You are scrolling via MQL5. You discover a low drawdown EA with 2% max DD and 300% returns. “Low danger, excessive reward.” The holy grail.
Besides profitability with out drawdown context usually means the other of what you suppose. That 2% drawdown on a low drawdown EA is likely to be probably the most harmful quantity on your entire web page.
Let me clarify why — and what to really search for when somebody guarantees you “low drawdown.”
“Low Drawdown” Is the Most Abused Phrase in EA Advertising
An ideal backtest in MT5 is virtually a confession that somebody is promoting you smoke. And “low drawdown” is the best metric to pretend. Here is how distributors do it:
Trick 1: Backtest-Solely Drawdown
Backtests are optimized. The developer runs hundreds of parameter mixtures, picks the one with the bottom drawdown and highest return, and publishes that curve. It isn’t prediction — it is curve becoming. After all the drawdown is low. They chose for it.
What to search for: Dwell verified outcomes. Myfxbook sign with actual cash. If the seller solely exhibits Technique Tester screenshots, stroll away.
Trick 2: Hidden Restoration Mechanism
Some “low drawdown” EAs use managed place scaling — after they lose, they open barely bigger positions to recuperate sooner. The fairness curve appears to be like easy as a result of losses are recovered rapidly. However the danger per commerce is rising precisely when the market has already confirmed it is towards you.
This works… till it would not. And when it fails, it fails catastrophically. The drawdown goes from 3% to 30% in a single unhealthy sequence.
Trick 3: Tiny Place Sizes
Any EA can present 2% drawdown if it trades 0.01 tons on a $10,000 account. The drawdown is “low” as a result of the EA barely trades relative to the account dimension. The returns are additionally minimal, however the vendor would not spotlight that half.
What to examine: What lot dimension was used relative to account steadiness? What is the danger per commerce? 2% drawdown on 0.01 tons is not conservative — it is irrelevant.
Trick 4: Brief Commentary Interval
A 30-day observe file with 2% drawdown means nothing. The EA merely hasn’t been via an actual stress check but. Run any EA for 3 years and the drawdown image adjustments dramatically. Markets cycle via situations {that a} 30-day window cannot seize.
Win Price would not prevent. Drawdown does:
What a Low Drawdown EA Ought to Truly Look Like
Actual low drawdown is not in regards to the quantity being small. It is in regards to the relationship between drawdown, danger, and your capability to remain within the recreation.
The great relationship with danger is not the one which maximizes potential revenue. It is the one that permits you to hold working with a transparent head. If a standard unhealthy week already places you psychologically outdoors the system, you are outsized — no matter what the backtest says.
Here is what to judge as a substitute of simply the drawdown proportion:
1. Drawdown Relative to Acquire
An EA with 15% achieve and eight% drawdown is a totally totally different beast from one with 300% achieve and eight% drawdown. The second is nearly definitely hiding danger someplace — both in place sizing, restoration logic, or cherry-picked backtest parameters.
A wholesome gain-to-drawdown ratio for a official EA: roughly 1.5:1 to three:1 on dwell outcomes. Something above 5:1 on a backtest ought to increase critical questions.
2. Drawdown Length
Was the max drawdown a fast spike that recovered in days, or a grinding decline over months? Each may present “8% max drawdown” however the psychological expertise is radically totally different.
A one-day spike to -8% from a information occasion is manageable. Three months of sluggish bleeding to -8% will make most merchants flip the EA off at -5% and miss the restoration.
3. Dwell vs Backtest Drawdown
If the backtest exhibits 3% drawdown however the dwell account exhibits 8%, that is not a catastrophe — that is actuality. Dwell execution has slippage, unfold variation, and actual market situations that backtests cannot absolutely replicate.
A dwell drawdown of 1.5-2x the backtest drawdown is regular. Greater than 2x means one thing is essentially totally different between check and dwell situations.
Alpha Pulse AI: 8.6% dwell drawdown. Verified.
+15.25% achieve, PF 1.29, 105 trades on Myfxbook. Not a backtest. Not a screenshot. Each quantity together with the losses.
The Psychology Lure of “Low Drawdown”
Here is the uncomfortable reality: you may have a look at an EA with genuinely low drawdown, purchase it, and nonetheless blow up. As a result of the drawdown quantity and your capability to deal with it are two utterly various things.
You may see one thing that appears nice on paper — low drawdown, constant returns, easy curve. However you have not lived via the unhealthy part. You have not felt what -6% appears to be like like by yourself cash at 3 AM when the market is transferring towards you.
