In this lesson
- find out what drawdown is normal for your strategy
- set tiered rules for reducing risk
- see in a simulation what such a rule costs and what it gains
You know from the previous course that a drawdown is harder to recover than to fall into, and you have daily and weekly loss limits. They protect you from one bad day. This lesson is about a worse enemy: a slow drawdown over weeks and months, when every single trade looks fine – and the account still keeps falling.
What drawdown is normal#
Every strategy has drawdowns. The question is not "whether" but "how big is still normal". A backtest gives you the answer, or a simulation like the one in the previous lesson.
Example: a strategy with a 40% win rate, a 2R win and a 2% risk per trade. In a simulation of 200 trades, the median largest drawdown is 23%, and only 5% of runs fall by more than 38%. The longest losing streak tends to be around 9 trades.
For this strategy, a 20% drawdown is unpleasant but normal. A 45% drawdown is no longer normal – something has most likely changed. Without this knowledge you would panic at 20%, or still be waiting at 45%.
Tiered rules#
The rules are set in advance and measured from the last peak of the equity curve:
| Drawdown from peak | What you do |
|---|---|
| 10% | Halve the risk per trade. Go through your journal: are you following rules? |
| 20% | Risk down to a quarter. Compare results with the backtest – win rate, wins, streaks. |
| Beyond the test limit | Stop. The strategy goes back to a forward test without money. |
| New peak | Full risk back. |
The "beyond the test limit" row is the switch-off rule from the forward-testing lesson: for example 1.5 times the largest drawdown in the backtest.
What the rule costs and what it gains#
A simulation with the same strategy (2% risk, 200 trades, 20,000 runs; only the 10% and 20% steps). Once the strategy works (40% win rate), once it has stopped working (33% win rate – the expected value is almost zero):
| Win rate | Rules | Account (median) | Drawdown exceeded by only 5% of runs | Chance of a 30%+ drawdown |
|---|---|---|---|---|
| 40% (works) | none | 2.04× | 38% | 19% |
| 40% (works) | tiered | 1.71× | 24% | almost none |
| 33% (does not work) | none | 0.89× | 61% | 75% |
| 33% (does not work) | tiered | 0.90× | 32% | 10% |
The rule is insurance: when the strategy works, it costs you part of the profit (2.04× against 1.71×), because after a drawdown you recover with smaller positions. When the strategy stops working, it prevents the worst – in the worst runs it turns a 61% drawdown into 32%. And as long as you do not know which of the two situations you are in, only a rule can decide.
A simulation, not a forecast
The numbers are a random simulation of a made-up strategy. They show how the rules work, not how much you will make. A real market also has slippage, price gaps and crashes the simulation does not know about.
Why not "just a bit longer"#
In a drawdown the brain switches: it wants to win the loss back fast (bigger positions) or not see it (stop looking and change nothing). Both are the opposite of what the rule does. That is why you:
- write the rules in advance into your trading plan, with numbers,
- calculate from the account, not from a feeling – once a week, note where the account is against its peak,
- do not change them during a drawdown. Make changes only after a review, calmly, and test them.
Common myths#
- "After a drawdown I have to risk more to catch up." That is exactly how accounts blow up.
- "Cutting risk = admitting the strategy does not work." No – it is insurance for when you do not know.
- "A 20% drawdown? The strategy is broken." It depends on what is normal for it. Find out in advance.
Summary
- Find out from a backtest or simulation what drawdown is normal for your strategy.
- Step down risk by rules set in advance: say half at 10%, a quarter at 20%, stop beyond the test limit.
- The rule costs part of the profit when everything works and saves the account when it does not.
- Do not change the rules during a drawdown.