Treating max DD as the worst that can happen
It is the worst that has happened in the measured sample. A longer sample almost always finds a deeper hole; the published figure is a floor for future expectations, not a ceiling.
Maximum drawdown
The largest peak-to-trough decline in account equity over a measurement period, as a percentage of the previous peak.
also: Max DD, Peak-to-trough drawdown, Worst drawdown
Updated
In plain English
The worst losing run the account has already lived through: how far it fell from its highest point before making a new high. It answers "how bad did it get?", not "how much did it make?".
Why it matters
Two strategies can post the same annual return and still be different products. The one that got there through a 10% dip is deployable; the one that got there through a 50% dip usually is not, because the account runs out of money or the trader quits at the bottom.
- It sets the account size. A common working rule is to fund 2–3× the published max drawdown, so a strategy with a 20% max DD is not run on the amount of capital you can only afford to lose once.
- It is the yardstick for a losing streak. If live drawdown passes the backtested maximum, the strategy has left its tested envelope, and that is a decision point rather than noise.
- It is path-dependent, so it captures something an average cannot: the single deepest hole, which is the one that closes accounts.
- The two figures a tester prints are not interchangeable. Across the fourteen runs published on this site, equity drawdown — the one that counts open floating loss — came out deeper than balance drawdown in fourteen cases out of fourteen. The gap is 4% to 18% on the twelve not read on MT5's Relative basis. The two read on it show 8% and 76%, and there the two figures pick different moments — so the 76% measures that mismatch, not a deeper hole.
How it is calculated
Max DD % = (Peak equity − Trough equity) ÷ Peak equity × 100 - Peak equity
- The highest account equity reached before the decline began.
- Trough equity
- The lowest equity reached after that peak and before a new peak.
The trough must come after the peak, and the peak is the highest equity reached up to that point — not the highest value in the whole period. A run of separate dips does not add up; only the deepest single peak-to-trough decline counts.
What counts as high or low
- under 10% Shallow — the strategy stayed close to its highs
- 10–20% Normal for a tested retail EA
- 20–35% Deep — needs a large account and a strong stomach
- over 35% The account is one bad run from being unusable
Read it together with
No single number describes a strategy. These metrics change how this one should be read.
Example
An account peaks at $10,000, falls to $8,500 during a losing run, then recovers. The drawdown is measured from the peak that preceded the fall, not from the starting balance.
- Peak equity before the decline
- $10,000
- Lowest equity reached afterwards
- $8,500
- A later, shallower dip does not replace this one.
- Maximum drawdown
- 15%
Calculation (10,000 − 8,500) ÷ 10,000 × 100 = 15%
Result 15% maximum drawdown
Illustrative figures — not the record of any listed EA.
How to interpret it
A drawdown figure means nothing without the window it was measured over and the kind of equity it was measured on. Read those two first, then the number.
| Range | What it means |
|---|---|
| Backtest max DD | Covers the historical window the strategy grew up on. It carries meaning only alongside the data range and the modelling quality. |
| Out-of-sample max DD | Covers data nobody tuned the rules on. It almost always runs deeper than in-sample, and the size of that gap is a curve-fitting signal. |
| Live / forward max DD | Comes from real or forward-tracked execution. Most relevant, smallest sample — a shallow figure early on mostly means the run is short. |
| Balance vs equity DD | Balance drawdown counts closed trades only; equity drawdown adds the open floating loss and comes out deeper — in every published run measured here. |
| Maximal basis vs Relative basis | MT5 prints both, and they can point at different moments of the same run. One listing here reads 4.94% balance / 5.43% equity on the Maximal basis and 5.15% / 5.56% on the Relative one — different peaks, not rounding. |
- Compare drawdowns only over comparable periods. A 12% max DD across seven years and a 12% max DD across seven months are not the same claim.
- Read it next to the recovery factor: net profit divided by max DD says whether the strategy earns enough to justify the hole it digs.
- If an EA quotes one drawdown figure without saying whether it is balance or equity, assume balance — the smaller of the two.
The fourteen runs published here span 0.67% to 14.52% balance drawdown, all measured on a 10,000 USD deposit at 0.1 lot. Eleven trade a single instrument — one of them four legs of it at once — and three spread across several, with the deepest figure belonging to the twenty-four-leg basket. So the spread compares how much sits open at once as much as it compares strategies. Each page states the window and the tester model beside the figure, and leads with equity drawdown wherever the report supplies it.
Common mistakes
Preferring one deep drawdown over many shallow ones on the number alone
A strategy with a single −25% decline in five years scores worse than one with twelve separate −8% declines, yet the second may be far harder to sit through. Max DD ignores frequency and duration entirely.
Comparing percentages measured on different equity bases
A drawdown taken from a peak that included open floating profit is not comparable with one taken from closed balance. Match the measurement, then compare.
Assuming recovery is proportional
A 20% drawdown needs 25% to get back; a 50% drawdown needs 100%. The deeper the hole, the more disproportionate the climb out of it.
Reading the percentage without the account it was measured on
A percentage silently carries the deposit and the lot size behind it. One listing here shows 9.92%, which is 1,136.67 USD at 0.1 lot on 10,000 USD — and 0.1 lot is what the distributed build starts with, so the same EA launched on a 1,000 USD account meets a drawdown already measured as larger than the whole balance.
In depth
Frequently asked questions
- Is maximum drawdown the same as my largest losing trade?
- No. A single losing trade is one position; maximum drawdown is the cumulative decline from a peak to the following trough, which can span dozens of trades and several weeks.
- What is an acceptable maximum drawdown for an EA?
- There is no universal figure — it depends on the account size and how long the strategy has been measured. As a working rule, fund an account with 2–3× the published maximum drawdown, and treat anything above roughly 35% as requiring capital most retail traders do not have.
- Why is the drawdown on my live account deeper than the backtest?
- Backtests price fills optimistically compared with live execution, and the live period contains market conditions the test window did not. A live drawdown moderately deeper than the backtest is normal; one far deeper means the strategy is operating outside its tested envelope.
Related articles
EAs where this matters
Listings whose published record is judged on Maximum drawdown.