I had this expertise myself with Alpha Pulse AI. The early outcomes have been thrilling — a few nice weeks. Then got here the tough weeks. The information was saying “that is inside parameters.” My intestine was saying “one thing is unsuitable.” The information was proper. My intestine was being a typical scared retail dealer.
The EA you may really maintain via a drawdown issues greater than the EA with the bottom drawdown. As a result of you may present somebody a system that recovers each time, they usually’ll nonetheless flip it off on the worst doable second. If you cannot survive the drawdown psychologically, the backtest is irrelevant. You will have already turned it off. I wrote a full information on what to really do (and never do) when your EA is in drawdown — as a result of the psychology of holding is a much bigger downside than the drawdown itself.
How I Consider Drawdown (Actual Framework)
Once I have a look at any EA’s drawdown numbers, this is what I really examine:
- Is it dwell verified? Myfxbook, FXBlue, or any third-party verification. If it is solely a Technique Tester screenshot, I am already performed. Choose it by dwell conduct, not by a reasonably tester
- What number of trades? Minimal 50 trades to contemplate the drawdown statistically significant. Minimal 100 to belief it. Fewer than 50 and the drawdown quantity is random noise
- What was the market doing? 8% drawdown throughout a peaceful market may be very totally different from 8% drawdown throughout a disaster. The EA that held 8% throughout March 2026’s volatility has confirmed one thing actual
- What is the danger per commerce? 2% danger per commerce with 8% drawdown means about 4 consecutive losers. That is regular for any system. 0.1% danger per commerce with 8% drawdown means 80 consecutive losers — one thing is badly unsuitable
- What is the restoration mechanism? Does the EA enhance place sizes after losses? In that case, the “low drawdown” interval is simply the calm earlier than the storm. Sensible restoration with limits is ok. Limitless scaling is a ticking bomb
Low-drawdown EA settings — dwell account proof:
A Portfolio Is the Actual “Low Drawdown” Technique
A single EA cannot diversify danger. It trades one logic on one instrument. If situations flip towards it, your whole account attracts down.
A portfolio of uncorrelated methods does what no single EA can: when one attracts down, others can compensate. The mixed drawdown is often decrease than any particular person technique’s drawdown. That is not magic — it is fundamental arithmetic of correlation.
As an alternative of looking for the one EA with impossibly low drawdown (which might be faking it), construct a portfolio with a number of methods which have trustworthy, manageable drawdowns. 3 methods with 10% particular person drawdown, correctly uncorrelated, can produce a portfolio drawdown of 6-8%.
For this reason a single EA all the time fails the psychology check. The portfolio method solves the drawdown downside at a structural stage, not with methods.
Construct your portfolio at zero price.
Obtain the free USDJPY technique module — your first diversification layer. Then add Alpha Pulse AI for Gold protection. Completely different markets, totally different logic, one portfolio.
FAQ
What’s a “secure” max drawdown for an EA?
There isn’t any common quantity. It is dependent upon your danger tolerance. 10% drawdown at 2% danger per commerce is conservative. 10% drawdown at 0.5% danger means one thing went sideways for weeks. The drawdown quantity with out risk-per-trade context is meaningless.
Ought to I solely have a look at EAs with lower than 5% drawdown?
No. An EA with 5% drawdown on a backtest will seemingly present 8-12% dwell. When you filter for <5% you will both get curve-fitted backtests or EAs buying and selling too small to matter. Real looking expectations: 8-15% max drawdown for a dwell EA with significant returns.
How do I do know if the drawdown is actual?
Third-party verification. If it is on Myfxbook with a verified actual account, the drawdown is actual. If it is solely a screenshot or a Technique Tester report, assume it is optimistic. Do not play solitaire with your self — if the proof would not have impartial verification, it would not depend.
My EA has “low drawdown” however I am nonetheless anxious. What’s unsuitable?
You are in all probability outsized. If a standard unhealthy week already makes you uncomfortable, the issue is not the EA — it is your place sizing. Cut back danger till you genuinely do not care if it wins or loses this week. That is the danger stage the place you may really maintain via drawdowns.
Sources
- Alpha Pulse AI: 105 Trades — Each Quantity Uncovered — Actual dwell drawdown information, not a backtest fairytale
- DoIt Alpha Pulse AI — AI-powered Gold EA. 8.6% max DD, PF 1.29. $397 one-time
- Free USDJPY Technique Module — First step towards actual portfolio diversification
- IC Markets — Uncooked spreads for execution-sensitive EAs
- E-newsletter — Weekly evaluation on EA efficiency, danger administration, and what I am really working
The following time an EA guarantees you “low drawdown,” ask: dwell or backtest? What number of trades? What is the danger per commerce? If they cannot reply all three, you’ve gotten your reply.
