How to trade gold (XAU/USD): a step-by-step guide

What gold is and how a gold CFD actually works

Gold is a physical metal traded globally as XAU/USD, the price of one troy ounce in US dollars. A contract for difference on gold gives you the price movement without owning metal: you and the broker exchange the difference between your entry and exit. Nothing is delivered, and your profit or loss is settled in cash, converted to rand on a ZAR account.

Because a CFD is a leveraged derivative, your exposure is larger than the cash you commit. That cuts both ways. A move in your favour multiplies against your margin, and a move against you can wipe out more than the deposit behind the position. Understanding that asymmetry is the whole reason the sizing rules later in this guide exist.

Lots and contract size: what one gold lot means

One standard lot of gold is 100 troy ounces, so a one-lot position moves $100 for every $1 move in the XAU/USD price. The pip on this instrument is 0.01, meaning a one-pip move on a standard lot is worth $1. Most retail accounts trade fractions of a lot, so a 0.10 lot controls 10 ounces and each pip is worth $0.10.

That contract size is the bridge between the chart and your money. Before you place any order, convert the distance in dollars into rand at your position size: ounces times the price move, then into ZAR. At a reference price near 4275.0, a 0.10-lot position is a 10-ounce exposure, so a $10 adverse move is roughly $100 before conversion. Knowing this number in advance is what separates a trade from a gamble.

Leverage and margin without the fantasy

Leverage is a cap, not a target. In South Africa, FxPro offers up to 1:200 for retail clients and up to 1:500 for eligible or professional clients depending on the instrument. The ratio only sets how much margin the broker holds aside; it does not change your profit or loss per pip, and it should never be read as permission to take a bigger position.

Margin is the cash reserved while the trade is open. At 1:200, a 0.10-lot gold position needs about $85.50 of margin, which is a modest sum against the exposure it controls. The gap between that margin and the full contract value is exactly where the risk sits. Treat free margin as a buffer for adverse movement, not as spare capital to deploy on another idea.

Sizing a position to a fixed risk

Fix your risk in rand before you look at the chart, then let the stop distance decide the lot size. If you are willing to lose, say, a set amount of your account on one trade, divide that rand figure by the rand value of your stop distance per lot. The answer is your maximum size. This order of operations stops the position from quietly growing every time you feel more confident.

The stop distance is the input you control, not the lot size. A wider stop on a volatile gold session means a smaller position, and that is the correct outcome, not a compromise. Decide the rand risk, measure the stop in dollars from your entry, convert to rand, and size down to fit. If the resulting lot is tiny, the trade is simply too risky for your account at that moment.

The real cost of a gold trade: spread and overnight swap

Every gold trade carries a spread and, if held past the daily rollover, an overnight swap. The spread is the difference between the bid and the ask and is charged the moment you open; it depends on the broker, the session and market liquidity, and it widens around major data releases and the daily rollover. You will not see a fixed figure here because it is not fixed.

The swap is the financing adjustment applied to positions held overnight, and on gold it can be positive or negative depending on direction and prevailing rates. Both costs are real and recurring, so a strategy that needs many small trades or long holds must clear them before it makes anything. Build the cost into your plan rather than discovering it in the statement.

Placing the stop and managing the trade

Place the stop-loss at the same time as the entry order, never afterwards. Set it at the level where your reason for the trade is proven wrong, then size the position so that level equals your fixed rand risk. A stop chosen for market structure and a size chosen to fit it is the only combination that survives a losing run without forcing you to widen anything.

Once the trade is live, management means following the plan you wrote, not improvising. Moving a stop further away to avoid a loss is the single most expensive habit on this instrument. You may move a stop to break-even or trail it in your favour under rules set in advance, but never in the opposite direction. If the setup invalidates, exit; do not negotiate.

Common beginner mistakes on gold specifically

The most common gold mistake is treating a fast, volatile instrument as a safe one because it is a familiar metal. Gold can travel a wide dollar range within a session, so a position sized for a quiet currency pair will be badly oversized here. Beginners also over-leverage because the margin looks small, forgetting that margin is not the same as risk.

The second cluster of errors is cost blindness and stop tampering. Traders hold leveraged positions for days without accounting for swap, or scalp repeatedly without counting the spread, then wonder why the account drifts down. Others widen stops or remove them altogether during a drawdown. Each of these turns a manageable loss into an uncontrolled one, and on gold the moves are large enough to make that fatal.

A realistic first gold trade, start to finish

A realistic first trade is small, planned and boring. You decide in advance how much of your account you will risk, note the reference price near 4275.0, and pick an entry with a stop a defined distance away. You convert that stop distance into rand at a 0.10-lot size, confirm the loss fits your limit, and place entry and stop together. No averaging, no adding.

