Rules & Alerts — Market Monitoring and Automatic Stress Scenarios

From market intelligence to stress testing

StressGen monitors market signals around the clock. Rules turn qualifying observations into alerts and stress scenarios, helping your team decide what to investigate and which portfolios to stress test.

Rules are checked around the clock, so your team hears about a qualifying move without watching screens. Each reading is as fresh as its source.

Create and enable a company rule

Open Settings → Rules & Alerts as a company admin. If beta access is disabled, contact your company admin or StressGen before planning a workflow around company rules.

1. 1Name the rule and select its source, source identifier, rule type, and threshold. Add a comparison window when required.
2. 2Choose severity, cooldown, and a generation owner. The owner defaults to the creator and must remain an active company admin with a seat.
3. 3Save the rule. New company rules start paused; inspect the configuration and run an observational test before enabling.
4. 4Explicitly enable the rule after reviewing automatic generation and its use of the full shared company allowance.

Admins can edit, test, pause, or delete rules. Pausing stops future scheduled triggers. A cooldown limits repeat triggers; each qualifying trigger episode produces one event rather than a new event on every scheduler tick.

Rule types and thresholds

The ten supported rule types are:

• Level above
• Level below
• Change above (window)
• Change below (window)
• Probability crosses above
• Probability crosses below
• Daily % change above
• Daily % change below
• Implied EV crosses above
• Implied EV crosses below

Level rules compare an observed value with a threshold. Change rules compare values over a configured window in days. Use the source's units, and preserve the sign of the move you want to detect.

Daily percentage-change rules require a market symbol, such as ^GSPC; a threshold of −5 means a 5% daily drop. Probability-crossing and implied expected-value rules require a prediction-market source. Probability thresholds run from 0 to 1 (0.7 means 70%); implied expected-value thresholds use the event's native unit, such as basis points. FRED rules use a series identifier, such as T10Y2Y. The form validates compatible rule types, sources, thresholds, and required windows.

Alert delivery

A qualifying company trigger creates a private event and notifies eligible company members through their configured in-app and email channels. Review notification preferences for immediate email and in-app delivery, severity filters, and quiet hours. Delivery timing follows those preferences. The event records the automatic generation outcome and links to the private scenario when available.

Automatic scenarios and company allowance

Each qualifying trigger requests one private hypothetical stress scenario, grounded in the trigger evidence and using the rule's severity. The selected generation owner provides the required company admin and seat eligibility. Automatic and manual generations draw from the same full company allowance; there is no separate automatic-generation cap.

If the allowance is exhausted or the owner becomes ineligible, monitoring and alerts continue. History records why generation was skipped. Skipped triggers are not automatically replayed when the allowance renews. Failed generations follow the existing billing reservation and refund lifecycle; retries reuse the original generation reservation rather than consuming a second allowance unit for the same event.

Scenario generation does not run a portfolio stress test automatically. Review the scenario, then apply it to your chosen portfolios through the stress testing workflow.

Check rule history and delivery

• Check that the rule is enabled and its cooldown has elapsed. Inspect the last evaluation time and trigger history.
• A recent evaluation is a scheduler heartbeat, not proof of fresh source data. Inspect observation details; missing data leaves the condition unknown.
• If the heartbeat is stale beyond the source's evaluation schedule, ask your admin to investigate monitoring availability.
• If an event exists but no message arrived, check immediate email or live-risk email preferences and severity filters. Quiet hours mute in-app alerts; they do not mute email. Company-rule alerts are not delivered through digests.
• If an alert has no scenario, inspect the recorded generation outcome, remaining shared allowance, and the owner's active admin and seat status.