performance-attribution

Decompose portfolio excess returns into allocation, selection, and interaction effects.

Updated May 5, 2026
One-click install
npx skills add https://github.com/wudye/traderAssistHK --skill performance-attribution-wudye
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: performance-attribution
Source: https://github.com/wudye/traderAssistHK/tree/main/backend/src/skills/performance-attribution
Command: npx skills add https://github.com/wudye/traderAssistHK --skill performance-attribution-wudye

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Performance attribution explains the sources of a portfolio’s returns, turning raw performance into actionable insight on allocation, selection, factor exposure, and timing relative to a benchmark.

Core Features & Use Cases

  • Brinson allocation/selection attribution: Quantifies how sector (or group) weight differences and within-group stock picking drive excess return, including interaction effects.
  • Factor alpha/beta decomposition: Separates market exposure and factor tilts using single- or multi-factor regressions (e.g., Fama-French style factors with interpretable proxies).
  • Market-timing evaluation & benchmark comparison: Tests timing skill with models like Treynor-Mazuy and Henriksson-Merton and summarizes results with risk-adjusted metrics and rolling stability.

Example

If a strategy shows strong monthly returns versus a benchmark, use this skill to determine whether the outperformance came mainly from sector allocation, stock selection, factor tilts (alpha vs beta), or genuine timing improvements.

Quick Start

Ask for a performance attribution report for your strategy and its benchmark that breaks excess return into Brinson effects, estimates factor betas/alpha, evaluates timing using Treynor-Mazuy and Henriksson-Merton, and summarizes results with Sharpe and information ratio.

Frequently Asked Questions about performance-attribution

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I decompose portfolio excess returns into allocation and selection effects?

Portfolio excess returns are decomposed using the Brinson model, which quantifies how sector weight differences and within-group stock picking drive performance, including interaction effects, relative to a benchmark.

What is factor decomposition in performance attribution and how does it separate alpha?

Factor decomposition in performance attribution uses single- or multi-factor regressions, such as Fama-French style factors, to separate genuine alpha from market exposure and factor tilts based on benchmark and portfolio return series.

Can I evaluate market timing skill using Treynor-Mazuy and Henriksson-Merton models?

Yes, market timing skill is evaluated using Treynor-Mazuy and Henriksson-Merton models to calculate TM/HM gamma parameters, requiring appropriate sample sizes of portfolio and benchmark returns to assess timing ability.

How do I run a multi-period performance attribution analysis for rolling evaluation?

Multi-period performance attribution applies Brinson allocation and selection effects across linked single periods, enabling rolling evaluation of sector and stock-level strategies in equity and multi-asset contexts using portfolio and benchmark weights.

Do I need factor return series to calculate risk-adjusted metrics and information ratio?

Factor return series are specifically required for factor alpha and beta decomposition, while risk-adjusted metrics like Sharpe and information ratio summarize overall attribution results from portfolio and benchmark return data.

Why does my strategy outperform the benchmark and how can I identify the source?

Strategy outperformance sources are identified by breaking down excess return into Brinson allocation effects, factor tilts, alpha generation, and genuine timing improvements, turning raw performance into actionable attribution insight.