Circuit breakers: two losses and max trades

Exchanges halt trading when prices move too far too fast, because they know that a market in that state makes worse decisions than a market that has been forced to pause. Your plan needs the same thing. Two rules do most of the work: what happens after two consecutive losses, and how many trades you may take in a day.

The two-loss rule

After two losses in a row, something changes. The plan says what.

The reason is not that two losses predict a third. At a 45% win rate, two losses in a row happen about 30% of the time and say nothing about the setup. The reason is that two losses change you. The third trade after two losses is taken by someone who wants to get back to flat, and that someone lowers the bar for "the setup is present", sizes up "to make it back faster", and widens the stop "to give it room". The journal will show this in the plan-follow score if you let it run. Better not to let it run.

Options for the rule, from strictest to loosest:

For a v1.0 plan, use the strictest one that fits your time frame. Loosen it at a quarterly review, if the journal shows that the trades after two losses were taken correctly and performed like the rest.

Max trades per day

Every setup fires a certain number of times per session, on average. A plan should allow roughly that many trades and no more, so that the extra trades, the ones that are not quite the setup, have nowhere to go.

If the setup produces one to two valid triggers a day, max trades is three. That leaves room for a valid third without leaving room for a sixth. If you regularly hit the cap, one of two things is true: the setup is firing more than you thought, in which case the filter is too loose, or you are taking trades that are not the setup, in which case the journal's setup-grade field will show it.

The cap counts entries, including scale-ins if you allow them. A trade closed and re-entered is two trades.

Key idea: Two losses in a row do not predict a third, but they change the person taking the third. The two-loss rule and the trade cap are there to stop that person from trading.

Daily and weekly loss limits

These come from the risk plan and they sit above the two-loss rule. The two-loss rule is about your state; the loss limit is about the account. A day can hit the two-loss rule without hitting the daily limit, if the second loss was small. It cannot hit the daily limit without also having crossed the two-loss rule, unless the plan allows very large single trades, which it should not.

When either fires, the platform closes. Not "watch without trading". Closes. Watching without trading is how a rule gets bent at 2:40pm.

The circuit breaker sheet

Put all the breakers in one place on the plan:

LIMITS   Max trades/day: 3        Max positions: 1
         After 2 consecutive losses: done for the day
         Daily stop 2R  ->  platform closed until tomorrow
         Weekly stop 4R ->  platform closed until Monday
         Monthly stop 8R -> platform closed; 1R halved next month
         Sleep < 5h or mood = angry/desperate: half size or no trading

Every line is a number and a consequence. Every consequence is something you do, not something you consider.

Reset conditions

A breaker that fires needs a reset rule, or it will be reset by mood. "Done for the day" resets at the next pre-market routine. "No new entries until review" resets when the review is done and logged, not when you feel better. "Half size for the rest of the month" resets on the first of the month, and only if the monthly review shows the plan was followed; if the losses came from plan-following errors, the next month starts at half size too.

The unusual day

Some days, a breaker fires because the market did something rare, not because you did. A gap through a stop, a halt, a broker outage. The breaker still fires; the platform still closes. The journal gets a tag, "external", and the monthly review decides whether the plan needs a rule for that situation. It does not get overridden on the day, because "this one doesn't count" is the sentence that precedes most of the large losses in most journals.

Try it: Write your LIMITS block using the format above. For each line, write the reset condition next to it. Then look at your last twenty trading days and count how many times each breaker would have fired. If the answer is zero for all of them, the limits are probably too loose; if it is more than five for any of them, the plan has a bigger problem than its limits.

Recap

• After two consecutive losses, something changes: done for the day, half size, or no new entries until review. Strictest that fits, for v1.0.
• Max trades per day is roughly the setup's typical frequency plus one. Hitting the cap regularly means the filter or the setup grading is off.
• Loss limits are about the account; the two-loss rule is about you. When either fires, the platform closes, not just the trading.
• Every breaker has a written reset condition so it cannot be reset by mood.
• Rare events still trip the breaker. "This one doesn't count" is not on the plan.