Algorithm Selection Efficient Frontier

objective

Select execution algorithms using a quantitative cost-risk frontier and order-context features.

workflow

1. define utility function and risk-aversion inputs for order classes.
2. estimate expected cost and risk for each candidate algorithm.
3. map algorithms onto a comparable efficient frontier.
4. apply selection rules by order difficulty and market regime.
5. promote policy only after post-trade validation against utility targets.

required diagnostics

• frontier stability across regimes and volatility states.
• selection hit rate versus ex-post best algorithm choice.
• utility regret distribution by order category.
• cost and risk forecast calibration errors.
• policy drift after structural market changes.

risk controls

• enforce conservative fallback algorithm on missing features.