portfolio-risk-exposure-controller

Submit the proposed trade along with current positions, exposure reports, and any factor or liquidity data. The review returns a control verdict first, followed by a nine-section analysis of incremental portfolio impact and a 1-to-5 risk scorecard.

Expected inputs include current positions, proposed trades, exposure and factor reports, volatility and correlation summaries, liquidity metrics, scenario results, loss limits, watchlists, and event calendars. Missing data is handled with clearly labeled narrow assumptions.

Every substantive review ends with one of four verdicts: Allow, Allow with size reduction, Allow only with hedge or control change, or Block. Non-allow verdicts state exactly what must change for the trade to become acceptable.

Yes, it supports three review modes: proposed trade review, strategy review, and existing book review. Book reviews check limit compliance, hidden accumulating exposures, and whether the portfolio could be de-risked under stress.

The review never invents exposures, correlations, or liquidity capacity; it uses only supplied data or explicitly labeled assumptions. When data is incomplete, it lists only the missing fields that would materially change the verdict.

No, out-of-scope questions are routed to specialist roles: execution feasibility to an Execution Quality and TCA Analyst, thesis questions to a Thesis Drift agent, macro context to a Macro and Market News Analyst, and limit breaches to a Trading Compliance and Surveillance Agent.