You then check the costs you will pay, including the spread on entry and the swap if you hold overnight, and accept them as part of the plan. If the trade goes against you, the stop closes it for the amount you chose. If it goes your way, you follow your pre-set exit rules. Afterwards you review the trade against the plan, not the outcome, and adjust only the process.

Your first week on the demo account: test execution, not prediction

Open the demo on the same platform you intend to use live and treat the first week as an execution audit, not a forecasting contest. In MT4, MT5, cTrader or FxPro Edge, place and cancel pending orders, close positions partially, and attach stop-loss and take-profit levels to a 0.10-lot XAU/USD position so you see how the platform handles each action. Log the time between clicking and the fill appearing, and note whether the fill matches the quote shown. You are testing the plumbing, not your view on gold.

Test the demo with the account size and leverage cap you will actually use, because a demo loaded with fictional capital teaches nothing about margin. Set the platform to reflect up to 1:200 for retail and note how much margin a 0.10-lot gold position consumes at that cap, which is about $85.50. Then deliberately trigger a margin warning by opening additional exposure, so you recognise the platform's alert before it ever appears on a funded account. The lesson is how the warning arrives and what the platform does next, not the profit or loss shown.

Spend the first week testing behaviour around the sessions when gold actually moves, and write down the clock times in South African time when spreads and fills change character. Note how the platform behaves during the rollover period and around major scheduled data, without assuming any particular spread or swap figure. Test your own reaction time too: can you place a stop within seconds of entry under pressure? A demo week is only useful if you finish it with a written list of platform quirks and personal habits that need fixing before rand is at risk.

What to write in a gold trade journal so it is worth rereading

Write down the reason you entered before you enter, in one sentence that names the price level and the condition that triggered it. A useful gold journal entry captures the date and time in South African time, the direction, the lot size, the entry price, the stop level and the target level, plus the planned risk in rand. Record the margin required at your leverage cap and the account balance at the time. If you cannot state the reason in a sentence before clicking, the trade is not ready to be taken.

Record the exit separately from the entry, including whether you closed at the stop, at the target, or manually, and what you were feeling when you did it. The emotional note is the part most traders skip and the part that explains repeated errors. Add the platform used, since behaviour can differ between MT4, MT5, cTrader and FxPro Edge, and note any slippage between the level you requested and the level you received. Do not record a spread or swap number unless you have taken it from your own statement.

Review the journal weekly and count patterns rather than judging individual trades: how many entries were taken without a written reason, how many stops were moved, how many positions were larger than the plan allowed. Compare the rand risk you intended with the rand risk you actually carried, and note any trade where the two diverged. Keep the journal in a format you will actually update, whether a spreadsheet or a notebook, and store it with your statements so the record of what you did survives the memory of what you meant to do.

Turning position sizing into a reflex you do not have to calculate twice

Make sizing a habit by fixing your rand risk per trade before the week begins and refusing to open the platform until that number is written down. Decide the amount in rand you are willing to lose on one gold trade, then let the stop distance determine the lot size rather than choosing a lot size first. Because one standard lot is 100 oz and one pip is 0.01, the arithmetic is mechanical once the stop is set. The habit is the order of operations: risk amount, then stop, then size, every time.

Anchor the habit to the margin figure you already know, so sizing becomes a check rather than a fresh calculation. At up to 1:200 for retail, a 0.10-lot gold position needs about $85.50 margin, which gives you a familiar reference point for how much room a small position takes. Before increasing size, confirm the new margin requirement against your available balance and against the rand risk you have set. If either check fails, the trade does not happen at that size, regardless of how strong the setup looks.

Reinforce the habit by reviewing size after the trade, not during it, and by treating any deviation as an error even when the trade wins. Keep a simple rule that no single gold trade may risk more than your written rand amount, and that the stop distance is set before the size is chosen. Practise the sequence on the demo until it takes under a minute, then repeat it on every live entry without exception. A habit only counts when it holds on the day you are most tempted to skip it.

The three costliest beginner errors on gold and the rule that blocks each

The most expensive beginner error is trading gold without a stop, or widening the stop after entry, because a single adverse move can erase many careful trades. The rule that prevents it is simple: no position is opened without a stop order placed at the same time, and that stop level is never moved further from the entry. Write the stop level in your journal before you click, and treat any urge to widen it as a signal to close the trade instead. The cost of a stop is known; the cost of removing one is not.

The second costly error is sizing by lot rather than by rand risk, which turns a normal stop-out into a balance-threatening loss. The rule that prevents it is to state the rand amount you will risk before you look at the chart, then let the stop distance set the lot size, checking the margin requirement at your leverage cap as you go. At up to 1:200 for retail, a 0.10-lot gold position needs about $85.50 margin, so a larger size demands a larger margin check. If the two do not fit, the size is wrong.

