Every strategy, even a great one, hits losing streaks. So the hardest judgment call in trading is: is this drawdown normal variance, or is my edge actually gone? Abandon a working strategy at the bottom of a variance dip and you quit right before it works; ride a genuinely broken one and you bleed out. Neither error is cheap, and there's no perfect answer — only a framework.
Start with the sample
First, is the drawdown even statistically unusual? A strategy with a 45% win rate will have losing streaks of 6, 8, even 10 regularly — that's expected variance, not a broken edge. If your drawdown is within the normal range your win rate implies, the answer is almost certainly “ride it.” Most strategy-abandonment happens inside the range of totally normal variance — traders quit noise.
Then look for a reason
Variance is random; a broken edge usually has a cause. Ask: has the market regime changed in a way that invalidates the strategy's premise (a range strategy meeting a trending market, a gamma regime flip)? Has the edge gotten crowded? Or has your execution degraded — are the losses from the strategy, or from you breaking its rules? A drawdown with an identifiable structural cause is different from a drawdown that's just a cold streak.
Variance has no reason; a broken edge usually does. If you can't name what changed, you're probably looking at a cold streak — and quitting cold streaks is how traders churn through good strategies.
A practical rule
Define, in advance, a drawdown threshold that triggers a review (not an automatic quit) — e.g., a drawdown beyond your historical norm, or a weekly-limit breach. At that point, size down and investigate: sample, regime, execution. Resume full size if it's variance with clean execution; retire or rework the strategy if the premise genuinely broke. Decide the rule when calm, so you're not making this call emotionally at the lows.