Is This Drawdown a Warning or Just Variance?
Posted: Thu Oct 29, 2026 8:24 am
Is This Drawdown a Warning or Just Variance?
A strategy tested with an expectancy of plus 0.25R and a standard deviation of 1.5R per trade. The historical maximum drawdown was 10R. Now, after 50 live trades, the trader has an average of minus 0.05R and is in a 12R drawdown. Should they stop?
Check against the predefined trigger. The rule says: review when drawdown exceeds 1.5 times the historical maximum, or 15R. At 12R, the trigger has not fired, but the trader is within 3R of it.
Check the average. The standard error of the mean after 50 trades is 1.5 divided by the square root of 50, which is 0.21R. The expected mean is 0.25, and the observed is minus 0.05, a gap of 0.30R, or about 1.4 standard errors. That happens by chance about 8 percent of the time in one direction. It is uncomfortable, but not conclusive.
Check process. Were the rules followed? If the trader broke rules on a third of trades, the cause is behavior. If the rules were followed and the market regime changed, the strategy may be out of conditions.
Decision: reduce risk by half, per the plan, continue for another 50 trades, and review. If the average remains near zero or negative after 100 trades, the evidence that the edge has vanished becomes stronger: the standard error is 0.15R, and a mean of minus 0.05 is 2 standard errors below the expectation.
This is the logic of staged response.
Exercise: define your own review triggers using drawdown and rolling averages.
A strategy tested with an expectancy of plus 0.25R and a standard deviation of 1.5R per trade. The historical maximum drawdown was 10R. Now, after 50 live trades, the trader has an average of minus 0.05R and is in a 12R drawdown. Should they stop?
Check against the predefined trigger. The rule says: review when drawdown exceeds 1.5 times the historical maximum, or 15R. At 12R, the trigger has not fired, but the trader is within 3R of it.
Check the average. The standard error of the mean after 50 trades is 1.5 divided by the square root of 50, which is 0.21R. The expected mean is 0.25, and the observed is minus 0.05, a gap of 0.30R, or about 1.4 standard errors. That happens by chance about 8 percent of the time in one direction. It is uncomfortable, but not conclusive.
Check process. Were the rules followed? If the trader broke rules on a third of trades, the cause is behavior. If the rules were followed and the market regime changed, the strategy may be out of conditions.
Decision: reduce risk by half, per the plan, continue for another 50 trades, and review. If the average remains near zero or negative after 100 trades, the evidence that the edge has vanished becomes stronger: the standard error is 0.15R, and a mean of minus 0.05 is 2 standard errors below the expectation.
This is the logic of staged response.
Exercise: define your own review triggers using drawdown and rolling averages.