
from finance_skills127
Decompose portfolio returns (Brinson, factor, fixed-income, currency) into allocation, selection, interaction and factor contributions to explain performance.
Provides detailed explanations and worked examples for portfolio performance attribution: single-period Brinson-Fachler, multi-period linking (Carino/Menchero), factor-based decomposition, fixed-income roll/curve/spread effects, and currency attribution. Includes formulas and sample calculations so agents can explain sources of active return and alpha.
Use when a user asks why a portfolio out- or under-performed, requests Brinson attribution, multi-period linking, factor contributions, fixed-income attribution, or currency effects. Useful for advisers preparing client review notes, performance reports, or for educational demos.
scripts/ folder demonstrating calculations and verification (has_scripts=true)--verify script mode for numeric checks.Best for Claude Code / Python-capable agent runtimes because the repo includes runnable Python demos and numeric examples.
Well-structured finance skill implementing Brinson-Fachler equity attribution, OLS factor-based attribution, and Carino multi-period linking. Script ran successfully with numpy (PEP 723 inline deps). The Carino linking demo shows a numerical mismatch between linked effects sum and geometric active return, indicating a minor implementation issue in the multi-period linker. Otherwise clean, well-documented code with proper type hints, docstrings, and a --verify mode.
Niche but competent finance skill. No security concerns whatsoever — pure computation, no network access, no credentials, no destructive operations. SKILL.md is thorough with worked examples matching the code. Minor numerical accuracy issue in the Carino linking demo output.