Skip to content
All library documents

Portfolio Performance and Risk Attribution Methods

Article BigQuant

Summary

The document explains how portfolio attribution compares a portfolio with its benchmark and decomposes excess return into drivers. It distinguishes return-based approaches, including Treynor-Mazuy, Henriksson-Merton, and related factor extensions, from holdings-based approaches such as Brinson attribution and multifactor models. Holdings-based analysis can describe investment decisions across more dimensions than return-based analysis alone.

It also separates risk attribution into ex-ante analysis, which relies on multifactor risk models, and ex-post analysis, which uses realized portfolio volatility. The document describes tools built around an A-share factor risk model: one analyzes periodic portfolio holdings, while another supports public-fund attribution. It provides no methodology details or empirical results for these tools. Model failure and shocks from extreme market conditions are stated limitations.

Key ideas

  • Performance attribution decomposes portfolio returns relative to a benchmark into decision-related sources.
  • Return-based and holdings-based attribution use different information and can reveal different aspects of manager decisions.
  • Brinson and multifactor models are cited as holdings-based approaches.
  • Ex-ante risk attribution uses factor models, while ex-post analysis uses realized volatility.
  • The described tools depend on an A-share risk model and are exposed to model failure and extreme markets.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.