Survival is a design constraint.
Risk is addressed across research validity, exposure, execution and operating authority before performance is considered credible.
A strong backtest can coexist with weak research controls. A strong signal can coexist with poor portfolio construction. We govern the whole system.
Four connected risk layers.
Research risk
Leakage, overfitting, unstable parameters, incomplete provenance and weak generalisation can invalidate apparent edge before capital is considered.
Market and portfolio risk
Exposure limits, concentration, liquidity, interactions and regime dependence determine how research behaves when assembled into a portfolio.
Execution and operational risk
Slippage, transaction costs, timing, reconciliation, incident handling and deployment controls matter because paper alpha that cannot survive implementation is not investable.
Governance risk
Authority should be explicit. Material state transitions, approvals and production changes are intended to be traceable, reviewable and reversible where practicable.
Return objectives remain subordinate to controlled exposure.
The firm’s risk philosophy prioritises continued operating capacity, liquidity and evidence quality over maximising short-term reported return.