A backtest can report an attractive return while concealing assumptions about how orders reach the market. The Expert Advisor may generate identical signals across several tests, yet its recorded trades can differ because prices, order processing and execution conditions are represented differently.
In mt5, strategy testing offers several ways to model market activity, including real ticks, generated ticks and opening prices. These settings are not interchangeable measures of execution quality. Their differences become especially important when a CFD strategy depends on intrabar movement, precise entry prices or timely exits.
Generated Ticks Can Invent a Different Intrabar Sequence
A one-minute candle records its opening, highest, lowest and closing prices, but those four observations do not establish the exact sequence of every transaction or quote change. Generated-tick testing reconstructs movements from available historical information, whereas real-tick testing uses recorded tick data when available.
A strategy placing both a stop and a profit target within the same candle may produce different outcomes depending on which price is reached first. The candle’s final appearance cannot resolve that question.
For short-duration CFD strategies, compare results using real ticks and generated ticks, paying particular attention to trades where both exit thresholds fall inside one bar.
Opening-Price Tests Can Overstate Execution Precision
The opening-prices-only mode is useful for strategies designed to make decisions exclusively when a new bar begins. It becomes less representative when an Expert Advisor reacts to changing prices throughout the candle.
An algorithm that continuously adjusts orders or evaluates intrabar conditions may not receive the event sequence it expects. Apparent improvements in profitability can therefore reflect fewer simulated opportunities for unfavorable execution rather than a stronger trading rule.
Check whether the program genuinely operates on completed bars before treating an opening-price test as meaningful evidence.
Spread Assumptions Can Change the Economics of a Trade
Imagine a hypothetical European packaging-sector index CFD trading near 3,200 points, with each contract worth €3 per index point. An automated strategy buys five contracts at 3,200.0 in a test and exits at 3,204.0, recording €60 in gross profit.
Under a different tick and spread environment, the same signal produces an executable entry at 3,201.2 and an exit at 3,203.1. The gain shrinks to 1.9 points, or €28.50 across five contracts. The difference is €31.50, even though the strategy’s directional forecast remains correct.
Such comparisons require care: recorded tick spreads and simulated pricing conditions depend on the historical data and tester configuration. A test using real ticks is not automatically a complete simulation of live broker execution.
Order-Filling Rules Can Change Whether Exposure Exists
An Expert Advisor may request immediate execution, but the instrument’s supported filling policies determine how volume requests are handled. Depending on the applicable policy, an order can require complete execution, permit partial execution or be canceled when its conditions cannot be met.
In mt5 testing, these distinctions affect strategies that assume every successful signal produces the requested position size. A rejected order may eliminate a profitable trade; a partially filled order can produce a smaller result than the strategy’s internal calculations anticipate.
Review the symbol’s execution and filling properties, then examine trade logs for rejected requests and unexpected volumes rather than comparing final account balances alone.
Simulated Delays Can Expose Fragile Order Timing
Order placement and execution are not instantaneous in live trading. Communication delays can separate the price observed when an Expert Advisor decides to trade from the price available when its request is processed.
The strategy tester supports execution-delay settings that help examine this sensitivity. A fixed or randomized delay can alter entry timing and expose assumptions that disappear in an idealized test.
More detailed simulation can sometimes produce a smoother equity curve, but greater apparent realism does not guarantee more conservative results. A particular delay may accidentally improve some fills while worsening others.
Before opening a CFD position, test the Expert Advisor with real ticks where available, compare alternative tick-generation modes, apply representative execution delays and inspect the order logs. Record changes in fill prices, rejected requests, position sizes and drawdowns before deciding whether the strategy remains viable.
