judge it against thresholds you wrote before looking.

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.
• Bailey, D. H., Borwein, J., López de Prado, M., & Zhu, Q. J. (2017). The Probability of Backtest Overfitting. Journal of Computational Finance, 20(4), 39–69.
• Moskowitz, T. J., Ooi, Y. H., & Pedersen, L. H. (2012). Time Series Momentum. Journal of Financial Economics, 104(2), 228–250.
• Pardo, R. (2008). The Evaluation and Optimization of Trading Strategies (2nd ed.). John Wiley & Sons.