Every trading strategy — manual or automated — advertises the same number: profit. Almost none lead with the number that actually predicts whether you'll still be trading next year: drawdown. Here's the math that makes it the most important metric in the report.
What drawdown measures
Drawdown is the fall from an equity peak to its lowest point before a new peak. Account goes $1,000 → $1,200 → $900 → recovers: that valley is a 25% drawdown. Maximum drawdown is the deepest valley in the whole record — the worst stretch you'd have had to sit through without switching the system off.
That last part matters more than people admit. Most strategies don't fail mathematically; they fail behaviourally — the human pulls the plug at the bottom of a drawdown the math would eventually have recovered from.
The asymmetry that changes everything
Losses and gains aren't symmetric:
| Drawdown | Gain needed to break even |
|---|---|
| 10% | +11% |
| 25% | +33% |
| 50% | +100% |
| 75% | +300% |
| 90% | +900% |
A system that halves the account must double just to get back to zero. This is why "avoid deep drawdowns" isn't a risk-appetite preference — it's arithmetic. Two systems with identical +60% annual returns are not equivalent if one did it with 12% max drawdown and the other with 55%.
The balance-vs-equity trick (how bad systems hide risk)
This is the part worth remembering if you evaluate EAs or algo strategies:
- Balance drawdown counts only closed trades.
- Equity drawdown includes floating losses on open positions.
Martingale and grid systems exploit the gap: they never close losers, so the balance curve stays beautifully smooth while equity sinks underwater. The record looks perfect — until one strong trend forces the floating loss to realise, and the account is gone in a day. Always judge equity drawdown. A smooth balance curve with hidden floating losses isn't low risk; it's deferred risk.
Working ranges (for gold EAs specifically)
Gold's volatility makes this sharper than on calmer instruments:
- <10–15% max equity DD — conservative, typical of fixed-risk hard-stop systems
- 15–30% — moderate, acceptable if returns justify it
- 30–50% — aggressive; most humans abandon ship mid-valley
- 50%+ — red flag; the recovery math is against you
How to actually check it
- Read maximal equity drawdown on a long, realistic backtest — then treat it as the minimum you'll see live.
- Prefer a third-party verified live record (Myfxbook-style) over any backtest.
- Confirm a hard stop-loss per trade + small fixed risk — that's what keeps drawdown bounded and visible instead of hidden in floating losses.
- Size the account for the worst streak, not the average month. There's a free drawdown/losing-streak simulator that does this math interactively.
I wrote the full guide with FAQ here: drawdown explained — with companion pieces on risk settings and honest backtesting.
TL;DR
- Profit is what happened; max equity drawdown is the risk of it not happening again
- Recovery is asymmetric: −50% needs +100%
- Judge equity, not balance — smooth curves can hide floating losses
- A hard stop + 1–2% risk per trade is what keeps drawdown survivable
Educational content only, not financial advice. Trading gold (XAUUSD) carries substantial risk. More free guides and calculators at xauusdrobot.com.
Top comments (0)