Portfolio attribution is a critical tool for investment managers, enabling them to dissect and understand the sources of portfolio performance relative to a benchmark. By breaking down returns into allocation and selection effects, attribution analysis provides transparency into which investment decisions are driving value and where adjustments may be needed.
Menchero’s multiperiod arithmetic attribution methodology is particularly important because it ensures that attribution effects are correctly linked over time, avoiding distortions that can arise from compounding returns. This method allows for a mathematically consistent and interpretable evaluation of performance, making it indispensable for asset managers aiming to refine their investment strategies and demonstrate accountability to stakeholders.
This explanatory note builds upon Menchero’s method by extending its application to stock-level performance attribution. While Menchero’s original work focused on sector-level effects, our enhancements ensure that stock-level attributions properly aggregate to sector-level results, providing a more granular and accurate view of portfolio performance.
Additionally, our extension incorporates trading profits into the attribution framework, ensuring a comprehensive evaluation of active management decisions. These refinements afford portfolio managers deeper insights into the drivers of excess returns and the effectiveness of their investment strategies.
We have developed Menchero’s original method and the extensions discussed here into a Python package to allow for the easy calculation of both stock and sector-level attributions. Additional information can be found at the following links:
Click Here to access the full paper.
Click Here to access the Python package (PyP.org).
Click Here to access the GitHub repository.