5% daily loss limit at 1% risk: odds a 6, 8 or 10 trade day reaches it at a 45% win rate

[Fairman]

5% daily loss limit at 1% risk: odds a 6, 8 or 10 trade day reaches it at a 45% win rate

A 5% daily loss limit with 1% risk per trade means five net losses in a day hit it. How likely is that when your win rate is 45% and winners pay 1.5R?

I computed the exact probabilities by tracking the day's running total after each trade. A breach means the total touches minus 5R at any point, and winners count as plus 1.5R, so they delay it.

6 trades in a day: 5.0%
8 trades: 8.5%
10 trades: 12.7%
14 trades: 15.1%
20 trades: 22.0%

Note that the number stops rising for a while, from 10 to 12 trades, because the next possible way to reach minus 5R takes 13 trades.

So a trader who takes 10 trades a day has roughly a one in eight chance of reaching the limit on any given day, despite a positive edge of 0.125R per trade. Over 20 such days the chance of at least one breach is 1 minus 0.873 to the power 20, which is 93%.

Halve the risk to 0.5% per trade and the same 5% limit needs ten net losses. For 20 trades the probability drops from 22.0% to 2.5%, and at 10 trades it's 0.25%.

If the rule is a hard fail, trade count and risk percentage decide your odds more than any entry method does. I'd pick them from a table like this before I looked at setups.

[Fairman]

The 45% win rate is a placeholder, and the table moves with it. At a 50% win rate with the same 1.5R winners the six trade breach probability falls to 3.1%, and for ten trades to 7.5%. At 40% it climbs. The point is the shape, not the single number. Your journal gives the real win rate, so run the calculation with that and with a pessimistic value 5 points lower, as the second figure will tell you how bad a poor month could be. Use a pessimistic value too, since a month with a 40% win rate is not rare in a real sample.

[Fairman]

A common safeguard is a personal daily limit well under the firm's. Stop after 3 net losses, which is 3% at 1% risk, and the firm's 5% limit can only be reached through a slipped stop or a gap. The cost is that you skip trades on days that would have recovered. How much that costs depends on how often it happens, so check your own history: count the days that touched minus 3R and then finished positive, and add up what the later trades on those days made. Your own history will tell you the real figure for that cost.

[Fairman]

Floating loss adds to the picture. If the limit counts open positions, the breach can happen while a trade is still open. After 4.4R of closed losses in a day, the next trade has only 0.6R of room, so if it's sized for 1R, a move 60% of the way to its stop already hits the limit. Treat the limit as closed losses plus the open trade's floating loss, and when the day is already down, cut the size so the full stop still fits inside what is left. That's a small habit and it prevents the breach that a single tick can cause.

[Fairman]

Losses can cluster, which the coin model doesn't allow for. A way to check is to count, in your journal, the days where at least four of the first six trades lost, and compare with the model. For six trades at a 55% loss rate, four or more losses happen on 44% of days. If your count is well above that, your losses are correlated, and the table is optimistic. If it's near 44%, the independent model is a fair first guide for sizing the day. Either way, the count takes ten minutes.