Skip to content
All library documents

Evaluating Equity Factors and Dynamically Combining Them by IC

Article BigQuant
Read the original

Summary

The document describes a framework for evaluating equity factors and combining selected factors into a portfolio. It estimates factor returns with periodic cross-sectional robust regressions, measures the relationship between factor exposures and subsequent returns using information coefficients, and uses portfolio sorts to examine whether returns change monotonically across exposure groups. Evaluation metrics include return significance, cumulative performance, IC and its information ratio, long-short returns, drawdown, and turnover.

The framework covers ten factor families, including value, size, growth, quality, momentum, volatility, technical, liquidity, and analyst-related measures. It proposes scoring factors across five measures—factor return, its t-statistic, IC, IR, and monotonicity—then dynamically adjusting portfolio weights based on factor IC histories. The report says a rolling 36-month approach with 150 holdings produced its strongest reported result, with an information ratio of 3.67 and annualized return of 31%, outperforming static weighting in its tests. These are reported historical findings; the document calls the work preliminary and does not provide the underlying report's detailed validation or implementation here.

Key ideas

  • Periodic cross-sectional robust regressions can estimate factor returns, while IC measures the association between factor exposures and next-period returns.
  • Portfolio sorts help test whether a factor's returns are monotonic across exposure groups.
  • Factor selection combines return, statistical significance, IC, IR, and monotonicity rather than relying on a single metric.
  • The proposed portfolio dynamically adjusts factor combination using the history of factor ICs.
  • The reported preferred settings and returns are historical test results, and the document describes the model as preliminary.

Tags

EquitiesFactor investingStatisticsBacktestingPortfolio construction

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