Portfolio-aware Sizing

What is Portfolio-aware Sizing?

Portfolio-aware sizing sets each position's size with reference to everything else the account already holds, rather than treating every trade as an island. Per-trade schemes like fixed fractional answer one question: how much to risk on this stop distance. Portfolio-aware schemes add the questions that follow: how correlated is the new trade to the open book, how much total risk is already deployed, and how concentrated the book is in one direction or factor.

Common implementations include correlation haircuts (shrink size when the new trade moves with existing positions, since correlated positions behave as one larger position), risk parity (allocate so each position contributes a similar share of portfolio volatility, extending volatility-targeted sizing to the book level), and capped ensembles, where several strategies each request size and an aggregate limiter scales the requests down so combined exposure stays inside a fixed budget. The shared premise is that risk lives at the portfolio level, and correlation is the thing per-trade math cannot see.

How traders use it

• As total-heat caps: sum the open risk of all positions (distance to stop times size) and refuse or shrink new entries once the sum reaches a preset ceiling, regardless of how good the next signal looks.
• As correlation-adjusted allocation: before adding a position, measure its correlation to the current book over a recent window and scale size down as correlation rises, so five correlated trades cannot quietly become one five-times-size trade.
• As an ensemble governor in multi-strategy systems: each sub-strategy proposes a size, and a portfolio layer rescales the set (risk-parity weights, or a Kelly-style cap on the aggregate fraction) so the combined book, not any single signal, defines the account's risk.