Risk modes: Chill to Extreme
The four risk modes and the hard numbers each one enforces: max leverage, risk per trade, daily loss lock, and the drawdown that auto-pauses the desk.
A risk mode is a contract, not a mood. Choosing one sets hard numbers the engine enforces on every single order; the AI cannot exceed them no matter how confident it claims to be. R-05 risk per trade and R-22 heat are Chill 10%, Normal 15%, Aggressive 20%, or Extreme 25%. R-07 daily-loss is 10%, 15%, 20%, 25%; R-09 auto-pause is 25%, 37.5%, 50%, 62.5%.
| Mode | Max leverage | Risk / trade | Daily loss lock | Auto-pause |
|---|---|---|---|---|
| Chill | 2x | 10% | 10% | 25% drawdown |
| Normal | 5x | 15% | 15% | 37.5% drawdown |
| Aggressive | 10x | 20% | 20% | 50% drawdown |
| Extreme | 20x | 25% | 25% | 62.5% drawdown |
How to read those numbers
- Risk per trade (R-05) is the most a single trade is sized to lose if its stop fills at the planned price, as a percentage of your equity, not of the position. A $500 Normal-mode account has a $75 planned stop-risk ceiling. Gaps, slippage, and liquidity can make realized loss larger.
- Daily loss lock: past this in one day, the desk stands down until the next trading day. No revenge trading, by construction.
- Auto-pause: past this peak-to-now drawdown, the desk locks and will not resume itself. Resuming after a locked drawdown is always a human decision.
- Leverage caps are per-asset too: the engine respects each market's venue limit. Isolated positions must keep the stop inside liquidation; cross positions are tested against projected maintenance after a correlated stop gap because their liquidation risk belongs to the whole collateral pool.
How the whole account participates
| Mode | R-22 portfolio heat | R-25 deployment target | Free-collateral reserve |
|---|---|---|---|
| Chill | 10% | 100% | 20% |
| Normal | 15% | 100% | 15% |
| Aggressive | 20% | 100% | 12% |
| Extreme | 25% | 100% | 10% |
All equity remains active collateral. The reserve is not removed from the account: it absorbs adverse P&L, fees, funding, and stop gaps. The engine may consume the rest as initial margin only when positive-growth setups exist. It never opens a weak trade merely to make a utilization gauge read 100%.
For autonomous and co-pilot orders, R-05 and R-22 are fixed by the selected profile. The current contract requests concentrated full-profile sizing by default, then R-22 is enforced both on the candidate DEX sleeve and across the whole wallet. This favors fewer meaningful positions over many venue-minimum fragments. R-25 is a 100% deployment objective, not permission to invent a trade.
Sizing still responds to the setup
Concentrated sizing requests the profile ceiling; it does not override it. Portfolio heat, factor concentration, venue collateral, maintenance stress, mandatory stops, liquidity, and execution feasibility may resize or veto. A managed ticker also needs at least $300,000 OI on mainnet or $100,000 on testnet, plus setup score 50. The enforced result is visible per trade in the verdict trail.
You pick a mode at desk creation (with a plain-words suggestion box if you're unsure) and can change it later from the mandate panel; changing it rewrites the hard contract and is confirmed explicitly.