Risk limits are the rules you write while you are calm so that the bot can hold the line when the market is not. Two of them do most of the work: a per-trade limit that caps what any single trade can cost you, and a daily limit that stops the bleeding when a session goes against you. Set both before you go live, and one bad trade — or one bad day — can no longer sink the account.
Why size the loss, not the win
Disciplined sizing starts from the downside. You decide the most you are willing to lose on a trade first, then let that decision determine the order size. Hope sizes positions by how much you might make; discipline sizes them by how much you can afford to lose. The second approach is the one that is still standing after a losing streak.
Steps
- Decide risk per trade. Pick the most you are willing to lose on any single trade, expressed as a small percentage of the account rather than a gut-feel dollar amount.
- Translate it into position size. Use your stop distance to convert that risk amount into an order size, so the math — not emotion — sets how big the position is.
- Set a daily loss limit. Choose a hard daily loss cap. Once hit, the bot stops trading for the day instead of chasing the loss back.
- Add a consecutive-loss cooldown. Force a pause after a defined run of losses, breaking the streak before it has a chance to compound.
- Confirm in paper first. Validate every limit in paper mode and review the audit trail before any of it touches live funds.
Limits are a system, not a single number
Per-trade sizing, the daily cap, and the cooldown reinforce one another: the first contains a single mistake, the second contains a bad day, and the third contains a bad mood. Behind all three sits the kill switch for anything that does not fit a tidy rule. Together they form a policy the bot enforces for you — which is far more reliable than willpower in the middle of a drawdown.
Protection, not a promise
Risk limits protect capital and enforce discipline. They do not make a strategy profitable and they do not promise returns — they simply make sure that when you are wrong, you are wrong by an amount you chose in advance. Nothing here is financial advice.
For the override that sits above these limits, read the emergency kill switch, explained. To see how limits apply at execution time, see paper mode vs live mode on a Kraken bot, and for the fundamentals behind sizing, the education pillar.