Broker and cost terms

Commission

Commission is a per-lot fee charged by the broker for executing a trade, typically on raw-spread accounts where the quoted spread is close to interbank.

Commission is usually quoted per lot per side or per lot round turn, and the two are frequently confused. A broker advertising "$3.50" may mean $7 for a complete trade. Read the specification rather than the marketing page, because the difference doubles your assumed cost.

The right way to compare a raw account against a standard one is to convert both to a single number: total cost per round turn at your typical size. A raw account with a $0.20 gold spread plus $7 per lot is not obviously cheaper or more expensive than a standard account with a $0.45 spread until you do that arithmetic.

For short-target gold strategies the total is decisive. On a $3 target at 0.10 lots the gross gain is $30, while a $0.30 spread costs $3 and $7 round-turn commission costs another $0.70 — more than 12% of the move gone before slippage. The gold profit calculator shows that share explicitly, because most calculators quietly report the gross figure alone.

ECN account

Also called: Raw spread account.

An ECN account routes orders to an electronic network of liquidity providers rather than filling them internally, charging a commission instead of a marked-up spread.

The practical differences are visible in your trade history rather than in the brochure. ECN routing produces variable spreads that tighten in liquid conditions and widen in thin ones, occasional positive slippage, and no requotes. Dealing-desk execution produces steadier spreads, no positive slippage, and requotes during fast markets.

ECN is generally the better fit for automated trading, mainly because the execution behaviour is consistent. A system tested against variable-but-honest pricing will behave similarly live; a system tested on a feed that is being quoted rather than routed can behave very differently once conditions change.

There is one specific quirk worth knowing if you are automating on such an account: many ECN brokers reject orders that arrive with a stop loss and take profit attached, requiring protection to be set as a separate modification after the fill. This is why our engine sends orders with zero protection and immediately modifies — and why a bot written for a dealing-desk broker often fails on its first ECN order.

Minimum lot size

The minimum lot size is the smallest position a broker will accept, usually 0.01 lots, and it puts a hard floor under how little you can risk per trade.

Because volume is quantised, position sizing cannot produce arbitrarily small positions. A calculation returning 0.004 lots must round to 0.01 or to zero, and every trading platform rounds up — which means your real risk exceeds the configured risk whenever the calculated size falls below the minimum.

On gold the effect is large enough to change what account sizes are viable. At 0.01 lots with a $10 stop, minimum risk is $10 per trade. Against 1.5% risk that implies an account of roughly $670 before the floor stops binding.

This is the honest answer to "what is the minimum balance for a gold bot", and it is why any single number given without a stop distance is meaningless. Tighten the stop and the required balance falls; widen it and it rises. Both levers are shown in the lot size calculator, and the question is answered in full here.

Negative balance protection

Negative balance protection is a broker policy or regulatory requirement that prevents a retail account from ending up owing money after a catastrophic move.

Without it, a gap large enough to blow through your stop and your remaining equity leaves a debt. With it, the account floors at zero and the broker absorbs the remainder. It is mandatory for retail clients in several regulated jurisdictions and offered voluntarily by many brokers elsewhere.

It is worth confirming rather than assuming, particularly on offshore entities of otherwise well-known brands. The same brand often operates several licences with materially different client protections, and the entity you are actually contracting with is stated in your account documents rather than on the homepage.

Treat it as a backstop against the extreme case, not a risk-management tool. It engages only after the account is already at zero, so it protects you from debt rather than from loss — the difference matters, and the risk disclosure spells out what it does and does not cover.

PAMM account

Also called: PAMM, Percentage allocation management module.

A PAMM account pools money from several investors into one trading account managed by a single trader, with profits and losses split by each investor's share of the pool.

It differs from copy trading in a way that matters legally as well as practically. Copy trading replicates trades onto accounts you still own; a PAMM moves your money into an account someone else operates. You hold a claim on a pool rather than a balance you control.

The practical consequences follow from that. Entry and exit usually happen only at defined rollover points rather than instantly, the manager's decisions apply to the whole pool at once, and your ability to withdraw mid-drawdown is limited by the structure. In exchange, execution is genuinely identical for every investor — there is only one account, so there is no copier delay and no slippage between provider and follower.

