Use this skill after single-name evidence and thesis work when the question is about portfolio fit. It does not execute trades or prescribe an order; it frames position-level risks, constraints, and watch items for the user.
Inputs
- target thesis, valuation scenario, market signals, and risk diligence where available
- current or proposed position size if the user provides it
- portfolio holdings, sector weights, factor exposures, cash, liquidity needs, or risk limits when available
- benchmark, mandate, time horizon, and tax or restriction context when relevant
Workflow
- Load the target archive and any supplied portfolio context.
- Identify whether the task is about a new idea, existing holding, watchlist candidate, trim/add decision, or portfolio-level risk review.
- Map exposures by issuer, sector, industry, geography, currency, factor, customer/supplier dependency, and catalyst overlap.
- Review position sizing context: concentration, liquidity, drawdown contribution, conviction, valuation range, risk severity, and time horizon.
- Identify correlated risks and duplicate bets across holdings or watchlist names.
- Produce user-controlled options such as watch, research more, add to watchlist, size smaller, rebalance candidate, or revisit after catalyst.
- Save the review under
research/targets/<target>/artifacts/portfolio-reviews/or a portfolio archive when one exists.
Read references/portfolio-review-framework.md before writing the review.
Output
Return:
- portfolio context used and missing context
- exposure map and concentration notes
- sizing considerations, not an autonomous trade instruction
- liquidity, correlation, drawdown, and catalyst-overlap risks
- fit with thesis, valuation, and risk evidence
- user-controlled options and watch triggers
- archive files created or updated
Quality Gate
Before finishing:
- state whether actual portfolio holdings were available or assumed absent
- do not recommend or execute trades
- distinguish target-specific risk from portfolio-level exposure
- label stale or missing prices, weights, liquidity, and benchmark data
- avoid precise sizing math without complete portfolio inputs
- preserve user control over any allocation decision