The third costly error is trading through the rollover and major data releases without knowing what your position will cost or how the platform will behave. The rule that prevents it is to check the platform clock and your open exposure before every scheduled event, and to reduce or close positions you cannot justify holding through it. Keep the check in your journal as a line item, and note the outcome afterwards without inventing figures. FxPro is licensed by the FCA (UK) and CySEC, and an FxPro entity holds an FSCA licence in South Africa — check which entity your own account is opened with.

Building a weekly review that catches drift before it costs money

A weekly review catches drift before it becomes a loss, so set a fixed time each week to read your journal and compare what you planned with what you did. Count the trades where the stop was placed at entry, the trades where size matched the written rand risk, and the trades where you entered without a reason. Do not judge the week by profit; judge it by how many of your own rules survived contact with the market. The review is a maintenance check on your process, not a scorecard on your predictions.

Review the platform side as well as the trading side, because execution problems hide in the details you skip when you are busy. Check whether fills matched the levels you requested, whether margin warnings appeared earlier than expected, and whether any position was held through a period you had intended to avoid. Note which platform you used, since MT4, MT5, cTrader and FxPro Edge each behave differently under stress. Keep the review to a single page so it stays a habit rather than a project.

End each review by writing one change for the coming week and nothing more, because a list of ten changes is a list you will ignore. The change should be a rule you can check, such as placing the stop before the entry or refusing any trade without a written reason. Carry the previous week's change forward until it holds without effort, then add the next. Over a few months this turns risk control into a routine, and the routine is what protects your rand when a gold trade goes wrong.

Your First Demo Week: Test Execution, Not Prediction

The first week on a demo account should test execution quality, not your ability to predict gold's direction. Open a demo with the same FxPro entity that will serve your live account, then place ten small XAU/USD orders at different times of day, including the Asian, London and New York sessions. Record the time from click to fill, the price you requested versus the price you got, and whether the platform accepted your stop and limit orders without rejection. On a 0.10-lot position, a one-pip difference is $0.10, so you can measure slippage in rand terms without risking capital.

Test how the platform behaves around news releases and rollover, because that is when execution gaps and spread widening appear. Place a pending order five minutes before a major scheduled release and note whether it fills at the requested price or slips. Check the swap charge on a position held overnight by comparing your account balance before and after 00:00 server time. Do this on a demo funded with virtual dollars, and convert the swap to rand at the prevailing exchange rate so the cost feels real. The goal is to see what FxPro's MT4, MT5, cTrader or FxPro Edge actually does with your order, not to double your demo balance.

End the week by testing the funding and withdrawal path you will use in South Africa. Use a local card or bank transfer in ZAR on the demo where the platform allows a simulated deposit, and note the steps, verification prompts and any currency conversion screens. Check that you can find the FSCA licence reference for the entity your demo is opened under, and confirm whether it matches the entity named in your welcome email. Write down the exact sequence of clicks for a deposit, a withdrawal and a stop-loss modification. That written sequence becomes your live-account checklist, and it removes the hesitation that causes mistakes when real rand are involved.

The Gold Trade Journal: What to Record So It Earns Its Keep

A gold trade journal earns its keep when it records decisions, not outcomes. For every XAU/USD trade, write the date and time, the direction, the entry price, the stop price and the target price before you place the order. Add the lot size and the rand risk you accepted, calculated from the stop distance in pips and the pip value. Note the reason you took the trade in one sentence, such as a break of a level or a reaction to a news release. Do not write what you hope will happen. The journal is a record of what you knew and chose at the moment of entry, and that record is the only thing you can review honestly later.

Record the execution details that a broker statement will not show you. Write the spread at entry, the swap charged if you held overnight, and any slippage between the price you requested and the price you received. For a 0.10-lot gold position, one pip is $0.10, so a two-pip slip is $0.20; convert that to rand and add it to the cost column. Note the platform you used, because MT4, MT5, cTrader and FxPro Edge can behave differently around pending orders. If you moved your stop or closed early, write the reason in the same sentence as the action. These details turn a list of trades into a map of your own behaviour.

Close the journal entry only after the trade is finished, and record the exit price, the rand result and whether you followed your original plan. Mark the trade as followed or deviated, and write one line on what you would repeat or change. Review the journal weekly and count how many trades were deviations rather than planned entries. If deviations dominate, the problem is process, not prediction. Keep the journal in a simple spreadsheet with columns for date, instrument, lot size, entry, stop, target, exit, costs in rand and a followed/deviated flag. That structure makes the weekly review a five-minute task instead of an hour of reconstruction.