The counterparty question is the one to settle first. In a PAMM your funds sit inside another party's trading account under the broker's rules, so the protections that apply depend on the broker's regulator and on the specific PAMM agreement. Read both before depositing, because this is the arrangement where "the strategy performed well" and "you got your money back" are genuinely separate outcomes.

Performance fee

Also called: Profit share, Success fee.

A performance fee is the share of profit a copy trading provider takes from subscribers, usually charged only on new highs so the same gains cannot be billed twice.

Typical rates run from ten to thirty percent of profit, sometimes to fifty. The important detail is not the rate but whether it is charged against a high water mark — the previous peak equity. With one, a provider who loses 20% must recover it before earning again. Without one, they can charge you on the recovery of losses they caused.

The incentive structure deserves more scrutiny than the number. A provider paid on profits but not charged for losses holds a one-sided bet, and the rational response to a one-sided bet is more risk than the subscriber would choose. This is the structural reason copy trading skews aggressive, independent of any individual provider's intentions.

Compare the total cost honestly against the alternative. A performance fee is a recurring share of everything the account ever makes; buying software is a fixed cost that does not scale with success. On a small account the fee is cheaper. On an account that grows, the software stops being the expensive option surprisingly quickly.

Spread

The spread is the difference between the bid and ask price, and it is the cost paid on entry to every trade regardless of whether the trade wins or loses.

Every position starts underwater by the spread. Buy at the ask, and you can only sell at the bid, so the trade must move by the spread before it breaks even. This is why spread matters far more to short-target strategies than to long-horizon ones: the same cost is a rounding error on a 200-point target and a third of the edge on a 30-point one.

On XAUUSD the spread varies enormously by account type and by time of day. Raw and ECN accounts often quote gold in the $0.10 to $0.40 range with a separate commission, while standard accounts fold a much wider spread in and charge nothing extra. Neither is automatically cheaper — the spread cost calculator compares them properly.

Spread also widens on schedule: at the daily rollover, at session opens, and around high-impact news. An automated system that does not check spread before entering will take its worst trades at exactly these moments. Our engine refuses to open a gold position when the spread exceeds $1.00. Measured over a year of bars on our own broker that threshold was reached in 3 of 70,546, so on this feed it is insurance rather than a daily filter.

Stop level

Also called: Minimum stop distance, Freeze level.

The stop level is the minimum distance from the current price at which a broker will accept a stop loss, take profit or pending order.

Brokers set this to stop clients placing orders so close to market that they would be triggered by the spread itself. It is expressed in points and can change with conditions — some brokers widen it during news, which means an order that was legal a minute ago is rejected now.

This single setting quietly breaks more automated systems than any other. A strategy that calculates a valid stop of $1.20 on gold will have every order rejected if the broker requires $2.00 of distance, and MetaTrader reports it as a generic invalid stops error rather than explaining the cause. If your EA opened positions but never attached protection, this is the first thing to check.

The workaround used by most well-built EAs, including ours, is to open the position with no protection attached and modify it immediately afterwards, because the constraint is often applied differently to a modification than to the original order. If that modification fails, the only safe response is to close the position rather than run it unprotected — which is what our engine does. See invalid stops.

Swap

Also called: Rollover, Overnight financing.

Swap is the interest debited or credited for holding a leveraged position overnight, applied at the broker's rollover time and tripled on one day of the week.

Because a leveraged position is effectively financed, holding it past the daily rollover incurs an interest adjustment. It can be positive or negative depending on the instrument and direction, though on most retail accounts and most instruments it is a cost in both directions once the broker's markup is applied.

One weekday carries triple swap to account for the weekend, commonly Wednesday. A strategy that holds positions for days will pay this repeatedly, and on a small account the accumulated cost is easy to underestimate — the swap calculator makes it concrete.

For intraday systems swap is close to irrelevant, and that is a genuine structural advantage rather than a detail. Our engine trades a session window and does not carry positions across the rollover, so swap does not enter the cost model at all. Any comparison between an intraday bot and a swing system should account for this rather than comparing gross returns.