Two buying and selling robots can end with the identical revenue and expose an account to very totally different floating losses. Earlier than evaluating headline returns, verify what the report measures, when every take a look at begins and the way the positions had been sized.
Watch: learn fairness drawdown first
Stability and fairness reply totally different questions
Stability displays closed outcomes. Fairness additionally displays open positions. A basket can present a clean steadiness curve whereas carrying an adversarial floating transfer. Learn most relative fairness drawdown and examine the fairness curve, particularly for methods that add positions or rebalance publicity.
For the printed comparisons in my EA Setup Hub, return means internet revenue divided by the preliminary take a look at deposit. Drawdown means most relative fairness drawdown as reported by MT5. Every separate comparability window begins from a brand new deposit; its outcomes can’t be added collectively as in the event that they had been one steady account.
A setting can change the technique you might be evaluating
Bitcoin Regime H1 Professional 2.00 has a reference profile with one permitted adverse-move add and a comparability profile with that add disabled. In matched historic assessments for 1 January 2020–13 July 2026, the reference profile produced USD 17,657.33 internet revenue with 26.03% most relative fairness drawdown. Disabling the add produced USD 3,446.07 internet revenue with 15.43% drawdown.
These runs use IC Markets (EU), BTCUSD H1, a USD 994 preliminary deposit and OHLC M1 modeling with 93% historical past high quality. They’re simulations. The reference profile was chosen utilizing historic outcomes, so this isn’t impartial potential validation. Neither consequence establishes what one other dealer or future interval will ship.
An replace doesn’t enhance each window
Gold Regime H1 Professional 4.10 diminished drawdown versus 4.00 within the three printed OHLC comparability home windows, with much less revenue over the longer durations. The latest 16 July–30 September 2026 window nonetheless misplaced 3.20%. June favored the sooner model. Displaying each variations and the dropping window provides you a extra helpful comparability than choosing solely a positive chart.
Six inquiries to put beside each report
- Is that this a historic simulation, a demo account or a reside account?
- What are the precise model, inputs and position-sizing guidelines?
- Which dealer, image contract, dates and modeling technique had been used?
- Are prices, execution delay and missing-data limitations said?
- Does drawdown embody floating positions, and are troublesome durations seen?
- Had been the settings chosen on these identical knowledge, or examined independently afterward?
Use protections with the proper interpretation
An fairness guard is a rule that detects a situation and makes an attempt an motion. It’s not a promise that loss can’t exceed a share. Different positions, market gaps, slippage, margin occasions and failed execution can have an effect on the result. Learn whether or not a guard screens the entire account and which positions it’s allowed to shut.
Obtain native stories and matching SET recordsdata from the Setup Hub. Use the free Market demo within the MT5 Technique Tester and examine a number of beginning dates. After buy or rental, validate on a demo account. Convey reproducibility inquiries to MQL5 Feedback or non-public messages, with account identifiers eliminated.
Leveraged buying and selling could cause substantial losses. Printed backtests are historic simulations, not a reside efficiency document; previous efficiency doesn’t assure future outcomes.
