Your trading strategy will fail. Here is how to plan for it.

Every systematic strategy needs a kill switch rule written before the strategy goes live. The kill switch is the rule that closes a strategy permanently when its behaviour exits the envelope you built it for.

Strategy decay does not announce itself. It arrives dressed as an ordinary drawdown. The kill switch is the rule that takes the strategy off the book when its behaviour leaves the envelope you built it for. Not a stop-loss on a trade. A stop-loss on the strategy itself.

Why does strategy decay cost you so much before you see it?

Strategy decay isn't a sudden event. It looks like normal drawdown, then it looks like a longer drawdown, then it looks like a drawdown that doesn't recover. By the time the equity curve has told you clearly that the edge is gone, you have already given back twelve to eighteen months of returns trying to ride it out.

Three components of a real kill switch rule

A statistical drawdown threshold tied to the strategy's backtested distribution. The rule is not 'kill at 20% drawdown' as a round number. The rule is 'kill when the strategy enters its backtested 95th-percentile drawdown'. Use the Monte Carlo bootstrap on the trade sequence to get the 95th-percentile max drawdown, and write that number into the kill switch. The threshold is strategy-specific because the distribution is strategy-specific.

A recovery-time threshold. Some strategies have wider drawdown distributions but shorter recovery times. The kill switch needs both. The rule is something like 'kill if drawdown exceeds the 95th-percentile AND recovery time exceeds 1.5x the 90th-percentile backtested recovery'. The two thresholds together catch the strategies that are merely unlucky differently from the strategies that have decayed.

A regime-mismatch flag. Run a rolling factor decomposition on the strategy's returns. If the factor exposures it carries today are materially different from the factor exposures it carried in the backtest, the strategy has drifted. The drift alone might not be a kill signal; combined with the drawdown and recovery thresholds, it's the third piece of evidence.

When the rule fires, the strategy comes off the book. Not paused. Off. Pausing is what discretionary traders do; it leaves the door open for the override that destroys systematic discipline.

The hard part isn't writing the rule

The hard part is honouring it when it fires.

Every systematic operator who has run real money knows the feeling: the strategy hits the kill threshold, you stare at the chart, you tell yourself the regime is about to shift, and you give it one more month. That month is where careers end. The whole point of the kill switch is to remove that decision from the operator in the moment it matters most.

The strategies I have killed are the ones that have funded the strategies that have compounded.

For the 28-strategy book I run, every strategy has all three thresholds written into the system before it goes live. The trigger is automated. The override is one-way, in writing, and reviewed quarterly. The strategies I have killed are the ones that have funded the strategies that have compounded.

If you don't have kill switches on your live book today, that's the highest-leverage piece of architecture you can add this month.