Use case · Monitor

Track an open spread's health and Greeks

compute_saved_spread_analytics returns live position Greeks for a saved spread; compute_health_components grades it against its plan and rolls the worst case into one status. This is the post-fill half of the engine — the same monitoring that drives the position views in Spread Foundry and MeridianScope.

When to use it

• You have an open, saved spread and need its current net and per-leg Greeks, theta decay, and progress toward target — not a fresh candidate score.
• You want a typed health readout that flags when a position needs attention (time stop, max loss, stale thesis, thinning liquidity).
• You need a single aggregate status and an alarm count to drive a dashboard or an alert.

Example

use ferro_spread::{
    compute_saved_spread_analytics, compute_health_components,
    aggregate_status, alarmed_count, HealthStatus,
};

// Position analytics for a saved spread
let pos = compute_saved_spread_analytics(&inputs)?;
println!("net delta {}  theta/day {}", pos.position.delta, pos.theta_decay_per_day);
println!("target progress {}", pos.target_progress);
println!("short-strike dist {}%", pos.short_strike_distance_spot_pct);
for leg in &pos.per_leg {
    // leg.delta / .gamma / .theta / .vega
}

// Health scoring against the position's plan
let health = compute_health_components(&hc_inputs);
let status = aggregate_status(&health);        // Ok | Watch | Alarm
if status == HealthStatus::Alarm {
    // surface the alarmed components (alarmed_count(&health)) to the UI
}
# Ok::<(), Box<dyn std::error::Error>>(())

Position analytics

Health components

compute_health_components returns a Vec<HealthComponent>, one per HealthComponentKind; aggregate_status reduces them to the worst status and alarmed_count counts the ones firing:

Notes

• Fail-loud, like the rest of the engine: a missing mark or IV yields an Unavailable confidence rather than a fabricated Greek. See the conventions.
• HealthStatus carries a .worse() combinator (Alarm > Watch > Ok), so aggregate_status always reflects the most urgent component.
• Monitoring consumes a saved spread — the output of the construct/rank path once a candidate has been taken.