Runtime Configuration
Set trading defaults for new deployments, then let account-wide risk limits (leverage cap, drawdown caps, per-strategy limits, loss-streak pauses) act automatically in code.
Configuration in Warren has two halves. The first is a set of trading defaults you choose: what a new bot or position should assume when you do not say otherwise. The second is a set of account-wide risk limits that you set once and then act automatically, in code, without waiting to be asked.
Configurable preferences
These defaults apply to new deployments and can be overridden per bot or per position:
- Preferred exchanges
- Preferred trading pairs
- Amount per trade (stake)
- Maximum open trades
- Trading mode preferences (paper or live)
- Spot or futures preferences
The platform also calculates required capital dynamically from your runtime configuration, so the stake, leverage, and open-trade settings together determine how much of your trading allocation a deployment reserves.
Account-wide risk limits
Beyond per-bot guards, Warren enforces a layer of account-wide rules that apply across every bot and across Warren's own discretionary book. You set them once; the engine applies them automatically. They live on the Risk screen, and when one is breached the affected trading auto-stops and you are alerted.
| Parameter | What it does | How it acts |
|---|---|---|
| Leverage cap | A ceiling on position leverage, capped at a small single-digit multiple, well below the venue's native 20–125×. | Enforced in code. Blocks risky configuration combinations before a bot can start, not after. |
| Max Drawdown | A ceiling on account-wide peak-to-trough loss (the total drawdown cap). | Auto-stops trading when cumulative loss hits the limit. |
| Max Daily Loss | A ceiling on realized loss inside a single day. | Auto-stops trading for the day when the day's loss reaches the limit. |
| Per-strategy capital limits | A cap on how much of the trading allocation any single strategy may commit. | Keeps one bot from monopolizing the shared capital pool. |
| Loss-streak pauses | A consecutive-loss circuit breaker. | Pauses a strategy after a run of losses so it can be reviewed before it keeps bleeding. |
| Max Open Trades | A cap on concurrent open positions. | Blocks new entries once the ceiling is reached. |
These limits are hard mechanisms, not advice. Over-leverage is the fastest path to liquidation: at high leverage the liquidation distance shrinks below normal market noise, so the leverage cap is what makes the rest of the risk framework meaningful. The cap and the auto-stops are enforced in code and cannot be talked past.
Per-bot guards (auto-stop after N trades, a take-profit target, a trailing stop, and a max-loss cap) stack on top of these account-wide limits. See Execution for the per-bot controls.
Strategy activation
After allocation is configured, Warren activates the selected strategy. The current runtime experience displays:
- Selected strategy
- Trading mode
- Allocation amount
- Connected exchange
- Selected pair
- Runtime status
- Performance visibility
Strategies immediately transition into monitoring after activation.
Continue
Capital Allocation
The three-bucket split your bots and Warren's book trade against.
Dashboard Overview
Portfolio metrics, regime visibility, and per-bot runtime state.
Exchange Connectivity
Non-custodial API connections and the leverage cap, in depth.
Next: Dashboard Overview
System Architecture
Four cooperating layers, designed so that honesty and safety are properties of the system, not a hope pinned on a prompt.
FAQ
Common questions about Warren: why it abstains, blocked and zero-trade backtests, the leverage cap, auto-stopped bots, marketplace vetting, custody, exchanges, authentication, and trading modes.