Performance Attribution Through a Factor Lens

We propose an adjustment of standard regression-based factor attribution to address a common issue: implementation constraints often mean that investors cannot realize the full potential of a factor strategy, but standard attribution analysis assumes that they can – leaving part of the portfolio return unexplained. Our alternative classifies stocks based on their factor exposures and identifies the segments most responsible for the unexplained portfolio return. The resulting nonlinear factor attribution better reconciles realized performance with the investment process, mitigating both the long-term average and short-term volatility of any residual. While our focus is on equity investing, the proposed methodology for factor attribution also applies to other asset classes