Month three of a drawdown, I opened the position size field and typed a number that wasn't in the plan.
Nothing in the data had changed. My mood had. That's the whole story of how a normal losing stretch becomes an expensive one.
The drawdown isn't the dangerous part
Every strategy has them. A trend system that wins around 35% of its trades will go through stretches of ten or twelve losses in a row, and the backtest shows it. The damage from those stretches is limited, because each loss is capped at the size you planned.
The damage you can't plan for is the one you add yourself. Drawdowns don't only cost money. They cost confidence, and confidence is what a discretionary trader is actually trading on.
Think about what discretion means in practice. Before every entry you ask whether it feels right. After eight losses, the honest answer is no. So you skip the next signal, which happens to be the one that would have paid for the whole stretch.
Three ways discretion shows up
Skipping trades is the quiet one. You tell yourself you're waiting for better conditions. What you're doing is filtering signals by how recent losses feel, which is the worst possible filter because the best trades tend to arrive right after the worst runs.
Resizing is the loud one. Cut size after losses and a normal recovery barely moves your equity. Raise it to win it back and one more bad run does real damage. Both versions are the same mistake: letting the last few outcomes set your risk.
Then there's tinkering. A new indicator, a tighter stop, a different timeframe. Each change feels like diligence. But if you edit the rules in the middle of a drawdown, you can't tell afterwards whether the system failed or you did.
What a rule does that willpower can't
A rule is decided once, on a day when you were calm and looking at a hundred trades instead of the last three. That's the whole advantage. It isn't smarter than you. It's you at your most rational, locked in before the pressure arrived.
When the drawdown hits, the rule doesn't get tired or scared. It takes the next signal at the planned size. Often that's the signal that ends the drawdown.
I'm not arguing that judgment is worthless. Choosing a market, setting a risk limit, deciding when a strategy is truly broken: those are judgment calls, and they deserve real thought. The point is when you make them. Do it in the calm of a weekly review, never in the middle of a Tuesday that's down 4%.
Separate the two decisions
There are two different questions hiding in every drawdown. Is this normal for the system? And should I keep running the system? Most traders answer both with a feeling.
Answer the first with numbers. Before going live, write down the worst drawdown and longest losing streak in your backtest, then add a buffer. If live results sit inside that range, the system is doing what it said it would. If they break well outside it, that's a real signal and it's time to stop and investigate.
Answer the second only on a scheduled review day, never in the moment. Write down the date and the threshold in advance. If you can't point to a pre-written reason to change something, don't change it.
The takeaway
Open your trading notes and look for a rule that says what you'll do after five losses in a row. If it isn't there, write it today. One line is enough: same size, same signals, next review on Sunday.
That one line removes the most expensive decision you'll face this year. If you'd like to see how a fully rule-based system behaves through its worst stretches, our complete backtest data and methodology are at v33systematic.com.