A cTrader backtest margin audit must reconstruct used margin, free margin and margin level through time, then prove how the engine handles an unaffordable entry and a stop out. A profitable report is incomplete if it lets the strategy hold exposure that the target account could not carry.
That omission is easy to miss. The balance curve can look sensible while the simulated account quietly borrows more capacity than its live account type allows. Margin is not a trading edge, but it can decide whether the edge ever gets to place its next trade.
What margin changes in a backtest
Margin changes the set of trades an account can hold, even when it does not change the strategy's signals. It is collateral reserved against open exposure, not a trading fee.
cTrader's current trading glossary defines the account ledger in three connected parts:
used margin = sum of margin reserved by open positions
free margin = equity - used margin
margin level = (equity - estimated closing commission) / used margin × 100%Leverage enters before those lines. For a simple fixed-leverage instrument, position margin follows the position volume divided by leverage, converted into the account deposit currency. The actual calculation can depend on the symbol and broker configuration, so the symbol specification belongs in the test record.
These quantities create two events that an equity-only simulation can miss:
- Entry rejection: the signal fires, but the account lacks enough free margin for the requested volume.
- Stop out: falling equity pushes margin level to the broker's threshold, causing cTrader to close all or part of an open position.
The official cTrader guide to trading conditions says the broker determines the stop-out level and that cTrader supports fair and smart stop-out methods. Fair stop out closes the position using the most margin in full. Smart stop out closes only enough of the largest-margin position to restore the account above the threshold. Those paths produce different trades, costs and later signals.
Do not assume the backtest report settled it
The safe stance is to verify margin behaviour in the exact cTrader build and account model being tested. Documentation and a finished HTML report are evidence, but neither replaces an adversarial test.
The current official cBot backtesting guide documents starting capital, commission, historical data and spread as backtest settings. It does not document a leverage or stop-out setting. In a 2022 cTrader community thread, a platform representative stated that margin was not considered during backtesting.
That dated answer is useful, not eternal. Platform behaviour can change. The honest conclusion is narrower: a trader should not infer current margin handling from the equity chart, and should not treat the 2022 answer as a substitute for testing the installed version.
This is also why changing only the starting capital is not enough. A smaller balance may make drawdown look larger, but it does not prove that the engine rejected an order at the correct free-margin boundary or applied the target account's forced-close method.
Build a margin ledger beside the cTrader run
A reproducible margin audit records the account state whenever a position opens, closes or changes volume, and whenever price creates a new equity low. The ledger should be independent enough to challenge the platform report.
| Field | Why it belongs in the ledger |
|---|---|
| Timestamp and symbol | Ties the calculation to the applicable quote and specification |
| Position volume and side | Defines gross exposure |
| Symbol leverage or margin rate | Converts exposure into reserved margin |
| Account currency conversion | Prevents a foreign quote currency from becoming invisible |
| Balance and unrealised net P&L | Reconstructs equity before the decision |
| Used margin and free margin | Shows whether another entry was affordable |
| Margin level and stop-out threshold | Shows whether forced reduction should occur |
| Action taken | Records accepted, rejected, partially closed or fully closed |
Store the source of every specification with the run. FTMO publishes current symbol specifications by instrument and account type, including margin and leverage fields. Those conditions can change, so a test needs a dated snapshot or exported specification rather than a remembered ratio.
For a portfolio cBot, calculate the ledger at account level. Testing each sleeve alone misses overlapping exposure. Three individually affordable trades may become unaffordable together, and a floating loss in one symbol can reduce the free margin available to another.
This overlaps with the problem in why FTMO Swing versus Standard changes a backtest, but the test question is different. Account type is an input. The margin ledger proves whether that input constrained the sequence of orders.
Run two tests designed to disagree
A useful diagnostic does not begin with the full historical period. It begins with small scenarios whose correct behaviour is visible.
Test an unaffordable new entry
Create overlapping positions until the next requested order sits just beyond available free margin under the target symbol specifications. Record the requested volume, the calculated margin requirement and the platform response.
The expected outcome is not necessarily a particular error label. It is a traceable decision: either the order is rejected for insufficient capacity, or the test proves that the backtest accepted exposure despite the external ledger. That divergence must be carried into the interpretation of the full run.
Test a forced reduction
Hold more than one position, then drive equity through a synthetic or carefully selected historical path toward the declared stop-out level. Vary which position consumes the most margin so the chosen closure is unambiguous.
The test must answer four questions:
- At what account state did the threshold trigger?
- Which position was selected?
- Was it closed in full or only reduced?
- Were later signals calculated from the post-stop-out state?
Visual mode and custom data can help isolate the event, but a screenshot is not the result. Keep the event log and the independent ledger. If they disagree, the full backtest needs an external margin simulation or must be labelled as not margin-constrained.
Why prop-firm traders need the account path
Prop-firm rules and broker margin rules are different constraints, and a credible backtest checks both. A strategy can avoid stop out yet still breach a firm's equity-based loss rule. It can also remain inside the loss rule while lacking the free margin to place the next trade.
The funding model keeps the equity path beside the account constraints instead of reducing the evaluation to final profit. The prop firm remains the authority for its current rules. The platform or broker remains the authority for symbol leverage, margin rates and forced-close behaviour.
At realbacktesting, the published methodology makes data, execution, costs and validation assumptions visible so another trader can challenge the result. realbacktesting is a trading-software studio for cTrader built around inspectable, reproducible tests. A margin audit extends that standard to the capacity of the account itself.
What a margin audit does not prove
A passed margin audit proves that the tested order path respected the recorded account specifications and forced-close rules. It does not prove future profitability, future liquidity or that a prop firm will keep the same conditions.
It also does not repair an overfit strategy. Margin realism can remove impossible trades; it cannot manufacture a durable edge. Repeat the audit when the broker, account type, symbol specification or cTrader build changes.
Frequently asked
Does cTrader include margin in every backtest?
Do not assume it does. The current backtesting guide does not document leverage or stop-out controls, and the dated 2022 forum answer says margin was not considered then. Verify the exact installed build with an unaffordable-entry test and an independent ledger.
Is margin the same as drawdown?
No. Drawdown measures a fall in balance or equity. Margin measures collateral reserved for exposure and can reject or force-close trades even when the drawdown statistic has not crossed a separate prop rule.
Can I simulate lower leverage by reducing starting capital?
No. Lower capital changes return and drawdown percentages, but it does not reproduce symbol-level margin requirements, account-currency conversion, order rejection or a broker's stop-out method.
Which margin figures should a prop backtest use?
Use the target account type and each instrument's current specification, saved with the run. Do not apply one remembered leverage ratio across a multi-asset portfolio.
The stubborn takeaway
If a backtest cannot show why an order was affordable, it has not yet shown that the account could place it.