Factor Regression

Factor Regression for Return Attribution and Risk Decomposition

Decompose asset or portfolio returns into factor exposures to understand what has been driving performance and what risks remain unexplained.

Factor Regression is for return attribution: how much of an asset or portfolio return is explained by common factors like market beta, size, value, momentum, and quality.

This page covers the regression approach, factor model options, and how to interpret factor loadings and residuals.

Open Factor RegressionRead Factor Regression Docs

What this tool explains

  • Factor exposures (betas) for individual assets or portfolios.
  • How much return variance common factors explain vs idiosyncratic risk.
  • Whether a portfolio is unintentionally concentrated in specific factor bets.

When to use factor regression

Use factor regression when you need to understand what is driving returns rather than just measuring them. It is especially useful before and after portfolio optimization to verify that intended exposures match actual exposures.

Inputs that shape the result

  • The factor model and regression window.
  • Whether the asset set covers the full portfolio or a subset.
  • How to interpret low R-squared or unstable factor loadings.

Complementary tools

  • PCA for structural concentration beyond named factors.
  • Portfolio Optimizer for targeting specific factor exposures.
  • Asset Analyzer for single-asset diagnostics.

FAQ

Does a high R-squared mean the model is correct?

Not necessarily. A high R-squared means the factors explain historical variance well, but it does not guarantee the same exposures will persist or that the factor model is complete.

Related Pages

Principal Component Analysis

Structural concentration beyond named factor models.

Factor Regression Docs

Model options, interpretation, and caveats.

Portfolio Optimizer

Target specific factor exposures in allocation design.