Risk framework

Risk controls are enforced by the worker on every order, not merely checked at configuration time.

Margin pre-flight

Before every order — and again between chunks — the worker performs a margin pre-flight against the live account state on the target venue. If placing the order would breach available margin, the order is not sent. Because the check uses live state rather than a cached snapshot, it accounts for funding debits, fees, and price moves that have occurred since the trade began.

Leverage caps

Margin is checked strictly per-exchange. Spreadr auto-detects the effective leverage available on each venue and treats the user's configured leverage as a cap, never an override: the effective limit is the lower of the venue's allowance and the user's setting. There is no cross-venue netting of margin — each leg must independently satisfy its own venue's requirement.

Slippage limits

Every hedge carries a user-defined slippage limit (see Execution model). This is the single most important user-facing risk parameter: it bounds the worst-case price at which Spreadr will complete a hedge, and therefore bounds the cost of resolving leg risk.

Kill switches

Spreadr provides two independent kill switches:

• A per-user kill switch, which halts all of a single user's activity.
• A global kill switch, which halts the entire platform.

Either can be triggered operationally in response to anomalous conditions, and both take effect on in-flight trades, not just new ones.

Disconnect handling

If a worker loses its WebSocket connection to a venue, it cancels resting orders on that venue, reconnects, reconciles against the venue's reported state, and resumes — rather than continuing to trade against a stale view of the book.