Losing Streaks: The Maths a Gold EA Survives
TL;DR A 50% win rate produces a run of seven losses somewhere in 200 trades more often than not - that is normal, not broken. At 1% risk a seven-loss run costs about 7%. At 5% risk it costs about 30%, and recovering 30% needs a 43% gain. The risk setting decides whether a normal streak is an inconvenience or an ending.
Streaks are not evidence of anything
Human intuition reads six losses in a row as proof that something has broken. Probability reads it as Tuesday.
At a 50% win rate, the chance of a specific run of six losses is 1 in 64 — but across 200 trades there are many places for such a run to start, and at least one usually does. Lower the win rate and the expected streak gets longer: a strategy winning 40% of the time will produce longer runs than one winning 60%.
The question worth asking is not "why did it lose six times" but "was six outside what this strategy should produce". Usually it was not.
What a streak costs at each risk level
Each loss comes off the remaining balance, so the damage compounds rather than adding up:
| Risk per trade | After 5 losses | After 7 losses | After 10 losses |
|---|---|---|---|
| 1% | -4.9% | -6.8% | -9.6% |
| 2% | -9.6% | -13.2% | -18.3% |
| 3% | -14.1% | -19.2% | -26.3% |
| 5% | -22.6% | -30.2% | -40.1% |
The 1% row is survivable while barely changing the plan. The 5% row ends most accounts, because of what comes next.
The recovery asymmetry
Losses and gains are not symmetrical, and the gap widens fast:
| Drawdown | Gain needed to recover |
|---|---|
| 10% | 11% |
| 20% | 25% |
| 30% | 43% |
| 50% | 100% |
| 70% | 233% |
A 10% hole is a few good weeks. A 50% hole requires doubling the account just to get back to even — which is why the only reliable drawdown strategy is not being in a deep one.
What this means for an EA setting
An EA does not get discouraged, which is its advantage, and it does not stop when something is wrong, which is its risk. It will trade the streak at whatever size you configured, all the way down.
So the risk percentage is not a performance dial. It is the answer to "how long can this be wrong before I cannot recover". Set it against the worst run the strategy has actually produced in testing, not against the average.
If the backtest shows a nine-loss run, the account has to survive nine losses with room to continue. At 1% that is routine. At 5% it is a 37% drawdown and a 59% gain required to recover.
Running the numbers on your own setup
Take the longest losing run from your backtest and extend it by half again — live runs are usually worse than tested ones, because spread and slippage turn marginal wins into losses.
Put that number into the risk calculator at your intended risk level and read the balance it leaves. If that figure would make you turn the EA off, the risk setting is too high — because turning it off at the bottom of a normal streak is how a working strategy gets abandoned.
Questions we get
How many losses in a row is normal for a gold EA?
At a 50% win rate, a run of six or seven appears somewhere in a couple of hundred trades more often than not. Lower win rates produce longer runs. A streak on its own is not evidence that anything has broken.
What does a 7-loss streak cost at 1% risk?
About 6.8% of the account, because each loss is taken from the reduced balance rather than the starting one. At 5% risk the same streak costs about 30%.
Why is recovering a drawdown harder than causing it?
Because the gain is calculated on a smaller balance. Losing 30% leaves 70, and getting 70 back to 100 is a 43% gain. At 50% down it takes a 100% gain just to break even.
Should I stop an EA during a losing streak?
Only if the streak is longer than anything the strategy produced in testing, or the market conditions it relies on have clearly changed. Stopping inside a normal run locks in the drawdown and forfeits the recovery the strategy was built to produce.
How do I pick a risk percentage for an EA?
Start from the longest losing run in testing, add a margin because live results are usually worse, and choose a risk level where that run still leaves an account you would be willing to keep trading.
Want a second read on your risk settings?
Send the backtest and the risk percentage and we will tell you what the worst run in it would have cost.
💬 Talk to Us on Telegram