Contents
The honest economics of gold scalping
Gold scalping is where most retail accounts are destroyed, and the reason is arithmetic rather than skill. A typical scalp aims for 5 to 15 points on XAUUSD. With a spread of 15 to 25 points on a standard account and a commission of around $3.50 per lot round turn, a $100,000 account trading 1 lot pays roughly $35 in costs per trade. If the gross profit per trade is 10 points, which is $10 per lot, you are losing money before you begin.
This is not a reason to avoid scalping. It is a reason to understand that the edge has to be large enough to cover the costs with room to spare, and that position size and lot specification matter enormously. A $500 pip on a micro account has a proportionally smaller spread in dollar terms, but the same percentage cost.
The practical conclusion is that scalping gold requires either a very large account relative to lot size, or a very tight spread from an ECN-style broker with a low commission. The relationship between spread and account size is covered in lowest spread gold brokers and fastest execution for gold.
When to scalp and when to stand aside
The only window worth scalping gold is the London-NY overlap, from 18:00 to 21:30 IST. The volume there is high enough that a genuine imbalance produces a real move rather than noise, and the spread is tight enough that the cost per trade is a manageable fraction of the target.
The London open at 12:30 IST is acceptable for a single early trade, but the first fifteen minutes should be observation. The Asian session is not scalpable at retail sizes, and the last hour before the New York close is where the market is most likely to reverse against a scalp's assumption of continuation.
Around scheduled US data the behaviour changes completely. The release spike is larger than any scalp target, and the spread can widen simultaneously. Stand aside for at least fifteen minutes after a major release, then reassess. We cover the specifics in trading gold through NFP.
A specific, testable setup
Here is one scalping setup stated precisely, so it can be tested rather than merely followed. Mark the 15-minute range at the start of the London-NY overlap. Wait for price to approach either edge of that range and then leave it. On the retest into the range edge, look for a diagonal imbalance on the 1-minute chart in the direction of the break, with a stacked imbalance at the retest level.
Enter on a 1-minute close beyond the retest level, in the direction of the break. Stop below the retest level for a long, or above it for a short. Target the midpoint of the range you broke, or the next significant volume node if it is closer.
That is the whole rule. It is unremarkable, which is a good sign, because the profitable strategies are the ones you can state in one paragraph. The value of the paragraph is that you can now test it, and testing is what separates this from another idea you will abandon in a fortnight. Out-of-sample testing covers the validation, and order flow confirmation covers the entry refinement.
Execution rules that protect the account
First, a hard daily loss limit, in currency terms, and you stop for the day when you hit it. Not “stop for the day unless I feel better”. A fixed rule, decided in advance, removes the decision from the moment you are least able to make it.
Second, a maximum number of attempts per session. Three losing attempts in the London-NY overlap means the conditions are wrong for your setup, not that the fourth will work. Risk management fundamentals covers this discipline.
Third, no trading in the first ten minutes after a data release, and no trading during the rollover. Fourth, position size computed from the stop, always, using the formula in the position sizing guide. Never size by conviction.
Fifth, and this is the one that separates survivors from the rest: when you take a trade that does not behave as expected within a few minutes, exit it. The premise is invalidated, and a scalp premise is time-sensitive. Holding a failing scalp because you are down is how a 1 percent risk becomes a 15 percent loss.
The mistakes specific to gold scalpers
Revenge scalping after a loss is the dominant failure mode. The psychological logic is that the market owes you a win, which is exactly the belief that produces the second and third loss.
Overtrading the good session is the second. The London-NY overlap produces a genuine opportunity, a trader takes it, and then the trader takes two more because the first one worked. Overtrading a good setup is a reliable way to give back a good day.
Ignoring the spread is the third, and we covered it above. The fourth is trading during news without a rule for it. The fifth is using a wide stop on a scalp, which turns a small loss into a large one and quietly converts a scalping account into a position-trading account with scalping costs. Stop loss placement covers the correct approach.