How to Survive Trading Drawdowns

At a glance

The core question: Is this variance, or is the edge gone?
How to answer it: Compare against the simulated distribution
When most people quit: The flat period after the decline, not the decline
Best preparation: Knowing the numbers before you need them

Key takeaways

A drawdown inside the range your simulation produced is information about variance, not about your strategy.
The abandonment point is usually the extended flat period afterwards, when losses have stopped but nothing is improving.
Three explanations exist for any drawdown: normal variance, execution failure, or edge decay, and each has a different remedy.
The response schedule must be written before the drawdown, because judgement during one is systematically poor.
Reducing size is almost always the right intermediate action, because it preserves the ability to participate in the recovery.

The only question that matters

During a drawdown there is exactly one question worth answering: is this within the range of outcomes my strategy produces, or is something structurally different? Everything else follows from the answer, and it is answerable with a number rather than a feeling.

From your Monte Carlo simulation, recorded BEFORE trading:

   Median max drawdown         -19%
   75th percentile             -24%
   95th percentile             -33%
   99th percentile             -41%
   Longest flat period (median)  8 months
   Longest flat period (95th)   19 months

Current situation: -22% drawdown, 6 months underwater

Conclusion: between the median and the 75th percentile
on depth, and below the median on duration.
This is an unremarkable outcome for this strategy.

If instead you were at -47% after 26 months, that falls
outside the simulated distribution entirely, which is
genuine evidence that something has changed.

The comparison that answers it.

The three possible causes

Cause	Evidence for it	Correct response
Normal variance	Drawdown within the simulated range; rules followed; no structural change	Continue; reduce size at predefined levels
Execution failure	Journal shows deviations; slippage above assumption; signals skipped	Fix the process, not the strategy
Edge decay	Outside the simulated range; win rate or payoff structurally changed; effect gone in other markets	Stop; re-examine; consider retirement

Most traders assume the third explanation when the first is correct, because the emotional experience of normal variance and of a broken strategy are identical. The journal and the simulation are what separate them, and without both you will guess, and you will guess wrong in the direction of abandonment.

The psychological trajectory

1
Early decline: confidence intact
Losses are within recent experience. The strategy is followed normally and the drawdown feels routine.
2
Deepening: doubt arrives
The drawdown exceeds anything experienced live. You begin re-examining the backtest and looking for what changed.
3
The trough: maximum pressure
Where most people consider stopping. Ironically, the trough is frequently the worst moment to stop, because the recovery begins from here.
4
Stabilisation: relief, then boredom
Losses stop but gains do not arrive. The account moves sideways.
5
The flat period: where abandonment happens
Months with no progress and no way to distinguish a normal stretch from a dead edge. This is where most strategies are abandoned.
6
Recovery: usually rapid and concentrated
Many strategies recover in a small number of favourable trades or weeks, which are missed entirely by anyone who stopped.

A written drawdown response schedule

Drawdown level	Action	Rationale
0 to 8%	No change	Within routine variance
8 to 15%	Execution audit: check adherence and slippage	Rules out the fixable cause first
15 to 20%	Reduce risk per trade by half	Extends survival; preserves participation
20 to 25%	No new positions; manage existing only	Stops the bleeding without full exit
Beyond 1.5x worst backtested drawdown	Stop entirely; full review before resuming	Evidence now suggests something changed
Outside the simulated 99th percentile	Treat as broken until proven otherwise	Statistically implausible under the original model

Decided before trading, applied mechanically.

The schedule reduces size progressively rather than stopping abruptly, which matters because stopping entirely at the trough converts a temporary decline into a permanent loss and forfeits the recovery.

Practical measures during a drawdown

Reduce size before you feel you need to. Halving risk halves the emotional intensity and preserves the ability to continue.
Reduce how often you check. Daily equity checking during a drawdown amplifies normal variance into a continuous stress signal.
Keep the journal current. It is the evidence that separates execution failure from variance, and it is most often abandoned exactly when it is most needed.
Do not change the strategy. Changes made during a drawdown are fitted to the recent losses and usually make the next period worse.
Do not add funds. Scheduled contributions are fine; reactive ones convert a bounded experiment into an unbounded one.
Re-read your own pre-drawdown notes. The version of you who wrote the expectations was better informed about the strategy than the version experiencing the decline.
Talk to someone who understands the base rates, rather than to someone who will encourage you to stop.

Frequently asked questions

How do I know if my strategy has stopped working?

Compare the current drawdown and its duration against the distribution from your Monte Carlo simulation. Inside that range, it is variance. Outside the 95th or 99th percentile, it is genuine evidence of change. Without the simulation, the comparison is unavailable and you will rely on how it feels, which is systematically misleading.

Should I stop trading during a drawdown?

Reduce size at predefined levels rather than stopping abruptly, and stop entirely only at a threshold you wrote down in advance. Stopping at an arbitrary point usually means stopping near the trough, which converts a temporary decline into a permanent loss and forfeits the recovery.

Why do traders quit during flat periods rather than crashes?

Because a sharp decline has an identifiable cause and a clear narrative, while a long flat period provides no information at all. Months of no progress with no way to distinguish normal variance from a dead edge is psychologically harder to sustain than a fast loss, and it lasts far longer.

How long do drawdowns usually last?

It depends entirely on the strategy. Short-term mean reversion recovers in weeks to months; diversified trend following has historically spent one to three years underwater; buy-and-hold equity drawdowns have taken five years or more. Knowing your own strategy’s historical distribution is essential preparation.

Should I change my strategy after a bad period?

Almost never during the period itself. Changes made in a drawdown are fitted to the specific losses that caused the discomfort and typically degrade future performance. Confine strategy changes to scheduled quarterly reviews, conducted against the full record rather than the recent experience.

Test this idea before you trade it

Describe the rules in plain language and AlgoTrader AI turns them into a structured strategy blueprint with a configurable historical backtest, cost assumptions, and exportable code.

Build a backtest

Keep reading

Referenced by

Educational use only. This guide explains how a strategy works. It is not investment advice, not a recommendation, and no result described here is a forecast. Test any approach on historical and out-of-sample data, size positions conservatively, and never risk money you cannot afford to lose.