Drawdown recovery math: why a 40% loss needs a 67% gain
Tóm tắt — Losses and gains are not symmetrical. A 20% drawdown needs a 25% gain to break even, a 40% drawdown needs 66.7%, and a 50% drawdown needs 100%. The formula, a full recovery table with realistic timelines, why prop firms cap drawdown near 10%, and the real relative-drawdown figures from our published EA reports — from 3.92% to 42.91%.
Drawdown recovery math: why a 40% loss needs a 67% gain
TL;DR — Losses and gains are not symmetrical, and that asymmetry is the whole reason drawdown limits exist.
- The gain needed to get back to breakeven is loss ÷ (1 − loss). A 20% drawdown needs 25%. A 40% drawdown needs 66.7%. A 50% drawdown needs 100%.
- At a steady 2% per month, recovering a 40% drawdown takes about 26 months — more than two years of good performance to erase one bad quarter.
- Prop firms cap drawdown near 10% because that is roughly the last point where recovery is still realistic.
- In our published reports, relative drawdown ranges from 3.92% (jpy-asia-ranger) to 42.91% (goldgridm1). Highest net profit and lowest risk are not the same EA.
What is drawdown recovery math?
It is the arithmetic that converts a percentage loss into the percentage gain required to get back to where you started. The formula is short:
required gain = drawdown ÷ (1 − drawdown)
The reason it is not simply "lose 40, make 40" is that the second percentage is calculated on a smaller base. Lose 40% of 10,000 USD and you have 6,000 USD. Making 40% on 6,000 gives 8,400 — still 1,600 short. You need 66.7% on the reduced balance to see 10,000 again.
| Drawdown | Gain needed to break even | Months at 2%/mo | Months at 1%/mo |
|---|---|---|---|
| 5% | 5.3% | 2.6 | 5.2 |
| 10% | 11.1% | 5.3 | 10.6 |
| 20% | 25.0% | 11.3 | 22.4 |
| 25% | 33.3% | 14.5 | 28.9 |
| 30% | 42.9% | 18.0 | 35.8 |
| 40% | 66.7% | 25.8 | 51.3 |
| 50% | 100.0% | 35.0 | 69.7 |
| 60% | 150.0% | 46.3 | 92.1 |
The recovery times assume the strategy keeps compounding at that rate with no further losing months — which never happens. Treat them as a floor, not a forecast.
Why do prop firms cap drawdown around 10 percent?
Because roughly 10% is the last drawdown level a trader can plausibly trade out of. An account down 10% needs 11.1% to recover — annoying, achievable. An account down 30% needs 42.9%, which for most retail-scale strategies is more than a full year of good returns. The firm is not being harsh; it is closing accounts at the point where the expected recovery cost exceeds the expected remaining value.
This also explains the two-limit structure most firms use: a daily loss limit (often 4–5%) stops a single catastrophic session, and a max drawdown (often 8–12%) stops slow bleed. We covered how those interact in daily loss limit vs max drawdown.
Which drawdown number should you look at in an EA backtest?
Relative drawdown, expressed as a percentage — not the absolute currency figure. Absolute drawdown depends entirely on the deposit and the risk setting used in the test; the percentage is the part that transfers to your own account size.
Here is that number across a sample of our published builds, with the recovery gain each one implies:
| EA | Symbol | Relative drawdown | Gain needed to recover | Recovery factor | Net profit (10k USD) |
|---|---|---|---|---|---|
| jpy-asia-ranger | USDJPY H1 | 3.92% | 4.08% | 2.59 | 1,013 |
| newsstraddle | XAUUSD M15 | 5.71% | 6.06% | 30.47 | 20,747 |
| indexorb | USTEC M15 | 6.42% | 6.86% | 3.86 | 2,723 |
| orderblocksmc | XAUUSD H1 | 6.77% | 7.26% | 1.40 | 1,029 |
| asianbreakout | USDJPY H1 | 10.29% | 11.47% | 0.85 | 940 |
| gold-trend | XAUUSD H1 | 12.98% | 14.92% | 0.77 | 1,000 |
| aurora-gold-breakout | XAUUSD H1 | 15.56% | 18.43% | 9.28 | 14,444 |
| goldgridm1 | XAUUSD M1 | 42.91% | 75.16% | 7.01 | 31,111 |
Read the last row against the first. goldgridm1 produced the largest net profit in the set, and it also spent part of the test needing a 75% gain to get back to its own equity peak. On a prop account with a 10% limit, that curve is not a big win — it is a closed account. Full reports are on the proof page.
One honest caveat about the source data: two of our builds report a relative drawdown of exactly 0.00 in the raw MT5 output while showing a non-zero absolute drawdown. We treat that as unreported, not as zero risk, and we have left them out of the table rather than publish a flattering number we cannot stand behind.
What is recovery factor, and why does it matter more than net profit?
Recovery factor is net profit divided by maximum drawdown. It answers the only question that matters when comparing two profitable systems: how much pain did you buy that profit with?
A recovery factor below 1.0 means the strategy has not yet earned back more than its own worst drawdown over the entire test window — asianbreakout at 0.85 and gold-trend at 0.77 are both in that category, and we publish them anyway. Above 3.0 is genuinely good. newsstraddle at 30.47 is an outlier driven by a narrow event-driven edge, and outliers deserve more scepticism, not less. You can line the builds up side by side on the compare page.
How do you keep an EA inside a drawdown limit?
Three controls, in order of effectiveness:
- Size from equity, not from a fixed lot. Percent-risk sizing shrinks position size automatically as the account falls, which flattens the tail of a losing streak instead of steepening it.
- Cap concurrent exposure. Two EAs both long gold is one position with two names on it. Correlated positions turn a 5% plan into a 12% outcome.
- Enforce a hard stop at the account level. An equity guardian that flattens and disables trading at a preset floor is the only control that survives an EA behaving unexpectedly. That is what PropGuard exists to do.
FAQ
How much do you need to gain to recover a 50% drawdown? 100%. You must double the remaining balance. This is the sentence worth remembering: a 50% loss and a 50% gain are not opposites — the loss takes half your account, the gain only takes it back to where it already was.
Is a 20% drawdown acceptable for an EA? For a personal account with a long horizon, possibly. For a prop-firm account, no — 20% breaches every mainstream firm's limit, and the 25% recovery it demands typically means a year of good months. Match the drawdown profile to the account rules before you look at profit.
Does a low drawdown in a backtest guarantee a low drawdown live? No. A backtest reports the worst drawdown that did happen in that specific window, not the worst that can happen. Live spreads, slippage, gaps and regime changes routinely produce drawdowns larger than the tested maximum. Plan for at least 1.5 times the tested figure.
Risk caveat: all figures above come from MetaTrader 5 strategy-tester reports on a 10,000 USD simulated account with percent-risk sizing, over the date ranges shown on each report. Backtest results are historical simulations and do not predict future performance. Trading leveraged products carries substantial risk of loss. Nothing here is financial advice — see our risk disclosure.